An Investigation of Item Parameter Invariance Using Focused Calibration Samples for MAP Growth

Autor/in	He, Wei
Institution	NWEA
Titel	An Investigation of Item Parameter Invariance Using Focused Calibration Samples for MAP Growth
Quelle	(2021), (23 Seiten) PDF als Volltext kostenfreie Datei Verfügbarkeit
Sprache	englisch
Dokumenttyp	gedruckt; online; Monographie
Schlagwörter	Achievement Tests; Test Items; Instructional Program Divisions; Difficulty Level; Item Response Theory; Computation; Elementary Secondary Education; Mathematics Tests; Reading Tests; Science Tests; Measures of Academic Progress + Suchen Sie Ihr Suchwort? Achievement test; Achievement; Testing; Test; Tests; Leistungsbeurteilung; Leistungsüberprüfung; Leistung; Testdurchführung; Testen; Test content; Testaufgabe; Schwierigkeitsgrad; Item-Response-Theorie; Lesetest
Abstract	New MAP® Growth™ assessments are being developed that administer items more closely matched to the grade level of the student. However, MAP Growth items are calibrated with samples that typically consist of students from a variety of grades, including the target grade to which an item is aligned. While this choice of calibration sample is reasonable given that the current MAP Growth tests are not limited to assessing grade-specific content, a question arose about the appropriateness of using the current MAP Growth item parameter estimates in the new MAP Growth assessments. In other words, are the existing MAP Growth item parameter estimates invariant across different calibration samples? To evaluate the MAP Growth item parameter invariance, two studies were conducted using focused calibration samples (i.e., a limited sample of examinees). Study 1 explored the degree to which on-grade only item parameter estimates were comparable to their original all-grade counterparts. Study 2 explored the degree to which item parameter estimates using responses from the target grade plus the adjacent grades were comparable to their on-grade and all-grade counterparts. While the item difficulty estimates of a few items from the more focused calibration samples had different parameter estimates from their original counterparts, the parameter estimates of most items were comparable regardless of the calibration sample. These findings, in conjunction with the result of a recent study by Wan and Thum (2021) that used differential item functioning (DIF) analyses to reveal that MAP Growth items perform comparably across states and grades, provide the quantitative evidence of MAP Growth item parameter estimate invariance to support the use of the existing parameter estimates in the new MAP Growth assessments. (ERIC).
Anmerkungen	NWEA. 121 NW Everett Street, Portland, OR 97209. Tel: 503-624-1951; Fax: 503-639-7873; Web site: http://nwea.org
Erfasst von	ERIC (Education Resources Information Center), Washington, DC
Update	2024/1/01