If you’ve applied for a job before, you’ve probably fretted over how to answer questions like “are you inclined to rely more on improvisation than on careful planning?” or “do you like to create challenges for yourself when you take on a new project?”. Companies commonly use personality tests filled with questions like these to assess the fit of a potential employees with the company. There’s a whole laundry list of these personality tests here. Some companies, like Google, have even developed their own.
All of these tests are sets of written questions meant to poke and prod at a candidates mind to get a real sense of their ability and personality. However, the meaning behind these questions is relatively transparent, motivating candidates to give the answers they think their employer wants to hear.
Startup Pairwise is taking a different approach to personality tests. Instead of words, Pairwise will use images to test a candidates mentality using data gleaned from their LikeBetter picture game. LikeBetter is a flash based game that shows you a series of pairs of images uploaded by users. For each pair, you pick which image you prefer. Based on the choices you make, LikeBetter makes a guess about your personal traits, which you then confirm or correct. As more people use the system, LikeBetter discovers the strong correlations between the choices people make and the attributes they express.
Based on this data, Pairwise creates a quiz using some of the most highly discriminating pairs, chosen to have the strongest and most confident predictive power across the broadest spectrum of personality traits. They can then track a candidate’s behavior through the test and make an educated guess about their personality based on the correlations they made in LikeBetter. Pairwise does their best to make the test harder to read into by being a completely image based test and using non-obvious pairs (no GI Joe vs. Barbie).
Under the hood, LikeBetter is using an iterative application of Bayes rule called Naive Bayesian inference. The method uses a lot of dense statistics involving proposing hypotheses and dependent probabilities. If you really want to learn about it, check out the Wikipedia entry. On the other hand, the employment quiz is not making and testing hypotheses, but comparing the user’s behavior with the statistics they collected through LikeBetter and determining the the applicants tendency toward either extreme of an attribute (i.e cleanliness vs. messiness).
Pairwise’s first customer is Y Combinator, for whom they crafted this little Y Combinator founder quiz based on personality tests done on all their current founders. Y Combinator will be using the test in their application drive ending April 2nd. We’ve included the test for you to take below. Here’s how I fared.
Update: The quiz has been revised and updated.
Y Combinator Founder Personality test. To take the test, pick which image you prefer in each pair:









Test A:
1. Are you young?
2. Are you smart?
3. Do you have any or want any long term commitments?
4. Do you not want to work?
5. Do you have a laptop and an Internet connection?
If yes, you are one. Can YCombinator give me some money?
You crack me up Alaska.
either a) you and i are identical twins/personalities or b) the result set is the same for all users and this beta is jacked.
i got the EXACT same results as you.
My results, too, were actually quite similar to Nick’s.
It said i am a girl….and definitely a girl.
I am not!
Crap 2.0
I got the same results as Nick too! Obviously, I’m destined to write for TechCrunch, or be the next big thing, or something badass.
Is it really so accurate that it’s able to properly identify that many TechCrunch readers are alike, or does it in fact suck?
Hmm, this test didn’t seem to do any better than a coin flip for me.
I got the same results as Nick also, something is not right, but it looks interesting.
i am i am pick me!
Kinda funny that the googlers are using windows xp
The test is flawed
took the test twice, the second time choosing the exact opposite.
but my results were exactly the same as nicks both times.
Where is the YCombinator test?…I can’t seem to find it the link or on pairwise.
Hmm. Did terribly for me. Gender wrong amongst other things. Tè
All these questionnaires are rot. In the end, the only question that matters is:
“Will your life feel incomplete if you don’t try to start a company?”
If the answer is no, then you are not an entrepreneur.
Looks like they need to refine their software a bit more before they start charging for it… it was way off for me on most things. Apparently I hate reading (nope), I’m very religious (nope again), am not very interested in politics (I work for a political news website…), I’m a morning person (I usually sleep 5am-noon), and I don’t have or want pets (my dog, Einstein, vocally disagrees — woof). Coin flip is right.
Some of their determinations are rather easy to spot. I’m labeled as a non-smoker presumably because I didn’t pick the picture of the person smoking a cigarette… that seems sort of silly. A lot of people smoke but that doesn’t necessarily me they like a picture of a cigarette more than one of, say, mountains (I can’t remember what it was).
Hey!
I got the same exact results as you.
I don’t think the test is working correctly.
Almost exactly the same as Nick for me too. But I do agree with the results for the most part. Expect for pets and religion which were the exact opposite.
nutcase,
The test doesn’t display is Safari which I assume you are using.
ok i don’t get it – clicked like 100 of them and they never stopped…..
