Multivariate Effects of Demand Characteristics on the Analogue Assessment of Heterosocial Competence

Abstract

The present study assessed the effects of a contextual demand manipulation (pretreatment assessment versus experiment) and an instructional demand manipulation (high demand for competent behavior versus neutral demand) on the analogue assessment of heterosocial competence. Also assessed was the interaction between each demand manipulation and subject characteristics (high-frequency dating/low heterosocially anxious versus low-frequency dating/high heterosocially anxious college males). Students were assigned randomly to one of four groups receiving differential information and instructions. Students completed a battery of five standard paper-and-pencil questionnaires, responded out loud to 10 taped social situations, and conversed with a female confederate while being videotaped for five minutes. Multivariate analyses revealed highly significant differences between subject groups and between contextual manipulation groups. Univariate analyses revealed that only 3 of the 16 dependent measures, including 2 of the self-reports, were significantly affected by the contextual demand. Results are discussed in light of previous studies of phobic behavior and social skills. Various avenues for future research were presented.

The Nature of Behavioral Assessment: A Commentary

“…Because of the possible reactivity of obtrusive assessment techniques, Kazdin (1979) advocates the use of unobtrusive measures. Martinez-Diaz and Edelstein (1979) found that instructions to act in a competent fashion versus neutral instructions did not affect heterosocial behavior, but portraying the measurement situation as pretreatment assessment versus an experiment did affect at least a few of the dependent variables. This points to the care with which behavior therapists must construct their assessment methods.…”

“…The assessment technique selected must suit the response system that is being measured and reveal its important dimensions. Hence, overt motor behavior is most frequently measured by direct observation by trained observers (e.g., Kent et al, 1979), by mediators, or by self-recorders (e.g., Williams, 1979) in either naturalistic (e.g., Greenwood, Greenwood et al, 1979) or analogue situations (e.g., Martinez-Diaz and Edelstein, 1979). Physiological-emotional responses are usually measured in analogue situations, due to instrumentation requirements.…”

“…For example, Henson et al (1979) used an erotic film while measuring women's sexual arousal, and Martinez-Diaz and Edelstein (1979) used a contrived conversation while measuring heart rate. Cognitive-verbal responses are measured via rating scales (e.g., Williams, 1979), checklists (e.g., Prinz et al, 1979), questionnaires (e.g., Martinez-Diaz and Edelstein, 1979), and academic tests (e.g., Filipczak et al, 1979).…”

“…Other techniques are being borrowed from traditional assessment, such as peer ratings and nominations (Foster and Ritchey, 1979;Greenwood et al, 1979). Complicated target behaviors are being assessed, such as social competence and interpersonal interaction (Foster and Ritchey, 1979;Greenwood et al, 1979;Martinez-Diaz and Edelstein, 1979;Williams, 1979).…”

“…This trend will create pressure for greater reliance on computers and complex statistical decision rules. In this issue, multivariate statistics were used by Martinez-Diaz and Edelstein (1979) to examine the effects of their independent variables on a large number of dependent variables, and by Prinz et al (1979) to determine which of their dependent variables best discriminated groups of distressed and nondistressed mother-adolescent dyads.…”