After considering these definitions, the committee decided a third cognitive skill, critical thinking, was not fully represented. The committee added critical thinking to the list of cognitive skills, since competence in critical thinking is usually judged to be an important component of both skills (Mayer, 1990). Thus, this chapter focuses on assessments of three cognitive skills: problem solving, critical thinking, and systems thinking.
DEFINING THE CONSTRUCT
One of the first steps in developing an assessment is to define the construct and operationalize it in a way that supports the development of assessment tasks. Defining some of the constructs included within the scope of 21st century skills is significantly more challenging than defining more traditional constructs, such as reading comprehension or mathematics computational skills. One of the challenges is that the definitions tend to be both broad and general. To be useful for test development, the definition needs to be specific so that there can be a shared conception of the construct for use by those writing the assessment questions or preparing the assessment tasks.
This set of skills also generates debate about whether they are domain general or domain specific. A predominant view in the past has been that critical thinking and problem-solving skills are domain general: that is, that they can be learned without reference to any specific domain and, further, once they are learned, can be applied in any domain. More recently, psychologists and learning theorists have argued for a domain-specific conception of these skills, maintaining that when students think critically or solve problems, they do not do it in the absence of subject matter: instead, they think about or solve a problem in relation to some topic. Under a domain-specific conception, the learner may acquire these skills in one domain as he or she acquires expertise in that domain, but acquiring them in one domain does not necessarily mean the learner can apply them in another.
At the workshop, Nathan Kuncel, professor of psychology with University of Minnesota, and Eric Anderman, professor of educational psychology with Ohio State University, discussed these issues. The sections below summarize their presentations and include excerpts from their papers,1 dealing first with the domain-general and domain-specific con-
1For Kuncel’s presentation, see http://www7.national-academies.org/bota/21st_Century_Workshop_Kuncel.pdf. For Kuncel’s paper, see http://www7.nationalacademies.org/bota/21st_Century_Workshop_Kuncel_Paper.pdf. For Anderman’s presentation, see http://www7.national-academies.org/bota/21st_Century_Workshop_Anderman.pdf. For Anderman’s paper, see http://nrc51/xpedio/groups/dbasse/documents/webpage/060387~1.pdf [August 2011].