structural invariance

Test washback is held to be complicated and multifaceted in that a host of cultural, social, individual, test, and institutional factors are involved in shaping it. Thus far, the majority of washback studies have had as their focus the role of teachers in test washback or washback to teachers. How educational environments or institutions might function in isolation or in interaction with other factors in shaping washback to the learners and test takers has not received adequate research attention. The current study examined the mediatory role of academic institutions in washback to learners' perceptions of test content and test preparation. To this aim, 86 senior English students from two universities, one a top tier and the other a low tier one, completed two questionnaires: one on test takers' preparation practices including test analysis, test taking skills, drilling target skills, and socio-affective strategies; and the other on test takers' construal of test demands and uses as well as their expectation of success on the test.  The data analyzed through partial least squares structural equation modeling revealed that a washback model based on expectancy-value theory explains a moderate amount of variance in test preparation. Further, for test takers from the low tier university, favorable perceptions of test content were associated with more value placed on test taking. However, Multi-group analysis pointed to group-invariance of the model across the two institutions, indicating a lack of strong evidence for the mediatory role of educational environments in washback to test takers’ perceptions and preparation.