Evaluation of validity and reliability of Pham and Taylor Academic Performance Questionnaire (1999)

Evaluation of the academic performance of students helps to increase the quality of teaching and learning. The information collected through performance assessment helps teachers to identify the problems of the students as well as the weak points of the programs and measure the degree of adaptation of the curriculum and educational approaches to the needs of students. These assessments also help students to know what mistakes they have made and how they can correct those mistakes. Questionnaire is one of the tools to collect information about students' academic performance. Questionnaire is one of the most widely used data collection tools in social science research. The main purpose of the questionnaire is to obtain relevant information in the most reliable way. Therefore, the accuracy and consistency of the questionnaire is an important aspect of the research method, which is known as validity and reliability. Validity and reliability increase transparency and reduce opportunities to introduce researcher bias into research. Reliability refers to stability. That is, if we use the same tool or test several times, we should get the same results. If the instrument is unreliable in this sense, then the data are considered irrelevant to the phenomenon or concept being measured. Validity means that a test or instrument measures exactly what the researcher intended it to measure.

This research is a psychometric type. The statistical population of this study was made up of all the male and female students of the second secondary school in Semnan city, numbering 3952 people. The sample size was determined with the help of Cochran's formula of 400 people. Then, using cluster random sampling method, sample people were selected and completed Pham and Taylor's academic performance questionnaire (EPT). This tool contains 48 items with a five-point Likert scale and it measures five dimensions of self-efficacy, emotional effects, planning, lack of outcome control, and motivation. In this scale, a score less than 120 means poor academic performance, higher than 175 means strong academic performance, and between 121-174 average academic performance. In order to analyze the data, the statistical methods of exploratory factor analysis, confirmatory factor analysis and internal consistency (Cronbach's alpha) were used.

Using the Kaiser-Meier-Oklin and Bartlett's sphericity tests, the adequacy of the sample and the factorial ability of the data were checked and the value of KMO statistic was equal to 0.710 and the chi-square value in Bartlett's test was 593 and in this way, the possibility of doing factor analysis on the variables was confirmed. The amount of variance extracted for all subscales was more than 0.5. Two factors with an eigenvalue greater than one were found and the Pebble diagram also confirmed the existence of two factors. The first factor could explain 51.508% and the second factor could explain 24.210% of the variance of all research variables. Self-efficacy, motivation and planning were the subgroup of the first final factor, and lack of control over the outcome and emotional effects were the subgroup of the second final factor. In justifying the first factor, it can be stated that familiarity with study skills and the use of learning strategies such as planning lead to academic success of students, and this positive experience affects self-efficacy beliefs and motivation to progress and regarding the justification of the second factor, it can be said that the training of emotion regulation skills changes the level of self-control of people. All the factor loadings after rotation (Varimax) were more than 0.7 and showed the appropriateness of the factor analysis and the extracted factors. The highest value of the factor load was related to planning (0.861) and the lowest value was related to emotional impact with a value of (0.837). The reliability of Pham and Taylor's academic performance scale was obtained through internal consistency using Cronbach's alpha of 0.904, which indicates the good reliability of this scale. Also, this result was obtained if questions 5, 7, 8, 23, 26, 33, 37 are removed, the reliability of the scale increases. Finally, Lisrel software and confirmatory factor analysis were used to check the relationships and fit of the extracted factors. The results indicated a good model fit (AGFI: 0.95, CFI: 0.96, IFI: 0.96, NFI: 0.94, GFI: 0.99). The RMSEA index was in the desired range (0.069) and the chi-square value to degrees of freedom was obtained less than 3 and 2.91.

In general, and based on the results, it can be said that the Iranian academic performance questionnaire of Pham and Taylor (1999) has good validity and reliability and can be used to achieve various goals. It is suggested to check the validity and reliability of this questionnaire for different educational levels and age groups with a larger number of samples or with other statistical methods.