Deborah D. Godwin

Estimating Housing Quality for Poverty and Development Policy Analysis: CWIQ in Ghana

Inadequate data on housing characteristics and lack of consensusregarding appropriate measures of housing quality characteristicsretard developments to meet housing needs of people in lessdeveloped countries. A constructed housing quality index is developed and Cronbachs alpha and factor analysis are employedeo evaluate its measurement properties. With data from the CoreWelfare Indicators Questionnaire (CWIQ) survey of 1997, theconstructed housing quality index identifies key factors relatedto housing quality in Ghana. Access to quality of life amenities contributes significant information and policy value to traditionalphysical attributes in modeling a housing quality index.

The Economics of household consumption

Preface Concepts, Theories, and Empirical Measurement The Household as an Economic System Theoretical Approaches to the Study of Household Consumption Methods of Studying Household Consumption Patterns and Trends in Household Consumption Household Consumption of Commodities in the United States Consumption Patterns in the United States Consumption in Other Countries Household Consumption in a Social Context Household Consumption and Public Policy Consumption Patterns of the Future Bibliography Index

Causal Modeling in Family Research.

Structural equations modeling or causal modeling is a strategy through which researchers enhance their ability to draw causal inferences for their empirical research. This type of modeling in family studies is useful when a researcher has the goal of determining whether certain cause and effect relationships among variables of interest exist in the real world of families. Godwin reviews the application of causal modeling in family research and the requirements for effectively implementing a causal modeling strategy. She concludes that a fundamental point that must be understood by critics of causal modeling with ex post facto data is that family researchers will never be able to address adequately the important causal questions of this discipline if researchers do not depart from the ideal experimental research design. That researchers will never know for sure whether a particular causal model is true is no reason not to try to discover whether it is a plausible representation of what is happening in the real world of families with as much certainty as possible. It is clearly superior to alternative methods such a relying on folklore armchair commentary or empirical tests using models and methods not designed for that purpose. Correlation does not imply causation but neither does the F test used with data from an experimental design taken alone imply causation. It is the logic of the theory the research design with its inherent advantage of creating ceteris paribus via random assignment and laboratory control the soundness of the measurement sampling and data collection plus the researchers careful consideration of group differences that collectively help the researcher infer causation. Regardless of what type of design is used being able to infer cause-and-effect in relationships from empirical data requires knowing about all the threats to such valid causal inference and ruling them out as systematically as possible. Causal modeling strategies should be viewed as 1 set of tools for aiding in that task. Causal modeling in family research has room for improvement. Until researchers begin thinking more comprehensively about the requirements for causal modeling and including those complexities in designs and empirical tests tests of causal models will be less than valid and understanding of families incomplete.

Longitudinal Research Methods and Family Theories

An outstanding characteristic of family relationships is their astonishing potential duration over time. A marital or parent-child dyad may persist for more than 60 years before being disrupted by death, and sibling relations may be still more long-lived. How such relations change over time, under what conditions they alter in specific ways, and what early factors shape subsequent developments in relationships are enduring questions for family researchers from various disciplines.