One nice thing about credit cards- they are not really that picky about disputing payments. They don't exactly go out of their way to promote the service, though.
So true. Whenever I see a DS in public, it's always in the hands of a 5-12 year old. Parents buy these things to keep their kids distracted in the car, out running errands, at a restaurant, etc. The DS does that job just fine, and no sane parent would put a $250 device in the hands of a child. Add to that the adult gamers who bought a PSP and were sorely disappointed with the titles available.
Product extensions on products with a bad reputation rarely fare well. Maybe I'm just bitter that I wasted my money buying a PSP...
It's a fair criticism, and I definitely do not want to paint myself in the corner by claiming that all tests accurately measure all aspects of a child's development. I also don't believe that a machine-readable test is the best way to gauge performance. I DO believe that, in general, kids who score better on tests have better mastery of the subject in question than kids who score worse. I DO believe that pushing a kid to score better on a test is a good thing and outweighs the negatives we can all think of.
Back to the topics you brought up. RE mastery of the English language- the SAT verbal section sets out to do exactly that through a diverse battery of questions involving sentence structure, reading comprehension, and a human-read series of essays. Of course, there is not perfect correlation between SAT Verbal scores and "mastery of the English language," but I'll bet it's within reason. A college looking to admit kids would be stupid to only look at SAT scores. But, it might raise red flags if a student has gotten great grades, but scores significantly poorer than his/her peers on the SAT. As a statistician, you might be willing to take the bet that this pattern is more likely to be caused by a non-challenging school district than by a flawed SAT.
Regardless, we're talking about measuring populations here (improve the performance/grades of all kids, not certain kids). If we threw out all ideas because they are not perfect, we would still be sitting around chewing on raw meat and complaining that the wheel is useless because I've got to build a road to use it.
Hey, I completely agree with you that teaching to the test is bad and a waste of time. But I do think that rewarding kids who get good grades is a good idea, and that folks react to it with a baby/bathwater mentality. If the problem is teachers teaching to the test, work on that problem (make it hard to guess what the test will be like). If the problem is lazy teachers who teach to the test instead of teaching concepts, make the test punish those teachers. Attack the root cause, not this methodology that appears to improve kids' grades. Grades aren't everything, but they are incredibly important.
The root cause of all this hubbub about standardized testing has nothing to do with the kids. It has everything to do with the teacher's union violently attacking any methodology that separates the good from the bad, and the hardworking teachers from the lazy. All unions are deathly afraid of anything that promotes competition between their members. There is a time-tested reason why unions promote seniority rather than performance.
I always get worried when people are willing to discredit measurable metrics in order to promote the unmeasurable (e.g. thinking skills- what does that even mean?). I come from the old school of "If you can't measure it, you can't improve it." Most people are simply not willing to change their minds on anything- when presented with overwhelming evidence in support of the opposition, they will often resort to whatever crutch works rather than doing choosing the logical side. Often this crutch involves throwing away solid data in order to take a position that cannot be disproved (or proved).
There are winners and losers in this world, but everyone thinks they are a winner. 90% of people think their sense of humor is above average; 100% of newlyweds think they have no chance of getting a divorce. Testing is hated because not as much because it is an inaccurate measure of ability (it is, within limits), but because it pops the illusion every parent has that their child is wonderful.
That study successfully showed that people who download a lot also buy a lot (presumably because of their high interest in music). But, it doesn't really answer the question we all want to know- "In the absence of filesharing, would these people have bought more or less music than they did?"
Someone needs to do a randomized test where they take N music consumers and split them into two groups: one that gets a $20 allowance (no strings attached), and another that gets a $20 allowance contingent on an enforceable pledge not to download music. Perhaps they could even expand the pledge a bit to disguise the intention of the study... At the end of a period of time, count up the value of purchases made, and tell us whether there was a statistically different measurement.
Until I see something like this, I personally don't feel that comfortable saying that more downloading LEADS TO (not- is correlated with) more purchases.
p.s. I tried really hard not to type CORRELATION != CAUSATION until now...
I don't think it's right to give a publisher $60 for a game that I will only get 20 hours play out of, regardless of whether or not I can sell it later. I buy a lot of WiiWare games simply because they are only $10. I really don't mind if I only play it for 5 hours when I only pay $10.
So you're willing to pay $2 an hour for entertainment, but not $3?
That's a pretty tight range. Perhaps you should buy a game for $60, play it for 20 hours, and sell it for $30, so that you can get things down to $1.50 an hour? OR, my favorite option- renting at $7.50/20 or ~$.38 an hour.
