Rod Ellis conducted a psychometric study to measure implicit and explicit knowledge in second language (L2) learning. The study aimed to establish operational definitions of these constructs and assess the validity of tests designed to measure them separately. Five tests were developed: an oral imitation test, an oral narration test, a timed grammaticality judgment test (GJT), an untimed GJT, and a metalinguistic knowledge test. Tests (a), (b), and (c) were designed to measure implicit knowledge, while tests (d) and (e) were designed to measure explicit knowledge. A principal component factor analysis revealed two clear factors: Factor 1, corresponding to implicit knowledge, and Factor 2, corresponding to explicit knowledge. The scores from tests (a), (b), and (c) loaded on Factor 1, while the scores from ungrammatical sentences in test (d) and total scores from test (e) loaded on Factor 2. The study also reported secondary analyses to support the interpretation of the construct validity of the tests. The study examined 17 English grammatical structures and found that native speakers (NSs) performed significantly better than L2 learners on most measures, except for the metalinguistic knowledge test and the ungrammatical sentences in the timed GJT. L2 learners scored highest on the untimed GJT measures. Both NSs and L2 learners scored higher on grammatical than on ungrammatical sentences in the timed GJT. L2 learners showed greater intergroup variance than NSs on all tests. Correlation analysis revealed strong correlations between the tests, with the metalinguistic knowledge test showing generally weaker correlations with the other tests. The study concluded that the tests provided relatively independent measures of implicit and explicit knowledge, supporting the construct validity of the tests.Rod Ellis conducted a psychometric study to measure implicit and explicit knowledge in second language (L2) learning. The study aimed to establish operational definitions of these constructs and assess the validity of tests designed to measure them separately. Five tests were developed: an oral imitation test, an oral narration test, a timed grammaticality judgment test (GJT), an untimed GJT, and a metalinguistic knowledge test. Tests (a), (b), and (c) were designed to measure implicit knowledge, while tests (d) and (e) were designed to measure explicit knowledge. A principal component factor analysis revealed two clear factors: Factor 1, corresponding to implicit knowledge, and Factor 2, corresponding to explicit knowledge. The scores from tests (a), (b), and (c) loaded on Factor 1, while the scores from ungrammatical sentences in test (d) and total scores from test (e) loaded on Factor 2. The study also reported secondary analyses to support the interpretation of the construct validity of the tests. The study examined 17 English grammatical structures and found that native speakers (NSs) performed significantly better than L2 learners on most measures, except for the metalinguistic knowledge test and the ungrammatical sentences in the timed GJT. L2 learners scored highest on the untimed GJT measures. Both NSs and L2 learners scored higher on grammatical than on ungrammatical sentences in the timed GJT. L2 learners showed greater intergroup variance than NSs on all tests. Correlation analysis revealed strong correlations between the tests, with the metalinguistic knowledge test showing generally weaker correlations with the other tests. The study concluded that the tests provided relatively independent measures of implicit and explicit knowledge, supporting the construct validity of the tests.