Cross-Cultural maps and Asch’s Conformity Experiment

So I’m go­ing through the se­quences (in AI to Zom­bies) and I get to the bit about Asch’s Con­for­mity Ex­per­i­ment.

It’s a good bit of writ­ing, but I mostly pass by with­out think­ing about it too much. I’ve been taught about the ex­per­i­ment be­fore, and while Eliezer’s point of whether or not the sub­jects were be­hav­ing ra­tio­nally is in­ter­est­ing, it kind of got swal­lowed up by his dis­cus­sion of lonely dis­sent, which I thought was more en­gag­ing.

Later, af­ter I’d passed the sec­tion on cult at­trac­tors and got into the sec­tion on let­ting go, a thought oc­curred to me, some­thing I’d never ac­tu­ally thought be­fore.

Eliezer notes:

Three-quar­ters of the sub­jects in Asch’s ex­per­i­ment gave a “con­form­ing” an­swer at least once. A third of the sub­jects con­formed more than half the time.

That an­swer is sur­pris­ing. It was sur­pris­ing to me the first time I learned about the ex­per­i­ment, and I think it’s sur­pris­ing to just about ev­ery­one the first time they hear it. Same thing with a lot of the psy­chol­ogy sur­round­ing heuris­tics and bi­ases, ac­tu­ally. For­get the In­qui­si­tion—no one saw the Stan­ford Pri­son Ex­per­i­ment com­ing.

Here’s the thought I had: Why was that re­sult so sur­pris­ing to me?

I’m not an ex­pert in his­tory, but I know plenty of re­li­gious peo­ple. I’ve learned about the USSR and China, about Nazi Ger­many and Jon­estown. I have plenty of available ev­i­dence of times where peo­ple went along with things they wouldn’t have on their own. And not all of them are nega­tive. I’ve gone to blood drives I prob­a­bly wouldn’t have if my friends weren’t go­ing as well.

When I thought about what my pre­dic­tion would be, had I been asked what per­centage of peo­ple I thought would dis­sent be­fore be­ing told, I think I would have guessed that more than 80% of sub­ject would con­sis­tently dis­sent. If not higher.

And yet that isn’t what the ex­per­i­ment shows, and it isn’t even what his­tory shows. For ev­ery dis­sen­ter in his­tory, there have to be at least a few thou­sand con­form­ers. At least. So why did I think dis­sent was the norm?

I no­tice that I am con­fused.

So I de­cide to think about it, and my brain im­me­di­ately spits out: you’re an Amer­i­can in an in­di­vi­d­u­al­is­tic cul­ture. Hy­poth­e­sis: you ex­pect peo­ple to con­form less be­cause of the cul­ture you live in/​were raised in. This begs the ques­tion: have their been cross-cul­tural stud­ies done on Asch’s Con­for­mity Ex­per­i­ment? Be­cause if peo­ple in China con­form more than peo­ple in Amer­ica, then how much peo­ple con­form prob­a­bly has some­thing to do with cul­ture.

A lit­tle googling brings up a 1996 pa­per that does a meta-anal­y­sis on stud­ies that re­peated Asch’s ex­per­i­ments, ei­ther with a differ­ent cul­ture, or at a later date in time. Their find­ings:

The re­sults of this re­view can be sum­ma­rized in three parts.

First, we in­ves­ti­gated the im­pact of a num­ber of po­ten­tial mod­er­a­tor vari­ables, fo­cus­ing just on those stud­ies con­ducted in the United States where we were able to in­ves­ti­gate their re­la­tion­ship with con­for­mity, free of any po­ten­tial in­ter­ac­tions with cul­tural vari­ables. Con­sis­tent with pre­vi­ous re­search, con­for­mity was sig­nifi­cantly higher, (a) the larger the size of the ma­jor­ity, (b) the greater the pro­por­tion of fe­male re­spon­dents, (c) when the ma­jor­ity did not con­sist of out-group mem­bers, and (d) the more am­bigu­ous the stim­u­lus. There was a non­signifi­cant ten­dency for con­for­mity to be higher, the more con­sis­tent the ma­jor­ity. There was also an un­ex­pected in­ter­ac­tion effect: Con­for­mity was higher in the Asch (1952b, 1956) paradigm (as was ex­pected), but only for stud­ies us­ing Asch’s (1956) stim­u­lus ma­te­ri­als; where other stim­u­lus ma­te­ri­als were used (but where the task was also judg­ing which of the three com­par­i­son lines was equal to a stan­dard), con­for­mity was higher in the Crutch­field (1955) paradigm. Fi­nally, al­though we had ex­pected con­for­mity to be lower when the par­ti­ci­pant’s re­sponse was not made available to the ma­jor­ity, this vari­able did not have a sig­nifi­cant effect.

The sec­ond area of in­ter­est was on changes in the level of con­for­mity over time. Again the main fo­cus was on the anal­y­sis just us­ing stud­ies con­ducted in the United States be­cause it is the chang­ing cul­tural cli­mate of Western so­cieties which has been thought by some to re­late to changes in con­for­mity. We found a nega­tive re­la­tion­ship. Levels of con­for­mity in gen­eral had steadily de­clined since Asch’s stud­ies in the early 1950s. We did not find any ev­i­dence for a curvil­in­ear trend (as, e.g., Larsen, 1982, had hy­poth­e­sized), and the di­rec­tion was op­po­site to that pre­dicted by Lamb and Alsi­faki (1980).

