Lesswrong 2016 Survey

It’s time for a new sur­vey!

Take the sur­vey now

The de­tails of the last sur­vey can be found here. And the re­sults can be found here.

I posted a few weeks back ask­ing for sug­ges­tions for ques­tions to in­clude on the sur­vey. As much as we’d like to in­clude more of them, we all know what hap­pens when we have too many ques­tions. The fol­low­ing graph is from the last sur­vey.



(Source: JD’s anal­y­sis of 2014 sur­vey data)

Two fac­tors seem to pre­dict if a ques­tion will get an an­swer:

  1. The position

  2. Whether peo­ple want to an­swer it. (Ob­vi­ously)

Peo­ple an­swer fewer ques­tions as we ap­proach the end. They also skip tricky ques­tions. The least an­swered ques­tion on the last sur­vey was—“what is your favourite lw post, provide a link”. Which I as­sume was mostly skipped for the amount of effort re­quired ei­ther in gen­er­at­ing a favourite or in find­ing a link to it. The sec­ond most skipped ques­tions were the digit-ra­tio ques­tions which re­quire more work, (get out a ruler and mea­sure) com­pared to the oth­ers. This is un­sur­pris­ing.

This year’s sur­vey is al­most the same size as the last one (though just a wee bit smaller). Pre­limi­nary es­ti­mates sug­gest you should put aside 25 min­utes to take the sur­vey, how­ever you can pause at any time and come back to the sur­vey when you have more time. If you’re in­ter­ested in helping pro­cess the sur­vey data please speak up ei­ther in a com­ment or a PM.

We’re fo­cus­ing this year par­tic­u­larly on get­ting a glimpse of the size and shape of the LessWrong di­as­pora. With that in mind; if pos­si­ble—please make sure that your friends (who might be less con­nected but still hang around in as­so­ci­ated cir­cles) get a chance to see that the sur­vey ex­ists; and if you’re up to it—en­courage them to fill out a copy of the sur­vey.

The sur­vey is hosted and man­aged by the team at FortFore­cast, you’ll be hear­ing more from them soon. The sur­vey can be ac­cessed through http://​​less­wrong.com/​​2016sur­vey.

Sur­vey re­sponses are anony­mous in that you’re not asked for your name. At the end we plan to do an opt-in pub­lic dump of the data. Be­fore pub­li­ca­tion the row or­der will be scram­bled, dat­es­tamps, IP ad­dresses and any other non-sur­vey ques­tion in­for­ma­tion will be stripped, and cer­tain ques­tions which are marked pri­vate such as the (op­tional) sign up for our mailing list will not be in­cluded. It helps the most if you say yes but we can un­der­stand if you don’t.

Thanks to Names­pace (JD) and the FortFore­cast team, the Slack, the #less­wrong IRC on freen­ode, and ev­ery­one else who offered help in putting the sur­vey to­gether, spe­cial thanks to Scott Alexan­der whose 2014 sur­vey was the foun­da­tion for this one.

When an­swer­ing the sur­vey, I ask you be helpful with the for­mat of your an­swers if you want them to be use­ful. For ex­am­ple if a ques­tion asks for an num­ber, please re­ply with “4” not “four”. Go­ing by the last sur­vey we may very well get thou­sands of re­sponses and clean­ing them all by hand will cost a for­tune on me­chan­i­cal turk. (And that’s for the ones we can put on me­chan­i­cal turk!) Thanks for your con­sid­er­a­tion.

The sur­vey will be open un­til the 1st of may 2016

Ad­den­dum from JD at FortFore­cast: Dur­ing user test­ing we’ve en­coun­tered re­ports of an er­ror some users get when they try to take the sur­vey which er­ro­neously re­ports that our database is down. We think we’ve fi­nally stamped it out but this par­tic­u­lar bug has proven re­silient. If you get this er­ror and still want to take the sur­vey here are the steps to miti­gate it:

  1. Refresh the sur­vey, it will still be bro­ken. You should see a screen with ques­tion ti­tles but no ques­tions.

  2. Press the “Exit and clear sur­vey” but­ton, this will re­set your sur­vey re­sponses and al­low you to try again fresh.

  3. Rinse and re­peat un­til you man­age to suc­cess­fully an­swer the first two ques­tions and move on. It usu­ally doesn’t take more than one or two tries. We haven’t re­ceived re­ports of the bug oc­cur­ring past this stage.

If you en­counter this please mail jd@fort­fore­cast.com with de­tails. Screen­shots would be ap­pre­ci­ated but if you don’t have the time just copy and paste the er­ror mes­sage you get into the email.

Take the sur­vey now

Meta—this took 2 hours to write and was re­viewed by the slack.

My Table of con­tents can be found here.