Holy cow – those results were pretty far off my own self knowledge. I either need counseling and strong medication or this test is flawed…
It’s a trick. If you have to take a test to find out, you don’t have what it takes.
Imagine if every VC made you take a personality test before they funded you???
Either great minds think alike. Just FYI, the Pairwise guys have been tweaking some things while this article was up. I’m going to try it again later this evening and I suggest you all do too. Till then, http://likebetter.com is fun to play with.
I found out that I am a gay, lazy, scared, messy, devil worshiper.
jim, i got the same thing.
http://prototyp..._combinator.gif
what gives?
well, I’m not gay, so that’s good because I am an only child and my parents really want natural grandchildren. I guess I’m more of a listener than a talker. hm. Oh, and I hate pets. And I am a very religious morning person.
http://www.like...;user_list_id=6
I am a very tidy man who hates reading…
My boyfriend is falling off his chair laughing, straight into one of my piles of crap and magazines that litter the loung/office…
Surely this should be your Sunday 1st post Michael?
I got the exact same results as you Mike. We are gay! Well, we are either gay or we are girls who want boyfriends.
I definitely think they got a line of code wrong somewhere; if they flip everything to opposite of what they have right now, they would much more accurate.
ie. Gender: Girl.
Yikes!
Perhaps to keep things simple they skipped the algorithm and just put up random results. Although random results should have been more accurate than what I got.
Sexual preference? Is it legal to be testing for that? Will this factor be used in determining who gets funded?
Also, if you click on someone else’s link and you’ve already taken the test, it will show *your* results not the other person’s.
Incase you’re wondering why Mike appears so similar to you;) Buggy bugger.
also got the same results.. wtf
This was fun, but ultimately useless. The results are interesting, but distorting. Their web site admits that they are not accurate. Plus, I’m not going to make any judgements on anyone until I see their abilities, potential, performance, and how they adapt to change. We do these things partly by learning about their experience, spending time with them, and immersing them into our environment to see how they react. I admire teams that have all kinds of people – the clean, the messy, the religious, the non religious, etc. Again, I think this is fun and interesting, but nothing more.
C’mon, this fails at the very first question. Can you imagine someone applying for Y-Combinator say they are only remotely interested in Tech? That’s what the average (green bar) seems to indicate though.
Are you willing to give up 10% of your company for $15,000?
Yes.
Congratulations! Welcome to YCombinator!
Oops, I take it back, was reading too fast. No direct questions, just pics, so the response to #1 is not above imagination
Zaid – ah. they need to fix that.
ok, looks like there’s an error in the embed. I think the problem is that everyone is taking the test on a single user account and it’s not resetting. The PW guys should have that fixed soon.
The “which superhero are you” test results were more accurate.
I do love their confidence.
Most tests are usually iffy and put things in the middle somewhere..but these guys are super confident that I’m a girl:)
-Zaid
I am a non-smoking, hate reading, morning person, who is very tidy who does not want a pet. ……. The test is completely bs,
Welcome to lazy recruitment 2.0 or lazy portfolio picker….
Well Im not suprised at the results I met one of the LikeBetter people at an event in SV. He seemed pretty clueless when I started to engage him about algorithims and DSS systems. It was pretty obvious he didnt have a GD clue about either.
Perhaps they need to try rereading the definitions of quantifacation and qualifacation before putting out a product that makes them look Jr. League.
Had to believe Y Comb would support this….um..questionable image based algorithim driven personality quantifacation system???…..
crap code in, crap code out…….
and here is the test to toss back at the investors to see if they have what it takes
http://electricpulp.com/vcat/
Sorry for the issues folks. We’ve addressed the “everyone’s like Nick” bug.
Please try again, it’ll be better.
It’s actually pretty cool!
Bryan, I thought everyone being like me was a feature?
Guys, the test is working properly now.
My new results: http://www.like...;user_list_id=6
I didn’t get pissed off until I was forced to choose between the Thundercats and The Sound of Music. WTF.
This is now about as accurate as randomly generated results would be.
I’ve taken tests from LikeBetter several times before, and this one works just as well: it’s noise on every attribute that I can authoritatively answer…
Why does anyone care about these guys?
Hah, this was complete garbage. I’m a serial entrepreneur who loves spicy food, is a complete night owl and has lived in major cities all his life. According to this test I am a highly risk averse morning person who hates spicy food and has a strong rural preference.
If any company anywhere starts using this to play a role in ANY recruitment decisions, they’re crazy. You might as well consult tea leaves.