Analysis: People Who Use Firefox Or Chrome Make Better Employees

HughPickens.com writes: In the world of Big Data, everything means something. Now Joe Pinsker reports that Cornerstone OnDemand, a company that sells software that helps employers recruit and retain workers, has found after analyzing data on about 50,000 people who took its 45-minute online job assessment, that people who took the test on a non-default browser, such as Firefox or Chrome, ended up staying at their jobs about 15 percent longer than those who stuck with Safari or Internet Explorer. They also tended to perform better on the job as well. Chief Analytics Officer Michael Housman offered an explanation for the results in an interview with Freakonomics Radio: "I think that the fact that you took the time to install Firefox on your computer shows us something about you. It shows that you're someone who is an informed consumer," says Housman. "You've made an active choice to do something that wasn't default." But why would a company care about something as seemingly trivial as the browser a candidate chooses to use? "Call centers are estimated to suffer from a turnover rate of about 45 percent annually (PDF), and it can cost thousands of dollars to hire new employees," says Pinsker. "Because of that, companies are eager to find any proxy for talent and dedication that they can."

Comments

And when the "default" is the preferred option?

[RogueyWon]

Well... up to a point. I can follow the logical connection that would suggest that people who act as informed consumers are likely to make better employees.

However, I've recently switched back to Internet Explorer after more than a decade with Firefox and a short experiment with Chrome. I did so because I find that comparing across the latest versions of all three, IE was my favourite in terms of performance and user-experience. So I made a reasonably informed decision to use it.

Making practical use of data like this would be more justifiable if there was a clear case that the "default" option was inferior (which in fairness, IE has sometimes been previously).

Re:And when the "default" is the preferred option?

[CastrTroy]

You see a lot of this kind of stuff with people using the Surface Pro. Chrome has some behaviours (bugs) that cause it to drastically increase the power usage causing the battery to drain quite quickly. A lot of people have switched back to IE, and found that the browser is quite good. The Metro/Modern version is basically the only browser that is optimized for touch screen, making it great for the Surface. Also, the desktop version is actually quite good and most people have no complaints. The biggest downfall is that IE doesn't have great adblocking, but it seems that adblock combined with Chrome is sure fire way to have a short battery life.

Re:And when the "default" is the preferred option?

[jpapon]

You're missing the point of the analysis (as is this Housman fellow). Which browser is the better choice is irrelevant. The analysis shows that better employees are currently more likely to use Chrome or Firefox. Whether or not Firefox/Chrome are better doesn't matter, all that matters is that, on average, better people use them (according to their measures).

Re:And when the "default" is the preferred option?

[serviscope_minor]

It doesn't matter: they're looking at the marginals. Switching to a non-default almost certainly means you have some level of competence. People using the default will contain a mixture of informed and uninformed people, and about 1/3 of the density of informed people as random selection (assuming informed people are split 3 ways between firefox/chrome/default).

If you have *no* other information, you will get better employees on average if you choose the non-default browser users.

If you've got some way of assessing competence, then you're no longer looking at the marginal distribution and so browser choice provides no further information over employee competence.

Or, looking at it another way, browser use is competence+noise. You can infer competence up to the noise by looking at the browser. However if you know their competence, then all browser use tells you is useless information about the noise.

Re:And when the "default" is the preferred option?

[meta-monkey]

Making practical use of data like this would be more justifiable if there was a clear case that the "default" option was inferior (which in fairness, IE has sometimes been previously).

No, it's justifiable by the fact that people who used a non-default browser stayed 15% longer. It doesn't matter if there are false positives (people who used Firefox installed by someone else) or false negatives (people who, like you, made an effort to choose and chose IE). On average, choosing someone who used Firefox or Chrome gets you an employee who stays15% longer (and "performed better" although I don't know the metric or methodology they used to determine that), and if that's what you want, you should do it.

You're concerned about false negatives, thinking people will be worse because they used IE, even though they chose IE because to them it is superior. But those people will be completely drowned out by the masses who use IE because it is the default on their computer.

