Self-assessment as a method for measuring proficiency in college-level language programs

We investigated the validity and utility of an oral proficiency self-assessment comprised of the ACTFL can-do statements (2015). We had two research questions: (1) What are the correlations among self-assessment results, oral proficiency scores, and course levels? (2) To what extent do oral proficiency scores and course levels predict self-assessment results?We had 863 college-level Spanish language learners take the ACTFL OPIc in Spanish. We excluded the data of 63 students because they failed to receive an interpretable OPIc score (i.e., they either received an AR, BR, or UR on the test). This resulted in the final dataset of 820 Spanish students, who were in 100-level (N = 134), 200-level (N = 257), 300-level (N = 344), and 400-level (N =85) courses when they were tested. Before taking the OPIc, all took our self-assessment based on the ACTFL can-do statements. Among the learners who achieved Intermediate-mid on the OPIc, 55% were predicted to reach Level 2, 18% were predicted reach Level 3, and 14% were predicted to reach Level 4 or higher on the self-assessment. Learners with higher OPIc scores were more likely to reach higher levels on the self-assessment. A strong contrast existed between the students at the top and those at the bottom of the OPIc scale. While only 1% or less of the students with a score of Novice-low or lower were predicted to reach Levels 3 to 5, 50% of those with a score of Advanced-mid or higher were expected to reach Level 5. Results indicate that the self-assessment measure performed well as a measure of program-level achievement and perhaps could be used to measure progress in tertiary language programs. However, if scores are needed at the individual level and for high-stakes decision-making purposes, the self-assessment scores may have to be interpreted with extreme caution.