Slashdot Mirror

In a news bulletin, University of California, Berkeley announces that its "Foundations of Data Science" course is "being offered free online this spring for the first time through the campus's online education hub, edX." From the report: The course -- Data 8X (Foundations of Data Science) -- covers everything from testing hypotheses, applying statistical inferences, visualizing distributions and drawing conclusions, all while coding in Python and using real-world data sets. One lesson might take economic data from different countries over the years to track global economic growth. The next might use a data set of cell samples to create a classification algorithm that can diagnose breast cancer. (Learn more from a video on the Berkeley data science website.) The online program is based on the Foundations of Data Science course that Berkeley launched on campus in 2015 and now has more than 1,000 students enrolling every semester. The Foundations of Data Science edX Professional Certificate program is a sequence of three five-week courses taught by three winners of Berkeley's top teaching honor, the Distinguished Teaching Award: DeNero, statistics professor Ani Adhikari and computer science professor David Wagner. The first of the three parts has already started (9 a.m. on April 2), but enrollment will remain open after the course begins. Furthermore, anyone in the world can enroll for free but those who want to earn the certificate will need to pay.

In a news bulletin, University of California, Berkeley announces that its "Foundations of Data Science" course is "being offered free online this spring for the first time through the campus's online education hub, edX." From the report: The course -- Data 8X (Foundations of Data Science) -- covers everything from testing hypotheses, applying statistical inferences, visualizing distributions and drawing conclusions, all while coding in Python and using real-world data sets. One lesson might take economic data from different countries over the years to track global economic growth. The next might use a data set of cell samples to create a classification algorithm that can diagnose breast cancer. (Learn more from a video on the Berkeley data science website.) The online program is based on the Foundations of Data Science course that Berkeley launched on campus in 2015 and now has more than 1,000 students enrolling every semester. The Foundations of Data Science edX Professional Certificate program is a sequence of three five-week courses taught by three winners of Berkeley's top teaching honor, the Distinguished Teaching Award: DeNero, statistics professor Ani Adhikari and computer science professor David Wagner. The first of the three parts has already started (9 a.m. on April 2), but enrollment will remain open after the course begins. Furthermore, anyone in the world can enroll for free but those who want to earn the certificate will need to pay.

In a news bulletin, University of California, Berkeley announces that its "Foundations of Data Science" course is "being offered free online this spring for the first time through the campus's online education hub, edX." From the report: The course -- Data 8X (Foundations of Data Science) -- covers everything from testing hypotheses, applying statistical inferences, visualizing distributions and drawing conclusions, all while coding in Python and using real-world data sets. One lesson might take economic data from different countries over the years to track global economic growth. The next might use a data set of cell samples to create a classification algorithm that can diagnose breast cancer. (Learn more from a video on the Berkeley data science website.) The online program is based on the Foundations of Data Science course that Berkeley launched on campus in 2015 and now has more than 1,000 students enrolling every semester. The Foundations of Data Science edX Professional Certificate program is a sequence of three five-week courses taught by three winners of Berkeley's top teaching honor, the Distinguished Teaching Award: DeNero, statistics professor Ani Adhikari and computer science professor David Wagner. The first of the three parts has already started (9 a.m. on April 2), but enrollment will remain open after the course begins. Furthermore, anyone in the world can enroll for free but those who want to earn the certificate will need to pay.

In a news bulletin, University of California, Berkeley announces that its "Foundations of Data Science" course is "being offered free online this spring for the first time through the campus's online education hub, edX." From the report: The course -- Data 8X (Foundations of Data Science) -- covers everything from testing hypotheses, applying statistical inferences, visualizing distributions and drawing conclusions, all while coding in Python and using real-world data sets. One lesson might take economic data from different countries over the years to track global economic growth. The next might use a data set of cell samples to create a classification algorithm that can diagnose breast cancer. (Learn more from a video on the Berkeley data science website.) The online program is based on the Foundations of Data Science course that Berkeley launched on campus in 2015 and now has more than 1,000 students enrolling every semester. The Foundations of Data Science edX Professional Certificate program is a sequence of three five-week courses taught by three winners of Berkeley's top teaching honor, the Distinguished Teaching Award: DeNero, statistics professor Ani Adhikari and computer science professor David Wagner. The first of the three parts has already started (9 a.m. on April 2), but enrollment will remain open after the course begins. Furthermore, anyone in the world can enroll for free but those who want to earn the certificate will need to pay.

Harvard computer science professor David J. Malan "is pretty amazing!" says long-time education-watcher theodp. And he's sharing a link to the online version of Malan's famous CS50 class, "if you can't pony up the estimated $63,025-a-year sticker price to take 'the quintessential Harvard (and Yale!) course' on campus."

KQED's education site "MindShift" reports: Malan's class attracts students who have never taken computer science before, as well as kids who have been coding a long time. His goal with this diverse group of learners is to create a community that's equal and collaborative. One way he does this is by asking students to self-identify by comfort level. Those groups become different section levels, and they sometimes get different homework, but harder assignments are not worth more credit. Malan said recently that the "less comfortable" group has dominated his 700-person course. "At the end of the day all students are treated with the same expectations," said Malan, speaking at the Building Learning Communities conference in Boston.

Students are graded based on each individual's growth; Malan and his team of teaching assistants don't use absolute measures when assigning grades. Instead, they look at scope, how hard the student tried, correctness, how right the work was, style, how aesthetic the code is, and design, which is the most subjective. When it's time to assign grades, Malan and his teaching fellows have lots of in-depth conversations about how each student has improved relative to where he or she started...
The course includes a tool that rewrites error messages to make them easier to understand, plus a code-checking tool which they're planning to open source. There's also a cloud-based IDE which "allows students to access their code from multiple locations," though students can also submit their code through GitHub. (The original submission complains that Harvard's students are "coddled.") But Malan says the class works partly because there's an intentionally social aspect to it -- including numerous teaching assistants holding office hours in public spaces and "the human structure within the course." Guest lecturers have even included Mark Zuckerberg and Steve Ballmer.

But all these technical details don't really capture the wild flavor of the course and all of its multimedia bells and whistles. Malan's fast-paced lectures often close with relevant clips from movies -- for example, a lecture on cryptography which ended with video from a movie you'd see "if you turn on your TV on December 24th."

Anant Agarwal is a professor of Electrical Engineering and Computer Science at MIT and the CEO of edX. A massive open online course platform founded by MIT and Harvard, edX offers numerous courses on a wide variety of subjects. As of 2014 edX had more than 4 million students taking more than 500 courses online. The organization has developed open-source software called Open edX that powers edX courses and is freely available online. Mr. Agarwal has agreed to take some time out of his schedule and answer your questions about edX and the future of learning. As usual, ask as many as you'd like, but please, one question per post.

Anant Agarwal is a professor of Electrical Engineering and Computer Science at MIT and the CEO of edX. A massive open online course platform founded by MIT and Harvard, edX offers numerous courses on a wide variety of subjects. As of 2014 edX had more than 4 million students taking more than 500 courses online. The organization has developed open-source software called Open edX that powers edX courses and is freely available online. Mr. Agarwal has agreed to take some time out of his schedule and answer your questions about edX and the future of learning. As usual, ask as many as you'd like, but please, one question per post.

Anant Agarwal is a professor of Electrical Engineering and Computer Science at MIT and the CEO of edX. A massive open online course platform founded by MIT and Harvard, edX offers numerous courses on a wide variety of subjects. As of 2014 edX had more than 4 million students taking more than 500 courses online. The organization has developed open-source software called Open edX that powers edX courses and is freely available online. Mr. Agarwal has agreed to take some time out of his schedule and answer your questions about edX and the future of learning. As usual, ask as many as you'd like, but please, one question per post.

theodp writes: This weekend, reports GeekWire, many of the original Excel team members are getting together to celebrate the 30th anniversary of the software's release. "We certainly ripped some stuff off," acknowledged Microsoft Excel 1.0 lead developer Doug Klunder, "but we also did some things that nobody else had done at the time and probably hasn't done since — some of which are really insane, and some of which turn out to be pretty handy." Klunder, who was responsible for Excel's killer "intelligent recalc" feature, quit his job after Bill Gates decided to shift the original Excel project from MS-DOS to the Mac, but ended up coming back and finishing the project after an ill-fated stint as a farm worker in the lettuce fields of California. "Just imagine having this product where one of the key components of it is really only understood by this guy who will quit routinely and go be a migrant farm worker down in California," said Excel 1.0 program manager Jabe Blumenthal. "It was not necessarily the most traditional or stable of environments." Many of the original Excel team members still use the program today — the RSVP sheet for this weekend's party was an Excel Online document. Before a professional naming firm came up with "Excel," the software was known by its code name "Odyssey", and other product names considered by Microsoft included "Master Plan" and "Mr. Spreadsheet." By the way, "Mr. Spreadsheet" makes his MOOC debut next week in edX's free-to-audit Excel for Data Analysis and Visualization course.

theodp writes: Those of you slugging your way through EdX's (free) Sabermetrics 101: Introduction to Baseball Analytics MOOC course might want to take a break from your R and SQL coding to check out Grantland's Before Beane, in which Ben Lindbergh tells the story of AVM Systems, the little-known company that jump-started sabermetrics and made Moneyball possible. Ken Mauriello, whose love-for-the-analytics-of-the-game led him to ditch a trading career to co-found AVM in the mid-90's, said of the early days, "Back in the day we weren't doing presentations [to skeptical MLB teams] with laptops. We were carrying around two enormous boxes with an enormous monitor and an enormous tower. It was like Planes, Trains & Automobiles traveling around with that stuff. Watching a great big Gateway box with your monitor come tumbling out upside down, and you pick it up and it's rattling. ... So we're in the hotel, saying, 'Please lord, let this thing work.'"

theodp writes: As one works his or her way through EdX's free The Analytics Edge, one finds oneself going back-and-forth between videos and R to complete the programming exercises associated with the lectures. While this can certainly be done on a cheap-o 13" laptop with a 6mbps connection by jumping around from the web-based videos to the client-based programming environment and to the web for help (god bless Stack Overflow), have you found (or do you dream of) a better setup for the MOOC programming courses offered by the likes of EdX, Udacity, and Coursera? Are you using multiple screens, split screens, touch screens, laptops/desktops/tablets, speakers, headphones, higher-speed connections? Anything else? Do you rely solely on the class materials and web-based resources, or do you purchase complementary books? Any thoughts on how to make the experience work best for those learning at home, in a classroom setting, on the road for business/travel, or during lengthy train commutes? Do you playback videos at faster speeds (e.g., 1.5x)? Any other tips?

Linux Foundation Executive Director Jim Zemlin doesn't use the phrase 'world domination' in these videos, but he could. He lists enough computing niches where GNU/Linux is the major player -- from supercomputers to the next generation of automotive systems -- that with or without world domination, Linux has obviously become an extremely important, widely used operating system that has grown amazingly since Linus Torvalds first shared his humble kernel with the world in 1991. With great popularity has come a great need for people who know how to administer and otherwise work with Linux, so the Linux Foundation is developing new courses in tandem with massive open online course (MOOC) provider edX. Unlike some of the Linux Foundation's previous course offerings, their edX ones are free to audit, and the cost for certification (if you want a cred, not just knowledge) is lower than many IT certification tests and certificates.

These videos (both visible today) were made remotely, with Timothy Lord at one end in Austin, TX, and Jim Zemlin at the other end in Tokyo, Japan. Their sound quality suffers from the distance involved, but they are generally intelligible -- and, of course, you can always choose to read the transcript instead of watching the videos.

theodp writes: "At edX," explains the upscale MOOC founded by MIT and Harvard, "we believe in offering the highest quality courses, created by schools and partners who share our commitment to excellence in teaching and learning, both online and in the classroom." You know, like Building Cloud Apps with Microsoft Azure (course trailer). On Tuesday, edX welcomed Microsoft as its first corporate member to offer MOOCs on edX.org. "Through this program," said edX, "Microsoft will offer the edX global learning community courses to acquire the core development skills needed to be successful in the cloud-first, mobile-first world." The new initiative, explained Microsoft, expands upon an existing Microsoft partnership with edX to create interactive online courses using Office Mix and PowerPoint 2013. Classes start March 31st.

theodp writes: "At edX," explains the upscale MOOC founded by MIT and Harvard, "we believe in offering the highest quality courses, created by schools and partners who share our commitment to excellence in teaching and learning, both online and in the classroom." You know, like Building Cloud Apps with Microsoft Azure (course trailer). On Tuesday, edX welcomed Microsoft as its first corporate member to offer MOOCs on edX.org. "Through this program," said edX, "Microsoft will offer the edX global learning community courses to acquire the core development skills needed to be successful in the cloud-first, mobile-first world." The new initiative, explained Microsoft, expands upon an existing Microsoft partnership with edX to create interactive online courses using Office Mix and PowerPoint 2013. Classes start March 31st.

theodp writes: "At edX," explains the upscale MOOC founded by MIT and Harvard, "we believe in offering the highest quality courses, created by schools and partners who share our commitment to excellence in teaching and learning, both online and in the classroom." You know, like Building Cloud Apps with Microsoft Azure (course trailer). On Tuesday, edX welcomed Microsoft as its first corporate member to offer MOOCs on edX.org. "Through this program," said edX, "Microsoft will offer the edX global learning community courses to acquire the core development skills needed to be successful in the cloud-first, mobile-first world." The new initiative, explained Microsoft, expands upon an existing Microsoft partnership with edX to create interactive online courses using Office Mix and PowerPoint 2013. Classes start March 31st.

theodp writes: "At edX," explains the upscale MOOC founded by MIT and Harvard, "we believe in offering the highest quality courses, created by schools and partners who share our commitment to excellence in teaching and learning, both online and in the classroom." You know, like Building Cloud Apps with Microsoft Azure (course trailer). On Tuesday, edX welcomed Microsoft as its first corporate member to offer MOOCs on edX.org. "Through this program," said edX, "Microsoft will offer the edX global learning community courses to acquire the core development skills needed to be successful in the cloud-first, mobile-first world." The new initiative, explained Microsoft, expands upon an existing Microsoft partnership with edX to create interactive online courses using Office Mix and PowerPoint 2013. Classes start March 31st.

samzenpus (5) writes "Alexander Stepanov is an award winning programmer who designed the C++ Standard Template Library. Daniel E. Rose is a programmer, research scientist, and is the Chief Scientist for Search at A9.com. In addition to working together, the duo have recently written a new book titled, From Mathematics to Generic Programming. Earlier this month you had a chance to ask the pair about their book, their work, or programming in general. Below you'll find the answers to those questions." Early Soviet Computing?
by eldavojohn

Alexander Stepanov, I have never had a chance to ask someone as qualified as you about this topic. I grew up on the opposite side of the Iron Curtain and have constantly wondered if (surely there must have been) alternative computing solutions developed in the USSR prior to Elbrus and SPARC. So my question is whether or not you know of any hardware or instruction set alternatives that died on the vine or were never mass fabricated in Soviet times? I don't expect to you to reveal some super advanced or future predicting instruction set but it has always disturbed me that these things aren't documented somewhere -- as you likely know failures can provide more fruit than successes. Failing that, could you offer us any tails of early computing that only seem to run in Russian circles?

If you can suggest references (preferably in English) I would be most appreciative. I know of only one book and it seems to be a singular point of view.

Alex: I'm not sure I have any unique knowledge, but can only describe my own experience. The first computer I used was a mainframe called M-20 (or one of its derivatives). My first programming exam was pass/fail, but I had to take it several times before I passed. I had no idea how to write code -- I didn't attend lectures or participate in labs, and thought I could just study the book and take the test. But programming isn't like that; the only way to learn is to do it.

Then in my first job, I participated in the design of a minicomputer called TA-100 used to control hydroelectric power stations. I was one of thekey designers of the realtime operating system, a contributor to the instruction set, and the lead designer of the programming tools (debugger, assembler, linker, etc.) -- all written in assembly language. The fact that I started at a very low level -- gates and instructions -- continues to be useful to my work even today. About that time, the Soviet Union started copying American designs, but I was very fortunate to be able to design something original from scratch. The head designer of the TA-100, Aleksandr Gurevich, was a great mentor to me. Two of my senior colleagues, Ilya Neistadt and Natalya Davydovskaya, also spent a lot of time trying to teach me all the things I didn't know.

