The Effect Of Sample Size And It's Selection Methods On
The Psycho-metric Properties Of Psycho-scales.
Abstract
This study aims to
investigate the effect of sample size and the way of selecting it on the
psychometric properties of the psychological scales and their items which are
characterized by the discrimination index of the items, their reliability
coefficient, scale reliability and validity, in addition to frequency
distribution.
To fulfill this
goal, four samples had been chosen randomly with different sizes (400, 200,
122, 30) members and by the occasional sample.
The study adopted (Goldberg’s)
Psychic Health Scale , where it is executed on random simple sample size and
occasional sample synchronized with Psychic Tension Scale. Then the sample members
responses were analyzed according to psychometric properties for the scale and
its items on all samples.
The study also shows
that (400) members in the simple random sample gave more better properties for
the standard compared with other sizes.
(200) size could be
used also in random simple sample, especially in calculating the psychometric
properties for the items.
The study recommends
the use of simple random sample in calculating items psychometric properties in
general.
As well as, it
recommends also with the possibility of occasional sample use, If it was small size around (30) member.
It is preferable
that the size of the psychometric sample properties in psychological scales not
less than (400) members , especially in calculating psychometric properties for
the items.
Key
Words: Sample Size , Selection Methods, Simple random
sample, Occasional sample , Psychometric Properties .
By:
Ramzi Hasan Al-shinbari
Supervisor:
Prof.D.: Kamil Thamir Al-kubaisi
ConversionConversion EmoticonEmoticon