Prediction models for peak expiratory flow rates in North Indian male population based on ordinary and weighted least square estimation

摘要

The present study was carried out to establish physiological norms (best regression equation) to predict peak expiratory flow values among the North Indian male population with a statistically appropriate model. The study aims to establish the best statistically sound multiple regression model for predicting peak expiratory flow rate in the North Indian healthy population considering age, height and weight as predictor variables. One hundred and thirty-seven normal male subjects aged between 20 and 69 years, who had come from different parts of Lucknow to attend a science exhibition, India were selected for the study. The ordinary least square multiple regression model was used for the study. Residuals in this model were heteroskedastic. Therefore, the model proposed by Prasad et al and the weighted least square models were also used for the study. All the models were compared statistically. The proposed weighted least square model was found to be best fitting model, which had minimum residual standard deviation, maximum explained variation and most precise regression coefficient estimates.