Article #20
Newsletter list 


Previous Article 

Next Article 


The Role of Cue Intercorrelations in Students' Judgments of Course Interest

James A. Athanasou & Olu Aiyewalehimi
Sydney, Australia

A person's interest in a subject acts as a key factor in many models of learning. Typically educational achievement and satisfaction are inferred from ratings of student interest, but these ratings involve perceptions and subtle decision-making about oneself. People employ such judgmental processes regularly in education and, in the case of interest, they may be based on the subject, the learning situation or many other factors. The purpose of this program of research is to study how people make judgments about how interested they are in a subject.

In the first phase, a case study of student judgment was undertaken that involved two undergraduate students from the University of Technology Sydney. Their judgments of interests were based on information that described six aspects of a subject rated from very low (0) to very high (9). The six aspects (cues) were whether the subject was challenging, fascinating, the quality of teaching, the usefulness of the text, the quality of the facilities, and the extent to which theory and practice were related. Participants were presented with 60 separate profiles comprising random values of the six cues. They studied the six cues in each profile and then judged their own personal level of interest on a scale from 0 (no interest) to 9 (very high interest) based on the information contained in each profile. Profiles (N=15) were repeated to determine consistency of judgment of interests. Results were analyzed using a single lens model that describes human judgment of interest. Results showed individual differences and complexity in judgment with an overall emphasis on the quality of teaching.

The second phase of the research program has investigated the role of cue intercorrelations in the judgment of interest. Students were presented with 75 profiles containing the same six cues, but in this case the intercorrelation between cues was varied. Four separate groups of adult students (N=28) were tested using cues that correlated 0.0, 0.3, 0.6 and 0.9. Students were presented with 75 profiles including 15 randomly selected and repeated profiles to determine test-retest reliability of judgments.

While previous research has investigated the effects of multicollinearity in multiple cue probability learning within a double lens model, this has been in contexts involving fewer than six cues and never in single lens models. However, much human judgment occurs under conditions that are consistent with the single lens model, that is, where the criterion is difficult to measure or unavailable.

Preliminary results indicate marked differences between groups in the multiple correlation of cues with the judgment and the cue utilities. The effect of increasing cue intercorrelations was to increase multiple correlations and cue utilities. Multicollinearity of cues in a single lens model affected judgments in a predictable and monotonically uniform manner. A summary of the results is available upon request.

Further studies that are planned for Phases 2, 3 and 4 of the project include focus on varying the content of the cues (cue labels).

Contact James A. Athanasou

Previous Article                Next Article

Newsletter list