Product Overviews Banner

CASAS Technical Manuals

Technical manuals document the research and development work in the item and test development process. Each manual provides information on a specific test series, the purpose of the series, and the tests in the series. Technical manuals are for state and local agency personnel as well as measurement specialists seeking technical information about the development and appropriate use of our tests. Just as we continually analyze our tests to ensure they remain reliable and valid, we update the technical manuals with the latest assessment research.

Although the development of our tests from multiple item banks is based on many traditional psychometric procedures including the preparation of test specifications, sound item writing practices, and both the pilot and field-testing of items using classical item analysis procedures, the underlying theoretical measurement foundation is IRT. The IRT model we employ to develop various item banks and the associated tests is the one parameter Rasch Model (1P-L).

The Development of our tests is described and includes the item writing process, the pilot and field-test process, the process to calibrate and assign items to forms, the results of fairness and sensitivity reviews, and psychometric analyses that examine the performance of test items and completed forms.

The Maintenance of our tests includes information about test publishing dates, steps taken to ensure score comparability across test forms, steps taken to maintain test security, and a history of test use.

Content Validitythe degree to which the content of the test matches the entire content domain associated with the construct, includes the match of the content to the NRS Educational Functioning Levels, the content standards and competencies measured by the test series, and the qualifications of the subject matter experts involved in content determination.

Standard-Setting Procedures  describes the establishment of cut scores for each NRS Educational Functioning Level and the standard error of measurement for each cut score.

Reliability, the consistency of test scores or the degree to which our tests measure the same way each time they are used under the same condition with the same subjects, includes a description of research designs used. Evidence of reliability includes the correlation of scores across alternate or parallel test forms and classification consistency into NRS Educational Functioning Levels (EFLs) when using alternate or parallel test forms.

Evidence of Construct Validity, the degree to which a test measures only the intended construct, compares our tests with other tests that assess educational gain for similar populations. Also included is additional evidence of convergent validity which shows the degree to which performance on CASAS assessments is correlated with other measures that it is theoretically predicted to correlate. Also include are analyses of practice effects.

The Other Information section covers the determination of test administration time, appropriate modifications for individuals who have a disability, and retesting recommendations.

For more information on CASAS technical manuals, contact Senior Research Associate Jared Jacobsen at jjacobsen@casas.org.

© 2021 by CASAS. All rights reserved.