Louis N. Gray

Measurement of Relative Variation: Sociological Examples

There exists a variety of situations in which the dispersion of data, rather than their mean or other central value, is of interest. The coefficient of variation, which is a measure of dispersion divided by the appropriate measure of central tendency, is preferable to a raw measure of dispersion for this purpose. However, these measures do not have a constant range, and the common practice of dividing the variation value obtained by 100 and expressing the result as a per cent is particularly inappropriate, as all have maxima larger than unity for n>2. The present paper provides standardization procedures for coefficients of variation. The resulting standardized coefficients (called S-measures) have a zero-to-one range, and the common practice of dividing the variation value obtained by 100 and deviation) for descriptive purposes is demonstrated. It is suggested that S-measures replace, or at least supplement, coefficients of variation and raw dispersion measures when the problem of interest is relative dispersion between groups.

The intergenerational transmission of marital conflict : Testing a process model

We describe and test a sequential process model derived from socialization theory and designed to expand on previous explanations of the transmission of marital conflict across generations. The model seeks to explicate the reported relationship between parental divorce and offsprings marital conflict by positing a five-phase process that combines in sequential order childhood experiences, self-images, trust, and magnitudes of disagreements. The first two factors are measured at the individual level; the last two at the couple level. The model is tested with data from a three-year panel study of newly married couples. We obtained good fits with Year 1 data. We then constructed a dynamic model to test data for all three years. The results show that the individual factors had significant direct effects only in the first year, whereas couple level factors appeared to mediate the relationship between parental divorce and couples marital conflict over all three years.

Modeling the Deterrent Effects of Punishment

Nonexperimental (homicide) and experimental data are used to study the effects of the certainty and severity of punishment. Many deterrence researchers, working without systematic theory of how punishments should operate to influence behavior, have opted for a simple additive model of certainty and severity. However, a satisfaction balance model-a social psychological theory representing a modification of the matching equation and positing interactive effects of certainty and severity-fits both sets of data well and substantially better than an additive model. The implications of the satisfaction balance model for a theory of deterrence are discussed

Goodman and Kruskal's Tau b: Multiple and Partial Analogs

Using the logic developed by Goodman and Kruskal, multiple and partial analogs for their measure Tau are suggested. These measures may be of particular utility to survey researchers in situations where the Goodman and Kruskal lambda measures are either inappropriate or inapplicable. This measure has been shown to have a clear proportional reduction in error interpretation. These analogs should be useful and necessary for detailed causal and/or multiple analysis with nominal-level data.

Crime and Deterrence: A Correlation Analysis

A correlation analysis of the severity and certainty of punish ment and offense rates for the major index crimes produces results consistent with the predictions of deterrence theory. Certainty of punishment proves to be the chief deterrent for most crimes. Homicide, however, is influenced by severity, pos sibly reflecting the differences between homicide and other of fenses. Little evidence of interaction is found between certainty and severity in effects on crime rate. A powerfunction proves to better describe the relationship between the punishment varia bles and crime rates than a rectilinear equation-a conclusion which, even apart from the date, appears more reasonable than the reverse.

Influence Attempts and Effective Power: A Re-Examination of an Unsubstantiated Hypothesis

This research examines a previously unsubstantiated hypothesis, proposed by French and Snyder: the amount of influence effected by a member of a group over other members increases with the amount of influence attempted. The research gives strong support to this hypothesis when influence is operationally defined in terms of control of overt behavior, and gives some information allowing conjecture concerning the basic form of the relationship. It is suggested that the original study failed to support the hypothesis due to the relative lack of clarity in the original statement and in the methods used to test the hypothesis.This research examines a previously unsubstantiated hypothesis, proposed by French and Snyder: the amount of influence effected by a member of a group over other members increases with the amount of influence attempted. The research gives strong support to this hypothesis when influence is operationally defined in terms of control of overt behavior, and gives some information allowing conjecture concerning the basic form of the relationship. It is suggested that the original study failed to support the hypothesis due to the relative lack of clarity in the original statement and in the methods used to test the hypothesis.

A theory of problem-solving behavior

In this paper we develop a formal, testable theory of problem-solving behavior with special relevance to individuals and small groups. The theory is consistent with principles drawn from operant behavior and social exchange theories but also incorporates elements of cognitive psychology. Problem solving is defined as a nonroutine activity oriented toward changing an undersirable state of affairs. The focus on change differentiates problem solving from coping, which is oriented toward relieving feelings of stress. A decision-making model is presented, which takes the problem-solving process through its latter stages

Rewards and punishments in complex human choices

Building on work on the matching law, this paper proposes a model of choice behavior that includes costs and benefits. We compare our model with alternative models in the behavioral literature, and use the results of two experiments with human subjects to assess the predictive accuracy of the various models.

On simple stochastic diffusion models

This paper expands the diffusion of information models developed by Funkhouser and McCombs (1972) to include situations involving simple interaction processes and more complicated situations involving both mass mediated messages and interactively mediated messages. This paper develops discrete time models of information diffusion as opposed to the continuous time models developed by Coleman (1964) and Bartholomew (1967) and others.

On the Social Law of Effect

Summary This paper suggests that exposure to task related stimuli prior to task interaction should result in greater discrimination in making task related decisions. This is explained in terms of a social version of the law of effect, and reflected in the value of the slope in a regression equation. Ss were college freshmen and sophomores enrolled in sociology classes. Ss formed a total of 40 dyads, controlled for sex. Twenty dyads (10 dyads of males and 10 of females) formed the experimental groups. The remaining 20 dyads (10 male and 10 female) formed the control group. Ss engaged in a simple task consisting of a word game which allowed the investigators to measure response and reinforcement rates. Results from the experiment were found to support the hypothesis at the .0001 level of significance. These results indicate that a social law of effect may be applied to relatively unstructured social situations, and that exposure to different types of environmental stimuli alters the slope of the regression ...