An eye fixation analysis of choice and judgment with multiattribute stimuli

The present paper examined the selection of processing heuristics for choice and judgment across a range of stimuli. Eye fixations were monitored while six subjects made choices or judgments of pairs of gambles, vacations, or gifts. Each stimulus was represented by three attributes that varied in the amount of dimensional interdependency. The two heuristics of interest were dimensional and holistic evaluation. Each manifested a characteristic pattern of eye fixations. Dimensional evaluation required alternating fixations from a single attribute of one stimulus to the same attribute of the other stimulus. Holistic evaluation was characterized by transitions from attribute to attribute within a single stimulus. The results demonstrated that when the stimulus attributes were either interdependent (gambles) or dissimilar (gifts), the processing heuristic was determined by stimulus characteristics. When the stimulus dimensions were neither interdependent nor dissimilar (vacations), the selection of a processing strategy was determined by the prescribed task. This study suggests that any global theory of choice or judgment must be validated over a wide range of stimuli.     

Self-consciousness, self-awareness, and self-attribution

Self-awareness was either manipulated by a mirror (experimental) or not (control). Subjects were selected for being high or low in private self-consciousness (disposition to attend to one's thoughts, feelings, motives). Private self-consciousness had a stronger effect on self-attributions than did self-awareness. These findings have implications for attribution, self-consciousness, and the relationship between manipulations and dispositions.     

Lump detection in simulated human breasts

Sixteen observers palpated silicone models of human breasts containing lumps 1.6-12.1 mm in diameter. Detectability depended on the size of the lump, producing a systematic psychometric function. In eight observers who participated in three or more sessions, performance improved with practice, with most improvement occurring within one or two 26-trial sessions. Three-week retention measures disclosed no appreciable decrease in performance, but a significant correlation was found between the number of lumps detected and duration of trial (p < .01). There was no difference in performance between four observers who used their preferred hands and four observers who used their nonpreferred hands. These data establish that examination of breast models for the detection of lumps simulating cancer is a task amenable to experimental analysis.     

Auditory evoked potentials during selective listening to dichotic speech messages

The sensitivity of the scalp-recorded, auditory evoked potential to selective attention was examined while subjects monitored one of two dichotically presented speech passages for content. Evoked potentials were elicited to irrelevant probe stimuli (vowel sounds) embedded in both the right- and the leftear’s message. The amplitude of the evoked potential was larger to probe stimuli embedded in the attended message than to probe stimuli in the unattended message. Recall performance was unaffected by the presence of the probes. The results are interpreted as supporting the hypothesis that this evoked potential sensitivity reflects an initial “input selection” stage of attention.     

Detection errors on the word the: evidence for reading units larger than letters.

In four experiments, subjects read 100-word passages and circled instances of the letter t. Subjects missed a disproportionate number of ts in the word the. Evidence was provided against explanations of this effect involving the location and pronunciation of the t in the and against an explanation in terms of the redundancy of the. Rather the high frequency of the appeared to be critical, and it was suggested that high-frequency words are read in terms of units larger than the letter.     

An examination of how one's attitudes toward women are reflected in one's defensiveness and self-esteem

The Berger Self-Esteem Measure, the Marlowe-Crowne Social Desirability Scale, and Spence's (1972) Attitudes Toward Women Scale (AWS) were used to test 96 undergraduates. Subjects were divided according to sex and median split scores on the personality measures. Analysis of variance revealed a significant main effect for sex and defensiveness and approached significance for the sex × self-esteem effect (p<.10). Males and females were found to respond significantly differently on the AWS scale. Highly defensive males were more discriminating toward women and self-concept appeared to be an influential variable. However, self-esteem alone did not seem to have any significant effects.