Abstract:
For the past thirty years the majority of the
research on the cognitive demands of learning computer
programming has assumed that successful computer
programmers display certain aptitudes. Some
researchers indicate that studies linking different
prerequisite profiles of abilities to different
programming outcomes are needed. Such research might
dispel many of these notions. There is evidence that
personality variables are related to overall
performance in computer programming. Only recently has
this type of research been conducted. The purpose of
this study was to investigate the influence of one
personality variable, global-analytic cognitive style,
on one of the component skills of programming,
debugging.
Thirty subjects were administered the Group
Embedded Figures Test (GEFT) to determine their
cognitive style. Instruments were developed to
measure subjects' skills in diagnosing both syntactic
and logic errors in COBOL programs. Weighting schemes,
derived from results of previous research, were
developed for scoring the diagnostic exams. Data were
analyzed using Spearmans Rank Correlation Coefficient.
Results of statistical analysis failed to support
hypotheses (at the .05 level of significance) stating
that GEFT scores were related to subjects' ability to
locate and correct either syntactic or logic errors in
computer programs. This result appears to indicate
that measures of individuals' cognitive style
represent an unsatisfactory means for predicting
success on one aspect of computer programming skill
for novices. However at higher alpha levels the
results indicate support of the hypotheses.