Re:And when the "default" is the preferred option?

[Drethon]

FF and Chrome have had performance and memory issues as of late but for me IE has too many other issues. Tabs locking up on a webpage that lockup all tabs in IE or even Windows. Tabs crashing periodically (this I think is more due to the lousy business apps). Freezing when creating new tabs and all initial attempts at typing an address going into previous tabs (often loading the new page in the old tab). I like Chrome and FF handling of tabs much better.

Re:And when the "default" is the preferred option?

[Anonymous Coward]

So... you are still using the same old default operating system?

Re:And when the "default" is the preferred option?

[tehlinux]

>The biggest downfall is that IE doesn't have great adblocking.

Tracking protection lists do a pretty good job of it for me.

Why are they talking about call center turn over?

[NeutronCowboy]

The reason call center turn over is so huge is because the job sucks. Low pay, tough hours, no control over what you do, little chance for success, and career means becoming a manager who has no training and needs to explain to his/her bosses why the peons are leaving in droves.

The summary shows the problem with big data: it's not the data that counts, it's what you do with it. And no algorithm in the world can make you make good decisions.

What if we use them all?

[Theovon]

I actually use all four of those browsers. I use Chrome for Google Docs, Firefox for high performance JavaScript applications and Safari for most everything else. Occasionally, I'm forced to boot Windows in a VM and use IE because of some idiots who tailored their legacy web app specifically for IE.

I've tried using just one, but each browser has or has had too many deal-breaker bugs. Actually, I used to use Firefox for everything, but there were too many problems with it, so I switched to Safari. It's improved a lot since then, but it doesn't integrate that well with Mac OS X or Google Docs, so I can't use it for everything. One reason I tend to avoid Chrome is that the developers are assholes. When I report bugs, they just argue with me and tell me I'm wrong. I actually formally studied HCI and cognitive engineering, so unlike those assholes, I know what I'm talking about.

Re:What if we use them all?

[Anonymous Coward]

If you've studied cognitive engineering, you should know by now that labeling a person an "asshole" is the best way to get them to seriously want to do your bidding.

Re:Take a look at the graph in TFA

[gnupun]

According to that graph, 5% more IE users are likely to quit compared to FF users. So hire only FF users to save money?

Re:No doubt:

[FyRE666]

If you're going to try a hackneyed troll like this, kid, learn to be a little more subtle. You're like a guy running naked through a shopping mall, screaming "I'm a troll, I'm a troll! Please give me some attention! Pleeeeease!!!"

i love this game

[nimbius]

Let me try my hand at this. I dont have a recruitment firm to clickbait but its worth a shot.
Analysis: people who eat marmite make better television repairmen
Analysis: people who ride motorcycles make better carpet salesmen
Analysis: people who chew on styrofoam and roll around in pickles make louder burps.

No Choice?

[Drethon]

What about those of us who don't have any choice in the matter and are forced to use IE? I suspect higher turnover due to annoyance with IT.

Re:How about Linux?

[rmdingler]

The study measured browsers used, regardless of machine.

Aside from that, it seems likely studies such as this will be (and probably already are) used to make decisions the facts are unqualified to support.

"Yes it true. According to our studies, people hired on Thursday will take more crap than your average employee."

45% turnover rate IS the problem

[duckintheface]

If an industry has a 45% turnover rate, as is cited for call centers, the problem is not the "talent and dedication" of the employees. The problem is that the job is structured in such a way that it is mind numbing, repetitive, and unsatisfying to the workers. And BTW, if you really want workers who can perform under such conditions, you are NOT looking for someone who wants control over their circumstances as indicated by the selection of a non-default browser.

Re:45% turnover rate IS the problem

[Kiaser Zohsay]

In this case, the "proxy for talent and dedication" would be "not working in a call center".

