Measurement Issues in the Assessment of the Experiences of Parenthood

Abstract

The widespread and dramatic changes in structural characteristics of contemporary American families have encouraged social scientists to rethink the family as a social form, debate characteristics and qualities of functional families, and work to dispel myths and monolithic view of contemporary life (e.g., Bernardes, 1993; Cheal, 1993; David, 1993). Yet, in spite of all attention given to complexity and diversity found within families, relatively little research exists that focuses on contemporary parents and their experiences of parenthood as unit of analysis (Goetting, 1986; Rossi, 1968). What little research there is has been criticized for employing measures that lack an adequate conceptual and/or psychometric foundation (Goetting, 1986). The notion that measurement is crucial to social sciences is a commonplace and unexceptional observation. While importance of good measurement is openly acknowledged by research community (Grotevant & Carlson, 1989), there is often a lack of systematic attention to practice of good measurement (Carmines & Zeller, 1979). Within parenthood research, for example, complex concepts, such as parent satisfaction, parental role strain, and parental competence, often are assessed without any effort to define and specify their meaning. Furthermore, these concepts typically are assessed using a limited number of operational indicators where no defensible argument is presented (or can be made]) for measure's content validity. These limitations make it all too clear that widespread acceptance of importance of good measurement is not always honored in practice of research. This article is based on premise that, in order for social scientists to address absence of information about contemporary parents, they must first direct attention to assessment of key parenthood constructs. The article presents a general discussion of conceptual and operational issues relevant to assessing parenthood constructs. This general discussion is followed by a review and critique of measures designed to assess concepts central to research on contemporary parents. Recommendations for ways to improve quality of measures designed to assess these concepts also are discussed. MEASUREMENT ISSUES IN THE RESEARCH ON PARENTHOOD Most of measures used in extant parenthood research lack conceptual precision and are characterized by a lack of operational comprehensiveness and logical consistency. These measurement issues are discussed in a general way as factors that undermine quality of information available on experiences of contemporary parents. Theory and Conceptual Precision The concepts of interest to parenthood researchers are abstractions; they do not exist in fact -- meaning that they are neither directly or indirectly observable (Kaplan, 1964). Insight into a concept is derived from measurement procedures used for linking these abstract concepts to empirical indicators (Carmines & Zeller, 1979). In order to link abstract concepts to empirical indicators, researchers must evolve an explicit and organized plan for classifying (and often quantifying) data that will be presented as representative of concepts (Riley, 1963). This explicit and organized plan requires that meaning of concepts be specified. The organized plan for assessing abstract concepts begins, thus, with a process that links theory and measurement -- namely, conceptualization process (Blalock, 1982). The failure to bridge gap between theory and measurement is seen as one of measurement error (Blalock, 1982). Measurement error exists in any and all efforts to assess constructs that are not and cannot be directly observable. The goal in construction of a measure is to reduce risk of measurement error. Theory is related to reduction of measurement error by clarifying parameters and dimensions of a concept. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 75 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.