This study introduces a new methodology of “ranking assessment” into pedagogical screening for the identification of second language (L2) readers requiring specific interventions. Results showed that a latent rank model has the potential to provide diagnostic information that determines what works for whom in L2 reading instruction.
BACKGROUND: This study introduces a new methodology of "ranking assessment" (Shojima, 2008) for diagnostic classification of second language (L2) learners. Pedagogical screening is conducted based on learners' individual differences (Ganschow & Sparks, 1991), such as L2 reading anxiety measured by the Foreign Language Reading Anxiety Scale (FLRAS; Saito, Garza, & Horwitz, 1999). In this study, a latent rank model is applied to the pedagogical screening for identification of learners requiring specific interventions.
METHOD: The CMC-FLRAS (Hamada & Takaki, 2021)-23 items with a 5-point Likert scale for cognitive, metacognitive, and classroom anxiety-was implemented for 633 Japanese learners of English. Statistical analyses by a graded latent rank model were conducted in the following three steps. The first step was to determine how many ordinal groups the learners were separated into by estimating latent cut-off points of the scale that indicated their risk of failure in L2 reading development. Second, the group characteristics were described based on structural equation modeling, and finally, the relationships between these group characteristics and learners' reading skills were examined.
RESULTS: The psychometrics (score range: 23–115) classified the learners into three groups with ordinal information. Group 1 (< 68: M = 56.70, SD = 8.45) exhibited good conditions in all three aspects of L2 reading anxiety. Group 2 (68–87: M = 79.23, SD = 6.90) reported high anxiety toward deficits in linguistic knowledge, while its metacognitive and classroom anxieties were undetected. Group 3 (> 87: M = 94.89, SD = 7.32) showed the highest anxiety about working memory capacity and classroom reading activities. Thus, the latent rank model has the potential to provide diagnostic information on what intervention works for whom in L2 teaching.