Sorry for the typo: I meant "where X is less than 100." I read a little bit more to satisfy my own curiousity and found this:
It has been observed that scores outside the range 55 to 145 must be cautiously interpreted because there are smaller numbers of respondents with which to make comparisons in those ranges. Moreover, at such extreme values, the normal distribution is a less accurate estimate of the true IQ distribution.
In actuality there is a higher percentage of the population measured at 3sd+ levels on the test than the probabilities of the normal distribution would predict. Some IQ scoring procedures may attempt to integrate such clusters of statistical outliers into the curve by adjusting the scores so that they better represent actual probabilities (according to Silverman) and in these cases, scores around 145 and above may actually have been notably higher, were they not so adjusted.
I'm not going to take sides in this pissing contest here, but this reminds me of one fallacy that people always make. There is no guarantee that any set of raw data (such as IQs) has a normal distribution. In fact, all the time I hear someone say such and such data set (height, weight, whatever) has a normal distribution when what it actually has is a distribution that is closely approximated by the normal distribution. It is often very convenient to use normal distribution tools for analysis of the raw data set, but it is only an approximation. The only thing that is guaranteed is that if you take a ton of MEANS of random samples from a population of raw data, those means will follow a normal distribution, regardless of how the data is distributed.
I skimmed through that wiki article and the language really doesn't say anywhere that IQ is a normal distribution- what is says is that "The median Full Scale IQ is centered at 100[2], with a standard deviation of 15. In a normal distribution, the IQ range of one standard deviation above and below the mean (i.e, between 85 and 115) is where approximately 68% of all adults would fall."
To me, that sounds like a curve that is approximately normal in the middle, but may or may not be normal as one gets outside of 1 or 2 standard deviations. In fact, I would imagine there are humps in IQ on the lower end of things where certain pathologies tend to cluster (EG Downs Syndrome has a cluster around X, where X 100). That could very well mean that an IQ of say, X has a higher probability than an IQ of X+1, which definitely doesn't meet the definition of a normal distribution.
I've forgotten the technical name for the term, but there's a well proven concept in behavioral theory that people do not proportionally associate the cost of payments with the actual cost of the payment. There's a fixed cost emotionally for any transaction, no matter how small the actual purchase. In simpler terms, people feel much less happy making 100 $1 payments than they do one $100 payment (all things being equal). I actually think this irrational behavior explains the attitude of buyers much better than the article does. I also think it is why micropayments are so hard to make work.
The author of the article is attempting to make the claim that games somehow fall into a special category of consumer behavior. I do not think this is the case at all. We have an irrational "fixed cost" for any transaction, regardless of price. If you buy a nickel gumball and it came out of the machine defective, you'd probably be upset- far more upset than the loss of a nickel actually justifies. Exactly the same way that the disappointment with a.99 game and demanding a refund is out of whack with the loss of a dollar. I personally believe that the disappointment of a failed transaction has to do with a perceived violation of trust (vital to any relationship or trade), which is far more upsetting to a social animal than the loss of an insignificant sum of money.
People are not at their best when they have to be creative, inventive, or thoughtful in real-time with an audience.
I'd be interested for you to elaborate on why you believe this. There's plenty of counter-evidence to this point in that the following practices are time-tested ways of creative thinking: - comedy writers bouncing ideas off each other to start penning a script or sketch - group brainstorming for new names of products and advertising ideas - new product ideation amongst engineers
I'm sure everyone is different, and some prefer quiet solitude to be creative, but it seems the exception rather than the rule in most organizations. I personally find that people tend to be much more cautious and defensive when they have time to craft a well thought out idea, as opposed to blurting out a potentially stupid/creative idea.
Well, since a big portion of IBM is consulting (50%?), it probably makes sense for them to hire consultants who are citizens of the countries they will be working in... How many Americans are fluent in a second language? How many are willing to relocate for 5+ years to Asia? I'd guess that hiring local gives IBM a GOOD name in those markets.
Plus, the people they are hiring are human beings too. Why is hiring an American a more noble thing than hiring an Indian or Chinese person? They need to feed their families too... On the converse side, would you take a pay cut to work for an American company instead of a Japanese company out of patriotism? An Amercian who coasted through high school makes more than a Chinese MD who was the best student in his class. Is that fair? The rest of the world sees American workers as being no better than the rich kid who was born into an aristocratic family and lives off his daddy's hard work.
