Table 10 Inferences, warrants, assumptions, and backing in the validity argument for the Criterion®
Inference | Warrant that justifies the inference | Assumption underlying the warrant | Backing for the assumption |
|---|---|---|---|
Generalization | Observed Criterion® scores are estimates of expected scores over the relevant parallel prompts. | Criterion® prompts are parallel in term of difficulty. | Significant but minor differences across prompts. |
Extrapolation | The construct of L2 writing proficiency as assessed by Criterion® accounts for the quality of linguistic performance in L2 writing contexts. | Criterion® holistic scores are related to indicators of L2 proficiency. | Relatively weak but positive correlations between Criterion® holistic scores and indicators of L2 proficiency. |
Utilization | Estimates of the writing proficiency obtained from Criterion® are useful for making decisions about long-term gains in L2 writing proficiency. | Criterion® holistic and writing quality scores can show changes in writing over time. | Significant and constant differences across three time points in Criterion® holistic scores, as well as scores of essay length, and syntactic complexity. |