The third and ma­jor area of in­ter­est was in the im­pact of cul­tural val­ues on con­for­mity, and speci­fi­cally differ­ences in in­di­vi­d­u­al­ism-col­lec­tivism. Analy­ses us­ing mea­sures of cul­tural val­ues de­rived from Hofst­ede (1980, 1983), Schwartz (1994), and Trompe­naars (1993) re­vealed sig­nifi­cant re­la­tion­ships con­firm­ing the gen­eral hy­poth­e­sis that con­for­mity would be higher in col­lec­tivist cul­tures than in in­di­vi­d­u­al­ist cul­tures. That all three sets of mea­sures gave similar re­sults, de­spite the differ­ences in the sam­ples and in­stru­ments used, pro­vides strong sup­port for the hy­poth­e­sis. More­over, the im­pact of the cul­tural vari­ables was greater than any other, in­clud­ing those mod­er­a­tor vari­ables such as ma­jor­ity size typ­i­cally iden­ti­fied as be­ing im­por­tant fac­tors.

Cul­tural val­ues, it would seem, are sig­nifi­cant me­di­a­tors of re­sponse in group pres­sure ex­per­i­ments.

So, while the pa­per isn’t defini­tive, it (and the pa­pers it draws from) show rea­son­able ev­i­dence that there is a cul­tural im­pact on how much peo­ple con­form.

I thought about that for a lit­tle while, and then I re­al­ized that I hadn’t ac­tu­ally an­swered my own ques­tion.

My con­fu­sion stems from the dis­par­ity be­tween my pre­dic­tion and re­al­ity. I’m not won­der­ing about the effect cul­ture has on con­for­mity (the ter­ri­tory), I’m won­der­ing about the effect cul­ture has on my pre­dic­tion of con­for­mity (the map).

In other words, do peo­ple born and raised in a cul­ture with col­lec­tivist val­ues (China, for ex­am­ple) or who ac­tu­ally do con­form be­yond the norm (peo­ple who are in a fly­ing-saucer cult, or the peo­ple ac­tu­ally liv­ing in a com­pound) ex­pect peo­ple to con­form more than I did? Is their map any differ­ent from mine?

Think about it—with all the differ­ent cult at­trac­tors, it prob­a­bly never feels as though you are vastly con­form­ing, even if you are in a cult. The same can prob­a­bly be said for any col­lec­tivist so­ciety. Imag­ine grow­ing up in the USSR—would you pre­dict that peo­ple would con­form with any higher per­centage than some­one born in 21st cen­tury Amer­ica? If you were raised in an ex­tremely re­li­gious house­hold, would you pre­dict that peo­ple would con­form as much as they do? Less? More?

How many times have I agreed with a ma­jor­ity even when I knew they prob­a­bly weren’t right, and never thought of it as “con­for­mity”? It took a long time for my be­lief in god to fi­nally die, even when I could ad­mit that I just be­lieved that I be­lieved. And why did I keep be­liev­ing (or keep try­ing to/​say­ing that I be­lieved)?

Be­cause it’s re­ally hard to ac­tu­ally dis­sent. And I wasn’t even lonely.

So why was my map that wrong?

What back­ground pro­cess or mo­ti­vated rea­son­ing or...what­ever caused that dis­par­ity?

One thing that, I think, con­tributes, is that I was gen­er­al­iz­ing from fic­tional ev­i­dence. Bat­man comes far more read­ily to my mind than Jon­estown. For that mat­ter, Bat­man comes more read­ily to my mind than the mil­lions of not-Bat­mans in Gotham city. I was also prob­a­bly not be­ing moved by his­tory enough. For ev­ery Spar­ta­cus, there are at min­i­mum hun­dreds of not-Spar­tuses, no mat­ter what the not-Spar­ta­cuses say when asked.

But to pre­dict that three-quar­ters of sub­jects would con­form at least once seems to re­quire a level of pes­simism be­yond even that. After all, there were no se­cret po­lice in Asch’s ex­per­i­ment; no one had emp­tied their bank ac­counts be­cause they thought the world was end­ing.

Per­haps I’m mak­ing a mis­take by putting my­self into the place of the sub­ject of the ex­per­i­ment. I think I’d dis­sent, but I would pre­dict that most peo­ple think that, and most peo­ple con­formed at least once. I’m also a rea­son­ably well-ed­u­cated per­son, but that didn’t seem to help the col­lege stu­dents in the ex­per­i­ment.

Has any re­search been done on peo­ple’s pre­dic­tion of their own and other’s con­for­mity, par­tic­u­larly across cul­tures or in groups that are “known” for their con­for­mity (com­mu­nism, the very re­li­gious, etc.)? Do peo­ple who are gen­uine dis­sen­ters pre­dict that more peo­ple will dis­sent than peo­ple who gen­uinely con­form?

I don’t think this is a use­less ques­tion. If you’re start­ing a busi­ness that offers a new solu­tion to a prob­lem where solu­tions already ex­ist, are you over­es­ti­mat­ing how many peo­ple will dis­sent and buy your product?