Clearly one has to balance lost sales from piracy vs. lost sales due to people unhappy with either the concept or inconvenience of DRM. If the 1:1000 ratio is accurate, it does make me think developers should be pushing much further along the spectrum in the direction of DRM-free or DRM-light.
I'm curious. Do you have any experience in this matter? I do, and I can assure you that the absolutes you've defined are not absolutes by any means. It really depends on the deal. Some companies buy to simply increase market share. Some think they can move a new product through their channel cheaply. Some think they can incorporate unique features into their existing product easily. Others want sales relationships (not in this case, from the sound of it).
There's no hard and fast rules to this. In deals I worked on, typically we wanted to use an acquisition as a seed to enter a new business and leverage our brand to grow it. You better believe we were interested in keeping the entrepreneurs happy. The problem is, once you dump a stack of money in an entrepreneur's lap, it's hard to keep them from moving on or retiring, even with an ironclad non-compete.
As you say, though, the only thing that matters is getting a good lawyer who has experience in acquisitions. They won't let you do anything stupid, but it's gonna be expensive. Probably something along the lines of 5% of the value of the deal. If the deal falls through, then you eat that cost.
In my experience, most entrepreneurs are WAY out of whack with regards to the value of their company. Most are convinced that BUYER X is interested in their company because their product/service/whatever is the wave of the future and has no possible substitutes. Usually, BUYER X has the means to develop whatever you're doing, but simply wants to get there faster.
Have you read Dawkin's "The Ancestor's Tale?" There's a tremendous amount of discussion on this issue, and it's one of the most fascinating books I've ever read.
You're right. The proper measurement is brain mass, not size. Unfortunately it's difficult to measure mass of soft tissue in fossils. Typically, there is a linear relationship between log(body mass) and log(brain mass). Humans and dolphins are the real outliers to the line, with primates outlying to a lesser extent. The animals we think of as being smart are always above average in log(brain mass)/log(body mass).
The way you explained it isn't completely true. The value of the options granted above and beyond the strike price is considered compensation, and therefore an expense. It comes off the bottom line for financial reporting purposes. It used to be the other way around for a long time, but FASB changed the rules based on the abuses of the 90's and early 00's. I think the rule changed around 2005? (Not an accountant).
That expression is meant to deride false patriotism, not to imply that all patriots are scoundrels. Quite the opposite, actually, but always misused to bash patriotism.
If the game requires any form of authentication to unlock any significant functionality then it's DRM'ed.
Finally, a co-conspirator! I too demand an end to Slashdot's and gmail's reign of DRM tyranny.
I've been lobbying for years for random people to be allowed to log onto my account to post and read my emails. However, the fatcats in Dexter and the Googleplex want to keep swimming in their moneybins and refuse to acknowledge my request.
Please communicate with me via email at nitpicker999@gmail.com.
Over the years, working with many companies, I've actually found managers at small businesses to be significantly more fast and loose with the law than at big businesses. Lawyers tend to go after deep pockets, so big businesses have HUGE infrastructures set up to prevent lawsuits from impacting them. For example, a lot of small company managers will make blatant racist/homophobic decisions and never suffer the consequences- at a big company, 18 internal policemen would descend on them, and they would be fired.
If you've ever wondered why no one creates a huge dry-cleaning business- it's precisely for this reason. Drycleaners violate a HUGE number of environmental regulations with the chemicals they work and working conditions they enforce. No lawyer will bother to sue even the most egregiously unethical Mom&Pop drycleaner- it's not worth his time. However, if there was a $1B corporation doing the exact same thing (or conducting even signficantly less harm), you can expect major news and major lawsuits.
My unprovable theory is that, in many cases, a $10B corporation will usually act more ethically than a hundred thousand cottage businesses adding up to $10B revenue simply because they tend to be more under the microscope.
Slashdot Mirror
One nice thing about credit cards- they are not really that picky about disputing payments. They don't exactly go out of their way to promote the service, though.
Seriously, way too overpriced compared to the DS.
So true. Whenever I see a DS in public, it's always in the hands of a 5-12 year old. Parents buy these things to keep their kids distracted in the car, out running errands, at a restaurant, etc. The DS does that job just fine, and no sane parent would put a $250 device in the hands of a child. Add to that the adult gamers who bought a PSP and were sorely disappointed with the titles available.
Product extensions on products with a bad reputation rarely fare well. Maybe I'm just bitter that I wasted my money buying a PSP...
