Spontaneous Trait Inferences

Definition

The notion of spontaneous trait inferences (STIs) refers to a frequently demonstrated empirical finding. Observing behaviors or reading behavior descriptions gives rise to immediate trait inferences, beyond the actually given information. Thus, somebody who steps on a partner's feet on the dance floor elicits the inference clumsy. Witnessing a student succeeding on a difficult task gives rise to the spontaneous inference clever. Such inferences take place even though the trait is not strictly implicated. Stepping on someone's feet can happen to nonclumsy people, just as even a dull student can solve a task under auspicious conditions. Logically, singular behaviors do not imply general traits. STIs are called spontaneous because they can be assumed to occur in the absence of explicit task instructions and deliberate intentions to think about the traits that correspond to a given behavior. In STI experiments, researchers make serious attempts to conceal their interest in trait inferences, ruling out demand characteristics that might account for controlled trait inferences.

Measurements

Several paradigms have been developed to investigate STI effects experimentally. In the original cued-recall paradigm, participants are exposed to a list of behavior descriptions (e.g., “Steven stepped on his partner's feet on the dance floor”). Then, on a so-called cued-recall test, their task is to recall the previously presented behavior descriptions based on variable cue words or phrases. The specific types of cues given in the recall test are manipulated. As it turns out, trait words that represent reasonable inferences but that never appeared in the original sentences (such as clumsy) provide more effective retrieval cues than such words or phrases that actually occurred in the list, suggesting that traits must have been inferred spontaneously. In another word-fragment completion paradigm, being exposed to behavior descriptions facilitates the subsequent generation of a corresponding trait concept from an incomplete letter string. For instance, a person who has been primed with the earlier sentence takes less time to generate the word concept clumsy from the word fragment c–m–y than does somebody who was not exposed to that behavior. The faster response latency provides evidence that the trait concept has already been inferred implicitly. Unlike the cued-recall method, this method warrants trait inferences that occur immediately after the behavior has been presented, ruling out inferences during a later recall stage.

In a picture-priming paradigm, behaviors are presented in pictures or moving pictures (film clips), and participants have to identify a trait word that is first hidden behind a mask and that appears only gradually (over 3 seconds or so) as small pieces of the mask are removed in random order. Again, an STI effect is evident in response speed. When the trait to be identified constitutes a reasonable inference from the behavior presented in the preceding picture or film, the identification time is slower than on nonmatching trials. This method has been extended to control for the mental activity during behavior presentation, by inserting a verification task. Thus, participants have to verify an aspect of the picture or film (e.g., whether the presented behavior is an instance of hitting or attacking or an instance of hostile), before the trait [Page 934]identification task (e.g., involving the trait aggressive) starts. In this fashion, the trait inference process can be guided or tuned experimentally.

...