NotesFAQContact Us
Search Tips
ERIC Number: ED424270
Record Type: Non-Journal
Publication Date: 1997-Aug
Pages: 26
Abstractor: N/A
Reference Count: N/A
Improving the Information Value of Performance Items in Large Scale Assessments. NAEP Validity Studies.
Pearson, P. David; Garavaglia, Diane R.
This essay explores what is known and what remains to be learned about the information value of performance items when they are used in large scale assessment. Within the context of the National Assessment of Educational Progress (NAEP), there is substantial interest in answering these questions as issues of national assessment are considered for the future. The concept of information value is operationalized in the essay. Information has value to the degree that it helps in decision making or drawing conclusions about issues that matter. In education the issues that matter usually take the form of questions about cognitive processes or curriculum. This essay is divided into four sections. The first considers the construct of information value in its broadest sense, and then describes ways to operationalize it. The second section reviews the available literature within each of the operational traditions of: (1) item response theory; (2) factor analysis; and (3) correlational studies. The third section considers some alternative versions of information value, based on more cognitive, conceptual, and pragmatic considerations. The final section outlines a series of studies that, it is asserted, should be supported by the National Center for Educational Statistics to answer the questions about how the NAEP can be modified to increase its information value. The question of what constructed response items add to the interpretation of NAEP results is critical at this point in the history of the NAEP. Critics assert that constructed response items should be eliminated in the interests of economy and precision. Until, and unless, it can be demonstrated that there are grounds, whether psychometric, conceptual, or pragmatic, for maintaining or expanding the emphasis on constructed response items, their use will be questioned and questionable. (Contains 26 references.) (SLD)
Publication Type: Reports - Evaluative
Education Level: N/A
Audience: N/A
Language: English
Sponsor: National Center for Education Statistics (ED), Washington, DC.
Authoring Institution: American Institutes for Research in the Behavioral Sciences, Palo Alto, CA.
Identifiers - Assessments and Surveys: National Assessment of Educational Progress