It's a fair criticism, and I definitely do not want to paint myself in the corner by claiming that all tests accurately measure all aspects of a child's development. I also don't believe that a machine-readable test is the best way to gauge performance. I DO believe that, in general, kids who score better on tests have better mastery of the subject in question than kids who score worse. I DO believe that pushing a kid to score better on a test is a good thing and outweighs the negatives we can all think of.
Back to the topics you brought up. RE mastery of the English language- the SAT verbal section sets out to do exactly that through a diverse battery of questions involving sentence structure, reading comprehension, and a human-read series of essays. Of course, there is not perfect correlation between SAT Verbal scores and "mastery of the English language," but I'll bet it's within reason. A college looking to admit kids would be stupid to only look at SAT scores. But, it might raise red flags if a student has gotten great grades, but scores significantly poorer than his/her peers on the SAT. As a statistician, you might be willing to take the bet that this pattern is more likely to be caused by a non-challenging school district than by a flawed SAT.
Regardless, we're talking about measuring populations here (improve the performance/grades of all kids, not certain kids). If we threw out all ideas because they are not perfect, we would still be sitting around chewing on raw meat and complaining that the wheel is useless because I've got to build a road to use it.
Hey, I completely agree with you that teaching to the test is bad and a waste of time. But I do think that rewarding kids who get good grades is a good idea, and that folks react to it with a baby/bathwater mentality. If the problem is teachers teaching to the test, work on that problem (make it hard to guess what the test will be like). If the problem is lazy teachers who teach to the test instead of teaching concepts, make the test punish those teachers. Attack the root cause, not this methodology that appears to improve kids' grades. Grades aren't everything, but they are incredibly important.
The root cause of all this hubbub about standardized testing has nothing to do with the kids. It has everything to do with the teacher's union violently attacking any methodology that separates the good from the bad, and the hardworking teachers from the lazy. All unions are deathly afraid of anything that promotes competition between their members. There is a time-tested reason why unions promote seniority rather than performance.
I always get worried when people are willing to discredit measurable metrics in order to promote the unmeasurable (e.g. thinking skills- what does that even mean?). I come from the old school of "If you can't measure it, you can't improve it." Most people are simply not willing to change their minds on anything- when presented with overwhelming evidence in support of the opposition, they will often resort to whatever crutch works rather than doing choosing the logical side. Often this crutch involves throwing away solid data in order to take a position that cannot be disproved (or proved).
There are winners and losers in this world, but everyone thinks they are a winner. 90% of people think their sense of humor is above average; 100% of newlyweds think they have no chance of getting a divorce. Testing is hated because not as much because it is an inaccurate measure of ability (it is, within limits), but because it pops the illusion every parent has that their child is wonderful.
That study successfully showed that people who download a lot also buy a lot (presumably because of their high interest in music). But, it doesn't really answer the question we all want to know- "In the absence of filesharing, would these people have bought more or less music than they did?"
Someone needs to do a randomized test where they take N music consumers and split them into two groups: one that gets a $20 allowance (no strings attached), and another that gets a $20 allowance contingent on an enforceable pledge not to download music. Perhaps they could even expand the pledge a bit to disguise the intention of the study... At the end of a period of time, count up the value of purchases made, and tell us whether there was a statistically different measurement.
Until I see something like this, I personally don't feel that comfortable saying that more downloading LEADS TO (not- is correlated with) more purchases.
p.s. I tried really hard not to type CORRELATION != CAUSATION until now...
Now it's like physics - value is never destroyed, it goes somewhere else.
He hasn't gotten to the chapter on entropy yet... It's going to blow his mind.
I don't think it's right to give a publisher $60 for a game that I will only get 20 hours play out of, regardless of whether or not I can sell it later. I buy a lot of WiiWare games simply because they are only $10. I really don't mind if I only play it for 5 hours when I only pay $10.
So you're willing to pay $2 an hour for entertainment, but not $3?
That's a pretty tight range. Perhaps you should buy a game for $60, play it for 20 hours, and sell it for $30, so that you can get things down to $1.50 an hour? OR, my favorite option- renting at $7.50/20 or ~$.38 an hour.
Keep on thinking that you're completely anonymous when you use a prepaid cellphone.
Can't you buy them with cash?
Or put it on your clients' credit card in exchange for a discount?
Or have a friend buy it?
I'm not using my cell phone for illegal purposes, but that's what I'd do if I were.
