follow CCP

Recent blog entries
« Okay, now *this* is the Liberal Republic of Science! | Main | New paper: Ideology, Motivated Reasoning, and Cognitive Reflection: An Experimental Study »
Friday
Nov302012

A surprising (to me) discovery: reflective Independents...

The analyses I did for my latest paper—on ideology, cognitive reflection, and motivated reasoning—really surprised me in one respect.

They didn’t surprise me altogether. Indeed, they corroborated the hypothesis (one that also was explored in the CCP study of science comprehension & climate change polarization) that people who are more disposed to use System 2 reasoning (conscious, analytical, reflective) are more likely to to selectively credit or discredit evidence in patterns that fit their ideological predispositions.  This is contrary to how most people thing heuristic-driven, System 1 reasoning contributes to public confusion and controversy on issues like climate change.

But what did surprise me was the finding that self-identified Independents are more reflective—more disposed to use System 2 rather than System 1 reasoning.  I assumed people who were in middle were just less reflective.  The difference isn’t huge (and actually, no one, of any particular political orientation or non-orientation demonstrates a high degree of reflection on Shane Frederick's gold-standard CRT test), but it’s there.

It also follows from the analyses that are in the paper that Independents display less motivated reasoning than partisans. Of course, that’s sort of a logical thing; if they don’t have a predisposition, they can’t be fitting their interpretation of evidence to it. But I think there’s more to it than that.

Why am I surprised? My experience in doing studies has caused me to form the impression that people who are “in the middle” on measures of cultural or ideological predispositions are sort of like statistical noise—random, unreliable--& not that important for figuring out what is going on, at least if the signal you get from people w/ a more choate sense of identity is a clear one.

Well, it looks anyway, like the Independents are not simply inert or confused. They are reflective people, engaging information of political significance in a non-ideological way.  That’s something to try to figure out, not dismiss.  What are they thinking? Who the hell are they?

At this point, I’m not suffering any great intellectual crisis.  I suspect if I thoughtfully engage the data a bit more, I’ll discover something that, without necessarily making this finding unimportant, reveals that it it poses no particular problem for the basic hypothesis behind the study (which is that individuals rationally engage information on societal risks in a manner that reflects their interest in forming and maintaining group connections).

But I’m curious. Also a bit excited and anxious; maybe I’m missing something really important.

What I’m going to do for now is think for a bit. Also read and re-read some other things (including John Bullock's great study on need for cognition and partisanship).  And try to form some interesting hypotheses about what the “Reflective Independent” datum might mean. Then I’ll see if there is a way to test those hypotheses, at least provisionally, with this data set.

Like I said, I don’t think I’ll find anything here that makes me think I have to adjust my thinking in a major way. But I want to approach this minor nuggest of surprise in a way that wouldn’t obscure the possibility that just beneath it is a deep deposit of information that would liberate me from the intellectually destitute state of unrecognized ignorance.

So to start this inquiry:

Do you have ideas about this little datum? What to make of it; how to explore its signficance?  

Fine to tell me, too, if you think this was “obvious” for reason x, y, & z; but do realize that you could have been assigned to one condition in a “many worlds” experimental design that includes another condition in which my doppelgänger has just blogged, “See! Independents are less reflective!,” and in which yours is typing up the response, “Of course, that was obvious!

PrintView Printer Friendly Version

EmailEmail Article to Friend

Reader Comments (8)

Are you possibly conflating partisanship with ideology? The former has a much more obvious connection to social or group loyalty than the latter, and you would lose this distinction if you lump them together. That is, an "ideologue", in the sense of someone who simply thinks more consistently about political matters ("System 2"), might well self-identify as an independent in terms of partisan politics.

December 1, 2012 | Unregistered CommenterMetamorf

I'm a "reflective independent." Also a big fan of Jon Haidt's work, which tries to factor notions of loyalty into hypotheses about how we REALLY think and reason. Somewhat similar to what you seem to be talking about, in other words.

I'm a longtime independent who is very interested in politics. So I've had regular occasion to talk to and disagree with both conservatives and liberals, and to be called one by the other. That's a continual learning experience.

There's really no reason, I think, for you to blow up your previous hypothesis. But you might need to think about how it's required to work in the real world. Here in America anyway, it's VERY easy to form a group identity and behave loyally by joining one of the two main schools of thought that correspond to the major parties, at least loosely. If you align with them, you quickly find many who agree with you, and a large number of fully formed moral narratives to soothe you. In other words, there are two big churches, so to speak, with ideologies that are fully mature and developed.

If you reject both of these as being no better than half true, and want something better, some sort of synthesis of thesis-antithesis between liberal and conservative, then there IS no well-funded, well-known church, and there is no bible and no gospel, no hymns, no soothing tales.

I suspect, as you suggest, that at least some of the folks like me would be very happy to become loyal and self-identify with a church, just like liberals and conservatives do. But as of now, we don't have one. We're largely crowded out. We have no apparatus, no party, no spokesmen in public, very few candidates to vote for.