Despite my personal experience (many details of which I've forgotten), I'm not actually an expert on the history of Soviet computing. But there is a good website containing many articles in English about early Soviet computers. One radically different approach was the "Setun" ternary computer. Unfortunately, there is no detailed treatment of Soviet-era computing at the level of detail and insight found in the second ("Computer Zoo") volume of Computer Architecture by Blaauw and Brooks, which provides an exhaustive treatment of Western designs. In general, computer history is an important field and requires great dedication. My friend Paul McJones does a fabulous job on history of programming languages and other software artifacts. See, for example, his history of FORTRAN site. (He also created sites for Lisp and ALGOL.) Sadly, there is no comparable effort on Soviet computing.



Successor to C++
by Anonymous Coward

I remember you wrote that STL has nothing to do with C++, it was a framework for generic programming and that C++ was chosen for its first implementation because it was less deficient for that purpose compared with other commercial programming languages. That implies you'd like to develop a programming language from scratch. Is that so? If so, how is it going?

Alex: In my first experiments in building a component architecture, I tried to design a language from scratch, called Tecton, with Deepak Kapur and Dave Musser. Tecton was my second system (i.e. it suffered from The Mythical-Man Month's "Second System Effect"). It was an extremely high level language indeed, and had concepts, but was unusable for anything practical. Then I implemented a version of the library in Scheme (together with Aaron Kerschenbaum and Dave Musser), and then another version in Ada (with Dave Musser). I was hired by Bell Labs to join their C++ library team in 1987, which was my first exposure to C++. My C and C++ mentor was Andy Koenig, who helped me understand the overall logic of the language. Unfortunately at that time, C++ was not ready for STL.

I returned to the library work in 1993 at HP Labs, together with Meng Lee. C++ had just gotten templates, and we were able to create a large generic library. At Andy Koenig's suggestion, we submitted a version of it for inclusion in the C++ language standard. This became STL.

After STL was accepted into the standard in 1994, I started thinking about designing a minimal programming language that will allow even more intimate access to hardware than C/C++ and also provide support for concepts and generic programming. I was hoping that somebody would fund such an activity. I interviewed with several companies, proposing such a design, but there was no interest. A senior VP at Microsoft told me: "We are not interested in innovating in the direction you suggest." They were "innovating" in the direction of C#, trying to displace Java. The situation now is not any better. There might be some interest eventually, but it will happen after my retirement: I am no longer in the game.



STL
by serviscope_minor

I'm a huge fan of the STL, and I think the design has stood the test of time amazingly well. That said, you now hae a bunch of hindsight. What would you do differently knowing what you know now. Also if you were doing it today and using today's languages, how do you think it would differ?

Alex: STL is the result of many compromises. There was a tension between my research goals for generic programming and getting something approved by the different constituents of the standards committee with diverse technical, business, and personal agendas. Such compromises are inevitable in real life.

Having said that, here are some of the things I would have preferred were different:

As we discuss in From Mathematics to Generic Programming, my original name for iterators was "coordinates" (or more specifically, "linear coordinates"). The standards committee people told me that there was already a name for this concept, "iterator," so I should use that term. They were wrong -- they were confusing my coordinates with heavyweight stateful iterators found in languages such as CLU and Alphard. This unfortunate terminology still often leads to misunderstanding about the concept of iterator in generic programming. Furthermore, as far back as 1987 I knew that linear coordinates (i.e. iterators) were only one kind of coordinate structures, data types that allow one to navigate through data structures. There are coordinate structures that deal with multidimensional arrays, trees, graphs, etc. (See, for example, chapters 7 and 8 of Elements of Programming.)

Also, there are many different types of containers and STL provided only a rudimentary classification. Moreover, containers with ownership semantics constitute only one way of dealing with data structures. There are others. A properly designed library would be based on a far larger set of data structures than what I could include into STL. There are also simple mistakes in algorithmic interfaces. Partition should place non-satisfying elements before satisfying elements. Copy_n should return a pair. I should have included algorithms dealing with integer concepts. I should have resisted the pressure to include allocators.

It would make perfect sense to redesign STL from scratch when they put concepts into C++. I would recommend that a person who decides to do it, should carefully study both Elements of Programming and From Mathematics to Generic Programming. Both of these books expand on these issues.



Re:STL
by Pseudonym

Related question: C++ was originally conceived as "C + Simula", but something that is interesting about the STL is how non-object-oriented it is, in particular using no inheritance. If we were designing a new "better C" today, one that you'd be happy to implement a STL-like system in, knowing what we know now, would we bother with Simula-style objects at all?

Alex: I am still convinced that Simula/C++/Java style inheritance is unsound. I do believe, however, that there is sometimes a need for run-time dispatch. But run time dispatch should be done as a run-time concept dispatch. Imagine, say, writing code in terms of a pointer to forward iterator. One should be able to obtain affiliated types at run time. Eventually languages will unify object-orientation and generic programming, but nobody seems to work on it now.

Dan: Bjarne Stroustrup describes C++ as a multi-paradigm language. The features that support object-oriented programming and the features that support generic programming are, for the most part, independent. That doesn't mean that both sets of features are not useful. Could Alex have designed STL for a language that doesn't have object-oriented features? Sure. But as a programmer, I'm happy that both sets of features are available. Just because object-oriented features are not needed to implement STL doesn't mean they provide no value in the language.

Alex: C++ has evolved over many years, and many of its features (inheritance, templates, exceptions, namespaces, etc.) were incorporated based on other work. As a result, they don't always work well together, and even when they do, it's in a baroque way. Now that we as a community have many years of experience with these features, we could design a minimal language from scratch that incorporates these features in a more concise and elegant way.



Hardware evolution
by jonkalb

The STL is about three decades old. In that time, we've seen both OS and hardware evolution. What is the impact of these changes on how the STL should be used? How would the STL be different if it where implemented targeting modern environments?

Alex: STL is "only" two decades old, but yes, there have been important changes during that period that would lead to some different decisions. STL was actually designed on a Leading Edge PC with no cache and 640K memory. (Our group at HP Labs didn't have enough money in the budget for an HP PC. When HP CEO Lew Platt came to visit me, HP Labs' director rushed in beforehand to hide the Leading Edge PC.)

One of the biggest changes since then has been the growth of caches. Cache misses are very costly, so locality of reference is much more important now. Node-based data structures, which have low locality of reference, make much less sense. If I were designing STL today, I would have a different set of containers. For example, an in-memory B*-tree is a far better choice than a red-black tree for implementing an associative container.

Another change is the increase in pipeline depth and support for unaligned reads. Today it is cheaper to read extra data rather than to have a branch.

Most processors today also support SIMD instructions. Libraries should take advantage of them whenever they can.

Modern applications such as search engines and databases also use lots of collections of very small data items that can be stored compactly without an extra level of indirection by using variable-sized encodings. It is essential that the libraries provide support for these variable-size entities. Dan and I, together with colleagues at A9, worked on this. Sadly enough, we were not able to finish our work, although you can see some relevant code snippets using variable-sized types and a new data structure called "tape" here.



Search seemingly getting worse over time
by TWX

This is more for Daniel Rose, but to what do you attribute the seeming decline in the quality of search results? I used Digital's Alta Vista search engine when it was fairly new and it seemed revolutionary and seemed to provide me with exactly what I wanted. Over time that declined and Alta Vista as it was ceased to be, and Google initially also seemed to provide me with exactly what I wanted. Now it seems like I have to put a whole lot of thought into faking Google into performing a somewhat-boolean-style search for me, and normal boolean expressions themselves no longer seem to work.

Is this the result of attempting to dumb-down the interface for tailored results, or something else or more insidious? Obviously the amount of content on the Internet is growing, but the computing power to process through all of it is growing too, so I would expect it wouldn't be getting this much worse, this quickly.

Dan: This is a huge question, which could be the subject of a whole book by itself. But the short answer is that there are several factors that have made the search experience be (or at least seem) worse. Here are a few:

1. Size of the problem. In the early days of AltaVista, there were around 100,000 web sites. Today there are around a billion. Assuming the number of web pages has grown proportionally to the number of sites, that's a factor of 10^4. Search ranking algorithms have actually been improved a lot -- they might even be 10x better than they were in 1995. But they haven't improved by 10^4x.

2. Complexity of the problem. Originally, web search engines dealt with static HTML pages. Now they are expected to work with many different types of documents in different formats, with users having a much wider variety of search goals. At A9, which provides the search engine for Amazon.com, we optimized the system specifically for product search. Web search engines have to work for all kinds of search.

3. Adversarial relationship between sites and engines. In the early days of web search engines, most web sites were purely informational, and many were run by nonprofit organizations like universities. Even when for-profit companies put up web sites, most were offered as an informational service to their customers -- it was a cost to the company, not a source of revenue. Obviously, all that has changed. Now it's in the interest of most web site providers to drive traffic to their sites. To do that they want to rank higher in search results -- often even for queries where their content is not relevant. So there is basically an arms race between search companies, which want to accurately rank results by relevance, and so-called search engine optimizers, who want their clients' pages to rank higher regardless of relevance. This leads to all kinds of spam.

4. Business model. The invention of search advertising by Overture, and its adoption by Google and others, meant that it was more profitable to show an ad than an organic search result. I know of specific instances where search engine companies chose not to deploy relevance improvements, because that would reduce revenue (more people would click on the results, and fewer on the ads). Even if a company tries to have a separation between their relevance and advertising teams, it is very hard to serve two masters.

Regarding the use of boolean expressions, there is evidence both from cognitive psychology and from information retrieval research that most people don't understand boolean logic, and that this misunderstanding leads to worse results. So I claim that Google's decision to stop interpreting certain words and symbols as boolean operators is good user-centered design, not "dumbing down" -- but I wish they still provided an advanced search option for those who want it.

If you're interested in learning more about search user issues in particular, here's a lecture I gave at UC Berkeley about 10 years ago on that topic.



Re:Search seemingly getting worse over time
by MouseTheLuckyDog

I was wondering something similar. Often times recent news tends to overshadow search results.

Let me give a practical example. Grand Jury. proceedings have undergone serious reform since the 70s. In some states a target can demand to appear before the Grand Jury. In some states a No Bill precludes the State from representing the case. In others there must be clear new evidence before a case can be represented. I know one state has a three strikes rules for GJ proceedings ( sorry don't remember which).

The day before the Michael Brown shooting, a search on Grand Jury Missouri would have found several articles on the specific laws to Grand Jury proceedings in Missouri. The day the DA announced he would present the case to a Grand Jury, the same search gets hundreds of articles on news story about Michael Brown and Grand Jury proceedings, but it becomes impossible to find those same scholarly articles about the peculiarity of Missouri Grand Jury proceedings. Not even the relevant statutes from the state website. What can be done to mitigate this effect?

Dan: This is a good illustration of what a complex problem search is. There are two issues here: First, should search results change in response to news events? I think so; in your example, it's almost certainly what the majority of users are looking for.

Second, how can the search engine make sure that *other* relevant results are also findable? One way is to make sure that the results address a diversity of user intents. When I was at AltaVista and Yahoo, we did some research on how to identify different user intents and how to make sure the results were not dominated by just one. The query "grand jury Missouri" has at least two obvious intents: "give me information about the grand jury system in Missouri" and "tell me what is going on with the particular grand jury investigating Michael Brown's death."

There are techniques that can do this diversification, and some search engines use some of them. But perhaps a better approach is to recognize that information-seeking is a process, not a magic oracle. Search engines should be designed to facilitate a kind of dialogue with the user. At AltaVista, we had a feature called "Prisma" that would show 12 related queries right below the search box -- not just queries that shared substrings with what the user has typed (like autocomplete), but queries that were about the range of different topics discussed in the actual pages. So for the query "grand jury Missouri," one suggestion might be "Michael Brown news" and another might be "grand jury statute Missouri".

My advice to intelligent search users is to to imagine what terminology would be used in a hypothetical good result on your intended topic, and use those words. If you want to find information about the legal basis of the grand jury system in Missouri, don't just type "grand jury Missouri," type "grand jury statute Missouri." When I do that query today on Google, I get a 2009 publication from the state of Missouri explaining the grand jury process, and the text of the actual statute, both on the first page of results.



What's your time like?
by mlheur

How much of your time do you dedicate to computing vs doing other things; what are your other hobbies or is the work you do also your play time?

Alex: Over the course of my life I have gradually narrowed my focus to spending time on the few items that, to me, are the essential examples of their category. These are things that stood the test of time; I re-read the books I already have read; I listen to music I have listened many times before; etc, etc. Yes, there is a chance that I will miss a new Mozart or Euclid, but it is a chance I am willing to take. Also, like the Pythagoreans of old, I view these as part of a unity: music reflects mathematics, literature is connected with history, etc. My work and my play and my life are inseparable. This unity is also reflected in From Mathematics to Generic Programming, which blends math, programming, history, and sometimes philosophy and art. Here are some of my favorites:

Literature: Greek and Roman classics: Homer, Plato, Ovid, Seneca; Bible; "modern" novels from Swift and Sterne to Dickens and Anthony Trollope. Math and science classics: Euclid, Euler, Gauss, Poincare. I still use printed books, not e-books.

Music: Bach, Mozart, Beethoven, Schubert, Wagner, and Mahler. I tend to listen to many different interpretations of the same piece. I do not use MP3s or streaming music, but CDs and, recently, SACDs.

Movies and TV: Chaplin, Marx Brothers, Kurosawa, Satyajit Ray, Kenneth Clark's Civilization, Peter Brook's Mahabharata, Brideshead Revisited, Royal Shakespeare Company production of Nicholas Nickelby, Maigret with Bruno Cremer. I am a blu-ray enthusiast. I do not use Netflix or Amazon Instant Video.

I love dogs, especially Welsh corgis; I spend 1-2 hours a day walking my dog Maxwell. I no longer eat meat or milk. I have been very happily married for 45 years; my wife Helen is my closest friend. We are practicing Roman Catholics, go to church on Sundays and holidays of obligation and try to keep the commandments. Our political views are in line with Pope Francis: we believe in having an economically just society.

Dan: Ironically, during much of my career as a researcher and engineering manager, I had fairly little time for programming. But now that I am not currently working full time, one of the things I've been doing for fun is programming -- learning iOS development and writing a musical iPhone app. I also enjoy playing very basic guitar and piano, reading, and lately, writing fiction. I try to alternate between reading nonfiction and fiction. The last books I read are The Swerve: How the World Became Modern by Stephen Greenblatt (about how the rediscovery of an ancient Roman poem helped spur the Renaissance) and Dave Eggers' novel The Circle (a cautionary tale about social networking and privacy, which should be required reading for everyone who works at Facebook, Google, and Apple).



Re:ack-nak
by blue trane

When will programming evolve to use subject-predicate syntax, rather than function-argument? Function-argument goes back (at least) to Frege, and his prejudices against subject-predicate syntax (which dominates natural languages). But isn't changePassword(a,b) more ambiguous than "change the password from a to b"? Don't we get an "information gain" effect from using a syntax we are familiar with outside of programming? When you first come to a function-argument command such as (in Oz, which is used in the Paradigms of Computer Programming MOOC) {Push S X}, there is maximum entropy as to whether S is pushed, or pushed onto. "Push X onto S" has no entropy; you know immediately, from the syntax alone, what is pushed onto what.

Dan: I think you need to decouple your argument about entropy with your argument about subject-predicate syntax. IIRC, stack-based languages like Forth and Postscript (and old HP scientific calculators) had completely unambiguous syntax. You either push something on the stack or perform an operation on the required number of arguments at the top of the stack. But these are not subject-predicate syntax languages. So there is more than one way to have what you call no-entropy syntax. Another way to avoid ambiguity is to require that argument names be part of the function name, as Smalltalk and Objective-C do. Then instead of your function call being changePassword(a, b), it's [foo changePasswordFrom:a to:b] (where foo is the object getting the message).

