Assessment of the generic problem-solving construct across different contexts

Year: 2018

Author: Nguyen Khoa, Lan Anh, Nguyen, Cuc, Adams, Raymond

Type of paper: Abstract refereed

Problem-solving, one of the generic skills in the so-called ‘21stcentury’ skill set, has been found to be critical to the personal and professional success of learners (Lynch, 1999; International Labour Office, 1998, pp. 41-43; Overtoom, 2000; Organisation for Economic Co-operation and Development (OECD), 2003a). Problem-solving skills are required in everyday activities, although we may not usually think of our activities in those terms, and are increasingly present in almost every subject in school nowadays.  
Numerous studies have been conducted worldwide to assess students’ problem-solving in a specific range of contexts. For example, studies have been conducted on mechanical problems within physics courses (Alagumalai, 1993; Chi, Feltovich, & Glaser, 1981; Docktor, 2009; Larkin, McDermott, D. Simon, & H. Simon, 1980), geometry and algebra problems within mathematics courses (Greeno, 1978; Khng & Lee, 2009; Schoenfeld, 1989; Schoenfeld & Herrmann, 1982; Wu & Adams, 2006), and chemistry and biotechnology problems with community college biotechnology students (Lavoie, 2003).
However, none of these studies has specifically identified generic problem-solving competence that is likely to transfer across different PS contexts and thus can be taught or learned easily. In particular, whereas evidence from empirical studies has shown that students were able to improve their problem-solving competence, none of the studies has assessed whether or not students were able to improve their skills in other situations (Harlim, 2012). Similarly, Billing (2007) raised the question of whether an individual problem-solving scale can be transferred from one educational setting to another, or whether such skills can be taught and assessed in one discipline and then applied in another. As Holmes (1998) stated, whether or how well those skills can be transferred is a matter of considerable debate.
The current study addresses these issues by drawing on publicly available PISA computer-based assessment 2012 secondary data. From the available literature and empirical studies on the generic problem-solving, we derive a definition and theoretically and empirically sound framework of the generic problem-solving construct.  Then we examine the conceptual meaning and the relationship between GPS construct and three domain-specific PS constructs by means of four Rasch models. Our results provide evidence for the knowledge of the GPS competence. Future empirical and practically studies can build on our findings and thus provide more useful results.