CLAP: A battery of measures to estimate language proficiency in native Chinese speakers

Due to increasing interest in investigating the effects of individual differences in language proficiency on word recognition, there is a growing need for objective, reliable, and valid proficiency measures for adult native speakers. While such measures are available for English (Andrews & Hersch, 2010), few comparable instruments are available for simplified Chinese, hindering rigorous investigation into individual differences in word recognition. To address this, we developed Chinese Language Assessments of Proficiency (CLAP). CLAP can be completed within 30 min and comprises: (1) a dictation test with 20 short-answer questions, (2) a vocabulary knowledge test with 30 multiple-choice questions, and (3) a reading comprehension test with 24 multiple-choice questions. Test development proceeded in two phases. In Phase 1, test items were developed and pilot-tested with 50 native Chinese participants. Item response theory analyses were performed to select items for the final tests. In Phase 2, the final tests were validated among 200 participants. The three tests demonstrated good reliability (McDonald's ω and Cronbach's α > 0.8), acceptable test homogeneity (average interitem correlations: .17 - .29), and good convergent validity, evidenced by significant, positive correlations with scores on two validation measures: (1) LEXTALE-CH, a character-based proficiency test (Chan & Chang, 2018); and (2) Hanyu Shuiping Kaoshi, a test assessing grammar, vocabulary and reading skills (HSK official website). Exploratory Factor Analyses demonstrated that CLAP tests and validation measures all loaded onto a single factor. CLAP provides a practical tool for studying individual differences in language proficiency or screening participants based on proficiency.