A cognitive mediation theory of task goals and human performance

A theory is proposed to explain the linkages between individual task goals and performance. Two cognitive constructs are postulated to mediate between task goals and performance:performance expectancy and performance valence. It is asserted that an individual's task goal has a positive influence on performance expectancy and a negative influence on performance valence. Performance expectancy is proposed to have a positive influence on performance while performance valence is proposed to have a negative influence on performance. Task ability is hypothesized to influence performance both directly and indirectly through its influence on performance expectancy. A laboratory experiment was designed to test the causal model proposed by the theory. A path analysis on the data from this experiment provides strong support for the model, with performance expectancy, performance valence, and task ability predicting 63% of the variance in performance on the laboratory task.     

Philosophy of medicine and other humanities: Toward a wholistic view

A less analytic and more wholistic approach to philosophy, described as ‘best overall fit’ or ‘seeing how things all hang together,’ is defended in recent works by John Rawls and Richard Rorty and can usefully be applied to problems in philosophy of medicine. Looking at sickness and its impact upon the person as a central problem for philosophy of medicine, this approach discourages a search for necessary and sufficient conditions for being sick, and instead encourages a listing of “true and interesting observations” about sickness which reflect the convergence of a number of different viewpoints. Among the relevant viewpoints are other humanities disciplines besides philosophy and the social sciences. Literature, in particular, provides insights into the meaning and the uniqueness of episodes of sickness in a way that philosophers may otherwise fail to grasp.     

Interviewer evaluations of interviewee nonverbal behavior

Personnel representatives (N = 52) were shown one of four videotaped job interviews in which the verbal content of the 16-min interview was identical, but the interviewee's nonverbal behavior was manipulated. A “low nonverbal” interviewee was defined by minimal eye contact, low energy level, lack of affect and voice modulation, and a lack of speech fluency. The “high nonverbal” interviewee demonstrated the opposite behavior on each of these components. The subjects were asked to rate the videotaped candidates on dimensions previously identified as critical in influencing a job interviewer's decisions. Nonverbal behavior was found to have a significant effect on almost every rating made by subjects in this study. After reviewing the entire 16-min interview, 23 of the 26 subjects who saw the “high nonverbal” candidate would have invited him/her for a second interview. All 26 of the subjects who saw the “low nonverbal” candidate would not have recommended a second interview.     

Mid-career change in academia: The decision to become an administrator

This study explored the reasons why professors would leave their teaching/ research occupations to begin new careers as administrators. Valence-Instrumentality-Expectancy theory of occupational preference and choice was used as a framework to examine mid-career change among professors and department chairmen at a major university. Data showed that professors' intentions to change careers and become administrators could be explained by the theory. Analyses of job characteristics demonstrated that professors who want a mid-career change are primarily attracted by the power and formal authority vested in the administrative position, while professors who prefer to remain in their career paths do so because of the autonomy it offers.     

Reinforcement and punishment in a closed system

Children chose between toys freely and with various restrictive contingencies. Sometimes play with a preferred toy was contingent on play with a nonpreferred toy. Sometimes play with a nonpreferred toy was contingent on play with a preferred toy. Another toy (free toy) was always available during contingency sessions. Because toy playing occupied the entire session the response system was closed. When one response increased in duration others had to decrease. An analysis of closed-system responding revealed that the restrictions of the closed system together with the restrictions imposed by the contingencies forced subjects to alter the proportion of time devoted to the three responses. A geometrical model representing the closed system and an analytical method based on distance are presented to describe the data.     

Tables of d′ for variable-standard discrimination paradigms

Signal detection theory (SDT) allows a bias-free measure of sensitivity, d′, to be simply estimated from discrimination data when certain conditions are met. The computation is not straightforward, however, in several popular discrimination designs, such as two- and four-interval same-different designs and the ABX design. The present tables, derived from the SDT models of Macmillan, Kaplan, and Creelman (1977), make possible the estimation of d’ from these complex discrimination designs.     

A deglitching circuit for the AA50 D/A converter

Switching transients or “glitches” are frequently encountered in signal processing applications involving digital-to-analog (D/A) conversion. A low-cost “deglitching” circuit is described and implemented for the Digital Equipment Corporation AA50 D/A converter option on the PDP-8/e minicomputer. Switching transients were eliminated using a noninverting, unity-gain sample-and-hold circuit (employing Datei System’s SHM-IC-1). Switching transients were reduced to below-noise levels across a wide range of output levels and sampling rates.     

Auditory streaming and vertical localization: Interdependence of “what” and “where” decisions in audition

When a sequence of two tones is presented over headphones in an ascending or descending order of pitch, it is heard as correspondingly ascending or descending in space. The illusion of spatial change that accompanies pitch change can be induced onto a pair of noise bursts by presenting them in synchrony with the tones. When cues known to produce stream segregation are introduced, the perceived position of the noises is less influenced by the tones. Stream organization is seen to be implicated in the ability to separately localize concurrent sources of sound. This suggests that “what” and “where” decisions are highly interactive and that neurological evidence that suggests separate pathways for these decisions must be interpreted with caution.     

On the nonautomaticity of “automatic” activation: Evidence of selective seeing

Early perceptual selection refers to a sequence of perceptual processing in which attentional selection of stimuli occurs before perceptual processing is complete. By contrast, late perceptual selection refers to a sequence in which all stimuli receive complete perceptual processing before selection takes place. An attempt was made to determine if early selection is possible. Three experiments were conducted in which subjects were presented with arrays containing a variable number of red items and a variable number of black items; all array items were randomly positioned around the circumference of an imaginary circle. The subjects task was to report the number of red items. Response time increased with the number of red items, but was essentially independent of the number of black items. The nature of the red and black items was also varied. When the to-be-attended red items were given “Stroop-like” quality (the symbols that were used were directly inconsistent with the correct response for that trial), mean reaction time was increased compared to a baseline condition; when the to-be-ignored black items were given “Stroop-like” quality, mean reaction time was unaffected. These results imply that the relevant (red) and irrelevant (black) items were processed differentially. We conclude that early selection is possible; late-selection models are not supported.