Hierarchical Multiple Regression Modelling on Predictors of Behavior and Sexual Practices at Takoradi Polytechnic, Ghana


  •  Anthony Turkson    
  •  James Otchey    

Abstract

INTRODUCTION: Various psychosocial studies on health related lifestyles lay emphasis on the fact that the perception one has of himself as being at risk of HIV/AIDS infection was a necessary condition for preventive behaviors to be adopted. Hierarchical Multiple Regression models was used to examine the relationship between eight independent variables and one dependent variable to isolate predictors which have significant influence on behavior and sexual practices.

METHODS: A Cross-sectional design was used for the study. Structured close-ended interviewer-administered questionnaire was used to collect primary data. Multistage stratified technique was used to sample views from 380 students from Takoradi Polytechnic, Ghana. A Hierarchical multiple regression model was used to ascertain the significance of certain predictors of sexual behavior and practices.

RESULTS: The variables that were extracted from the multiple regression were; for the constant; Beta=14.202, t=2.279, p=0.023, variable is significant; for the marital status; Beta=0.092, t=1.996, p<0.05, variable is significant; for the knowledge on AIDs; Beta=0.090, t=1.996, p<0.05, variable is significant; for the attitude towards HIV/AIDs; =0.486, t=10.575, p<0.001, variable is highly significant. Thus, the best fitting model for predicting behavior and sexual practices was a linear combination of the constant, one’s marital status, knowledge on HIV/AIDs and Attitude towards HIV/AIDs.,

Y(Behavior and sexual practies)= Beta0+Beta1(Marital status)+Beta2(Knowledge on HIV/AIDs issues)+Beta3(Attitude towards HIV/AIDs issues)

Beta0, Beta1, Beta2 and Beta3 are respectively 14.201, 2.038, 0.148 and 0.486; the higher the better.

CONCLUSIONS: Attitude and behavior change education on HIV/AIDs should be intensified in the institution so that students could adopt better lifestyles.



This work is licensed under a Creative Commons Attribution 4.0 License.