Re:45% turnover rate IS the problem

[jellomizer]

However there is a skill, in finding empowerment even in mind numbing jobs. Installing Chrome or Firefox, is usually one way, as it is one of those things that are normally "officially" against the rules, but you do it anyways, because you know your browsing experience will be a little bit safer. Knowing when to bend/break the rules, and when follow them is an important skill.
I see too many people who just suffer their job and their performance hinders, because they so cautious on following the rules, that they cannot break out of the hum-drum activity. I also see people get fired for just going too gung ho and broke the rules just because they didn't like them.

But there is a line, It is one thing to use a different internet browser, then to say use a different email client.

Re:45% turnover rate IS the problem

[GameboyRMH]

Hey there are lots of talented and dedicated people who end up settling for shitty jobs - we aren't all privileged hipsters with flawless resumes living in big US coastal cities with hot startup scenes.

Re:45% turnover rate IS the problem

[pr0fessor]

Installing Chrome or Firefox, is usually one way, as it is one of those things that are normally "officially" against the rules, but you do it anyways, because you know your browsing experience will be a little bit safer.

Really? Most of the time when I catch someone using an unauthorized software it's to get around security features like a white listed proxy.

Re:45% turnover rate IS the problem

[roc97007]

There seems to be a movement in industry to *make* jobs mind numbing, repetitive and unsatisfying. Currently we're going through the Lean "Standardized Work" process, which in itself isn't necessarily bad, but the stated goal of management in this case is to use the data to reduce jobs to the simplest, most repetitive tasks possible. [1] This goes along with the whole "replace admins with 'best shore' operators" mentality, which at its root has the belief that in IT jobs there is no actual thinking involved, just getting a message, looking the message up in the procedures, and then pressing the button indicated by the procedure, where anything outside procedure results in a call to the vendor. As those of us here well know, there's all kinds of reasons why this doesn't work in practice, and in some cases leads to longer, more severe outages, but besides that, the long range effect is to drive out the best and brightest. I guess they get a job with the vendor?

(But why don't we replace management with offshore operators using Decision Management software? Because management gets to make the decision.)

[1] Or at very least, simple enough that management can understand them.

I must be in a bad mood today.

Re:Why are they talking about call center turn ove

[Capt.Albatross]

The summary shows the problem with big data: it's not the data that counts, it's what you do with it. And no algorithm in the world can make you make good decisions.

So the problem with pens is that no writing tool in the world can make you a good writer?

You had me for a second, but this is not a valid analogy. The valid analogy would require someone to say that their analysis has found the pen that will make you a good writer.

Re: Why are they talking about call center turn ov

[ceoyoyo]

Writers don't look for magic pens to make them good writers. Managers seem to look for magic tools (like big data) to make them good managers.

Re:How about Linux?

[serviscope_minor]

"Yes it true. According to our studies, people hired on Thursday will take more crap than your average employee."

Except there is no reasonably way to imagine that's the case. Whereas if in some weird-ass world you have no information on which to base hiring decisions except for browser choice, you'll be better on average with non-default browser users.

browser use = competence + noise

Incompetent always use the default, competent people don't always. Therefore if you have only browser use, you can infer competence up to the noise level. However if you already have a separate measure of competence, then all you can infer by browser use is what the noise is.

But this will probably be misused, like the Casio terrorist watch statistic.

Wait, call centers?

[facetube]

Sounds like the Safari and IE users are the better-informed ones.

Is Cornerstone OnDemand full of racists?

[WaffleMonster]

If a group, race or gender 'x' can be statistically shown to be more 'y' or less 'z' then it is ok to use generalities about a group to make judgments about individuals?

This is very same error in judgment routinely used by racists and crackpots to justify all kinds of craziness.

Correlation does not imply causation!

[serviscope_minor]

Oddly enough people on slashdot harp on about "correlation is not causation" pretty much all the time, except oddly this thread where it's a classic example, much better, in fact than all the times it does come up.

Use of alternative broswers is correlated with doing good work, because BOTH are correlated with being competent.

But correlation ISN'T causation and in this case using an alternative browser does NOT cause good work to be done. Neither does doing good work CAUSE you to use an alternative browser.

Example: competent person decides based on relative merits to use default browser.