The role of procedural similarity, self-explanation and self-constructed diagrams in analogical problem solving

Saifaddin, Huda B. (2011) The role of procedural similarity, self-explanation and self-constructed diagrams in analogical problem solving. PhD thesis, University of Nottingham.

PDF - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (8MB) | Preview


This study aimed to investigate the precise role of self-support methods, such as self-explanation and self-constructed diagrams, as an alternative to external methods in enhancing the cognitive processes considered crucial for effective transfer performance in analogical problem-solving that depicts a multi-step process involving source problems and target problems. This was achieved by systematically examining how type of representation (Verbal & Pictorial) and levels of similarity (Principle, Strategy, and Procedural) interact with self-support methods (Self-explanation (SE) and Self Constructed Diagrams (SCD)) in influencing transfer performance. Three experiments were conducted each addressing a set of issues related to the purpose of the study.

Experiment 1 (N = 48) was conducted to identify the cognitive processes and their sub-processes involved in analogical problem solving using pictorial representation and also investigated the specific effects of the self-explanation method on transfer process. This experiment consisted of two experimental conditions; self-explanation (SE) (expermintal group) and verbalization (VB) (control group), and three levels of similarity (i.e., procedural, strategy, and principle). Procedural similarity combined with the SE method was found to have a positive significant influence on the transfer process compared to the principle and strategy levels and VB condition. However, the verbal protocols also revealed that despite the inherent advantages of SE the percentage of complete solvers was low. This was attributed to some difficulty arising from adapting information from a pictorial source to solve a verbal target.

Experiment 2 (N = 84) investigated the effect of verbal and pictorial types of representation on transfer performance in a within-subjects design, where each participant solved a pictorial source (PS) and verbal source (VS) problem, and their verbal target analogues. The mean performance of the pictorial representation was higher compared to verbal representation. Transfer performance was higher in the procedural level than the strategy level. This indicated that information from PS tends to be utilized more effectively than VS in retrieving and applying that information to the target problem. Thus having ensured that pictorial representation was an advantage in problems depicting a multistep to be implemented, Experiment 3 was conducted.

Experiment 3 (N = 160) aimed at finding whether self-constructed diagrams (SCD) are a better alternative to external support in facilitating the cognitive processes crucial for transfer in analogical problem-solving. As predicted, a significant difference was found between the experimental (SCD) and No Diagrams (ND) control groups in the transfer performance. No significant within subject difference in the transfer performance of verbal and pictorial source representations was found in the SCD condition. An interesting finding was that transfer performance was significantly higher in the verbal representation and strategy level of similarity in the SCD condition than ND. Theoretically, this suggests that because visual memory is more easily accessible than auditory memory, SCD may play a critical role in creating accessible information from the source problem for effective feedback to help solve the target problem.

It was concluded that explaining by diagrams helps in identifying the various elements of the problem that stimulate the memory and motivate the person to recall what he drew earlier while solving the target problem. This study contributed to the field of research on the cognitive processes involved in problem-solving by analogy. The methodology employed in each of the experiments was unique in terms of coding and scoring the protocols, which generated strong and reliable results. The outcome of the study was a dynamic model “The Generative Procedural Model of Analogical Problem-solving” which contributed to our understanding of not only how information is processed from verbal and pictorial representations during problem-solving by analogy but also the potential of a self-method in optimizing the processes of noticing, retrieving, and implementing a learned solution process successfully.

Item Type: Thesis (University of Nottingham only) (PhD)
Supervisors: O’Malley, C.E.
Keywords: analogy, analogical problem solving, self-explanation, verbalization, self-constructed diagrams
Subjects: B Philosophy. Psychology. Religion > BF Psychology
Faculties/Schools: UK Campuses > Faculty of Science > School of Psychology
Item ID: 11997
Depositing User: EP, Services
Date Deposited: 24 Oct 2011 11:10
Last Modified: 13 Oct 2017 21:10

Actions (Archive Staff Only)

Edit View Edit View