Testing Reliability, Validity of questionnaire and identifying the factors affecting academic performance of high school students

Objective: In present study we want to Check the reliability and validity of the questionnaire and identify the student and school related factors that affect academic performances of secondary school students. Methods: Stratified random sampling technique and primary data are collected from the 1000 class X students. Cronbach’s Alpha coefficient check the internal consistency; convergent and discriminant validity of the questionnaire is checked by using multi-trait multi item correlation matrix. After excluding the invalid items regression analysis is used to identify the effect of different factors on academic performance and for this we use SPSS software. Findings: The Cronbach’s Alpha give a good result for all the items included in student scales and school factors having value greater than 0.70. The items whose own scale correlation coefficient is greater than 0.40 follow convergent validity. The items Q.6, Q.17, Q. 31, Q.32, and Q.40 refuse to follow convergent validity with own scale correlation coefficient <0.40. Again the items Q.17 , Q.23 , Q.53, Q.54and Q.56 did not satisfy the discriminant validity as the own scale correlation coefficient is not significantly higher than competing scales. Excluding irrelevant items in the questionnaire and performing multiple regression (R^2=812) it was observed that the private and government school, school in urban area, teachers good personality traits, teaching and school , students time spend on study , presence of father at home, classroom Study and related factors, Parent education and facilities available at home have significant positive impact on academic performance of students. On the other hand, school logistics, strictness and drawback, students bad habits- problems faced in study have significant negative impact on academic performances of students. Novelty : Use of proper sampling technique, personal collection of data, checked the reliability-validity of the questionnaire. Exclusion of invalid items in regression analysis make the study strong.