Dual attitudes
From PsychWiki - A Collaborative Psychology Wiki
The concept of dual attitudes is based on the notion that within the mind there can be different evaluations of the same attitude object. The first of these evaluations is called an implicit attitude, and is an automatic response to the attitude object; it is nonconscious, and uncontrolled. The second evaluation is called an explicit attitude, and is a response that is both controlled and conscious (Baumeister and Bushman, 2008).
Most social psychologists believe that these evaluations can occur separately and independently of one another, and can serve different functions, though there are some that believe there is a connection between the two. Regardless of the relationship between the two attitudes, it is possible for individuals to have conflicting implicit and explicit attitudes. It is rare for individuals to be aware of the conflict; the implicit attitude often goes unnoticed, and individuals consequently may believe that the conscious, explicit attitude is the only one present (Baumeister and Bushman, 2008).
Contents |
Measuring Implicit Attitudes
While explicit, conscious attitudes can be fairly easily measured using standard Likert scales and related methodologies, it is much harder to measure nonconscious, implicit attitudes. Over the past few years, however, social psychologists have developed several methods to try to accomplish this goal. One way to measure implicit attitude is through response-latency measures, which involve the measurement of the time delay that occurs before a response. It is unclear whether this is an effective way to measure implicit responses, due to a lack of conclusive evidence as to whether there is a correlation between implicit attitudes and time (Cunnningham, Preacher, and. Banaji, 2001). Fazio, Sanbonmatsu, Powell and Karde (1986) came up with a new technique, naming it the Evaluative Priming Technique, which was a variation of the semantic priming technique (Cunningham et al., 2001). In 1998, Anthony Greenwald, Debbie McGhee, and Jordan Schwartz developed the Implicit Association Test (IAT), which compares the relative strengths of associations between the object eliciting the attitude, and the ideas of both “good” and “bad.” There are many different versions of the IAT, each testing different domains for implicit attitudes.
Experimental Findings
A significant amount of research has been done on implicit and explicit attitudes, although a clear conclusion has not yet been reached as to whether the two occur independently, or are somehow linked. Cunningham and his associates attempted to prove that the two attitudes occur independently, while Hofmann and his associates set up a study in which they tried to prove the opposite.
Cunningham et al., (2001).
This study used racial attitudes to explore different methods of measuring the beliefs that lie outside of conscious thought, and then compared them to explicit attitudes. Three different methods of measuring implicit attitudes were used: (1) the response-window procedure, which increases the degree of both subliminal and supraliminal priming effects; (2) the Implicit Association Test (IAT), and (3) an overlapping method that had a response-window included in the IAT. Explicit attitudes were measured by asking questions taken from the Modern Racism Scale; participants were asked to rate their level of agreement to each of the statements, a 1 being strong disagreement, a 6 strong agreement.
In each of the sessions various techniques were used, in accordance with the method being tested. The basic idea, however, was to test the accuracy and response time that subjects had when comparing words as “good” or “bad,” and faces as “white” or “black.” In half of the trials, “good” and “white” (and therefore “bad” and “black”) were marked on the same key, while in the other half of the trails, “bad” and white” (and “good” and “black”) were linked. For the response-window method, the independent variable was the race of the face shown, and the dependent variable was whether or not the participant could correctly identify the following as positive or negative. For the IAT method, the independent variable was the combination of face and word type per key, and the dependent variable was the response time of the participant. For the response-window IAT session, the independent variable was the same as in the IAT session, but the dependent variable became whether the participant could correctly identify the faces and words correctly.
The results of the IAT session showed that on average, participants took longer to respond to trials with a white face and negative word combination than to a white face and positive word combination. The response-window method showed that participants made more errors when white faces were matched with negative words than when white faces were matched with positive words. The experimenters determined that this meant that participants were more likely to pair white faces with positive words, and correspondingly, black faces with negative words. The response- window IAT session also showed that there were more errors when white faces were paired with negative words. All of these results showed that participants preferred white faces to black faces.
The results of the Modern Racism Scale showed that, on average, participants disagreed with prejudicial statements made against African-Americans, and therefore did not have prejudicial explicit attitudes. This directly goes against the results of the other three sessions of the experiment, which measured the implicit attitudes to be pro-white. This showed clear evidence for the separation of implicit and explicit attitudes, and how they can directly conflict with one another.
Hofmann, Gawronski, Gschwendner, Le, and Schmitt (2005)
This purpose of this study was to explore the correlation between measurements taken with the IAT and self-reports for explicit attitudes. This study revealed that there was a positive correlation of .24 between the measurements collected through the IAT and those compiled through the self-reports. It was concluded that 56% of the variability found between the IAT and the explicit self-reports could be accounted for by certain moderator variables, such as spontaneity, the level of similarity between the topics, and certain aspects of the IAT having to do with the methods used. The results showed that the more a subject though about a specific topic, the greater the decrease in correlation (between IAT and self-reports); on the other hand, when subjects relied on gut reactions, the correlation increased. The study produced no evidence to support the ideas that the correlation is influenced by introspection or social desirability, but that does not mean that these variables do not affect correlations on an individual level. This study concluded that the IAT and explicit self-tests are related in a systematic manner.
Critiques of the IAT
One common critique of the IAT is that it can be manipulated by unrelated factors, such as the media or other social influences (Baumeister and Bushman, 2008). This theory is brought into deeper detail by Fiedler, Klaus, Messner, Claude, Bluemke, and Matthias (2006), who provide a list of problems with the IAT. They argue that the IAT has five main flaws when it was used to make diagnostic inferences: 1. There is a difference between casual and diagnostic inferences, and it is difficult to tell which one of these the Implicit Association Test makes. 2. There is uncertainty as to the capability and strength of the more fundamental association model that it was based on, calling this the second flaw. 3. There is not a more basic testable model with which the inferences can be compared. 4. It is very challenging to interpret the scores produced by the IAT. 5. The test can easily be manipulated so that the processing can be faked or changed (Fiedler et al., 2006). The conclusion that is reached is that there are many existing factors independent of whatever attitude may be under assessment by the IAT, and these factors can significantly manipulate the scores and results of the test. This is not to say, however, that the test is unusable, or entirely inaccurate.
References
Baumeister, Roy F., and Brad J. Bushman. Social Psychology and Human Nature. United States: Thomson Wadsworth, 2008. 226-227.
Cunningham, William A., Kristopher J. Preacher, and Mahzarin R. Banaji. "Implicit Attitude Measures: Consistency, Stability, and Convergent Validity." Psychological Science 12 (2001): 163-170.
Fiedler, Klaus, Messner, Claude, Bluemke, and Matthias. "Unresolved Problems with the “I”, the “a”, and the “T”: a Logical and Psychometric Critique of the Implicit Association Test (IAT)." European Review of Social Psychology 17 (2006): 74-147.
Hofmann, Wilhelm, Bertram Gawronski, Tobias Gschwendner, Huy Le, and Schmitt Manfred. "A Meta-Analysis on the Correlation Between the Implicit Association Test and Explicit Self-Report Measures." Personality and Social Psychology Bulletin 31 (2005): 1369-1385.
Rudman, Laurie A. "Sources of Implicit Attitudes." Current Directions in Psychological Science 13 (2004): 79-82.