Sorry for the typo: I meant "where X is less than 100." I read a little bit more to satisfy my own curiousity and found this:
It has been observed that scores outside the range 55 to 145 must be cautiously interpreted because there are smaller numbers of respondents with which to make comparisons in those ranges. Moreover, at such extreme values, the normal distribution is a less accurate estimate of the true IQ distribution.
In actuality there is a higher percentage of the population measured at 3sd+ levels on the test than the probabilities of the normal distribution would predict. Some IQ scoring procedures may attempt to integrate such clusters of statistical outliers into the curve by adjusting the scores so that they better represent actual probabilities (according to Silverman) and in these cases, scores around 145 and above may actually have been notably higher, were they not so adjusted.
I'm not going to take sides in this pissing contest here, but this reminds me of one fallacy that people always make. There is no guarantee that any set of raw data (such as IQs) has a normal distribution. In fact, all the time I hear someone say such and such data set (height, weight, whatever) has a normal distribution when what it actually has is a distribution that is closely approximated by the normal distribution. It is often very convenient to use normal distribution tools for analysis of the raw data set, but it is only an approximation. The only thing that is guaranteed is that if you take a ton of MEANS of random samples from a population of raw data, those means will follow a normal distribution, regardless of how the data is distributed.
I skimmed through that wiki article and the language really doesn't say anywhere that IQ is a normal distribution- what is says is that "The median Full Scale IQ is centered at 100[2], with a standard deviation of 15. In a normal distribution, the IQ range of one standard deviation above and below the mean (i.e, between 85 and 115) is where approximately 68% of all adults would fall."
To me, that sounds like a curve that is approximately normal in the middle, but may or may not be normal as one gets outside of 1 or 2 standard deviations. In fact, I would imagine there are humps in IQ on the lower end of things where certain pathologies tend to cluster (EG Downs Syndrome has a cluster around X, where X 100). That could very well mean that an IQ of say, X has a higher probability than an IQ of X+1, which definitely doesn't meet the definition of a normal distribution.
I've forgotten the technical name for the term, but there's a well proven concept in behavioral theory that people do not proportionally associate the cost of payments with the actual cost of the payment. There's a fixed cost emotionally for any transaction, no matter how small the actual purchase. In simpler terms, people feel much less happy making 100 $1 payments than they do one $100 payment (all things being equal). I actually think this irrational behavior explains the attitude of buyers much better than the article does. I also think it is why micropayments are so hard to make work.
The author of the article is attempting to make the claim that games somehow fall into a special category of consumer behavior. I do not think this is the case at all. We have an irrational "fixed cost" for any transaction, regardless of price. If you buy a nickel gumball and it came out of the machine defective, you'd probably be upset- far more upset than the loss of a nickel actually justifies. Exactly the same way that the disappointment with a .99 game and demanding a refund is out of whack with the loss of a dollar. I personally believe that the disappointment of a failed transaction has to do with a perceived violation of trust (vital to any relationship or trade), which is far more upsetting to a social animal than the loss of an insignificant sum of money.
People are not at their best when they have to be creative, inventive, or thoughtful in real-time with an audience.
I'd be interested for you to elaborate on why you believe this. There's plenty of counter-evidence to this point in that the following practices are time-tested ways of creative thinking:
- comedy writers bouncing ideas off each other to start penning a script or sketch
- group brainstorming for new names of products and advertising ideas
- new product ideation amongst engineers
I'm sure everyone is different, and some prefer quiet solitude to be creative, but it seems the exception rather than the rule in most organizations. I personally find that people tend to be much more cautious and defensive when they have time to craft a well thought out idea, as opposed to blurting out a potentially stupid/creative idea.
Not fastidious in using the spell check.
Or perhaps a fascicular error between the brain and typing fingers.
I don't think it's an astroturf by a factitious poster.
Satan worshippers believe in -1 gods.
Well, since a big portion of IBM is consulting (50%?), it probably makes sense for them to hire consultants who are citizens of the countries they will be working in... How many Americans are fluent in a second language? How many are willing to relocate for 5+ years to Asia? I'd guess that hiring local gives IBM a GOOD name in those markets.
Plus, the people they are hiring are human beings too. Why is hiring an American a more noble thing than hiring an Indian or Chinese person? They need to feed their families too... On the converse side, would you take a pay cut to work for an American company instead of a Japanese company out of patriotism? An Amercian who coasted through high school makes more than a Chinese MD who was the best student in his class. Is that fair? The rest of the world sees American workers as being no better than the rich kid who was born into an aristocratic family and lives off his daddy's hard work.
