Ethical issues in a study of Internet use: uncertainty, responsibility, and the spirit of research relationships

In this article we explore ethical issues arising in a study of home Internet use by low-income families. We consider questions of our responsibility as educational researchers and discuss the ethical implications of some unanticipated consequences of our study. We illustrate ways in which the principles of research ethics for use of human subjects can be ambiguous and possibly inadequate for anticipating potential harm in educational research. In this exploratory research of personal communication technologies, participants experienced changes that were personal and relational. These unanticipated changes in their way of being complicated our research relationships, testing the boundaries of our committment to the principle of trustworthiness and forcing us to reevaluate our responsibilities.     

Influence of subject and interviewer characteristics on the reliability of young adults' self-reports of drinking

The influence of subject and interviewer characteristics on the reliability of self-reports of alcohol consumption among young adults was investigated. Data were gathered from black and white college students of both sexes (N's=24) using a time-line procedure. The results snowed that these young adults provided highly reliable self-reports regarding their use of alcohol. Test-retest correlations for a criterion interval of 90 days were .96, .93, and .97 for the numbers of abstinent, moderate-drinking, and heavy-drinking days, respectively. Analyses also showed that white female subjects generally provided more reliable reports of abstinent and heavy-drinking days and that white female interviewers gathered more reliable reports across the three drinking disposition categories. These findings suggest that nonalcoholic young adults' retrospective reports of their drinking behavior can be reliably assessed using the time-line methodology. Future research is required to determine the validity of these self-reports and to understand the differential influence of subject and interviewer characteristics on the levels of reliability found.     

Behavioral contrast in short term memory: Serial binary memory models or parallel continuous memory models?

This paper develops a model wherein STM primacy as well as recency effects can occur. The STM primacy effects can be used to generate correct immediate recall of short lists that have not been coded in LTM. The properties of the model are interpreted in terms of explicit neural mechanisms. The STM primacy effect is a behavioral contrast effect that is analogous to the behavioral contrast that can occur during discrimination learning. The adaptational mechanism that accounts for these effects is also implicated in data on reaction time, retinal adaptation, ratio scales in choice behavior, and von Restorff-type effects. Its ubiquitous appearance is due to the fact that it solves a universal problem concerning the parallel processing of patterned data by noisy cells with finitely many excitable sites. It is argued that the STM primacy effect is not measured in interference experiments because it is masked by competitive STM interactions. These competitive interactions do not prevent the LTM primacy effect from influencing performance. The paper criticizes recent models of STM that use computer analogies to justify binary codes, serial STM buffers, and serial scanning procedures. Several deficiencies of serial models in dealing with psychological and neural processing are overcome by a model in which continuous STM activities and parallel real-time operations play an important role.     

An eight-channel computer-controlled scanning electromyograph

A measurement system is described for the computerized detection, digitizing, and storage of electromyographic (EMG) signals from up to eight muscle groups. The use of a channel scanning technique allows almost simultaneous multichannel recording with minimal circuit redundancy. Specifically, the scanning electromyograph allows several EMG preamplifiers to share expensive isolation, signal-conditioning, and interface circuitry. The design incorporates monolithic integrated circuit preamplifiers and includes optical isolation of the preamplifier circuitry for protection of the subject. Hybrid active filter modules allow convenient selection of system bandpass characteristics. Key measurement parameters are software programmable. The scanning electromyograph is inexpensively constructed from readily available componentry. Specifications for all functional blocks within the instrument are comparable to those of commercial instrumentation.