Home > Development and validation of intelligence test for Pakistani adult population
Development and validation of intelligence test for Pakistani adult population
Thesis Info
Author
Nighat Gul
Supervisor
Sonia Shagufta
Department
Psychology
Program
PhD
Institute
Shaheed Benazir Bhutto women university Peshawar Khyber Pakhtunkhwa Pakistan
Institute Type
Public
City
Peshawar
Province
Khyber Pakhtunkhwa Pakistan
Country
Pakistan
Degree Starting Year
2019
Degree End Year
2024
Viva Year
2024
Thesis Completing Year
2024
Thesis Completion Status
Completed
Page
417
Subject
Psychology/ Psychometric Psychology
Language
English
Other
This PhD research aimed to develop a reliable and validated intelligence test for the Pakistani population. A new intelligence test, the Pakistan Multiple Intelligence Scale, was constructed and administered to a large, provincially representative sample (N = 4650). The results showed high test-retest reliability (0.90), alpha reliability (0.80-0.97), and convergent validity. The test demonstrated a significant relationship with academic performance, indicating high concurrent validity. Norms were established for linguistic, logical, mathematical, and spatial intelligence across different age groups. The study concluded that the newly developed intelligence test is a reliable and valid measure of intellectual abilities for the Pakistani population, with implications for clinical, educational, organizational, and counseling psychology.
Intelligence is an intellectual ability of an individual to perform well in a particular area, judge well, and solve problems in the best way possible. In Pakistan, most of the intelligence tests used 0f foreign countries which were not translated and adopted in our host culture. There was no any specific multiple intelligence tests for native home land. Thus, for doctoral research, the PhD scholar thought to construct a reliable and validated test for the Pakistan population. For new intelligence test development, researcher prepared four pools of items consisting on varied number of items. Through the committee approach, judgmental and face validity was assessed with the help of subject experts, psychology teachers, psychometric psychologists, and bilingual experts. They rated each item individually for preparing the best items for each pool of items. The present study consisted of different phases. After the generation of intelligence test items, the first phase started. For initial study (n = 1000), people were recruited to attain the particular research purpose. During the pilot study, item analysis technique item difficulty, item discrimination was utilized for rejection and acceptance of the suited items for advance study. Those items difficulty level was 0.5 and less than 0.80 were selected for further testing. The acceptable item discrimination level range was 0.4 –0.66. There (N = 1000) participants were recruited randomly separately for test-retest reliability.. The Pakistan Multiple Intelligence Scale for Pakistani Adults was administered twice with an appropriate time interval to the same participants. The results revealed very high test-retest reliability of 0.90 for the full intelligence scale, and the subscales of the intelligence scale showed test-retest reliability within the range of 0.84 to 0.90.
In the final tryout, a large provincially representative sample (N = 4650) was recruited from all over Pakistan by using multistage sampling technique. A newly constructed intelligence test was used for data collection across the age groups of target population. The total number of participants was recruited from Khyber Pukhtoon Khawa (n = 1150), Punjab (n = 1150), Sindh (n = 850), Baluchistan (n = 350), Islamabad (n= 900) and Kashmir (n = 250). Statistical calculation was used to determine the number of the participants across the age groups and regions. After data collection, an editing, management, and organization process were adopted to clean the data. Statistical analyses included descriptive statistics, items discrimination, items difficult level method, inter items correlation within the items of all subtypes of intelligence and full scale of intelligence, alpha reliability, split half reliability, test retest reliability, total item correlation, percentiles, convergent reliability, concurrent and construct reliability. To construct new reliable and valid intelligence test, items difficulty level and items discrimination across the age groups were determined. Furthermore, norms were established for the new developed intelligence test according to the different age groups.
Items analysis was made in second tryout through items difficulty, items discrimination, and items correlations. Research findings revealed that item difficulty was ranged between 0.60 and 0.79, whiles items discrimination limit was between 0.4 and 0.55 for the new developed intelligence test items. Alpha reliability, split half reliability and test retest reliability were found very higher for intelligence test and sub scales ranged between 0.80 and 0.97. Results revealed that inter items correlation within sub sets and sub scales of intelligence test was significantly higher (p < .001). Positive significant correlation (p < 0.001) within items provide evidences for the convergent validity of the currently developed intelligence test at p <0.001. In the present research, a significant direct relationship of intelligence and examinees academic performance is presented, indicating high concurrent validity for the newly developed intelligence test. Norms were developed in standard scores, likewise percentiles, for linguistic intelligence, logical intelligence, mathematical intelligence, and spatial intelligence across the age groups and shown in norms development section. Statistical evidences supported that the newly constructed standard intelligence test was a reliable and valid test for measuring the intellectual abilities of the sample behavior. This test was equally important and applicable for clinical psychology, educational psychology, organizational psychology, and counseling psychology, etc.