International Journal of Applied Science and Technology

ISSN 2221-0997 (Print), 2221-1004 (Online) 10.30845/ijast

APPLICATION OF RESPONSE SURFACE METHODOLOGY FOR CAPTURING OPTIMUM RESPONSE IN A LONGITUDINAL SURVEY
O.M. Olayiwola, G.N. Amahia, A.A. Adewara, A. U. Chukwu

 

Abstract
Non-response rates in surveys have been recognized as important indicators of data quality since they introduce bias in the estimates which increases the mean square error. This study was designed to apply Response Surface methodology in a Longitudinal survey to reduce non response and capture optimum response. Seven hundred and fifty (750) households in Oyo town were randomly . House-heads were interviewed in five waves. An interviewer-administered questionnaire was used to collect data on demographic characteristics and response predictors. Demographic characteristics were analyzed using summary statistics. Multi-way contingency tables were constructed to establish relationships and dependence structures among the variables under investigation. A log-linear model was fitted to constructed contingency tables to capture significant predictors of response. Using demographic characteristics, a Response Surface Model (RSM) was constructed and subjected to canonical analysis for the characterization of its turning point and to capture the combination of levels of response predictors that produced optimum response. Log-linear model showed that family size (x1), duration of interview (x2), and their interaction (x1 , x2) significantly (p < 0.05) determined response rate. The RSM has an adjusted R2 = 0.722. Canonical analysis of the RSM gave eigen-values -0.007 and -0.002. The turning point of the RSM was a maximum implying the point for optimum response. The response was optimum when the family size was three and duration of interview was twelve minutes.

Full Text: PDF