Résumé:
The study aims to calibrate the items of Multiple Intelligences test, by using Rasch
model, and identifying the psychometric characteristics ( validity reliability), as well as
finding the table of percentile ranks , T- scors, deviation QI, for the males and females.
To achieve the objectives, we use the descripive method, were the researcher
applied the M ultiple Intelligences test prepared by(Nabil Ibrahim), on a sample of (305)
students representing (36,7%) of the study population.
The study data were processed using the following stastistical methods: measures
of central tendency, measures of dispersion correlation coefficients ,Alfa Cponbah
coefficient, factor analysis ,fit estimation , difficulty estimation, ability estimation,
percentile ranks, T- scores, deviation QI.
The following results were concluded:
1-An observed data concordance with Rasch model.
2- Multiples Intelligences test items were calibrated according to their difficulty.
3- Items difficulty ranged from(–0,83)to (1,14)Logit in test of Linguistic Intelligence for
example.
4- Ability Pupils to every possible raw score, in the test of linguistic Intelligence are
between (–5,39) to (4,95)Logit.
5-Coefficient reliability of the students ability range from (0,79) to (0,86) in Multiples
Intelligences test, where as the of items difficulty is between (0,88) to (0,98), this indicates
that the scale is highly reliable.
6- The study proved the validity of the Multiples Intelligences test, through Rasch Model
unidimentionality.
7- The level of student ability can be judged and explained compared to his peers in male
and female, this can be achieved through the table of percentile ranks norms, T-score and
deviationQI.
