"Low-effort" thought? Not so fast...

This was a response to a left-leaning friend’s comment about his exchange with a conservative former friend, wherein he referred to the recently published study “Low-Effort Thought Promotes Political Conservatism”

I think there is a substantial paucity of critical thinking across the spectrum, and is not confined to the extremes. On the contrary, I suspect (this is my personal observation, not a scientific study, and I have no studies to support this) that lack of critical thought is the rule in general, rather than the exception. That being said, I read the study referred to by Andrew entitled “Low-Effort Thought Promotes Political Conservatism” (http://psp.sagepub.com/content/early/2012/03/16/0146167212439213.full.pdf+html ) and it is an interesting study that examines the assessment of various political ideas by controlling for thought effort in several different ways (alcholol consumption, multi-tasking, limiting time available to draw conclusions).

What it does not do is to examine the thought-effort level of those who generally hold conservative views with those who hold more liberal views. That is, it makes no comparative statement at all about those who might hold conservative or liberal views in general, or the effort they expend forming their political conclusions. In fact, in study 2 they compared groups under load with groups under no load, but present no baseline of political attitudes for ther two groups.The “change” observed under load in this study may be more repesentative of the baseline attitudes in these groups, where N was small (total N = 38) and therefore it is easy for them to be heterogeneous. The same could be said for studies 3 and 4, which had similar design and number of participants.

Critically reviewing a study such as this requires one to consider several potential biases in the data. As usual, the paper does not provide enough raw information for an outside observer to assess this, such as the actual questions used on their surveys (were they unfairly skewed or worded to promote the authors’ hypothesis?), the baseline views of the participants, the script followed by those running the sessions and how closely it was followed, the potential for approval bias (the pariticpants strive to give answers the interviewer is looking for), etc.

Probably the most glaring potential for bias in this study is selection bias – how were participants chosen to be included in the stdy, and what may be the potential impact on the results. This study, with the exception of New-England local bar patrons, studied about 100 undergraduate students enrolled in introductory psychology courses. I can think of any number of reasons why this group may not be representative of the general population with respect to how they process political ideas, not the least of which is age (what was the range and distribution of ages? not reported). Gender may also play a role in that in studies 2-4 (excluding the alcohol arm) 69% of participants are female. Accordig to a 2009 Gallup sudy (http://www.gallup.com/poll/120857/conservatives-single-largest-ideological-group.aspx) 62% of women self-describe as liberal or moderate (rather than conservative) versus 55% of men. Similarly, 70% those in the 18-29 year age range self-describe as liberal or moderate versus 57% in the 30-49 year age range. So on it’s face, this study seems to be evaluating the effects of low-effort thought on the conclusions drawn by a group more likely to be composed of those self-described as liberal. Combine this with a recent UCLA study (http://heri.ucla.edu/pr-display.php?prQry=88) that demonstrated that among students, even those who self-describe as conservative have increasingly left-leaning attitudes with regards to same-sex marriage, affirmative action, and access to higher education by undocumented students, and we have a study with a potentially very limited applicability.

None of these thoughts will keep the punditry, the media, and those who are close-minded from using the headline to denigrate and attempt to deligitimize those with a different point of view, rather than researching the issue and forming a salient argument.

Finally, Andrew asked his former friend to illustrate his point with at least 50/120,000 incidents – or 0.04% – which is indeed very small. Given that about 80% of the US population is 15 years old or greater (CIA World Factbook), and the population of the US is 313,472,000, there are 2.5 Million adults in the US, 0.04% of which is 100311 – much greater than the 108 undergarduate psychology course students included in the study. While the study does show a likelihood that the results did not occur by chance, I think there is ample reason to believe that it may not be representative of the population in general, and as I noted at the outset, draws no conclusion as to the relative thought-effort of those who lean left or right by any measure.