Doctoral Dissertations

Date of Award

8-2011

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

School Psychology

Major Professor

R. Steve McCallum

Committee Members

Sherry M. Bell, Vena M. Long, Michael K. Smith

Abstract

The study investigated the psychometric properties of a newly developed math curriculum-based measure, the Monitoring Academic Progress: Mathematics (MAP:M), through examination of its internal consistency, alternate-form, slope, and test-retest reliability and validity. Participants included 1688 first through third-grade students from a school district in Northeast Tennessee. Application of Generalizability Theory produced reliability coefficients, score variances, and standard-error-of-measures (SEM) for both absolute and relative decisions based on a particular number of probes. MAP:M reliability coefficients for relative decisions ranged from .67 to .97 across eleven probes. The highest percentage of score variance at all three grades was attributed to the Person facet (academic growth), 58% to 66%, while variation in the Item facet (test difficulty) contributed the least score variance (11% to 16%). The SEMs for relative decisions ranged from 7.06 after one probe to 1.70 after eleven probes. SEMs were higher for absolute decisions. Results indicate that one MAP:M probe provides adequate reliability, although two probes provide good reliability estimates, exceeding .80. Alternate-form reliabilities’ ranged from .65 to .83 at first grade, .51 to .79 at second grade, and .59 to .80 at third grade. Lower reliabilities are best explained by increased time between administrations. Test-retest reliabilities were .81, .79, and .76 for first, second, and third grades, respectively. Correlation coefficients between the MAP:M and Monitoring Basic Skills Progress: Computation (MBSP) show adequate concurrent validity estimates ranging from .58 to .75 for first and third grades. Second-grade concurrent validity estimates were somewhat lower and ranged from .37 to .52. A stepwise multiple regression indicates that the MAP:M provides stronger relative predictive power than does the MBSP when predicting end-of-year STAR Math scores, , R2 =.33, F(28) = 13.11, p < .01. Results support adequate to strong reliability and validity of the MAP:M, in addition to successful implementation in a school environment.

Hopkins_Dissertation_Final_Submission_Version(1).doc (1209 kB)
Final Word Document w/o ETD and Appendix B

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS