Keeping the importance on intelligence test in minds, the researcher decided to standardize an intelligence test to the students on middle level (i.e. Class VI and VII, age group 11+ and 12+).
The objectives of the study was to analyze each item through traditional method in order to determine the validity and reliability of the test and to fix the person-ability and item difficulty by using the latest technique of item analysis i.e. Rasch Model.
For the present study, the students of each sex of class VI and VII in the Government Elementary, Secondary and Higher Secondary Schools of Multan, Bahawalpur, D.G, Khan and Sargodha divisions of Punjab were selected as the population of the study for administering the test.
Stratified random sampling technique was applied by the researcher. 10% schools and 5% students from the Elementary, Secondary , and Higher Secondary schools of Multan, Bahawalpur, D.G Khan and Sargodha divisions were selected.
After pilot study ten test, prepared by researcher and two Australian standardized test, were administered on sample of 12120 boys and girls of Class VI and 9645 boys and girls of Class VII from 335 schools of Bahawalpur, D.G.Khan, Multan and Sargodha divisions.
Thurston‚€™s model of Multiple-factor Theory of intelligence test was used for this study. The researcher developed ten tests(i.e. five for class VI and five for class VII) in Urdu.
The research personally administered the test in different urban and rural schools.
Relevant information‚€™s were collected through answer sheets and results were tabulated with help of computer.
With the help of computer, total score in each test was recorded for item analysis. The data was analyzed to compute facility Index (F%), Discriminatory Index (D) and power of discrimination (Ō†).
Co-efficient of correlation was computed for internal reliability of the test.
The find the reliability of the test, the researcher used the Kudar and Richardson‚€™s formula KR#20 and KR#21 for class VI and VII.
For each class and for each test identical procedure was adopted. After organizing and editing the data matrix (identifying item score and person scores), final item calibration and person measurement estimates were obtained. The values of probability of success of different persons on different items were also calculated.
As a result of research study, it was observed that:
1All tests are reliable for Class VI and VII. 2One item was to be discarded. 3Five items were to be improved.