Model building in family planning research

Jimlar Mutane. 1976 Sep;1(2):101-16.

Abstract

PIP: In reviewing the effort of research workers in model testing with special reference to determinants of contraceptive usage, this paper draws on 2 fertility case studies in urban Ghana and central India. A brief description of path analysis technique, a special form of regression analysis, is given (weights are assigned to each variable for optimal prediction). In applying this technique to the case studies, the author notes some similarities in the adoption of family planning. This conclusion applies to education and income as a factor in the Indian case, and education and household status factors in the case of Ghana. In the case of India the effect is direct; for Ghana it is considered indirect. The advantages of path analysis technique include: 1) its flexibility in analyzing data in the most frequently arising circumstances, namely where an interval scale or dichotomous (dependent) variable is to be understood; 2) its ability to bridge the gap between sociological theory and statistical analysis. However, drawbacks in using regression models in testing social theories and models must be considered. Multiple regression assumes a 1 way scheme of cause and effect; a linear relationship between variables is also assumed but not always substantiated by reality e.g., family size may not change with socioeconomic status; types of data suitable for use is usually limited to ratio or interval data but social scientists frequently analyze nominal data such as religion, ethnic group or residence. In both the Ghanaian and Indian study, the data were not originally collected with the intention of performing a path analysis. However with both studies an initial attempt to test models in fertility/family planning behavior was made that is useful to the understanding of how the hypothesized relationships between variables in these models stand up to sophisticated statistical procedures. Tables accompanying the text include Mishler and Westoff's conceptual model for fertility research; Hill, Back and Stycos' model; the path analytic model, and its application to contraceptive use in urban Ghana and central India; and variables used in these case studies.

MeSH terms

  • Demography
  • Fertility
  • Ghana
  • India
  • Models, Theoretical*
  • Population
  • Population Dynamics
  • Regression Analysis*
  • Research
  • Sexual Behavior*
  • Statistics as Topic*