This study found that preventing 1000 pirated copies results in an additional 1 sale.
http://www.gamasutra.com/php-bin/news_index.php?story=17350
Clearly one has to balance lost sales from piracy vs. lost sales due to people unhappy with either the concept or inconvenience of DRM. If the 1:1000 ratio is accurate, it does make me think developers should be pushing much further along the spectrum in the direction of DRM-free or DRM-light.
First, it doesn't say that the RIAA "stopped" doing anything. To "discontinue" does not mean to "stop," it means "to break the continuity of."
Quickly! We need to repaint all the octagonal red signs in this country to say "DISCONTINUE MOTION."
Otherwise I predict severe, permanent traffic problems.
I'm curious. Do you have any experience in this matter? I do, and I can assure you that the absolutes you've defined are not absolutes by any means. It really depends on the deal. Some companies buy to simply increase market share. Some think they can move a new product through their channel cheaply. Some think they can incorporate unique features into their existing product easily. Others want sales relationships (not in this case, from the sound of it).
There's no hard and fast rules to this. In deals I worked on, typically we wanted to use an acquisition as a seed to enter a new business and leverage our brand to grow it. You better believe we were interested in keeping the entrepreneurs happy. The problem is, once you dump a stack of money in an entrepreneur's lap, it's hard to keep them from moving on or retiring, even with an ironclad non-compete.
As you say, though, the only thing that matters is getting a good lawyer who has experience in acquisitions. They won't let you do anything stupid, but it's gonna be expensive. Probably something along the lines of 5% of the value of the deal. If the deal falls through, then you eat that cost.
In my experience, most entrepreneurs are WAY out of whack with regards to the value of their company. Most are convinced that BUYER X is interested in their company because their product/service/whatever is the wave of the future and has no possible substitutes. Usually, BUYER X has the means to develop whatever you're doing, but simply wants to get there faster.
Have you read Dawkin's "The Ancestor's Tale?" There's a tremendous amount of discussion on this issue, and it's one of the most fascinating books I've ever read.
You're right. The proper measurement is brain mass, not size. Unfortunately it's difficult to measure mass of soft tissue in fossils. Typically, there is a linear relationship between log(body mass) and log(brain mass). Humans and dolphins are the real outliers to the line, with primates outlying to a lesser extent. The animals we think of as being smart are always above average in log(brain mass)/log(body mass).
The way you explained it isn't completely true. The value of the options granted above and beyond the strike price is considered compensation, and therefore an expense. It comes off the bottom line for financial reporting purposes. It used to be the other way around for a long time, but FASB changed the rules based on the abuses of the 90's and early 00's. I think the rule changed around 2005? (Not an accountant).
That expression is meant to deride false patriotism, not to imply that all patriots are scoundrels. Quite the opposite, actually, but always misused to bash patriotism.
If the game requires any form of authentication to unlock any significant functionality then it's DRM'ed.
Finally, a co-conspirator! I too demand an end to Slashdot's and gmail's reign of DRM tyranny.
I've been lobbying for years for random people to be allowed to log onto my account to post and read my emails. However, the fatcats in Dexter and the Googleplex want to keep swimming in their moneybins and refuse to acknowledge my request.
Please communicate with me via email at nitpicker999@gmail.com.
No That's a different category from cash. Since they are marked to the market, they can swing drastically based on market fluctuations.
For example, if Microsoft bought a 12% share in Google, it wouldn't be classified as cash even though it could be sold on the market immediately.
Over the years, working with many companies, I've actually found managers at small businesses to be significantly more fast and loose with the law than at big businesses. Lawyers tend to go after deep pockets, so big businesses have HUGE infrastructures set up to prevent lawsuits from impacting them. For example, a lot of small company managers will make blatant racist/homophobic decisions and never suffer the consequences- at a big company, 18 internal policemen would descend on them, and they would be fired.
If you've ever wondered why no one creates a huge dry-cleaning business- it's precisely for this reason. Drycleaners violate a HUGE number of environmental regulations with the chemicals they work and working conditions they enforce. No lawyer will bother to sue even the most egregiously unethical Mom&Pop drycleaner- it's not worth his time. However, if there was a $1B corporation doing the exact same thing (or conducting even signficantly less harm), you can expect major news and major lawsuits.
My unprovable theory is that, in many cases, a $10B corporation will usually act more ethically than a hundred thousand cottage businesses adding up to $10B revenue simply because they tend to be more under the microscope.