Off-campus FHSU users: Please use the following link to log into our proxy server and download this work.
Date of Award
Master of Science (MS)
A review of the literature concerning the Mathematics Anxiety Rating Scale (MARS) revealed that the MARS is currently being employed as a measure of therapy progress despite a lack of conclusive research to justify this procedure. Many scales of a similar nature have experienced considerable variability when used as a gauge of therapy progress. In addition, considerable pre- to posttest variability within the MARS had been previously reported. The purposes of this study were to identify these sources of variability, and attempt to stabilize the MARS by manipulating the serial order of the items. Three experimental versions of the MARS were employed: the original MARS (MARS-O), a reversed item sequence version (AARS-R), and a stratified version (MARS-S). Each version was administered according to standard instructions on the pretest, and re-administered four weeks later. Results showed a significant pre- to posttest reduction of group scores on the MARS-O and the MARS-S. Female Ss scored consistently higher than male Ss, and also demonstrated a significant pre- to posttest change on each of the three scale versions. Considerable item scoring and item discrimination variability was noted both within and across scales. Although the test-retest reliability of each of the scale versions was comparable to previously reported MARS data. The individual indices of change suggested a more pronounced variability than the group statistics implied. Attempts to stabilize the MARS through item manipulation were not successful. Implementation of the MARS as a pre- to posttest index of therapy change was not supported by this study.
Copyright 1974 Michael J. Hannafin
Hannafin, Michael J., "An Investigation of the Item Stability of the Mathematics Anxiety Rating Scale (Mars) : A Study of the Individual and Group Reliability Components of the Mars" (1974). Master's Theses. 1508.