Separate from the entropy issue, is there a cognitive benefit from having programming languages use syntax familiar from natural languages? Perhaps, but which natural language's syntax will you use? Many languages (e.g. Japanese) use a subject-object-verb syntax, while English uses subject-verb-object. Romance languages use SOV some of the time (e.g. with pronouns) and SVO the rest of the time. Talk about ambiguous argument order!

Furthermore, natural languages have evolved to convey all kinds of nuances and deliberate ambiguities that make it hard to specify anything precisely. As a small example, the English meaning of "and" and "or" is quite different from their Boolean interpretation. (If the waiter says that breakfast comes with juice or coffee, getting both is not an option.)

The business application language COBOL (the most popular language of the 1970s) was supposed to have "English-like" syntax, with expressions like "add 1 to x." I'm skeptical that this syntax made programming any easier, but it did lead to this old joke: "Did you hear about the new version of COBOL? It's called ADD 1 TO COBOL."

My opinion is that we will always have different languages with different styles of syntax, to meet the needs of different communities of programmers.



Why is Generic Programming often second class?
by Anonymous Coward

We see many programming languages with at least some support for Generics, but usually as a second class citizen, and often added as an afterthought in later releases, and subordinate to some other programming paradigm. Java is primarily OO, with generics added later. C# is also primarily OO, though with generic support. It took C++ several iterations to get generics, and C++ is "multi paradigm". Go doesn't have generics, and doesn't seem like it will not a while.

It seems to me like generic programming is sufficiently powerful as a paradigm to not need other paradigms like OO in the same language. In fact, in many ways, OO, which ties together data and algorithms, seems antithetical to generic programming. So, do you see a possibility of a programming language whose primary paradigm is generic programming? Why do language designers not get generics into the first releases of their languages, even now, when the issues would seem to be well known? What would such a language look like?

Alex: To design a language for generic programming, one needs to learn to program generically. One has to write lots of code before things become clear. In the Appendix B of Elements of Programming, Sean Parent and Bjarne Stroustrup outlined a minimal language needed for programming. The appendix is about 8 pages long. To make it real, it probably needs to grow by a factor of 3. So, something like 25 pages should be sufficient. I am too old to do it, but I wish that someone would try.

A more difficult problem is not to design the language: C++, after all, contains most of the things needed. The problem is to teach programmers to think abstractly. And that is a very difficult task. I do not know a single university where one could even learn the preliminaries: understanding the machine, and understanding abstract mathematics. Our new book, From Mathematics to Generic Programming, is an attempt to sketch what is needed. Hopefully some school will try to teach both assembly level programming and abstract algebra.

An even harder problem is to convince the software industry to build software out of carefully designed components. What I see, however, is the movement in the opposite direction. Hand-crafted, one-off, undisciplined code is impossible to replicate. Adobe did a fabulous job specifying Postscript; that allowed Peter Deutsch to single-handedly produce Ghostscript. Now Adobe is not going to specify Photoshop's behavior. Let the Gimp guys try to replicate it. While Linus Torvalds was able to replicate Unix from the carefully written System V interface definitions, no one could replicate Windows: being nonstandard creates barriers to entry. There are grave economic reasons making any progress unlikely while undisciplined programmers generate huge amount of capital. It's analogous to the programmer whose terrible spaghetti code gives him job security, since no one else can understand it.

Dan: The idea that object-oriented programming and generic programming are competing paradigms is, in my opinion, mistaken. They are really orthogonal approaches. As we discuss in our book, generic programming is really an attitude. This attitude is useful whether you are using an object-oriented approach or not.

I would love to have a real-world, efficient, popular language that supports generic programming -- including concepts, in particular -- as first-class features. But I see no reason why this language shouldn't also support OOP.

An anonymous reader writes "The Linux Foundation has teamed up with MOOC provider edX to teach an introduction to Linux class. Quoting the course description: 'This course explores the various tools and techniques commonly used by Linux programmers, system administrators and end users to achieve their day-to-day work in a Linux environment. It is designed for experienced computer users who have limited or no previous exposure to Linux, whether they are working in an individual or Enterprise environment.' The course begins on August 1st. In addition to the free version of the course, a verified track is available for students who want a credential with more weight (for a nominal price)." Update: As many have pointed out Linus just did an intro for the class. Headline corrected accordingly.

theodp writes "In 'The Online Education Revolution Drifts Off Course,' NPR's Eric Westervelt reports that 2013 might be dubbed the year that online education fell back to earth. Westervelt joins others in citing the higher failure rate of online students as evidence that MOOCs aren't all they're cracked up to be. But viewed another way, the ability to try and fail without dire debt or academic consequences that's afforded by MOOCs could be viewed as a feature and not a bug. Being able to learn at one's own pace is what Dr. Yung Tae Kim has long argued is something STEM education sorely lacks, and MOOCs make it feasible to allow students to try-try-again if at first they don't succeed. By the way, if you couldn't scrape together $65,000 to take CS50 in-person at Harvard this year, today's the first day of look-Ma-no-tuition CS50x (review), kids!"

A few weeks ago you had the chance to ask the President of edX, Anant Agarwal, about the future of education and the growth of MOOCs. Below you'll find his answers to your questions. Could someone set up an archive site ?
by Taco Cowboy

The other day I was looking up an open course offered by Harvard that I had meant to try my hands on for the past few years (I know, I know, I procrastinated). I googled it up and clicked on the link - and long and behold, the Harvard server told me that the course had been deleted, due to some "incompatibility" of the video format and their new hardware, or something like that. I did not take that course. I have no idea if it was good or not. What if it was an excellent course ? Now that that particular course is gone (a few lessons still can still be had on youtube), the opportunity cost for many people does accumulate. If there was only an archive site for all the open-courses, wouldn't that be great ?

Agarwal: I know, why do today what you can put off till tomorrow :-) Seriously, a large number of our learners like you just want to audit a course and prefer it to be always available. For that reason, a significant number of our previous courses are indeed offered as "past courses" or "archived courses". You will see this on the announcements page when you go into such an archived course: "This is a past/archived course. Certain features of this course may not be active, but we still invite you to explore the available materials." To see archived courses you can click on courses at edx.org, select courses, and then select past courses. For example, the justice course from HarvardX is available as an archived course at: edx.org. In general, we strongly encourage our university partners to offer their past courses as archived courses. A few courses (like the one you tried perhaps) are not offered as an archived course. There are several reasons why a university might choose not to offer their course as an archived course. The next version of the course might have started, and so they might take down the archived course because they reuse some of the questions which have their answers available as handouts. Or, the course might need significant resources to run, for example, special grading servers may need to be maintained, for which they may not have resources.



Professors
by pyrognat

Dear Anant,

I am a young researcher at your own institution. One might think that online courses (such as those offered by edX) will make professors (at least those who teach) obsolete. What role do you see professors playing in the future of education? As someone on that career path, I am particularly interested in your views.

Sincerely,
Nathaniel Stapleton

Agarwal: First, let me point out that professors have been using various forms of online learning for decades, so what edx and others are doing with moocs is simply taking what we have been doing to the next level in terms of scalability and quality.

Think of online learning and platforms to create great online course material not as making teachers obsolete, but as tools for teachers, which will enable them to do a lot more than they could previously with the time they had. Sadly, among the few teaching tools we have given professors in the past few centuries has been a piece of chalk. Blogging sites and online news did not make journalists obsolete, rather created 10 million journalists, while making news much more interactive and exciting. In much the same manner, online learning has the potential to transform education. Online learning and moocs can give existing professors new tools to create additional modalities of teaching. Professors and soon-to-be professors like yourself that are quick to experiment with these new tools, are likely to find success, creating new ways of engaging our millennial generation of students who are perfectly comfortable texting, tweeting, blogging, emailing, moocing, youtubing, slashdotting ( :-) ) and snapchating.

We have also been experimenting with blended models of education where you combine the best of online and in person. Initial results are promising. In one these blended model experiments in which we collaborated with san jose state university, the students watched videos, did online exercises and virtual labs on the edx online platform, and then came to class for in-person activity including working in small groups and interacting with the professor to get a deeper understanding of the material. The pass rate for this course went from a traditional average of about 60% to 90% with the blended class. Research by Brewlow et al (you can see paper at edx.org also shows that student success in a mooc course was positively correlated with working with an instructor or mentor.



Making money
by peter303

How do the VCs plan to make back their investments? That was not clear to me.

Agarwal: edX is a non-profit startup, but we still need to be self-sustaining. So we are exploring several business models spanning both B2C and B2B. Unlike other MOOC providers, since edX is a non-profit, we do not have VC investors, rather we have institutional funders including MIT and Harvard. Many of our university partners such as the UT system have also contributed $s to our cause. One can view these initial contributions as an early round of investment in edX. In the B2C space, we use a freemium model. Students can take our courses for free, and they can even get an honor code certificate for free. However, we charge a small fee for ID verified certificates. In addition to the fee for the ID verified certificate (typically between $25 and $100 for a course), we offer the students the opportunity to contribute voluntarily to our cause by paying more than the minimum fee.

In the B2B space, we work with corporations, NGO, governments and state university systems. Here, edX hosts the platform for the institution, and the institution offers its courses on our platform for their employees or for other targeted learners. They pay edX for various services such as hosting, education services, training, etc. We have a number of such partnerships underway, including one with IMF.



What about moving to an badges based system
by Joe_Dragon

Where courses can be more right sized and not jammed / padded out into the older collgle time table system. Where you don't have to take a big 2-4-6+ year block of time to get something that says to you know some thing. It can also make ongoing education / learning new skills have more meaning as well. What about merging Professional certification systems into an over all badges based system? Do you think this is an good idea?

Agarwal: This is a good idea.

Good old degrees with 4 years in college, in stove-piped departments are an antediluvian concept. Why 4 years? My bachelor's degree at IIT madras was a 5-year program, and I probably use 20% of that in my job at this point in my life. Why specialize in one field? In today's fast moving, nimble world, learners and employers are looking for multidisciplinary education. Further, they are looking to refresh their skills as the workplace needs change.

A degree, fundamentally, is a signaling mechanism. It tells an employer that the holder likely has a set of skills. In the modern world, employers are looking for a diverse set of skills. So a promising alternative is one where learners acquire a traditional degree for 2 years (certainly less than the traditional 4 years), so that they can satisfy the bean-counting HR folks that need to see something that spells degree on the resume (and that will go away too with time). Then, the learner augments that with various other signals -- call them badges, or mooc-style certificates, so that employers can now see the whole portfolio.

A couple of weeks ago, edX also announced another interesting signaling mechanism, an XSeries certificate. Here a learner can take a sequence of courses in a given discipline on edx, and by passing all the courses in the sequence, earn an XSeries certificate. MIT and edX announced XSeries certificates in foundations of computer science and supply chain management for a start. Some of you have commented that individual courses do not mean much to employers, but mastery of a discipline is akin to a mini-masters or a minor program, and is meaningful. We believe these XSeries credentials from edX will have strong signaling value to employers and take MOOC credentialing and badging one additional step forward.

I can go on and on about this. I believe that in the future degree transcripts will be replaced by a colorful portfolio of credentials, including badges and certificates. OK OK, and degrees too, for backwards compatibility.



Why not get rid of the Honor Code?
by blue trane

The Honor Code seems like a holdover from obsolete old educational methods. It seeks to make the free and open sharing of information somehow dishonorable.Often students want to help each other in the forums. The quizzes and exercises can provide interesting applications that the instructor didn't go over in the videos. Why censor a student who, of his own free will, wants to help out another student?

The Honor Code, in forbidding explicit help to questions on assignments, encourages deviousness and obfuscation in the forums. Often, posts will be made deliberately vague, so that one has to make guesses, or "read between the lines", or try to mind-read. Wouldn't it be better to encourage clear, simple explanations on the forums? Students are sometimes as (or more) knowledgeable than the instructors, and can explain things in a better, simpler way. Often the instructors have been at the subject so long that they've forgotten what it's like to look at the material for the first time. Other students can fill in the gaps. But the Honor Code works against this type of peer-helping-peer interaction, because often the most interesting applications of the subject are in the exercises.

When I've argued for the dissolution of the Honor Code before, one response has been: you just have to wait until after the deadline. However this response is not adequate, because often the deadlines are a few weeks off. When a student is engaged in a particular problem, that is the most opportune time for him to learn. I've had questions I couldn't answer, and haven't gone back to check how to do them after the deadline passes, because I'm now involved in something else...

I think the Honor Code works against the spirit of openness and freedom of speech that the internet was founded on. What kind of skills are you trying to teach, by enforcing the Honor Code? Does a client care whether you "cheated" by looking up the answer to a programming problem on the internet, when you're writing a program for him?

I think there are better technological solutions than enforcing an archaic Honor Code. Can you put a "spoiler" tag on posts that reveal how to do an assignment question, and reward those students who don't click on those posts? You're supposed to be tracking our every click...

Agarwal: The question about the honor code brings up fundamental issues about teaching and learning and the answer is not obvious. Our current education approach, whether online or in person, attempts to address two distinct issues: learning and student assessment.

Assessment attempts to verify that the student has learned the given material and has achieved a certain level of mastery in the subject. EdX and our partner universities give an honor code certificate to students who pass the course, as measured by assessments like homeworks and exams. These certificates are highly valued by a significant proportion of our learners. They often serve as signaling credentials to employers for example.

For the honor code certificate granted to an individual student to mean something as a signaling mechanism indicating mastery of the material in the course, the assessment method must represent the student’s own level of accomplishment (at least to the extent allowed by the honor code) and not that of the collaboration group to which the student belonged, or the person from whom the student may have gotten an answer. Without an honor code, there would be no honor code certificate. And hence the honor code.

Of course, we have a number of continuing learners to whom the honor code certificate, and hence the honor code, is not important. They are simply interested in learning the material, and do not care about earning a signaling mechanism. For this class of learners, we make our courses available as archived courses once their final assessments have passed their due date. Archived courses have all the answers readily available, but do not offer a certificate. Here, students can discuss answers without any worry about the honor code because there is no certificate attached.



Prestige of Online Degrees
by Calsar

I received an under graduate and master’s degree from traditional universities. I also received two masters’ degrees through on-line classes. In my opinion programs like edX are the future of education, but on-line degrees are still not regarded with the same level of prestige as those received through traditional education. In part this has been due to questionable practices of some on-line educational institutions. How can this perception be changed and do you have and do you have any plans in that regard?

Agarwal: Indeed, online education has been around for several decades. But online degrees and credentials were previously not as highly regarded as those obtained through campus education. There were a number of reasons for this beyond those that you have identified including the poorer quality of the education and relatively low success rates compared to campus programs.

This trend has changed completely in the past year for several reasons. First, reputable non-profit institutions like MIT, Harvard, Stanford, and Berkeley and their star faculty are embracing MOOCs and online learning wholeheartedly, and validating online education. To be sure, these institutions were also experimenting with online learning for decades, but their recent public endorsement and huge investments in MOOCs and rigorous online education has really helped the field gain a new respect. Imagine, you can now take online one of the world’s best AI course from Dan Klein and Pieter Abbeel from Berkeley.

Second, recent advances in computing and communications have completely altered online education and dramatically improved its quality. Content distribution networks have transformed high quality video delivery. Cloud computing enables elastic scaling to hundreds of thousands of learners and provide instant feedback for learning exercises. Social networking and discussion forums enable peer to peer learning at a global scale. And massive amounts of computing power in the palm of our hand has enabled us to offer simulation based virtual laboratories with online game-like-interaction latencies. Of course, getting large numbers of the world’s great instructors to experiment with online teaching helps online education quality more than any technology. New technologies like virtual labs and the efforts of leading instructors have enabled us to offer courses whose rigor is similar to that of campus courses.

Third, reputed universities are offering credentials like free honor code certificates for passing online courses. The introduction of ID verified certificates for a small fee adds even more to the credibility of the credential.

These are some examples of approaches that have improved the perception of online education. However, we can do a lot more. We are working with campuses to blend online learning with in-person education. Initial results in student outcomes at institutions like San Jose State University, Mass Bay Community College, and Bunker Hill Community College, are very promising. As more of these blended experiments yield good results, the perception of online learning will continue to improve, and be viewed not as an alternative to on-campus learning, but rather as complementary to it.



Motivation in Online Courses
by Eric Hosman

Motivation plays a large role in any educational setting, but this is especially true in online courses. How do you best maintain a learner's motivation after the initial novelty has worn off? Online educational opportunities attract a wide array of learners, and we can't expect them all to be intrinsically motivated at a level consistent enough to complete a course, even if they are taking it for college credit or future growth opportunities. What are the best techniques to keep as many learners as possible engaged throughout an online course.

Agarwal: As we listen to our students who are taking MOOCs, or understand the correlators to success from researchers who are studying the big data of learning, we are seeing a number of approaches to increase student motivation and improve outcomes. Here are some of them.

A recent study by Lori Breslow et al edx.org found that students were motivated and had a better chance at success if they worked in groups. Anecdotal data also points to this result. I was chatting with a learner who was taking an edX course in Krgezistan and he had to drop out half way. Later he told me that he discovered that there were three other students taking the same class in his town close to where he lived, and had he known that, he would have connected with them and stayed the course. Following these learnings, we are encouraging our students to form study groups through meetups, and facilitating this with links to meetup sites in the course introduction email.

We are hearing from our learners that the certificate is a big part of the motivation. Although you can audit a course, we are giving free honor code certificates, and also ID verified certificates to our students for a small fee. EdX also provides an online link with a unique identifier for each of these certificates so that an employer can verify that the certificate is indeed authentic. Students are linking these certificates on their various social pages and on their resumes.

Many learners and employers have also asked us to create sequences of courses, so they have a program that they can follow. In response to this demand, we launched XSeries, a sequence of courses in a discipline. A student can also obtain a certificate for an XSeries. We believe that programs such as this will increase the motivation of students to complete their existing course, so they can go on to the next one, eventually earning an XSeries certificate to indicate mastery in a given discipline.



MOOCs more about Teachers than Learners
by IRGlover

As social creatures much of our knowledge is built from social interactions, where we integrate our own experiences and beliefs with that of others to build new knowledge and understanding (i.e. Social Constructivism). The current dominant MOOC model is extremely procedural, teacher-centered and discourages these types of social interactions. While this works well for some subjects (particularly at introductory levels), it is much less effective in other situations. How can the large MOOC platforms, and EdX in particular, encourage a more social method of learning?

Agarwal: I am not sure I agree with your thesis that the current dominant MOOC model is extremely procedural. Although many components are procedural, social and peer learning is a central part of MOOC platforms. Take the example of the discussion forum. Learners post questions on the forum, and a majority of the answers and discussion comes from peer learners. Many students are telling us that they are learning by teaching on the forum.

Based on feedback from many of our colleagues in the humanities where the small discussion group is a central part of the learning process, we created the concept of cohorts in the discussion forum. Cohorts enable us to create many smaller groups and assign each student to a specific cohort. A number of courses such as the Greek Heroes course from HarvardX and the Copyright Law course also from HarvardX used this mechanism during Spring 2013. Harvard went a step further and engaged their alumni to help as community moderators in each of these cohorts.

Many courses are taking advantage of the international diversity of the students on the platform to conduct course related studies and discussions. The Public Health course from Harvard for example did a study where they surveyed their worldwide students in the course in how they did certain things in their part of the world, and then discussed it on the forums. Such diversity and numbers are simply not possible in a campus class.

We have a wiki on the platform as part of each course where students work together and contribute material. In fact, in some of the courses the wiki notes produced by students are so good that it would not take the group much additional effort to turn the set of notes into a textbook.

For question types where the computer is not great at grading, we support peer grading, in which students grade each other’s work. Here again, the peer group is key to the learning and assessment process.

That said, we can and should do even more to take advantage of the peer learning effects that are possible with MOOCs. Here is but one example of the many that we have been thinking about.

It turns out that we can post some of our simulation laboratories on the wiki where students can collaborate on labs. Imagine asking the students to work together to solve a lab problem working jointly to design a circuit. In fact, when we posted our simulation lab tool on the wiki just for fun in the 6.002x circuits course, we had a student who had begun the design of a circuit on the wiki, and had then stepped out for a break. When he returned, he noticed his circuit had changed and he reported to us that the tool was buggy. Little did he realize that his circuit had been modified by an unknown collaborator probably from around the world!



Are teaching faculty dead?
by Anonymous Coward

Currently, many 2nd and 3rd tier universities rely heavily on (cost-effective) adjuncts and teaching-only "lecturers" for the majority of the instructional duties. To further maximize revenue, these schools could replace adjuncts and lecturers with MOOCs taught by professors at 1st tier universities. Are lecturers and adjuncts dead?

Agarwal: I believe the exact opposite will happen! I predict that 10 years from now there will be a much greater number of adjuncts and lecturers, and a much greater choice of courses for students in 2nd and 3rd tier universities. Similarly, the quality of education will also increase in all universities, whether 1st tier or 3rd tier. Read on.

Think of online learning and platforms to create great online course material not as making teachers obsolete, but as tools for teachers, which will enable them to do a lot more than they could previously with the time they had. Sadly, one of the few teaching tools we have given professors in the past few centuries has been a piece of chalk. Blogging sites and online news did not make journalists obsolete, rather created 10 million journalists, while making news much more interactive and exciting. In much the same manner, online learning has the potential to transform education. Online learning and moocs can give existing professors new tools to create additional modalities of teaching. Instructors that are quick to experiment with these new tools are likely to find success, creating new ways of engaging our millennial generation of students who are perfectly comfortable texting, tweeting, blogging, emailing, moocing, youtubing, slashdotting (:-) ) and snapchating.

We have also been experimenting with blended models of education where you combine the best of online and in person. This model seems perfect for courses which are taught by an adjunct, where the adjunct has the choice to leverage online material from a MOOC course much like a new-age textbook, and improve the overall student experience. The blended approach works great even for courses taught by a professor at a top tier university. My colleagues and I used this model ourselves for an experimental group of 20 students at MIT for the MITx 6.002x circuits and electronics, our first MOOC course on edX, and came away with a good experience (see the paper by Afridi et al at edx.org).

In one these blended model experiments in which we collaborated with San Jose State University, the students watched videos, did online exercises and virtual labs from the 6.002x course on our online platform, and then came to class for in-person activity including working in small groups and interacting with a professor to get a deeper understanding of the material. Armando Fox from Berkeley has termed such courses SPOCs (small private online courses). The pass rate for this course went from a traditional average of about 60% to 90% with the blended class. Research by Brewlow et al (you can see paper athttps://www.edx.org/research[edx.org]) also shows that student success in a mooc course was positively correlated with working with an instructor or lecturer.

This may seem surprising, but I believe that the opportunity to use MOOCs in the SPOC model will result in universities hiring more adjunct faculty to offer wider choices of quality courses for their students. This will be a good thing for both students and adjuncts.

A few weeks ago you had the chance to ask the President of edX, Anant Agarwal, about the future of education and the growth of MOOCs. Below you'll find his answers to your questions. Could someone set up an archive site ?
by Taco Cowboy

The other day I was looking up an open course offered by Harvard that I had meant to try my hands on for the past few years (I know, I know, I procrastinated). I googled it up and clicked on the link - and long and behold, the Harvard server told me that the course had been deleted, due to some "incompatibility" of the video format and their new hardware, or something like that. I did not take that course. I have no idea if it was good or not. What if it was an excellent course ? Now that that particular course is gone (a few lessons still can still be had on youtube), the opportunity cost for many people does accumulate. If there was only an archive site for all the open-courses, wouldn't that be great ?

Agarwal: I know, why do today what you can put off till tomorrow :-) Seriously, a large number of our learners like you just want to audit a course and prefer it to be always available. For that reason, a significant number of our previous courses are indeed offered as "past courses" or "archived courses". You will see this on the announcements page when you go into such an archived course: "This is a past/archived course. Certain features of this course may not be active, but we still invite you to explore the available materials." To see archived courses you can click on courses at edx.org, select courses, and then select past courses. For example, the justice course from HarvardX is available as an archived course at: edx.org. In general, we strongly encourage our university partners to offer their past courses as archived courses. A few courses (like the one you tried perhaps) are not offered as an archived course. There are several reasons why a university might choose not to offer their course as an archived course. The next version of the course might have started, and so they might take down the archived course because they reuse some of the questions which have their answers available as handouts. Or, the course might need significant resources to run, for example, special grading servers may need to be maintained, for which they may not have resources.



Professors
by pyrognat

Dear Anant,

I am a young researcher at your own institution. One might think that online courses (such as those offered by edX) will make professors (at least those who teach) obsolete. What role do you see professors playing in the future of education? As someone on that career path, I am particularly interested in your views.

Sincerely,
Nathaniel Stapleton

Agarwal: First, let me point out that professors have been using various forms of online learning for decades, so what edx and others are doing with moocs is simply taking what we have been doing to the next level in terms of scalability and quality.

Think of online learning and platforms to create great online course material not as making teachers obsolete, but as tools for teachers, which will enable them to do a lot more than they could previously with the time they had. Sadly, among the few teaching tools we have given professors in the past few centuries has been a piece of chalk. Blogging sites and online news did not make journalists obsolete, rather created 10 million journalists, while making news much more interactive and exciting. In much the same manner, online learning has the potential to transform education. Online learning and moocs can give existing professors new tools to create additional modalities of teaching. Professors and soon-to-be professors like yourself that are quick to experiment with these new tools, are likely to find success, creating new ways of engaging our millennial generation of students who are perfectly comfortable texting, tweeting, blogging, emailing, moocing, youtubing, slashdotting ( :-) ) and snapchating.

We have also been experimenting with blended models of education where you combine the best of online and in person. Initial results are promising. In one these blended model experiments in which we collaborated with san jose state university, the students watched videos, did online exercises and virtual labs on the edx online platform, and then came to class for in-person activity including working in small groups and interacting with the professor to get a deeper understanding of the material. The pass rate for this course went from a traditional average of about 60% to 90% with the blended class. Research by Brewlow et al (you can see paper at edx.org also shows that student success in a mooc course was positively correlated with working with an instructor or mentor.



Making money
by peter303

How do the VCs plan to make back their investments? That was not clear to me.

Agarwal: edX is a non-profit startup, but we still need to be self-sustaining. So we are exploring several business models spanning both B2C and B2B. Unlike other MOOC providers, since edX is a non-profit, we do not have VC investors, rather we have institutional funders including MIT and Harvard. Many of our university partners such as the UT system have also contributed $s to our cause. One can view these initial contributions as an early round of investment in edX. In the B2C space, we use a freemium model. Students can take our courses for free, and they can even get an honor code certificate for free. However, we charge a small fee for ID verified certificates. In addition to the fee for the ID verified certificate (typically between $25 and $100 for a course), we offer the students the opportunity to contribute voluntarily to our cause by paying more than the minimum fee.

In the B2B space, we work with corporations, NGO, governments and state university systems. Here, edX hosts the platform for the institution, and the institution offers its courses on our platform for their employees or for other targeted learners. They pay edX for various services such as hosting, education services, training, etc. We have a number of such partnerships underway, including one with IMF.



What about moving to an badges based system
by Joe_Dragon

Where courses can be more right sized and not jammed / padded out into the older collgle time table system. Where you don't have to take a big 2-4-6+ year block of time to get something that says to you know some thing. It can also make ongoing education / learning new skills have more meaning as well. What about merging Professional certification systems into an over all badges based system? Do you think this is an good idea?

Agarwal: This is a good idea.

Good old degrees with 4 years in college, in stove-piped departments are an antediluvian concept. Why 4 years? My bachelor's degree at IIT madras was a 5-year program, and I probably use 20% of that in my job at this point in my life. Why specialize in one field? In today's fast moving, nimble world, learners and employers are looking for multidisciplinary education. Further, they are looking to refresh their skills as the workplace needs change.

A degree, fundamentally, is a signaling mechanism. It tells an employer that the holder likely has a set of skills. In the modern world, employers are looking for a diverse set of skills. So a promising alternative is one where learners acquire a traditional degree for 2 years (certainly less than the traditional 4 years), so that they can satisfy the bean-counting HR folks that need to see something that spells degree on the resume (and that will go away too with time). Then, the learner augments that with various other signals -- call them badges, or mooc-style certificates, so that employers can now see the whole portfolio.

A couple of weeks ago, edX also announced another interesting signaling mechanism, an XSeries certificate. Here a learner can take a sequence of courses in a given discipline on edx, and by passing all the courses in the sequence, earn an XSeries certificate. MIT and edX announced XSeries certificates in foundations of computer science and supply chain management for a start. Some of you have commented that individual courses do not mean much to employers, but mastery of a discipline is akin to a mini-masters or a minor program, and is meaningful. We believe these XSeries credentials from edX will have strong signaling value to employers and take MOOC credentialing and badging one additional step forward.

I can go on and on about this. I believe that in the future degree transcripts will be replaced by a colorful portfolio of credentials, including badges and certificates. OK OK, and degrees too, for backwards compatibility.



Why not get rid of the Honor Code?
by blue trane

The Honor Code seems like a holdover from obsolete old educational methods. It seeks to make the free and open sharing of information somehow dishonorable.Often students want to help each other in the forums. The quizzes and exercises can provide interesting applications that the instructor didn't go over in the videos. Why censor a student who, of his own free will, wants to help out another student?

The Honor Code, in forbidding explicit help to questions on assignments, encourages deviousness and obfuscation in the forums. Often, posts will be made deliberately vague, so that one has to make guesses, or "read between the lines", or try to mind-read. Wouldn't it be better to encourage clear, simple explanations on the forums? Students are sometimes as (or more) knowledgeable than the instructors, and can explain things in a better, simpler way. Often the instructors have been at the subject so long that they've forgotten what it's like to look at the material for the first time. Other students can fill in the gaps. But the Honor Code works against this type of peer-helping-peer interaction, because often the most interesting applications of the subject are in the exercises.

When I've argued for the dissolution of the Honor Code before, one response has been: you just have to wait until after the deadline. However this response is not adequate, because often the deadlines are a few weeks off. When a student is engaged in a particular problem, that is the most opportune time for him to learn. I've had questions I couldn't answer, and haven't gone back to check how to do them after the deadline passes, because I'm now involved in something else...

I think the Honor Code works against the spirit of openness and freedom of speech that the internet was founded on. What kind of skills are you trying to teach, by enforcing the Honor Code? Does a client care whether you "cheated" by looking up the answer to a programming problem on the internet, when you're writing a program for him?

I think there are better technological solutions than enforcing an archaic Honor Code. Can you put a "spoiler" tag on posts that reveal how to do an assignment question, and reward those students who don't click on those posts? You're supposed to be tracking our every click...

Agarwal: The question about the honor code brings up fundamental issues about teaching and learning and the answer is not obvious. Our current education approach, whether online or in person, attempts to address two distinct issues: learning and student assessment.

Assessment attempts to verify that the student has learned the given material and has achieved a certain level of mastery in the subject. EdX and our partner universities give an honor code certificate to students who pass the course, as measured by assessments like homeworks and exams. These certificates are highly valued by a significant proportion of our learners. They often serve as signaling credentials to employers for example.

For the honor code certificate granted to an individual student to mean something as a signaling mechanism indicating mastery of the material in the course, the assessment method must represent the student’s own level of accomplishment (at least to the extent allowed by the honor code) and not that of the collaboration group to which the student belonged, or the person from whom the student may have gotten an answer. Without an honor code, there would be no honor code certificate. And hence the honor code.

Of course, we have a number of continuing learners to whom the honor code certificate, and hence the honor code, is not important. They are simply interested in learning the material, and do not care about earning a signaling mechanism. For this class of learners, we make our courses available as archived courses once their final assessments have passed their due date. Archived courses have all the answers readily available, but do not offer a certificate. Here, students can discuss answers without any worry about the honor code because there is no certificate attached.



