Cognitive Factors Related to Computer Programming and Software Productivity *

The significance of software productivity is discussed and methods to increase computer programming productivity are reviewed. The range of performance capabilities of computer programmers in academic work environments based on several cognitive variables are delineated. Results reveal that for 90 percent of the population these ranges can vary from a low of 1:2 to a high of 1:99. Mathematical abilities are shown to predict up to 25 percent of the variance accounting for computer programming proficiency. Also combined chemistry and mathematical skills can predict up to 40 percent of the variance in computer programming proficiency. Basic Cognitive Skills, as measured by standard cognitive ability tests, are not shown to be helpful in predicting programming proficiency, whereas chunking appears to be a useful method for determining computer proficiency.