Author/Authors :
RAHMAN, ATTA UR Institute of Management Sciences, Pakistan , HAYAT, YOUSAF Agricultural University - Department of Mathematics, Statistics and Computer Science, Pakistan , HAYAT, SHAUKAT Agricultural University - Institute of Business and Management Sciences, Pakistan , ZAMAN, QAMRUZ University of Peshawar - Department of Statistics, Pakistan , IQBAL, MUHAMMAD University of Peshawar - Department of Statistics, Pakistan , RAHMAN, ABDUR Agricultural University - Department of Livestock Management, Indonesia
Abstract :
The present study was conducted to investigate the socio-economic factors affecting education and to evaluate their significance in the presence and absence of outliers. A sample of 840 respondents was taken from seven randomly selected union councils of district Peshawar, using a multistage sampling technique in February, 2010. A logistic regression technique was applied for data analysis and observed that only two main effect factors i.e. father income (FI) and father education (FE) and two factors interaction (FE×ME), (FE×PE) and (HHS×PE) significantly (P 0.05) affect the education. It was also found that the presence of outliers in the data can seriously affect the model parameters and their removal can increase the model adequacy and forecasting ability.
Keywords :
Socio , economic factors , education , logistic regression , outlier analysis
