The Effect of the Omission of Intercept term in a Linear Regression Model
The effect of the omission of intercept term in a linear regression model is considered for a simple linear regression with pooled explanatory variable, simple linear regression with single explanatory variable (sugar level) and multiple linear regression with two explanatory variables (age and sugar level). The study considered estimation of the intercept and slope for a simple linear model with intercept term and the estimation of slope for a model without intercept. SPSS software 23 version was applied in the analysis. Student t-test was conducted to ascertain the best model. Coefficient of determination and correlation were also applied. The model with intercept was observed to be significant in modelling the Blood pressure and the explanatory variables for the 100 persons. Age proved to be contributing more to the Blood pressure of human beings as compared to the sugar level. As the intercept is omitted from a model, the coefficient of determination, correlation and the sum of square error increases. The plots in Appendix G, H and I show that there exist unstable relationships between the response variable (blood pressures) and the explanatory variables (heights weights, ages, sex, sugar levels etc) for the 100 persons studied.