The other day, Larry Bartels posted a Monkey Cage article entitled “Here’s how a cartoon smiley face punched a big hole in democratic theory,” with the subtitle, “Fleeting exposure to ‘irrelevant stimuli’ powerfully shapes our assessments of policy arguments.”

Some people asked what I thought. Mark Palko e-mailed:

I started to do a post entitled ‘Return of the Hidden Persuaders?’… but this is way out of my field so I decided to go slow. I will say that this makes me nervous. I have no trouble believing that irrelevant stimuli and unconscious thought processes influence decisions but getting this magnitude of response to a smiley face is hard to believe.

I am under the impression that, after James Vicary, researchers and advertisers chased the subliminal philosopher’s stone for decades and never had much luck. How could they have missed something this big and easy too spot?

Any thoughts?

I responded that I, too, was skeptical, even though when I clicked quickly on the post I saw no smoking gun that would reveal any obvious problem with the cited research. After getting Palko’s e-mail, I decided to look more carefully.

Here’s the background. Bartels’s post reports on two experiments by political scientist Cengiz Erisen. Bartels wrote:

In one experiment, 224 subjects read and responded to policy statements regarding illegal immigration and energy security. For example, they might be asked to respond to the statement: “The number of illegal immigrants coming to the United States will drastically increase in six years.” Independent coders rated each response as a positive or negative thought.

Not surprisingly, subjects’ responses to the policy prompts tended to reflect their preexisting views about each issue. For example, pro-immigration subjects offered 34% fewer negative thoughts and 38% more positive thoughts than those who opposed immigration. However, that difference was dwarfed by the difference between subjects with similar views exposed to different “irrelevant stimuli.” Erisen’s data indicate that subjects exposed to positive stimuli offered 42% fewer negative thoughts and 160% more positive thoughts about illegal immigration than those exposed to negative stimuli.

Bartels continues:

What were these powerful “irrelevant stimuli” that were outweighing the impact of subjects’ prior policy views? Before seeing each policy statement, each subject was subliminally exposed (for 39 milliseconds — well below the threshold of conscious awareness) to one of three images: a smiling cartoon face, a frowning cartoon face, or a neutral cartoon face. According to Lodge and Taber, “Fleeting images of cartoon smiley faces have a larger effect than prior immigration attitudes on the valence of thoughts in response to illegal immigration policy prompts.”

And here’s more:

In a second experiment, 125 subjects read and responded to six policy statements regarding illegal immigration (three pro and three anti, presented in random order). Again, the subliminal cartoon faces substantially altered their assessments of the policy statements — and the resulting negative and positive thoughts produced substantial changes in policy attitudes measured up to 45 minutes after the original presentation of pro- and anti- statements.

Okay, it’s me again. When I read all this, I was skeptical, partly because, like Palko, it’s hard for me to believe that subliminal messages could be so effective, given all the stories I’ve heard about subliminal advertising not really working (see, for example, here). On the other hand, I’ve never looked closely at the claims that subliminal ads don’t work, and maybe I’m biased because I’ve published a paper arguing that voters make decisions based on “enlightened preferences,” so maybe I resist claims that public opinion is so irrational as to be swayed by smiley faces.

Bartels does not link to the research paper in his post, but via some Googling and following of links, I found it. The paper — by Cengiz Erisen, Milton Lodge and Charles Taber — is called “Affective contagion in effortful political thinking,” and it appeared in the journal Political Psychology this year.  It appears to be based on a paper presented at the American Political Science Association meeting in 2008.

The paper has some path diagrams that I could do without, so I skipped straight to the experiments and the data. From experiment No. 1, it’s clear that when the students were exposed to positive priming, they expressed more positive thoughts:

Screen Shot 2014-09-06 at 6.09.33 PM

I’d prefer a graph, but . . . whatever.  It’s clear what’s going on here. Indeed, I almost never see data with such a clear pattern. The transition from positive to neutral to negative is monotonic in all eight cases. You just don’t get cleaner than that.

So on to experiment No. 2. In this case, they do the same smiley/neutral/sad-face priming, but this time they also prime their students with some different statements about illegal immigration. The outcome of interest was “evaluations of the six policy proposals for illegal immigration,” as measured by coding of positivity or negativity of expressed thoughts.

Unfortunately they don’t give the data or any clear summary of the data from experiment No. 2, so I can’t evaluate it. I respect Larry Bartels, and I see that he characterized the results as the “subliminal cartoon faces substantially altered their assessments of the policy statements — and the resulting negative and positive thoughts produced substantial changes in policy attitudes.” But based on the evidence given in the paper, I can’t evaluate this claim.  I’m not saying it’s wrong. I’m just saying that I can’t express judgment on it, given the information provided.

In short, this experiment seems to show that, in the words of Erisen et al, “negative primes promote negative thoughts and inhibit positive thoughts, while positive primes trigger positive and inhibit negative thoughts.” But I don’t see how they make the leap to their next statement, that these cartoon faces “significantly and consistently altered [students’] thoughts and considerations on a political issue.” I don’t see a change in the number of positive and negative expressions as equivalent to a change in political attitudes or considerations.

The effect of the subliminal smiley or frowny face on expressed thoughts — that already is interesting and surprising to me.  I’ll accept that this finding certainly jumps out of the statistics, even though, for the reasons Palko gives, I still wonder what’s going on. But I can’t see how this result leads to concerns about democracy expressed by Bartels. I don’t see any evidence that the subliminal images affected political attitudes. It might well be that I’m missing something here. I’d like to see the data.