Résumé:
This study entitled «study compare between classical theory and Rasch
model in choice the items of Minnesota Multiphasic Personality Inventory
2(M.M.P.I.2) with pupils phase shighschools in Algeria-Afield study in some
secondry school in the wilaya of Mostaganem», we propose to identify the
psychometric characteristics of Minnesota Multiphasic Personality Inventory 2
according to classical measurement ,and according to Rash mode.To realize
this objective, we used the descriptive methodology with the measures of
central tendency, measures of variability, exploratory factor analysis, correlation
cofficents, T test, Cronbach's Alpha, fit statistics, difficulty estimation, ability
estimation. The study was conducted on a sample consists to (300) pupils (,
who wobble their age between(16-18 ) year. in the school year
(2015/2016).The sample of the study was selected: as a multi-stage cluster
deliberately.
The study yielded the following results:
1.Concerningclassical theory
1. 1.The value of arithmetic means are higher than of the median and the
mode; which refers to normality of the distribution.
1.2 The validity of the test:
1.2.1.Exploratory factor analysis shows :
_Three Factors for validity scales and clinical scales:
The first factor Bad adaptation general or psychopathology .
The second factor The disorders neuroticism.
the third factor social incompatibility.
_ Three Factors for supplementary scales:
The first factor Psychogical incompatibility.
The second factor Social adaptation
the third factor Neurotic symptom
_ Three Factors for content scales
The first factor Psychological confliet.
The second factor Mood and behavior and relation disorders
the third factor social poor adaptation.
1.2.2.Existence of the items are significantly correlated to the total of the
scales at the levels 0,01 and at 0,05 ;and there are som items not
significantly correlated to the scales.
1.2.3.Differences of statistical significance between the least third and
supreme third in the scales raw score estimated; and then we can say the
test has a good diagnosis ability.
1.3.The reliability of the test:
1.3.1.Internal consistency factor by Chronbach's Alpha method varied between
[0,13 and below - 0,60 and above].
1.3.2.The test – retest method show a factor variation between [0,13 and
below - 0,60 and above].
2.Concerning Rasch model
2.1.The study proved the validity of the scales Minnesota Multiphasic
Personality Inventory 2 , because Rasch model provides the indicators of
conformity infit indices and outfit indices, and unidimensionality in
measurement.
2.2.Coefficient of reliability of items between [0,95 ] to[0,99 ] and that of
persons between [0,13 and below - 0,60 and above]. This indicates that
som scales of this test are acceptable reliable.
