Sociological Methods & Research

The distinctions between nominal, ordinal, interval, and ratio measurement were popularized by S. S. Stevens. Unfortunately, positions taken by Stevens have often been disseminated without criticism. One problem is the common assumption that "ordinal" statistics are the best statistics to use for presumed noninterval continuous social variables, when, in fact, they use addition, subtraction, and division, which make the measurements interval by definition. Additionally, the relationship of the normal distribution to interval measurement is commonly misunderstood, the latter existing by definition if a normal distribution exists. Concern with levels of measurements may mislead persons into attending to issues other than maximizing (utility, given) the particular limits of the state of the measurement art in the social sciences.

This study examines the accuracy of a survey question about the number of times a subject has been arrested. Specifically, the study answers questions dealing with extent of response error, random versus systematic response errors, the reasons for response errors (lapses in memory and motivational factors that lead to under- reporting or overreporting), and the extent to which the survey response is an adequate substitute for actual data. The results demonstrate that there is a considerable amount of response error in the measure and that both memory lapses and motivational factors contribute to it. Some strategies for dealing with the problem are explored.

An important class of applications of measurement error or constrained factor analytic models consists of comparing models for several populations. In such cases it is appropriate to make explicit statistical tests of model similarity across groups and to constrain some model parameters to be equal across groups using a priori substantive information. This article discusses a statistical model devel oped by Jöreskog for these purposes. The model is applied to children's and parents' reports of parental socioeconomic statuses for several grade levels.

This article investigates sex differences in the accuracy of young adults' retrospective reports of parental status using Joreskog's general framework for the simultaneous covariance structure analysis of multiple populations. Results indicate that young women's reports of maternal education are significantly more reliable than are young men's reports and that the reliabilities of reports ofpaternal traits are similar for young men and women.

This article deals with the problem of analyzing sets of ipsative variables using the common factor model. We demonstrate that the usual assumptions of the common factor model, especially the assumption of uncorrelated disturbances, are not appropriate for sets of ipsative variables. We develop a common factor model that takes into account the ipsative properties of such data and show how this model can be applied to any set of ipsative measures using the methods of confirmatory factor analysis. We then suggest that the application of this model may be useful in modeling the latent content of sets ofrankings and other measures that have the ipsative property as a result of the measurement procedure. Finally, we apply the model to Kohn's measures of parental values, using sample data from the General Social Surveys.

Recent social research shows a renewed interest in detailed characteristics of jobs. The Dictionary of Occupational Titles (DOT) and census classification systems as bases for measures of job features are reviewed. The article reports on a set of characteristics that have been estimated for detailed 1960 and 1970 census categories, including those generated by Temme and new indicators for routiniza tion and closeness of supervision in work. The report includes information on how the measures were generated, evidence on their validity as assessed against corresponding individual-level measures, and an assessment of the measurement slippage involved in moving between the different occupation classification systems.