A Good Tool?

Tuesday, November 02, 2010

Dear Uncle Gus,

Are you familiar with the [concept] of fundamental attribution error in social psychology? I'd like to know if you think this could ever be [a] valid a tool in analyzing people's behavior.

Skeptic Individualist

Dear Skeptic Individualist,

I don't think you meant to ask me whether an undetected error could be useful in any kind of analysis. Rather, I think you are asking whether the idea of fundamental attribution error can be useful in the analysis of other people's behavior.

My answer, as someone who is not a social psychologist and had to dig around a little bit to learn just what the fundamental attribution error is supposed to be, is a cautious and heavily-qualified, "maybe."

The Wikipedia link above takes us to an article that defines the error as, "the tendency to over-value dispositional or personality-based explanations for the observed behaviors of others while under-valuing situational explanations for those behaviors." [links dropped] (I would have to add "philosophical" and "sense-of-life-based" to the laundry list of non-situational "explanations" in that definition to regard it as useful.) The article then goes on to describe what immediately struck me as a highly flawed demonstration of the error.

Based on an earlier theory developed by Edward E. Jones and Keith Davis, Jones and Harris hypothesized that people would attribute apparently freely-chosen behaviors to disposition, and apparently chance-directed behaviors to situation. The hypothesis was confounded by the fundamental attribution error.

Subjects read pro- and anti-Fidel Castro essays. Subjects were asked to rate the pro-Castro attitudes of the writers. When the subjects believed that the writers freely chose the positions they took (for or against Castro), they naturally rated the people who spoke in favor of Castro as having a more positive attitude towards Castro. However, contradicting Jones and Harris' initial hypothesis, when the subjects were told that the writer's positions were determined by a coin toss, they still rated writers who spoke in favor of Castro as having, on average, a more positive attitude towards Castro than those who spoke against him. In other words, the subjects were unable to see the influence of the situational constraints placed upon the writers; they could not refrain from attributing sincere belief to the writers. [link dropped]
Suppose half of the pro- (and anti-) Castro speakers actually believed their positions and half did not. Even if a subject knew the side a speaker took was randomly determined, he would still also know that some speakers really believed what they were saying and some did not. It would not be unreasonable for him to to conclude on such a basis that a more convincing or passionate speaker got to speak his own mind.

In other words, even in an experimental situation, if someone is asked to guess whether someone is speaking his own mind, he has more context to consider than just the fact that a coin toss selected the speaker's position. Indeed, the very idea of attempting to guess the speaker's motivation for expressing an opinion about Castro seems to me to practically call for such speculation and, far from showing that the subjects lacked good judgment, indicated that they had a better implicit grasp than the scientists did of the role of context in an individual's actions. (I have not seen the original study, but I wonder whether asking, "Do you know this person's position with certainty?" would have resulted in more or most participants saying, "I really don't know.")

I took a look at a paper (PDF) that reviewed fundamental attribution error (aka correspondence bias) and found that I was right. (Of course, this is framed as the subjects "failing to realize how motivating a debate coach could be.")
Jones and Harris's (1967) subjects, for example, had failed to realize how motivating a debate coach could be when he ordered a debater to defend an unpopular position [e.g., deliver a pro-Castro speech --ed]. Similarly, subjects in Festinger and Carlsmith's (1959) classic dissonance study had failed to realize how much pressure an experimenter could exert by politely asking them to tell a little white lie. Bierbrauer's (1979) subjects failed to realize how intimidating Milgram's experimenter could be when he donned a white lab coat and commanded one person to electrocute another. And so on. In each of these cases, subjects had mistaken a strong situation for a relatively weak one. They had mistaken highly constrained actors for lightly constrained actors and, as such, made the kinds of inferences about the former that one usually reserves for the latter.
Other, more subtle errors, theoretical and experimental, still need to be resolved, too.

That said, note before I go on that I am speaking about a field I am unfamiliar with, but which I suspect (1) may be rife with determinism; (2) is in need of a better accounting of how mental activity and survival are related (as with related fields); and (3) is in need of a better accounting for the role of philosophic ideas (and related psychological factors) as drivers of human action. That said, I think that the basic idea is a good one. I take that basic idea to be: "When attempting to understand why someone else acts as he does, make sure you have a full grasp of that person's situational context (as well as of his character)." That is, it is an attempt to better understand how one must gauge the effects of a person's situation when attempting to judge his motives for an action.

In the broad sense, the general idea of taking someone's full context into consideration is useful right now. And while I think the work in social psychology can offer some useful insights into the problem, I am not convinced that the term "fundamental attribution error" is necessarily useful outside that field. I don't think the term is that common and, more important, we can use ordinary terminology to caution against drawing mistaken conclusions about the character or motivations of another due to failing to adequately account for the situational context of an action of his.

-- CAV

If you'd like to ask a question, just type it into the box at the upper right labeled, "Ask Uncle Gus."

Updates

Today: Cut out two sentences on whether fundamental attribution error is a legitimate concept.

2 comments:

mmigawa said...

I have understood the fundamental attribution error in a bit of a different sense. For example, let suppose you and someone else gets a DUI. You would have a tendency to explain this in terms of your situation (I just happen to get distracted and have an extra glass of wine, etc), while the tendency would be to explain the other’s DUI in terms of their disposition (they are careless, reckless or an alcoholic.) For a positive occurrence, let’s say winning a race, one would use the opposite tact. You would now be the fast, awesome athlete (disposition) after your win, while the other person, should they win, would just be lucky, or have a better trainer, or gotten better night’s sleep.
In my experience this often holds; one attributes negatives to oneself situationally, and to others dispositionally. For positives it is the opposite. I find it is useful to keep this tendency in mind when forming judgements.

Gus Van Horn said...

The contrasting judgments of self and others you describe can involve the fundamental attribution error, the kind of error described in the Fable of the Alms Bag, or both.