Methods Employed in Studies Identifying "Levels" of Test Anxiety in University Students: A Systematic Review.

Test anxiety research has been dominated by attention to theory building and examining the causes and consequences of this construct. However, recently, considerable attention has been turned toward using test anxiety as a diagnostic indicator of students who are at risk of underperforming in educational settings. This systematic review of the literature was focused on (a) describing the strategies used in the field, (b) highlighting the benefits and limitations of those approaches, and (c) offering guidance in creating a framework for appropriate methods when identifying severity levels on measures of test anxiety for university students. The results confirmed that the vast majority of studies on test anxiety have no formal "leveling" approaches (maintaining test anxiety as a continuous variable). However, when researchers do employ "leveling" strategies, the majority adopt inappropriate methods (e.g., single-sample splits). However, there are exemplars that demonstrate statistically sound procedures for identifying distinct profiles of test anxiety that may form a basis to build consensus around a classification method for elevated test anxiety.