Errors of Measurement, Theory, and Public Policy. William H. Angoff Memorial Lecture Series
Saved in:
| Title: | Errors of Measurement, Theory, and Public Policy. William H. Angoff Memorial Lecture Series |
|---|---|
| Language: | English |
| Authors: | Kane, Michael, Educational Testing Service |
| Source: | Educational Testing Service. 2010. |
| Availability: | Educational Testing Service. Rosedale Road Mailstop 19R, Princeton, NJ 08541-0001. Tel: 609-921-9000; Fax: 609-734-5410; Web site: http://www.ets.org |
| Peer Reviewed: | N |
| Page Count: | 40 |
| Publication Date: | 2010 |
| Document Type: | Opinion Papers Speeches/Meeting Papers |
| Descriptors: | Error of Measurement, Scores, Public Policy, Test Theory, Reliability, Certification, High Stakes Tests, Testing, Federal Legislation, Educational Legislation, Test Validity |
| Laws, Policies and Program Identifiers: | No Child Left Behind Act 2001 |
| Abstract: | The 12th annual William H. Angoff Memorial Lecture was presented by Dr. Michael T. Kane, ETS's (Educational Testing Service) Samuel J. Messick Chair in Test Validity and the former Director of Research at the National Conference of Bar Examiners. Dr. Kane argues that it is important for policymakers to recognize the impact of errors of measurement on test scores and on average scores for groups. He asserts that a clear understanding of the magnitude of errors of measurement can have at least two benefits: identifying where measurement procedures need to be improved, and improving policy decisions by reducing the tendency to interpret and act on score differences that may, in fact, be meaningless. Dr. Kane grounds his argument in a discussion of the origins of errors of measurement, their role in conceptual frameworks and their control. He then discusses the relationship between definitions of constructs (variables of interest) and individuals' conception of errors of measurement. Understanding this relationship helps users of test scores decide whether a score difference should be interpreted as a difference in the construct being measured or should be attributed to error. (Contains 6 footnotes and 1 table.) [This lecture was presented at Educational Testing Service, Princeton, New Jersey, on November 19, 2008.] |
| Abstractor: | ERIC |
| Number of References: | 29 |
| Entry Date: | 2010 |
| Accession Number: | ED509385 |
| Database: | ERIC |
| Abstract: | The 12th annual William H. Angoff Memorial Lecture was presented by Dr. Michael T. Kane, ETS's (Educational Testing Service) Samuel J. Messick Chair in Test Validity and the former Director of Research at the National Conference of Bar Examiners. Dr. Kane argues that it is important for policymakers to recognize the impact of errors of measurement on test scores and on average scores for groups. He asserts that a clear understanding of the magnitude of errors of measurement can have at least two benefits: identifying where measurement procedures need to be improved, and improving policy decisions by reducing the tendency to interpret and act on score differences that may, in fact, be meaningless. Dr. Kane grounds his argument in a discussion of the origins of errors of measurement, their role in conceptual frameworks and their control. He then discusses the relationship between definitions of constructs (variables of interest) and individuals' conception of errors of measurement. Understanding this relationship helps users of test scores decide whether a score difference should be interpreted as a difference in the construct being measured or should be attributed to error. (Contains 6 footnotes and 1 table.) [This lecture was presented at Educational Testing Service, Princeton, New Jersey, on November 19, 2008.] |
|---|