I highly doubt that reflective independents have transcended human nature, somehow "curing" or evolving beyond a desire to be loyal to something greater than the single self. We'd wear shoes to protect our feet, just like everyone else, if we could only find a pair that fits, available and for sale. Hope that helps.

December 1, 2012 | Unregistered CommenterBrian Keegan

Is it possible that there are other distinct political ideologies besides the main two? My first thought was 'libertarians', but there are other options. They would not identify as either of the main parties, and would switch to whichever was currently the closer match to their views, but in fact they would have a separate, well-defined position and cultural in-group. What's more, the more reflective people who think carefully about their political position may be more likely than average to pick a minority party than those who go with the flow.

The independents could thus be a mixture of wishy-washy types who had no strong feelings or deep knowledge about politics and voted on impulse, and strong ideologues with very definite and carefully considered ideas, but who neither of the two main parties served well.

It could be tested by asking about their views on individual policies rather than party collectives, and asking them where they get the information that informs their political views from. Is there a community of the like-minded, distinct from the two main parties, that they are motivated to fit in to?

December 2, 2012 | Unregistered CommenterNiV

@Metamorph: Indeed, I'm equating. The paper forms a continuous measure of "right/left" disposition by aggregating subjects' responses to a partisan self-identification (strength of attachment to one party or other) and a liberal-conservative ideology measure. They form a nice scale--psychometrically speaking -- which fits the supposition that they are measuring the same thing. I am also assuming that people who share value orientations tend to belong to groups that hold those values and that tend to exert an important influence on them; the groups themsleves might not be all that political but there's something about them that makes them cohere w/ those political outlooks & affiliations. So "ideology" and "party" are both being used as indicators to measure something elese-- a latent or unobserved group attachment. What's more there are bound to be other indirect measures out there that will work too. Indeed, I don't myself think party affiliation & "liberal-conservative" ideology are all that good in comparison to cultural worldviews. I can tell you, however, that in this paper I'd get same result if I ran analysis using only party or only liberal-conservative by itself-- the precision of the estimates would merely be smaller (because the aggregated scale is a better measure of the latent disposition). But I think I'm likely not even answering your question ... I agree with you there is a difference between the latent variable of group connection that I'm intersted in and logic/coherence of ideological thought. I'm mainly interested in former. Indeed, I don't think measure of ideology I'm using is likely to be strongly correlated w/ coherence of ldeological thinking *excdept* for people who are high in "political knowledge..." Boy I'm still probably not answreing! But what do you think?

@Brian: That's super interesting. My thought, actually (as said above), is that the affinity groups in question probably aren't super political; they are more likely to be things like churches, workplaces, localities, etc,. But (1) there will be a correlatoin w/ politics and these various gorups and (2) political positions will become -- now & again, anyway -- powerful markers of affiliation w/ the groups. Your point, then, makes me wonder whether we ought to be trying to find groups that function in relatoin to members' risk perceptions in way that the groups featured in our theory but that unlike the gorups in our theory don't share values or share values of the sort measured by our cultural worldview scales (see above; I prefer the worldview scales in part b/c I think they correlate better w/ nonpolitical groups that happen to share broadly political values than do party id or lib-con ideology).

@Niv: Yes, that seems very powerful. If (a) independents are as likely as partisans to show credit or discredit CRT conditional on relationship between experimental assignment and independents' views on climate change & (b) that effect is magnified by CRT, that would consistent with your hypothesis -- that they belong to other groups that aren't correlated w/ party id but that still make position on climate change importnat for id. Would also be consistent, though, I suppose w/ more generic self-serving form of bias (it would be unself-flattering to learn you hold view of people who tend to be unreflective). Either way, the Independents would come off looking not much better than partisans in putting presuppostiions aside when looking at political evidence..,. thanks!

December 2, 2012 | Registered CommenterDan Kahan

Do Independents show a more idiosyncratic series of responses to the CCR group-grid scale? Do their answers fail to “match up” with a particular cultural viewpoint to a greater degree than those who identify as Democrat or Republican? Or do their answers appear specifically contradictory? {For instance, do they say that feminists have gone too far and society is sexist? Or that too many people want their needs met and society should meet their basic needs?} Do they earn less income than one would expect from their education levels? That might suggest a lack of fit with the dominant cultures in the US. Is there gender asymmetry in who is willing to describe themselves as Independent? Like NiV, I wonder, could it be Libertarians masquerading as Independents? As thee notes in thy recent working paper, Iyer et al found Libertarians to have slightly higher CRT scores.

Perhaps people ready, willing and able to identify themselves as Independents include some number of people who have rejected (or been rejected by) the dominant cultural viewpoints, and are thereby environmentally forced to be more reflective? Muslims come to mind as a group, for instance, who might have cultural commitments that might not match up well with the current cultural viewpoint/political alignments. {What if no one seems to speak from my cultural perspective? What if it seems to me that all four cultural emperors have no clothes? That might encourage more reflective thinking in general.}

Perhaps they are evidence for the mythical fifth dimension of Cultural Theory: the Hermits? People who say a pox on all four of thy cultural houses?