Prestige of Online Degrees
by Calsar

I received an under graduate and master’s degree from traditional universities. I also received two masters’ degrees through on-line classes. In my opinion programs like edX are the future of education, but on-line degrees are still not regarded with the same level of prestige as those received through traditional education. In part this has been due to questionable practices of some on-line educational institutions. How can this perception be changed and do you have and do you have any plans in that regard?

Agarwal: Indeed, online education has been around for several decades. But online degrees and credentials were previously not as highly regarded as those obtained through campus education. There were a number of reasons for this beyond those that you have identified including the poorer quality of the education and relatively low success rates compared to campus programs.

This trend has changed completely in the past year for several reasons. First, reputable non-profit institutions like MIT, Harvard, Stanford, and Berkeley and their star faculty are embracing MOOCs and online learning wholeheartedly, and validating online education. To be sure, these institutions were also experimenting with online learning for decades, but their recent public endorsement and huge investments in MOOCs and rigorous online education has really helped the field gain a new respect. Imagine, you can now take online one of the world’s best AI course from Dan Klein and Pieter Abbeel from Berkeley.

Second, recent advances in computing and communications have completely altered online education and dramatically improved its quality. Content distribution networks have transformed high quality video delivery. Cloud computing enables elastic scaling to hundreds of thousands of learners and provide instant feedback for learning exercises. Social networking and discussion forums enable peer to peer learning at a global scale. And massive amounts of computing power in the palm of our hand has enabled us to offer simulation based virtual laboratories with online game-like-interaction latencies. Of course, getting large numbers of the world’s great instructors to experiment with online teaching helps online education quality more than any technology. New technologies like virtual labs and the efforts of leading instructors have enabled us to offer courses whose rigor is similar to that of campus courses.

Third, reputed universities are offering credentials like free honor code certificates for passing online courses. The introduction of ID verified certificates for a small fee adds even more to the credibility of the credential.

These are some examples of approaches that have improved the perception of online education. However, we can do a lot more. We are working with campuses to blend online learning with in-person education. Initial results in student outcomes at institutions like San Jose State University, Mass Bay Community College, and Bunker Hill Community College, are very promising. As more of these blended experiments yield good results, the perception of online learning will continue to improve, and be viewed not as an alternative to on-campus learning, but rather as complementary to it.



Motivation in Online Courses
by Eric Hosman

Motivation plays a large role in any educational setting, but this is especially true in online courses. How do you best maintain a learner's motivation after the initial novelty has worn off? Online educational opportunities attract a wide array of learners, and we can't expect them all to be intrinsically motivated at a level consistent enough to complete a course, even if they are taking it for college credit or future growth opportunities. What are the best techniques to keep as many learners as possible engaged throughout an online course.

Agarwal: As we listen to our students who are taking MOOCs, or understand the correlators to success from researchers who are studying the big data of learning, we are seeing a number of approaches to increase student motivation and improve outcomes. Here are some of them.

A recent study by Lori Breslow et al edx.org found that students were motivated and had a better chance at success if they worked in groups. Anecdotal data also points to this result. I was chatting with a learner who was taking an edX course in Krgezistan and he had to drop out half way. Later he told me that he discovered that there were three other students taking the same class in his town close to where he lived, and had he known that, he would have connected with them and stayed the course. Following these learnings, we are encouraging our students to form study groups through meetups, and facilitating this with links to meetup sites in the course introduction email.

We are hearing from our learners that the certificate is a big part of the motivation. Although you can audit a course, we are giving free honor code certificates, and also ID verified certificates to our students for a small fee. EdX also provides an online link with a unique identifier for each of these certificates so that an employer can verify that the certificate is indeed authentic. Students are linking these certificates on their various social pages and on their resumes.

Many learners and employers have also asked us to create sequences of courses, so they have a program that they can follow. In response to this demand, we launched XSeries, a sequence of courses in a discipline. A student can also obtain a certificate for an XSeries. We believe that programs such as this will increase the motivation of students to complete their existing course, so they can go on to the next one, eventually earning an XSeries certificate to indicate mastery in a given discipline.



MOOCs more about Teachers than Learners
by IRGlover

As social creatures much of our knowledge is built from social interactions, where we integrate our own experiences and beliefs with that of others to build new knowledge and understanding (i.e. Social Constructivism). The current dominant MOOC model is extremely procedural, teacher-centered and discourages these types of social interactions. While this works well for some subjects (particularly at introductory levels), it is much less effective in other situations. How can the large MOOC platforms, and EdX in particular, encourage a more social method of learning?

Agarwal: I am not sure I agree with your thesis that the current dominant MOOC model is extremely procedural. Although many components are procedural, social and peer learning is a central part of MOOC platforms. Take the example of the discussion forum. Learners post questions on the forum, and a majority of the answers and discussion comes from peer learners. Many students are telling us that they are learning by teaching on the forum.

Based on feedback from many of our colleagues in the humanities where the small discussion group is a central part of the learning process, we created the concept of cohorts in the discussion forum. Cohorts enable us to create many smaller groups and assign each student to a specific cohort. A number of courses such as the Greek Heroes course from HarvardX and the Copyright Law course also from HarvardX used this mechanism during Spring 2013. Harvard went a step further and engaged their alumni to help as community moderators in each of these cohorts.

Many courses are taking advantage of the international diversity of the students on the platform to conduct course related studies and discussions. The Public Health course from Harvard for example did a study where they surveyed their worldwide students in the course in how they did certain things in their part of the world, and then discussed it on the forums. Such diversity and numbers are simply not possible in a campus class.

We have a wiki on the platform as part of each course where students work together and contribute material. In fact, in some of the courses the wiki notes produced by students are so good that it would not take the group much additional effort to turn the set of notes into a textbook.

For question types where the computer is not great at grading, we support peer grading, in which students grade each other’s work. Here again, the peer group is key to the learning and assessment process.

That said, we can and should do even more to take advantage of the peer learning effects that are possible with MOOCs. Here is but one example of the many that we have been thinking about.

It turns out that we can post some of our simulation laboratories on the wiki where students can collaborate on labs. Imagine asking the students to work together to solve a lab problem working jointly to design a circuit. In fact, when we posted our simulation lab tool on the wiki just for fun in the 6.002x circuits course, we had a student who had begun the design of a circuit on the wiki, and had then stepped out for a break. When he returned, he noticed his circuit had changed and he reported to us that the tool was buggy. Little did he realize that his circuit had been modified by an unknown collaborator probably from around the world!



Are teaching faculty dead?
by Anonymous Coward

Currently, many 2nd and 3rd tier universities rely heavily on (cost-effective) adjuncts and teaching-only "lecturers" for the majority of the instructional duties. To further maximize revenue, these schools could replace adjuncts and lecturers with MOOCs taught by professors at 1st tier universities. Are lecturers and adjuncts dead?

Agarwal: I believe the exact opposite will happen! I predict that 10 years from now there will be a much greater number of adjuncts and lecturers, and a much greater choice of courses for students in 2nd and 3rd tier universities. Similarly, the quality of education will also increase in all universities, whether 1st tier or 3rd tier. Read on.

Think of online learning and platforms to create great online course material not as making teachers obsolete, but as tools for teachers, which will enable them to do a lot more than they could previously with the time they had. Sadly, one of the few teaching tools we have given professors in the past few centuries has been a piece of chalk. Blogging sites and online news did not make journalists obsolete, rather created 10 million journalists, while making news much more interactive and exciting. In much the same manner, online learning has the potential to transform education. Online learning and moocs can give existing professors new tools to create additional modalities of teaching. Instructors that are quick to experiment with these new tools are likely to find success, creating new ways of engaging our millennial generation of students who are perfectly comfortable texting, tweeting, blogging, emailing, moocing, youtubing, slashdotting (:-) ) and snapchating.

We have also been experimenting with blended models of education where you combine the best of online and in person. This model seems perfect for courses which are taught by an adjunct, where the adjunct has the choice to leverage online material from a MOOC course much like a new-age textbook, and improve the overall student experience. The blended approach works great even for courses taught by a professor at a top tier university. My colleagues and I used this model ourselves for an experimental group of 20 students at MIT for the MITx 6.002x circuits and electronics, our first MOOC course on edX, and came away with a good experience (see the paper by Afridi et al at edx.org).

In one these blended model experiments in which we collaborated with San Jose State University, the students watched videos, did online exercises and virtual labs from the 6.002x course on our online platform, and then came to class for in-person activity including working in small groups and interacting with a professor to get a deeper understanding of the material. Armando Fox from Berkeley has termed such courses SPOCs (small private online courses). The pass rate for this course went from a traditional average of about 60% to 90% with the blended class. Research by Brewlow et al (you can see paper athttps://www.edx.org/research[edx.org]) also shows that student success in a mooc course was positively correlated with working with an instructor or lecturer.

This may seem surprising, but I believe that the opportunity to use MOOCs in the SPOC model will result in universities hiring more adjunct faculty to offer wider choices of quality courses for their students. This will be a good thing for both students and adjuncts.

A few weeks ago you had the chance to ask the President of edX, Anant Agarwal, about the future of education and the growth of MOOCs. Below you'll find his answers to your questions. Could someone set up an archive site ?
by Taco Cowboy

The other day I was looking up an open course offered by Harvard that I had meant to try my hands on for the past few years (I know, I know, I procrastinated). I googled it up and clicked on the link - and long and behold, the Harvard server told me that the course had been deleted, due to some "incompatibility" of the video format and their new hardware, or something like that. I did not take that course. I have no idea if it was good or not. What if it was an excellent course ? Now that that particular course is gone (a few lessons still can still be had on youtube), the opportunity cost for many people does accumulate. If there was only an archive site for all the open-courses, wouldn't that be great ?

Agarwal: I know, why do today what you can put off till tomorrow :-) Seriously, a large number of our learners like you just want to audit a course and prefer it to be always available. For that reason, a significant number of our previous courses are indeed offered as "past courses" or "archived courses". You will see this on the announcements page when you go into such an archived course: "This is a past/archived course. Certain features of this course may not be active, but we still invite you to explore the available materials." To see archived courses you can click on courses at edx.org, select courses, and then select past courses. For example, the justice course from HarvardX is available as an archived course at: edx.org. In general, we strongly encourage our university partners to offer their past courses as archived courses. A few courses (like the one you tried perhaps) are not offered as an archived course. There are several reasons why a university might choose not to offer their course as an archived course. The next version of the course might have started, and so they might take down the archived course because they reuse some of the questions which have their answers available as handouts. Or, the course might need significant resources to run, for example, special grading servers may need to be maintained, for which they may not have resources.



Professors
by pyrognat

Dear Anant,

I am a young researcher at your own institution. One might think that online courses (such as those offered by edX) will make professors (at least those who teach) obsolete. What role do you see professors playing in the future of education? As someone on that career path, I am particularly interested in your views.

Sincerely,
Nathaniel Stapleton

Agarwal: First, let me point out that professors have been using various forms of online learning for decades, so what edx and others are doing with moocs is simply taking what we have been doing to the next level in terms of scalability and quality.

Think of online learning and platforms to create great online course material not as making teachers obsolete, but as tools for teachers, which will enable them to do a lot more than they could previously with the time they had. Sadly, among the few teaching tools we have given professors in the past few centuries has been a piece of chalk. Blogging sites and online news did not make journalists obsolete, rather created 10 million journalists, while making news much more interactive and exciting. In much the same manner, online learning has the potential to transform education. Online learning and moocs can give existing professors new tools to create additional modalities of teaching. Professors and soon-to-be professors like yourself that are quick to experiment with these new tools, are likely to find success, creating new ways of engaging our millennial generation of students who are perfectly comfortable texting, tweeting, blogging, emailing, moocing, youtubing, slashdotting ( :-) ) and snapchating.

We have also been experimenting with blended models of education where you combine the best of online and in person. Initial results are promising. In one these blended model experiments in which we collaborated with san jose state university, the students watched videos, did online exercises and virtual labs on the edx online platform, and then came to class for in-person activity including working in small groups and interacting with the professor to get a deeper understanding of the material. The pass rate for this course went from a traditional average of about 60% to 90% with the blended class. Research by Brewlow et al (you can see paper at edx.org also shows that student success in a mooc course was positively correlated with working with an instructor or mentor.



Making money
by peter303

How do the VCs plan to make back their investments? That was not clear to me.

Agarwal: edX is a non-profit startup, but we still need to be self-sustaining. So we are exploring several business models spanning both B2C and B2B. Unlike other MOOC providers, since edX is a non-profit, we do not have VC investors, rather we have institutional funders including MIT and Harvard. Many of our university partners such as the UT system have also contributed $s to our cause. One can view these initial contributions as an early round of investment in edX. In the B2C space, we use a freemium model. Students can take our courses for free, and they can even get an honor code certificate for free. However, we charge a small fee for ID verified certificates. In addition to the fee for the ID verified certificate (typically between $25 and $100 for a course), we offer the students the opportunity to contribute voluntarily to our cause by paying more than the minimum fee.

In the B2B space, we work with corporations, NGO, governments and state university systems. Here, edX hosts the platform for the institution, and the institution offers its courses on our platform for their employees or for other targeted learners. They pay edX for various services such as hosting, education services, training, etc. We have a number of such partnerships underway, including one with IMF.



What about moving to an badges based system
by Joe_Dragon

Where courses can be more right sized and not jammed / padded out into the older collgle time table system. Where you don't have to take a big 2-4-6+ year block of time to get something that says to you know some thing. It can also make ongoing education / learning new skills have more meaning as well. What about merging Professional certification systems into an over all badges based system? Do you think this is an good idea?

Agarwal: This is a good idea.

Good old degrees with 4 years in college, in stove-piped departments are an antediluvian concept. Why 4 years? My bachelor's degree at IIT madras was a 5-year program, and I probably use 20% of that in my job at this point in my life. Why specialize in one field? In today's fast moving, nimble world, learners and employers are looking for multidisciplinary education. Further, they are looking to refresh their skills as the workplace needs change.

A degree, fundamentally, is a signaling mechanism. It tells an employer that the holder likely has a set of skills. In the modern world, employers are looking for a diverse set of skills. So a promising alternative is one where learners acquire a traditional degree for 2 years (certainly less than the traditional 4 years), so that they can satisfy the bean-counting HR folks that need to see something that spells degree on the resume (and that will go away too with time). Then, the learner augments that with various other signals -- call them badges, or mooc-style certificates, so that employers can now see the whole portfolio.

A couple of weeks ago, edX also announced another interesting signaling mechanism, an XSeries certificate. Here a learner can take a sequence of courses in a given discipline on edx, and by passing all the courses in the sequence, earn an XSeries certificate. MIT and edX announced XSeries certificates in foundations of computer science and supply chain management for a start. Some of you have commented that individual courses do not mean much to employers, but mastery of a discipline is akin to a mini-masters or a minor program, and is meaningful. We believe these XSeries credentials from edX will have strong signaling value to employers and take MOOC credentialing and badging one additional step forward.

I can go on and on about this. I believe that in the future degree transcripts will be replaced by a colorful portfolio of credentials, including badges and certificates. OK OK, and degrees too, for backwards compatibility.



Why not get rid of the Honor Code?
by blue trane

The Honor Code seems like a holdover from obsolete old educational methods. It seeks to make the free and open sharing of information somehow dishonorable.Often students want to help each other in the forums. The quizzes and exercises can provide interesting applications that the instructor didn't go over in the videos. Why censor a student who, of his own free will, wants to help out another student?

The Honor Code, in forbidding explicit help to questions on assignments, encourages deviousness and obfuscation in the forums. Often, posts will be made deliberately vague, so that one has to make guesses, or "read between the lines", or try to mind-read. Wouldn't it be better to encourage clear, simple explanations on the forums? Students are sometimes as (or more) knowledgeable than the instructors, and can explain things in a better, simpler way. Often the instructors have been at the subject so long that they've forgotten what it's like to look at the material for the first time. Other students can fill in the gaps. But the Honor Code works against this type of peer-helping-peer interaction, because often the most interesting applications of the subject are in the exercises.

