The Challenge A principal adaptive feature of the ACTFL Oral Proficiency Interview—computer (OPIc) is a self‐assessment instrument through which test takers select a level‐appropriate test form. How accurately do test takers select level‐appropriate forms on the OPIc? Would assistance with the test form selection process optimize the testing experience?

The present eye-tracking study tests the assumption that online techniques can tap implicit knowledge and reduce access to explicit knowledge by triangulating eye movements with a post-reading questionnaire. L1 English learners of L2 Spanish and Spanish native speakers (NSs) read sentences embedded with violations of determiner-noun and noun-adjective agreement followed by comprehension questions. They then completed a post-reading questionnaire that measured to what extent they were aware of the violations in the experimental stimuli. In the eye-tracking experiment, both L2 learners and Spanish NSs were sensitive to violations of determiner-noun agreement, but only Spanish NSs were sensitive to violations of noun-adjective agreement. On the post-reading questionnaire, most participants were also aware of the grammatical gender agreement violations, with determiner-noun agreement violations generally being more salient than noun-adjective violations. These results indicate that the violation detection paradigm implemented in this experiment did not entirely obviate explicit knowledge. Results are discussed in terms of methodological considerations researchers should keep in mind when designing and implementing studies that investigate grammatical gender agreement and other linguistic phenomena.