The main purpose of this study is to explore the predictors of programming achievement. With this aim in mind, the students' success in the programming courses is specified as the dependent variable and creativity, problem solving, general aptitudes, computer attitudes and mathematics achievement are specified as the independent variables. A correlational design was used to explain the relations between dependent and independent variables. The study group consists of 48 high school students in Profilo Anatolia Technical High School, Istanbul. At the end of the study, significant relations were found between the students' programming achievement and their general aptitudes and mathematics achievement. Also, in order to determine the predictors of the students' programming achievement, multiple regression analysis was applied. The findings reveal that only one variable that significantly predicts the students' programming achievement is general aptitude.
There have been many studies in recent years into academic success in computer programming (McNamarah & Pyne, 2004; Byrne & Lyons, 2001; Begum, 2003; Fowler and et al., 2002). Today, industry is keen to accept as many graduates as the academic institutions can produce, and there is an assumption that any bright student can be successful in computer programming. However, experience in the classroom would suggest that this is not true. Students who are proficient in many other subjects sometimes fail to achieve success in programming (Byrne & Lyons, 2001), because programming is different from other discipline.
The developments of programming languages and methods, and the teaching of them, have up to now hardly been linked to a psychological study of the activity of programming. Psychology must go beyond the procedural aspect of programming; because it is becoming more and more important nowadays due to the variety of applications and the training that programmers receive (Hoc and et al., 1990). Prior research indicates that standardized measures of aptitude (e.g. SAT and ACT scores), prior academic performance (e.g. high school GPA) and effort or motivation explain a significant portion of the variation in student performance (Eskew & Faley, 1981; Hostetler, 1983; Goold & Rimmer, 2000).
In a review of studies attempting to predict programming achievement done up to 1990, Hostetler and Corman make a specific case for the inclusion of cognitive factors in any study of this kind (Hostetler, 1983; Corman, 1986). They found that some of the demographic, academic, computer exposure or cognitive variables were particularly strong predictors of class performance. According to Taylor and Mounfield (1989) prior experience in programming provides a significant predictor of how students perform in the programming courses. They founded that prior exposure whether at the high school or college level is an important factor to students' success in computer programming.
Also, the link between mathematics ability and programming is widely accepted. Several of the reviewed studies showed that success in Mathematics was a good predictor of success in computer science (Byrne & Lyons, 2001; Fowler and et al., 2002; Werth, 1986; Campbell, 1984; Chmura, 1998). There is a belief that the concepts which a student has to comprehend in order to master mathematics problems are similar to those for programming (Byrne & Lyons, 2001; Werth, 1986).
Besides those there appears to be a number of factors which influence the success in the programming. In general the reviewed research found correlation between computer attitudes and computer programming (Dey & Mand, 1986; Austin, 1987). Also, earlier studies indicated that demographic data impacted on programming success (Byrne & Lyons, 2001; Goold & Rimmer, 2000; Grant, 2003). Five factors were reviewed as potentially predictive to success in programming. …