This bachelor's thesis focuses on verifying the convergent,
differential and construct validity of a diagnostic computer
game for measuring spatial abilities (specifically
visualization), called Lin and space navigation. The game is
based on the CHC theory of intelligence and uses elements of
gamification. The game is meant for pupils of the 1st to 6th
grade of elementary school, and the research population of this
work consists of pupils of the 4th and 5th grade (N = 86). The
pupils were administered the game Lin in a group, as well as
three individually administered subtests from the WJ IV COG
battery (COG 2, COG 7 and COG 14). Data analysis was carried
out using corre-lation analyses, reliability analysis and
factor analysis. The results of the correlation analysis
provide sufficient evidence of the convergent validity of the
game based on the correlation of the game and the COG 7 subtest
(r = 0.44; p < 0.001), which measures visualization, but we
failed to provide sufficient evidence of the differential
validity based on the correlation of the game with COG 2
subtest (r = 0.41; p < 0.01), which measures quantitative
reasoning. The analyses of the relationship between the game
and the COG 14 subtest (r = 0.27; p = 0.012), which measures
short-term visual memory, suggests that this ability does not
manifest significantly during the game. We did not verify the
construct validity of the game. One-factor model had bad fit
with our data, but at the same time we did not obtain
sufficient evidence for the existence of a second significant
factor. The reliability of the game is relatively high (α =
0.74). All in all, this thesis successfully verified the
convergent validity of the game and estimated its reliability
as satisfactory but failed to verify its differential and
construct validity.