A non-linear approach to solving linear algorithmic problems

Solving algorithmic problems and implementing them in code is a major theme in introductory computer science courses. Problems related to string manipulation are common and of great importance in various domains. In the context of computer science education, solving algorithmic problems of this type may be very challenging for novices, even for the more competent students. Analysis of novice students´ solutions to a specific class of problems dealing with string manipulation revealed that a widespread intuitive and non-linear approach was used to solve these problems, which led to cumbersome and erroneous solutions. On the other hand, a straightforward linear approach that leads to shorter and more elegant solutions is uncommon. Uncovering students´ common and unsatisfactory approaches in problem solving have significant implications on instruction.