Or perhaps they have had cross-cultural-viewpoint experiences that cause them to be “bicultural” in ways that mirror actual bicultural perspective differences like those described by Heine et al. in “What’s Wrong With Cross-Cultural Comparisons of Subjective Likert Scales?: The Reference-Group Effect” Living as a closet conservatives in academia Or perhaps Lesbian Republicans?

Swimming against the cultural tides, or swimming in multiple cultural tides, or simply refusing to swim in any particular political-cultural tide pool perhaps strengthens some of the mental muscles measured by CRT.

So, like NiV, I wonder if perhaps these Independents have a less inchoate sense of identity than thee supposed—just one formed outside of (or in opposition to) the dominant cultural viewpoints? {I was tempted to say, “a more choate sense of identity,” but I can’t quite bear to bring myself to use the word “choate,” per Scalia diatribes described here}

Again, like NiV, I wonder if rather than “engaging information of political significance in a non-ideological way,” perhaps Independents are engaging information of political significance in an ideological way that differs from the ways statistical Democrats and Republicans engage that information. One problem, it seems to me, is that if there are a number of ways to “be” a Democrat or Republican, which it seems to me there are, even though they are relatively organized groups, how many ways would there be to choose to identify as an Independent? While there seem to be identifiable individualist versus communitarian ways up the Democrat and Republican mountains, there is no guarantee any two random Independents are even attempting the same mountain. One can only know what mountains they aren’t attempting.

December 3, 2012 | Unregistered CommenterIsabel Penraeth

@Isabel: You & NiV definitely are thinking in reinforcing patterns & helping me to see things that I find clarifying -- or at least as potentially clarifying subject to what some more analyses reveal.

I think there are 2 sets of questions you are raising:

1. Is it possible partisan id is obscuring group-affinity based form of motivated reasoning that use of the culture scales would pick up. For sure that's the first thing to try! Afterall, the reason for using cultural worldviews rather than party id & libcon ideology measures is that the latter are a lot less discerning than the former. Indeed, I already know that "the typical" hierarch individualist & typical "egalitarian communitarian" *are* independents who "lean" Republican & Democrat, respectively. So I'll construct an appropriate model using the cultural worldview predictors, which I imagine will reproduce the same effects in the experiment as did right-left orientation -- except *bigger*, precisely b/c cultural worldivew is a more discerning measure. Then I'll see if that *culturally* motivated reasoning effect disappears for Independents. I predict it won't!

2. Second set of questions offer conjectures on what it might be that makes Independents different. I think the value of exploring those questions is conditional on what exploring first set reveals. That is, if it turns out Independents display culturally motivated reasoning coparable to people who identify with either of the major political parties, the motivation to see *why* they are different essentially disappears -- or waits for some other source of *how* they are different to re-motivate it.

So I'm going to do the first sort of analysis -- in a new episode of WSMD? JA! dedicated to you & NiV.

December 3, 2012 | Registered CommenterDan Kahan

Thanks for the response, Dan, and sorry I missed it until now. My point was simply that if you're puzzled that independents don't seem any less reflective in their political reasoning than overt partisans, one reason may have to do with the lumping together of the tribal loyalties of partisans, the group adhesion associated with simple politcal labels (liberal or conservative), and the reflective politics of those more concerned with intellectual cohesion -- i.e., the real ideologists -- who may well be more likely to reject identification with political parties.

I see now though that you've broadened your test for "culturally motivated reasoning", and have found what you were looking for -- namely that "independents" display such reasoning as well. Mystery solved, but possibly at the cost of loss of significance. That is, of course everyone, even those who've tried to reflect on, or think through, their politics in some sort of coherent way, will be biased by just those very efforts -- even scientists, even you (even me) -- and those efforts commonly align with other, similar efforts to make up so-called worldviews. Not to be so biased would almost be to abandon coherent thought altogether, unless of course you think that things divide neatly into syllogisms and facts on the one hand, and emotional wishes and wants on the other. Outside of that positivist neatness, what seems like fact or logic within the context of one worldview can often break down in the context of another. And those who think they are without a wordview would do well to reconsider the famous quote of Keynes: "Practical men, who believe themselves to be quite exempt from any intellectual influence, are usually the slaves of some defunct economist." And it's not just in economics.

December 8, 2012 | Unregistered CommenterMetamorf

@Metamorf: I don't things are so bleak as you say. My "Is cultural cognition a bias" parts one and two explain why. Also why I think, though, that making things better *does* require not misleading ourselves about why and how things are bad.... Gotta say, makes me feel sort of disappointed in myself to think I give people the impression that I would take any pleasure in discovering confirmation of distortions in people's thinking; I did give you cause to believe that certainly...

December 8, 2012 | Unregistered Commenterdmk38

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
Post:
 
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>