Abstract
Investigated maze test comparability and parallel form reliability; investigated whether different teachers could construct comparable maze tests over the same reading passage when given the same directions. Twenty-two teachers independently prepared two maze tests using two second reader passages. Three of the 22 sets of maze tests were then randomly selected for use with 246 second grade students. Each subject completed two maze forms, one of three possible forms for each of the two reader passages. The findings revealed that the teachers were able to prepare reliable, congruent maze tests of comparable difficulty if they properly followed the directions provided for test construction. With one exception the maze forms showed overall high reliability and congruence as well as comparability. An inspection of the items on the single maze form which was less reliable and which was neither congruent nor comparable revealed that the teacher who prepared the test did not follow directions and failed to compose a number of valid options.