When I've argued for the dissolution of the Honor Code before, one response has been: you just have to wait until after the deadline. However this response is not adequate, because often the deadlines are a few weeks off. When a student is engaged in a particular problem, that is the most opportune time for him to learn. I've had questions I couldn't answer, and haven't gone back to check how to do them after the deadline passes, because I'm now involved in something else...

I think the Honor Code works against the spirit of openness and freedom of speech that the internet was founded on. What kind of skills are you trying to teach, by enforcing the Honor Code? Does a client care whether you "cheated" by looking up the answer to a programming problem on the internet, when you're writing a program for him?

I think there are better technological solutions than enforcing an archaic Honor Code. Can you put a "spoiler" tag on posts that reveal how to do an assignment question, and reward those students who don't click on those posts? You're supposed to be tracking our every click...

Agarwal: The question about the honor code brings up fundamental issues about teaching and learning and the answer is not obvious. Our current education approach, whether online or in person, attempts to address two distinct issues: learning and student assessment.

Assessment attempts to verify that the student has learned the given material and has achieved a certain level of mastery in the subject. EdX and our partner universities give an honor code certificate to students who pass the course, as measured by assessments like homeworks and exams. These certificates are highly valued by a significant proportion of our learners. They often serve as signaling credentials to employers for example.

For the honor code certificate granted to an individual student to mean something as a signaling mechanism indicating mastery of the material in the course, the assessment method must represent the student’s own level of accomplishment (at least to the extent allowed by the honor code) and not that of the collaboration group to which the student belonged, or the person from whom the student may have gotten an answer. Without an honor code, there would be no honor code certificate. And hence the honor code.

Of course, we have a number of continuing learners to whom the honor code certificate, and hence the honor code, is not important. They are simply interested in learning the material, and do not care about earning a signaling mechanism. For this class of learners, we make our courses available as archived courses once their final assessments have passed their due date. Archived courses have all the answers readily available, but do not offer a certificate. Here, students can discuss answers without any worry about the honor code because there is no certificate attached.



Prestige of Online Degrees
by Calsar

I received an under graduate and master’s degree from traditional universities. I also received two masters’ degrees through on-line classes. In my opinion programs like edX are the future of education, but on-line degrees are still not regarded with the same level of prestige as those received through traditional education. In part this has been due to questionable practices of some on-line educational institutions. How can this perception be changed and do you have and do you have any plans in that regard?

Agarwal: Indeed, online education has been around for several decades. But online degrees and credentials were previously not as highly regarded as those obtained through campus education. There were a number of reasons for this beyond those that you have identified including the poorer quality of the education and relatively low success rates compared to campus programs.

This trend has changed completely in the past year for several reasons. First, reputable non-profit institutions like MIT, Harvard, Stanford, and Berkeley and their star faculty are embracing MOOCs and online learning wholeheartedly, and validating online education. To be sure, these institutions were also experimenting with online learning for decades, but their recent public endorsement and huge investments in MOOCs and rigorous online education has really helped the field gain a new respect. Imagine, you can now take online one of the world’s best AI course from Dan Klein and Pieter Abbeel from Berkeley.

Second, recent advances in computing and communications have completely altered online education and dramatically improved its quality. Content distribution networks have transformed high quality video delivery. Cloud computing enables elastic scaling to hundreds of thousands of learners and provide instant feedback for learning exercises. Social networking and discussion forums enable peer to peer learning at a global scale. And massive amounts of computing power in the palm of our hand has enabled us to offer simulation based virtual laboratories with online game-like-interaction latencies. Of course, getting large numbers of the world’s great instructors to experiment with online teaching helps online education quality more than any technology. New technologies like virtual labs and the efforts of leading instructors have enabled us to offer courses whose rigor is similar to that of campus courses.

Third, reputed universities are offering credentials like free honor code certificates for passing online courses. The introduction of ID verified certificates for a small fee adds even more to the credibility of the credential.

These are some examples of approaches that have improved the perception of online education. However, we can do a lot more. We are working with campuses to blend online learning with in-person education. Initial results in student outcomes at institutions like San Jose State University, Mass Bay Community College, and Bunker Hill Community College, are very promising. As more of these blended experiments yield good results, the perception of online learning will continue to improve, and be viewed not as an alternative to on-campus learning, but rather as complementary to it.



Motivation in Online Courses
by Eric Hosman

Motivation plays a large role in any educational setting, but this is especially true in online courses. How do you best maintain a learner's motivation after the initial novelty has worn off? Online educational opportunities attract a wide array of learners, and we can't expect them all to be intrinsically motivated at a level consistent enough to complete a course, even if they are taking it for college credit or future growth opportunities. What are the best techniques to keep as many learners as possible engaged throughout an online course.

Agarwal: As we listen to our students who are taking MOOCs, or understand the correlators to success from researchers who are studying the big data of learning, we are seeing a number of approaches to increase student motivation and improve outcomes. Here are some of them.

A recent study by Lori Breslow et al edx.org found that students were motivated and had a better chance at success if they worked in groups. Anecdotal data also points to this result. I was chatting with a learner who was taking an edX course in Krgezistan and he had to drop out half way. Later he told me that he discovered that there were three other students taking the same class in his town close to where he lived, and had he known that, he would have connected with them and stayed the course. Following these learnings, we are encouraging our students to form study groups through meetups, and facilitating this with links to meetup sites in the course introduction email.

We are hearing from our learners that the certificate is a big part of the motivation. Although you can audit a course, we are giving free honor code certificates, and also ID verified certificates to our students for a small fee. EdX also provides an online link with a unique identifier for each of these certificates so that an employer can verify that the certificate is indeed authentic. Students are linking these certificates on their various social pages and on their resumes.

Many learners and employers have also asked us to create sequences of courses, so they have a program that they can follow. In response to this demand, we launched XSeries, a sequence of courses in a discipline. A student can also obtain a certificate for an XSeries. We believe that programs such as this will increase the motivation of students to complete their existing course, so they can go on to the next one, eventually earning an XSeries certificate to indicate mastery in a given discipline.



MOOCs more about Teachers than Learners
by IRGlover

As social creatures much of our knowledge is built from social interactions, where we integrate our own experiences and beliefs with that of others to build new knowledge and understanding (i.e. Social Constructivism). The current dominant MOOC model is extremely procedural, teacher-centered and discourages these types of social interactions. While this works well for some subjects (particularly at introductory levels), it is much less effective in other situations. How can the large MOOC platforms, and EdX in particular, encourage a more social method of learning?

Agarwal: I am not sure I agree with your thesis that the current dominant MOOC model is extremely procedural. Although many components are procedural, social and peer learning is a central part of MOOC platforms. Take the example of the discussion forum. Learners post questions on the forum, and a majority of the answers and discussion comes from peer learners. Many students are telling us that they are learning by teaching on the forum.

Based on feedback from many of our colleagues in the humanities where the small discussion group is a central part of the learning process, we created the concept of cohorts in the discussion forum. Cohorts enable us to create many smaller groups and assign each student to a specific cohort. A number of courses such as the Greek Heroes course from HarvardX and the Copyright Law course also from HarvardX used this mechanism during Spring 2013. Harvard went a step further and engaged their alumni to help as community moderators in each of these cohorts.

Many courses are taking advantage of the international diversity of the students on the platform to conduct course related studies and discussions. The Public Health course from Harvard for example did a study where they surveyed their worldwide students in the course in how they did certain things in their part of the world, and then discussed it on the forums. Such diversity and numbers are simply not possible in a campus class.

We have a wiki on the platform as part of each course where students work together and contribute material. In fact, in some of the courses the wiki notes produced by students are so good that it would not take the group much additional effort to turn the set of notes into a textbook.

For question types where the computer is not great at grading, we support peer grading, in which students grade each other’s work. Here again, the peer group is key to the learning and assessment process.

That said, we can and should do even more to take advantage of the peer learning effects that are possible with MOOCs. Here is but one example of the many that we have been thinking about.

It turns out that we can post some of our simulation laboratories on the wiki where students can collaborate on labs. Imagine asking the students to work together to solve a lab problem working jointly to design a circuit. In fact, when we posted our simulation lab tool on the wiki just for fun in the 6.002x circuits course, we had a student who had begun the design of a circuit on the wiki, and had then stepped out for a break. When he returned, he noticed his circuit had changed and he reported to us that the tool was buggy. Little did he realize that his circuit had been modified by an unknown collaborator probably from around the world!



Are teaching faculty dead?
by Anonymous Coward

Currently, many 2nd and 3rd tier universities rely heavily on (cost-effective) adjuncts and teaching-only "lecturers" for the majority of the instructional duties. To further maximize revenue, these schools could replace adjuncts and lecturers with MOOCs taught by professors at 1st tier universities. Are lecturers and adjuncts dead?

Agarwal: I believe the exact opposite will happen! I predict that 10 years from now there will be a much greater number of adjuncts and lecturers, and a much greater choice of courses for students in 2nd and 3rd tier universities. Similarly, the quality of education will also increase in all universities, whether 1st tier or 3rd tier. Read on.

Think of online learning and platforms to create great online course material not as making teachers obsolete, but as tools for teachers, which will enable them to do a lot more than they could previously with the time they had. Sadly, one of the few teaching tools we have given professors in the past few centuries has been a piece of chalk. Blogging sites and online news did not make journalists obsolete, rather created 10 million journalists, while making news much more interactive and exciting. In much the same manner, online learning has the potential to transform education. Online learning and moocs can give existing professors new tools to create additional modalities of teaching. Instructors that are quick to experiment with these new tools are likely to find success, creating new ways of engaging our millennial generation of students who are perfectly comfortable texting, tweeting, blogging, emailing, moocing, youtubing, slashdotting (:-) ) and snapchating.

We have also been experimenting with blended models of education where you combine the best of online and in person. This model seems perfect for courses which are taught by an adjunct, where the adjunct has the choice to leverage online material from a MOOC course much like a new-age textbook, and improve the overall student experience. The blended approach works great even for courses taught by a professor at a top tier university. My colleagues and I used this model ourselves for an experimental group of 20 students at MIT for the MITx 6.002x circuits and electronics, our first MOOC course on edX, and came away with a good experience (see the paper by Afridi et al at edx.org).

In one these blended model experiments in which we collaborated with San Jose State University, the students watched videos, did online exercises and virtual labs from the 6.002x course on our online platform, and then came to class for in-person activity including working in small groups and interacting with a professor to get a deeper understanding of the material. Armando Fox from Berkeley has termed such courses SPOCs (small private online courses). The pass rate for this course went from a traditional average of about 60% to 90% with the blended class. Research by Brewlow et al (you can see paper athttps://www.edx.org/research[edx.org]) also shows that student success in a mooc course was positively correlated with working with an instructor or lecturer.

This may seem surprising, but I believe that the opportunity to use MOOCs in the SPOC model will result in universities hiring more adjunct faculty to offer wider choices of quality courses for their students. This will be a good thing for both students and adjuncts.

A few weeks ago you had the chance to ask the President of edX, Anant Agarwal, about the future of education and the growth of MOOCs. Below you'll find his answers to your questions. Could someone set up an archive site ?
by Taco Cowboy

The other day I was looking up an open course offered by Harvard that I had meant to try my hands on for the past few years (I know, I know, I procrastinated). I googled it up and clicked on the link - and long and behold, the Harvard server told me that the course had been deleted, due to some "incompatibility" of the video format and their new hardware, or something like that. I did not take that course. I have no idea if it was good or not. What if it was an excellent course ? Now that that particular course is gone (a few lessons still can still be had on youtube), the opportunity cost for many people does accumulate. If there was only an archive site for all the open-courses, wouldn't that be great ?

Agarwal: I know, why do today what you can put off till tomorrow :-) Seriously, a large number of our learners like you just want to audit a course and prefer it to be always available. For that reason, a significant number of our previous courses are indeed offered as "past courses" or "archived courses". You will see this on the announcements page when you go into such an archived course: "This is a past/archived course. Certain features of this course may not be active, but we still invite you to explore the available materials." To see archived courses you can click on courses at edx.org, select courses, and then select past courses. For example, the justice course from HarvardX is available as an archived course at: edx.org. In general, we strongly encourage our university partners to offer their past courses as archived courses. A few courses (like the one you tried perhaps) are not offered as an archived course. There are several reasons why a university might choose not to offer their course as an archived course. The next version of the course might have started, and so they might take down the archived course because they reuse some of the questions which have their answers available as handouts. Or, the course might need significant resources to run, for example, special grading servers may need to be maintained, for which they may not have resources.



Professors
by pyrognat

Dear Anant,

I am a young researcher at your own institution. One might think that online courses (such as those offered by edX) will make professors (at least those who teach) obsolete. What role do you see professors playing in the future of education? As someone on that career path, I am particularly interested in your views.

Sincerely,
Nathaniel Stapleton

Agarwal: First, let me point out that professors have been using various forms of online learning for decades, so what edx and others are doing with moocs is simply taking what we have been doing to the next level in terms of scalability and quality.

Think of online learning and platforms to create great online course material not as making teachers obsolete, but as tools for teachers, which will enable them to do a lot more than they could previously with the time they had. Sadly, among the few teaching tools we have given professors in the past few centuries has been a piece of chalk. Blogging sites and online news did not make journalists obsolete, rather created 10 million journalists, while making news much more interactive and exciting. In much the same manner, online learning has the potential to transform education. Online learning and moocs can give existing professors new tools to create additional modalities of teaching. Professors and soon-to-be professors like yourself that are quick to experiment with these new tools, are likely to find success, creating new ways of engaging our millennial generation of students who are perfectly comfortable texting, tweeting, blogging, emailing, moocing, youtubing, slashdotting ( :-) ) and snapchating.

We have also been experimenting with blended models of education where you combine the best of online and in person. Initial results are promising. In one these blended model experiments in which we collaborated with san jose state university, the students watched videos, did online exercises and virtual labs on the edx online platform, and then came to class for in-person activity including working in small groups and interacting with the professor to get a deeper understanding of the material. The pass rate for this course went from a traditional average of about 60% to 90% with the blended class. Research by Brewlow et al (you can see paper at edx.org also shows that student success in a mooc course was positively correlated with working with an instructor or mentor.



Making money
by peter303

How do the VCs plan to make back their investments? That was not clear to me.

Agarwal: edX is a non-profit startup, but we still need to be self-sustaining. So we are exploring several business models spanning both B2C and B2B. Unlike other MOOC providers, since edX is a non-profit, we do not have VC investors, rather we have institutional funders including MIT and Harvard. Many of our university partners such as the UT system have also contributed $s to our cause. One can view these initial contributions as an early round of investment in edX. In the B2C space, we use a freemium model. Students can take our courses for free, and they can even get an honor code certificate for free. However, we charge a small fee for ID verified certificates. In addition to the fee for the ID verified certificate (typically between $25 and $100 for a course), we offer the students the opportunity to contribute voluntarily to our cause by paying more than the minimum fee.

In the B2B space, we work with corporations, NGO, governments and state university systems. Here, edX hosts the platform for the institution, and the institution offers its courses on our platform for their employees or for other targeted learners. They pay edX for various services such as hosting, education services, training, etc. We have a number of such partnerships underway, including one with IMF.



What about moving to an badges based system
by Joe_Dragon

Where courses can be more right sized and not jammed / padded out into the older collgle time table system. Where you don't have to take a big 2-4-6+ year block of time to get something that says to you know some thing. It can also make ongoing education / learning new skills have more meaning as well. What about merging Professional certification systems into an over all badges based system? Do you think this is an good idea?

Agarwal: This is a good idea.

Good old degrees with 4 years in college, in stove-piped departments are an antediluvian concept. Why 4 years? My bachelor's degree at IIT madras was a 5-year program, and I probably use 20% of that in my job at this point in my life. Why specialize in one field? In today's fast moving, nimble world, learners and employers are looking for multidisciplinary education. Further, they are looking to refresh their skills as the workplace needs change.

A degree, fundamentally, is a signaling mechanism. It tells an employer that the holder likely has a set of skills. In the modern world, employers are looking for a diverse set of skills. So a promising alternative is one where learners acquire a traditional degree for 2 years (certainly less than the traditional 4 years), so that they can satisfy the bean-counting HR folks that need to see something that spells degree on the resume (and that will go away too with time). Then, the learner augments that with various other signals -- call them badges, or mooc-style certificates, so that employers can now see the whole portfolio.

A couple of weeks ago, edX also announced another interesting signaling mechanism, an XSeries certificate. Here a learner can take a sequence of courses in a given discipline on edx, and by passing all the courses in the sequence, earn an XSeries certificate. MIT and edX announced XSeries certificates in foundations of computer science and supply chain management for a start. Some of you have commented that individual courses do not mean much to employers, but mastery of a discipline is akin to a mini-masters or a minor program, and is meaningful. We believe these XSeries credentials from edX will have strong signaling value to employers and take MOOC credentialing and badging one additional step forward.

I can go on and on about this. I believe that in the future degree transcripts will be replaced by a colorful portfolio of credentials, including badges and certificates. OK OK, and degrees too, for backwards compatibility.



Why not get rid of the Honor Code?
by blue trane

The Honor Code seems like a holdover from obsolete old educational methods. It seeks to make the free and open sharing of information somehow dishonorable.Often students want to help each other in the forums. The quizzes and exercises can provide interesting applications that the instructor didn't go over in the videos. Why censor a student who, of his own free will, wants to help out another student?

The Honor Code, in forbidding explicit help to questions on assignments, encourages deviousness and obfuscation in the forums. Often, posts will be made deliberately vague, so that one has to make guesses, or "read between the lines", or try to mind-read. Wouldn't it be better to encourage clear, simple explanations on the forums? Students are sometimes as (or more) knowledgeable than the instructors, and can explain things in a better, simpler way. Often the instructors have been at the subject so long that they've forgotten what it's like to look at the material for the first time. Other students can fill in the gaps. But the Honor Code works against this type of peer-helping-peer interaction, because often the most interesting applications of the subject are in the exercises.

When I've argued for the dissolution of the Honor Code before, one response has been: you just have to wait until after the deadline. However this response is not adequate, because often the deadlines are a few weeks off. When a student is engaged in a particular problem, that is the most opportune time for him to learn. I've had questions I couldn't answer, and haven't gone back to check how to do them after the deadline passes, because I'm now involved in something else...

I think the Honor Code works against the spirit of openness and freedom of speech that the internet was founded on. What kind of skills are you trying to teach, by enforcing the Honor Code? Does a client care whether you "cheated" by looking up the answer to a programming problem on the internet, when you're writing a program for him?

I think there are better technological solutions than enforcing an archaic Honor Code. Can you put a "spoiler" tag on posts that reveal how to do an assignment question, and reward those students who don't click on those posts? You're supposed to be tracking our every click...

Agarwal: The question about the honor code brings up fundamental issues about teaching and learning and the answer is not obvious. Our current education approach, whether online or in person, attempts to address two distinct issues: learning and student assessment.

Assessment attempts to verify that the student has learned the given material and has achieved a certain level of mastery in the subject. EdX and our partner universities give an honor code certificate to students who pass the course, as measured by assessments like homeworks and exams. These certificates are highly valued by a significant proportion of our learners. They often serve as signaling credentials to employers for example.

For the honor code certificate granted to an individual student to mean something as a signaling mechanism indicating mastery of the material in the course, the assessment method must represent the student’s own level of accomplishment (at least to the extent allowed by the honor code) and not that of the collaboration group to which the student belonged, or the person from whom the student may have gotten an answer. Without an honor code, there would be no honor code certificate. And hence the honor code.

Of course, we have a number of continuing learners to whom the honor code certificate, and hence the honor code, is not important. They are simply interested in learning the material, and do not care about earning a signaling mechanism. For this class of learners, we make our courses available as archived courses once their final assessments have passed their due date. Archived courses have all the answers readily available, but do not offer a certificate. Here, students can discuss answers without any worry about the honor code because there is no certificate attached.



Prestige of Online Degrees
by Calsar

I received an under graduate and master’s degree from traditional universities. I also received two masters’ degrees through on-line classes. In my opinion programs like edX are the future of education, but on-line degrees are still not regarded with the same level of prestige as those received through traditional education. In part this has been due to questionable practices of some on-line educational institutions. How can this perception be changed and do you have and do you have any plans in that regard?

Agarwal: Indeed, online education has been around for several decades. But online degrees and credentials were previously not as highly regarded as those obtained through campus education. There were a number of reasons for this beyond those that you have identified including the poorer quality of the education and relatively low success rates compared to campus programs.

This trend has changed completely in the past year for several reasons. First, reputable non-profit institutions like MIT, Harvard, Stanford, and Berkeley and their star faculty are embracing MOOCs and online learning wholeheartedly, and validating online education. To be sure, these institutions were also experimenting with online learning for decades, but their recent public endorsement and huge investments in MOOCs and rigorous online education has really helped the field gain a new respect. Imagine, you can now take online one of the world’s best AI course from Dan Klein and Pieter Abbeel from Berkeley.

Second, recent advances in computing and communications have completely altered online education and dramatically improved its quality. Content distribution networks have transformed high quality video delivery. Cloud computing enables elastic scaling to hundreds of thousands of learners and provide instant feedback for learning exercises. Social networking and discussion forums enable peer to peer learning at a global scale. And massive amounts of computing power in the palm of our hand has enabled us to offer simulation based virtual laboratories with online game-like-interaction latencies. Of course, getting large numbers of the world’s great instructors to experiment with online teaching helps online education quality more than any technology. New technologies like virtual labs and the efforts of leading instructors have enabled us to offer courses whose rigor is similar to that of campus courses.

Third, reputed universities are offering credentials like free honor code certificates for passing online courses. The introduction of ID verified certificates for a small fee adds even more to the credibility of the credential.

These are some examples of approaches that have improved the perception of online education. However, we can do a lot more. We are working with campuses to blend online learning with in-person education. Initial results in student outcomes at institutions like San Jose State University, Mass Bay Community College, and Bunker Hill Community College, are very promising. As more of these blended experiments yield good results, the perception of online learning will continue to improve, and be viewed not as an alternative to on-campus learning, but rather as complementary to it.



Motivation in Online Courses
by Eric Hosman

Motivation plays a large role in any educational setting, but this is especially true in online courses. How do you best maintain a learner's motivation after the initial novelty has worn off? Online educational opportunities attract a wide array of learners, and we can't expect them all to be intrinsically motivated at a level consistent enough to complete a course, even if they are taking it for college credit or future growth opportunities. What are the best techniques to keep as many learners as possible engaged throughout an online course.

Agarwal: As we listen to our students who are taking MOOCs, or understand the correlators to success from researchers who are studying the big data of learning, we are seeing a number of approaches to increase student motivation and improve outcomes. Here are some of them.

A recent study by Lori Breslow et al edx.org found that students were motivated and had a better chance at success if they worked in groups. Anecdotal data also points to this result. I was chatting with a learner who was taking an edX course in Krgezistan and he had to drop out half way. Later he told me that he discovered that there were three other students taking the same class in his town close to where he lived, and had he known that, he would have connected with them and stayed the course. Following these learnings, we are encouraging our students to form study groups through meetups, and facilitating this with links to meetup sites in the course introduction email.

We are hearing from our learners that the certificate is a big part of the motivation. Although you can audit a course, we are giving free honor code certificates, and also ID verified certificates to our students for a small fee. EdX also provides an online link with a unique identifier for each of these certificates so that an employer can verify that the certificate is indeed authentic. Students are linking these certificates on their various social pages and on their resumes.

Many learners and employers have also asked us to create sequences of courses, so they have a program that they can follow. In response to this demand, we launched XSeries, a sequence of courses in a discipline. A student can also obtain a certificate for an XSeries. We believe that programs such as this will increase the motivation of students to complete their existing course, so they can go on to the next one, eventually earning an XSeries certificate to indicate mastery in a given discipline.



MOOCs more about Teachers than Learners
by IRGlover

As social creatures much of our knowledge is built from social interactions, where we integrate our own experiences and beliefs with that of others to build new knowledge and understanding (i.e. Social Constructivism). The current dominant MOOC model is extremely procedural, teacher-centered and discourages these types of social interactions. While this works well for some subjects (particularly at introductory levels), it is much less effective in other situations. How can the large MOOC platforms, and EdX in particular, encourage a more social method of learning?

Agarwal: I am not sure I agree with your thesis that the current dominant MOOC model is extremely procedural. Although many components are procedural, social and peer learning is a central part of MOOC platforms. Take the example of the discussion forum. Learners post questions on the forum, and a majority of the answers and discussion comes from peer learners. Many students are telling us that they are learning by teaching on the forum.

Based on feedback from many of our colleagues in the humanities where the small discussion group is a central part of the learning process, we created the concept of cohorts in the discussion forum. Cohorts enable us to create many smaller groups and assign each student to a specific cohort. A number of courses such as the Greek Heroes course from HarvardX and the Copyright Law course also from HarvardX used this mechanism during Spring 2013. Harvard went a step further and engaged their alumni to help as community moderators in each of these cohorts.

Many courses are taking advantage of the international diversity of the students on the platform to conduct course related studies and discussions. The Public Health course from Harvard for example did a study where they surveyed their worldwide students in the course in how they did certain things in their part of the world, and then discussed it on the forums. Such diversity and numbers are simply not possible in a campus class.

We have a wiki on the platform as part of each course where students work together and contribute material. In fact, in some of the courses the wiki notes produced by students are so good that it would not take the group much additional effort to turn the set of notes into a textbook.

For question types where the computer is not great at grading, we support peer grading, in which students grade each other’s work. Here again, the peer group is key to the learning and assessment process.

That said, we can and should do even more to take advantage of the peer learning effects that are possible with MOOCs. Here is but one example of the many that we have been thinking about.

It turns out that we can post some of our simulation laboratories on the wiki where students can collaborate on labs. Imagine asking the students to work together to solve a lab problem working jointly to design a circuit. In fact, when we posted our simulation lab tool on the wiki just for fun in the 6.002x circuits course, we had a student who had begun the design of a circuit on the wiki, and had then stepped out for a break. When he returned, he noticed his circuit had changed and he reported to us that the tool was buggy. Little did he realize that his circuit had been modified by an unknown collaborator probably from around the world!



Are teaching faculty dead?
by Anonymous Coward

Currently, many 2nd and 3rd tier universities rely heavily on (cost-effective) adjuncts and teaching-only "lecturers" for the majority of the instructional duties. To further maximize revenue, these schools could replace adjuncts and lecturers with MOOCs taught by professors at 1st tier universities. Are lecturers and adjuncts dead?

Agarwal: I believe the exact opposite will happen! I predict that 10 years from now there will be a much greater number of adjuncts and lecturers, and a much greater choice of courses for students in 2nd and 3rd tier universities. Similarly, the quality of education will also increase in all universities, whether 1st tier or 3rd tier. Read on.

Think of online learning and platforms to create great online course material not as making teachers obsolete, but as tools for teachers, which will enable them to do a lot more than they could previously with the time they had. Sadly, one of the few teaching tools we have given professors in the past few centuries has been a piece of chalk. Blogging sites and online news did not make journalists obsolete, rather created 10 million journalists, while making news much more interactive and exciting. In much the same manner, online learning has the potential to transform education. Online learning and moocs can give existing professors new tools to create additional modalities of teaching. Instructors that are quick to experiment with these new tools are likely to find success, creating new ways of engaging our millennial generation of students who are perfectly comfortable texting, tweeting, blogging, emailing, moocing, youtubing, slashdotting (:-) ) and snapchating.

We have also been experimenting with blended models of education where you combine the best of online and in person. This model seems perfect for courses which are taught by an adjunct, where the adjunct has the choice to leverage online material from a MOOC course much like a new-age textbook, and improve the overall student experience. The blended approach works great even for courses taught by a professor at a top tier university. My colleagues and I used this model ourselves for an experimental group of 20 students at MIT for the MITx 6.002x circuits and electronics, our first MOOC course on edX, and came away with a good experience (see the paper by Afridi et al at edx.org).

In one these blended model experiments in which we collaborated with San Jose State University, the students watched videos, did online exercises and virtual labs from the 6.002x course on our online platform, and then came to class for in-person activity including working in small groups and interacting with a professor to get a deeper understanding of the material. Armando Fox from Berkeley has termed such courses SPOCs (small private online courses). The pass rate for this course went from a traditional average of about 60% to 90% with the blended class. Research by Brewlow et al (you can see paper athttps://www.edx.org/research[edx.org]) also shows that student success in a mooc course was positively correlated with working with an instructor or lecturer.

This may seem surprising, but I believe that the opportunity to use MOOCs in the SPOC model will result in universities hiring more adjunct faculty to offer wider choices of quality courses for their students. This will be a good thing for both students and adjuncts.

A few weeks ago you had the chance to ask the President of edX, Anant Agarwal, about the future of education and the growth of MOOCs. Below you'll find his answers to your questions. Could someone set up an archive site ?
by Taco Cowboy

The other day I was looking up an open course offered by Harvard that I had meant to try my hands on for the past few years (I know, I know, I procrastinated). I googled it up and clicked on the link - and long and behold, the Harvard server told me that the course had been deleted, due to some "incompatibility" of the video format and their new hardware, or something like that. I did not take that course. I have no idea if it was good or not. What if it was an excellent course ? Now that that particular course is gone (a few lessons still can still be had on youtube), the opportunity cost for many people does accumulate. If there was only an archive site for all the open-courses, wouldn't that be great ?

Agarwal: I know, why do today what you can put off till tomorrow :-) Seriously, a large number of our learners like you just want to audit a course and prefer it to be always available. For that reason, a significant number of our previous courses are indeed offered as "past courses" or "archived courses". You will see this on the announcements page when you go into such an archived course: "This is a past/archived course. Certain features of this course may not be active, but we still invite you to explore the available materials." To see archived courses you can click on courses at edx.org, select courses, and then select past courses. For example, the justice course from HarvardX is available as an archived course at: edx.org. In general, we strongly encourage our university partners to offer their past courses as archived courses. A few courses (like the one you tried perhaps) are not offered as an archived course. There are several reasons why a university might choose not to offer their course as an archived course. The next version of the course might have started, and so they might take down the archived course because they reuse some of the questions which have their answers available as handouts. Or, the course might need significant resources to run, for example, special grading servers may need to be maintained, for which they may not have resources.



Professors
by pyrognat

Dear Anant,

I am a young researcher at your own institution. One might think that online courses (such as those offered by edX) will make professors (at least those who teach) obsolete. What role do you see professors playing in the future of education? As someone on that career path, I am particularly interested in your views.

Sincerely,
Nathaniel Stapleton

Agarwal: First, let me point out that professors have been using various forms of online learning for decades, so what edx and others are doing with moocs is simply taking what we have been doing to the next level in terms of scalability and quality.

Think of online learning and platforms to create great online course material not as making teachers obsolete, but as tools for teachers, which will enable them to do a lot more than they could previously with the time they had. Sadly, among the few teaching tools we have given professors in the past few centuries has been a piece of chalk. Blogging sites and online news did not make journalists obsolete, rather created 10 million journalists, while making news much more interactive and exciting. In much the same manner, online learning has the potential to transform education. Online learning and moocs can give existing professors new tools to create additional modalities of teaching. Professors and soon-to-be professors like yourself that are quick to experiment with these new tools, are likely to find success, creating new ways of engaging our millennial generation of students who are perfectly comfortable texting, tweeting, blogging, emailing, moocing, youtubing, slashdotting ( :-) ) and snapchating.

We have also been experimenting with blended models of education where you combine the best of online and in person. Initial results are promising. In one these blended model experiments in which we collaborated with san jose state university, the students watched videos, did online exercises and virtual labs on the edx online platform, and then came to class for in-person activity including working in small groups and interacting with the professor to get a deeper understanding of the material. The pass rate for this course went from a traditional average of about 60% to 90% with the blended class. Research by Brewlow et al (you can see paper at edx.org also shows that student success in a mooc course was positively correlated with working with an instructor or mentor.



Making money
by peter303

How do the VCs plan to make back their investments? That was not clear to me.

Agarwal: edX is a non-profit startup, but we still need to be self-sustaining. So we are exploring several business models spanning both B2C and B2B. Unlike other MOOC providers, since edX is a non-profit, we do not have VC investors, rather we have institutional funders including MIT and Harvard. Many of our university partners such as the UT system have also contributed $s to our cause. One can view these initial contributions as an early round of investment in edX. In the B2C space, we use a freemium model. Students can take our courses for free, and they can even get an honor code certificate for free. However, we charge a small fee for ID verified certificates. In addition to the fee for the ID verified certificate (typically between $25 and $100 for a course), we offer the students the opportunity to contribute voluntarily to our cause by paying more than the minimum fee.

In the B2B space, we work with corporations, NGO, governments and state university systems. Here, edX hosts the platform for the institution, and the institution offers its courses on our platform for their employees or for other targeted learners. They pay edX for various services such as hosting, education services, training, etc. We have a number of such partnerships underway, including one with IMF.



What about moving to an badges based system
by Joe_Dragon

Where courses can be more right sized and not jammed / padded out into the older collgle time table system. Where you don't have to take a big 2-4-6+ year block of time to get something that says to you know some thing. It can also make ongoing education / learning new skills have more meaning as well. What about merging Professional certification systems into an over all badges based system? Do you think this is an good idea?

Agarwal: This is a good idea.

Good old degrees with 4 years in college, in stove-piped departments are an antediluvian concept. Why 4 years? My bachelor's degree at IIT madras was a 5-year program, and I probably use 20% of that in my job at this point in my life. Why specialize in one field? In today's fast moving, nimble world, learners and employers are looking for multidisciplinary education. Further, they are looking to refresh their skills as the workplace needs change.

A degree, fundamentally, is a signaling mechanism. It tells an employer that the holder likely has a set of skills. In the modern world, employers are looking for a diverse set of skills. So a promising alternative is one where learners acquire a traditional degree for 2 years (certainly less than the traditional 4 years), so that they can satisfy the bean-counting HR folks that need to see something that spells degree on the resume (and that will go away too with time). Then, the learner augments that with various other signals -- call them badges, or mooc-style certificates, so that employers can now see the whole portfolio.

A couple of weeks ago, edX also announced another interesting signaling mechanism, an XSeries certificate. Here a learner can take a sequence of courses in a given discipline on edx, and by passing all the courses in the sequence, earn an XSeries certificate. MIT and edX announced XSeries certificates in foundations of computer science and supply chain management for a start. Some of you have commented that individual courses do not mean much to employers, but mastery of a discipline is akin to a mini-masters or a minor program, and is meaningful. We believe these XSeries credentials from edX will have strong signaling value to employers and take MOOC credentialing and badging one additional step forward.

I can go on and on about this. I believe that in the future degree transcripts will be replaced by a colorful portfolio of credentials, including badges and certificates. OK OK, and degrees too, for backwards compatibility.



Why not get rid of the Honor Code?
by blue trane

The Honor Code seems like a holdover from obsolete old educational methods. It seeks to make the free and open sharing of information somehow dishonorable.Often students want to help each other in the forums. The quizzes and exercises can provide interesting applications that the instructor didn't go over in the videos. Why censor a student who, of his own free will, wants to help out another student?

The Honor Code, in forbidding explicit help to questions on assignments, encourages deviousness and obfuscation in the forums. Often, posts will be made deliberately vague, so that one has to make guesses, or "read between the lines", or try to mind-read. Wouldn't it be better to encourage clear, simple explanations on the forums? Students are sometimes as (or more) knowledgeable than the instructors, and can explain things in a better, simpler way. Often the instructors have been at the subject so long that they've forgotten what it's like to look at the material for the first time. Other students can fill in the gaps. But the Honor Code works against this type of peer-helping-peer interaction, because often the most interesting applications of the subject are in the exercises.

When I've argued for the dissolution of the Honor Code before, one response has been: you just have to wait until after the deadline. However this response is not adequate, because often the deadlines are a few weeks off. When a student is engaged in a particular problem, that is the most opportune time for him to learn. I've had questions I couldn't answer, and haven't gone back to check how to do them after the deadline passes, because I'm now involved in something else...

I think the Honor Code works against the spirit of openness and freedom of speech that the internet was founded on. What kind of skills are you trying to teach, by enforcing the Honor Code? Does a client care whether you "cheated" by looking up the answer to a programming problem on the internet, when you're writing a program for him?

I think there are better technological solutions than enforcing an archaic Honor Code. Can you put a "spoiler" tag on posts that reveal how to do an assignment question, and reward those students who don't click on those posts? You're supposed to be tracking our every click...

Agarwal: The question about the honor code brings up fundamental issues about teaching and learning and the answer is not obvious. Our current education approach, whether online or in person, attempts to address two distinct issues: learning and student assessment.

Assessment attempts to verify that the student has learned the given material and has achieved a certain level of mastery in the subject. EdX and our partner universities give an honor code certificate to students who pass the course, as measured by assessments like homeworks and exams. These certificates are highly valued by a significant proportion of our learners. They often serve as signaling credentials to employers for example.

For the honor code certificate granted to an individual student to mean something as a signaling mechanism indicating mastery of the material in the course, the assessment method must represent the student’s own level of accomplishment (at least to the extent allowed by the honor code) and not that of the collaboration group to which the student belonged, or the person from whom the student may have gotten an answer. Without an honor code, there would be no honor code certificate. And hence the honor code.

Of course, we have a number of continuing learners to whom the honor code certificate, and hence the honor code, is not important. They are simply interested in learning the material, and do not care about earning a signaling mechanism. For this class of learners, we make our courses available as archived courses once their final assessments have passed their due date. Archived courses have all the answers readily available, but do not offer a certificate. Here, students can discuss answers without any worry about the honor code because there is no certificate attached.



Prestige of Online Degrees
by Calsar

I received an under graduate and master’s degree from traditional universities. I also received two masters’ degrees through on-line classes. In my opinion programs like edX are the future of education, but on-line degrees are still not regarded with the same level of prestige as those received through traditional education. In part this has been due to questionable practices of some on-line educational institutions. How can this perception be changed and do you have and do you have any plans in that regard?

Agarwal: Indeed, online education has been around for several decades. But online degrees and credentials were previously not as highly regarded as those obtained through campus education. There were a number of reasons for this beyond those that you have identified including the poorer quality of the education and relatively low success rates compared to campus programs.

This trend has changed completely in the past year for several reasons. First, reputable non-profit institutions like MIT, Harvard, Stanford, and Berkeley and their star faculty are embracing MOOCs and online learning wholeheartedly, and validating online education. To be sure, these institutions were also experimenting with online learning for decades, but their recent public endorsement and huge investments in MOOCs and rigorous online education has really helped the field gain a new respect. Imagine, you can now take online one of the world’s best AI course from Dan Klein and Pieter Abbeel from Berkeley.

Second, recent advances in computing and communications have completely altered online education and dramatically improved its quality. Content distribution networks have transformed high quality video delivery. Cloud computing enables elastic scaling to hundreds of thousands of learners and provide instant feedback for learning exercises. Social networking and discussion forums enable peer to peer learning at a global scale. And massive amounts of computing power in the palm of our hand has enabled us to offer simulation based virtual laboratories with online game-like-interaction latencies. Of course, getting large numbers of the world’s great instructors to experiment with online teaching helps online education quality more than any technology. New technologies like virtual labs and the efforts of leading instructors have enabled us to offer courses whose rigor is similar to that of campus courses.

Third, reputed universities are offering credentials like free honor code certificates for passing online courses. The introduction of ID verified certificates for a small fee adds even more to the credibility of the credential.

These are some examples of approaches that have improved the perception of online education. However, we can do a lot more. We are working with campuses to blend online learning with in-person education. Initial results in student outcomes at institutions like San Jose State University, Mass Bay Community College, and Bunker Hill Community College, are very promising. As more of these blended experiments yield good results, the perception of online learning will continue to improve, and be viewed not as an alternative to on-campus learning, but rather as complementary to it.



Motivation in Online Courses
by Eric Hosman

Motivation plays a large role in any educational setting, but this is especially true in online courses. How do you best maintain a learner's motivation after the initial novelty has worn off? Online educational opportunities attract a wide array of learners, and we can't expect them all to be intrinsically motivated at a level consistent enough to complete a course, even if they are taking it for college credit or future growth opportunities. What are the best techniques to keep as many learners as possible engaged throughout an online course.

Agarwal: As we listen to our students who are taking MOOCs, or understand the correlators to success from researchers who are studying the big data of learning, we are seeing a number of approaches to increase student motivation and improve outcomes. Here are some of them.

A recent study by Lori Breslow et al edx.org found that students were motivated and had a better chance at success if they worked in groups. Anecdotal data also points to this result. I was chatting with a learner who was taking an edX course in Krgezistan and he had to drop out half way. Later he told me that he discovered that there were three other students taking the same class in his town close to where he lived, and had he known that, he would have connected with them and stayed the course. Following these learnings, we are encouraging our students to form study groups through meetups, and facilitating this with links to meetup sites in the course introduction email.

We are hearing from our learners that the certificate is a big part of the motivation. Although you can audit a course, we are giving free honor code certificates, and also ID verified certificates to our students for a small fee. EdX also provides an online link with a unique identifier for each of these certificates so that an employer can verify that the certificate is indeed authentic. Students are linking these certificates on their various social pages and on their resumes.

Many learners and employers have also asked us to create sequences of courses, so they have a program that they can follow. In response to this demand, we launched XSeries, a sequence of courses in a discipline. A student can also obtain a certificate for an XSeries. We believe that programs such as this will increase the motivation of students to complete their existing course, so they can go on to the next one, eventually earning an XSeries certificate to indicate mastery in a given discipline.



MOOCs more about Teachers than Learners
by IRGlover

As social creatures much of our knowledge is built from social interactions, where we integrate our own experiences and beliefs with that of others to build new knowledge and understanding (i.e. Social Constructivism). The current dominant MOOC model is extremely procedural, teacher-centered and discourages these types of social interactions. While this works well for some subjects (particularly at introductory levels), it is much less effective in other situations. How can the large MOOC platforms, and EdX in particular, encourage a more social method of learning?

Agarwal: I am not sure I agree with your thesis that the current dominant MOOC model is extremely procedural. Although many components are procedural, social and peer learning is a central part of MOOC platforms. Take the example of the discussion forum. Learners post questions on the forum, and a majority of the answers and discussion comes from peer learners. Many students are telling us that they are learning by teaching on the forum.

Based on feedback from many of our colleagues in the humanities where the small discussion group is a central part of the learning process, we created the concept of cohorts in the discussion forum. Cohorts enable us to create many smaller groups and assign each student to a specific cohort. A number of courses such as the Greek Heroes course from HarvardX and the Copyright Law course also from HarvardX used this mechanism during Spring 2013. Harvard went a step further and engaged their alumni to help as community moderators in each of these cohorts.

Many courses are taking advantage of the international diversity of the students on the platform to conduct course related studies and discussions. The Public Health course from Harvard for example did a study where they surveyed their worldwide students in the course in how they did certain things in their part of the world, and then discussed it on the forums. Such diversity and numbers are simply not possible in a campus class.

We have a wiki on the platform as part of each course where students work together and contribute material. In fact, in some of the courses the wiki notes produced by students are so good that it would not take the group much additional effort to turn the set of notes into a textbook.

For question types where the computer is not great at grading, we support peer grading, in which students grade each other’s work. Here again, the peer group is key to the learning and assessment process.

That said, we can and should do even more to take advantage of the peer learning effects that are possible with MOOCs. Here is but one example of the many that we have been thinking about.

It turns out that we can post some of our simulation laboratories on the wiki where students can collaborate on labs. Imagine asking the students to work together to solve a lab problem working jointly to design a circuit. In fact, when we posted our simulation lab tool on the wiki just for fun in the 6.002x circuits course, we had a student who had begun the design of a circuit on the wiki, and had then stepped out for a break. When he returned, he noticed his circuit had changed and he reported to us that the tool was buggy. Little did he realize that his circuit had been modified by an unknown collaborator probably from around the world!



Are teaching faculty dead?
by Anonymous Coward

Currently, many 2nd and 3rd tier universities rely heavily on (cost-effective) adjuncts and teaching-only "lecturers" for the majority of the instructional duties. To further maximize revenue, these schools could replace adjuncts and lecturers with MOOCs taught by professors at 1st tier universities. Are lecturers and adjuncts dead?

Agarwal: I believe the exact opposite will happen! I predict that 10 years from now there will be a much greater number of adjuncts and lecturers, and a much greater choice of courses for students in 2nd and 3rd tier universities. Similarly, the quality of education will also increase in all universities, whether 1st tier or 3rd tier. Read on.

Think of online learning and platforms to create great online course material not as making teachers obsolete, but as tools for teachers, which will enable them to do a lot more than they could previously with the time they had. Sadly, one of the few teaching tools we have given professors in the past few centuries has been a piece of chalk. Blogging sites and online news did not make journalists obsolete, rather created 10 million journalists, while making news much more interactive and exciting. In much the same manner, online learning has the potential to transform education. Online learning and moocs can give existing professors new tools to create additional modalities of teaching. Instructors that are quick to experiment with these new tools are likely to find success, creating new ways of engaging our millennial generation of students who are perfectly comfortable texting, tweeting, blogging, emailing, moocing, youtubing, slashdotting (:-) ) and snapchating.

We have also been experimenting with blended models of education where you combine the best of online and in person. This model seems perfect for courses which are taught by an adjunct, where the adjunct has the choice to leverage online material from a MOOC course much like a new-age textbook, and improve the overall student experience. The blended approach works great even for courses taught by a professor at a top tier university. My colleagues and I used this model ourselves for an experimental group of 20 students at MIT for the MITx 6.002x circuits and electronics, our first MOOC course on edX, and came away with a good experience (see the paper by Afridi et al at edx.org).

In one these blended model experiments in which we collaborated with San Jose State University, the students watched videos, did online exercises and virtual labs from the 6.002x course on our online platform, and then came to class for in-person activity including working in small groups and interacting with a professor to get a deeper understanding of the material. Armando Fox from Berkeley has termed such courses SPOCs (small private online courses). The pass rate for this course went from a traditional average of about 60% to 90% with the blended class. Research by Brewlow et al (you can see paper athttps://www.edx.org/research[edx.org]) also shows that student success in a mooc course was positively correlated with working with an instructor or lecturer.

This may seem surprising, but I believe that the opportunity to use MOOCs in the SPOC model will result in universities hiring more adjunct faculty to offer wider choices of quality courses for their students. This will be a good thing for both students and adjuncts.

Anant Agarwal is a professor of Electrical Engineering and Computer Science at MIT and the President of edX. A massive open online course platform founded by MIT and Harvard, edX offers numerous courses on a wide variety of subjects and is affiliated with 29 different institutes of higher education. Mr. Agarwal has agreed to take some time out of his schedule and answer your questions about edX and the future of learning. As usual, ask as many as you'd like, but please, one question per post.

Anant Agarwal is a professor of Electrical Engineering and Computer Science at MIT and the President of edX. A massive open online course platform founded by MIT and Harvard, edX offers numerous courses on a wide variety of subjects and is affiliated with 29 different institutes of higher education. Mr. Agarwal has agreed to take some time out of his schedule and answer your questions about edX and the future of learning. As usual, ask as many as you'd like, but please, one question per post.

The edX project today announced that they are joining forces with Stanford and releasing the source to edX on June 1st. As part of the platform going Free, Stanford will be integrating features from their Open Source Class2Go project. From Stanford: "Mitchell said that Stanford's Class2Go platform development team has been in contact with the edX team for a number of months, and that much code is already synchronized so that the collaboration between the two platforms will be a smooth one. The advantage will then be 'a larger team building one strong open source platform, rather than two competing open source platforms, which we think will be more desirable for universities around the world,' Mitchell added."