Czech version of JEPQ was submitted to confirmatory factor analysis us
ing data on 1184 girls and 1237 boys, age range 11 - 14 years. Using s
tandardized regression equations from congeneric analyses 21 item from
the 81 original ones were removed in order to reduce measurement nois
e. The overall confirmatory factor analysis based on 60 x 60 polychori
c matrix, using full LISREL 7 (ML) procedure, showed that GFI was 0.90
, AGFI 0.893, (RMR = 0.105), and generally supported expected four-fac
tor structure of the JEPQ instrument, but RMR index (0.105) indicated
that certain amount of variance remained unexplained suggesting that f
inding should be interpreted cautiously. Internal consistency of JEPQ
scales (normalized Cronbach coefficient alpha) ranged from 0.47 to 0.9
6 a stability in time (ten-week-interval) range was r=0.64 - 0.74. Tre
nd analysis, using boys and girls responses separately, supported the
hypothesis of linearity of changes in the JEPQ scores during four-year
-interval. Apart from L scale, equality of slopes (parallelism), for b
oys and girls data was confirmed for all scales. The findings are very
similar to that obtain on the British population and the shortened Cz
ech JEPQ version can be recommended for experimental use, especially E
and N scales.