How much questionnaire detail is needed?

As part of the Q-BEx validation study, we have determined the minimal amount of questionnaire detail required to produce reliable predictors of proficiency in each language. You can read about it in our new preprint:

• We compare alternative measures of language exposure as predictors of language proficiency.
• We identify the optimal ones using information-theoretic and cross-validation methods.
• Cumulative exposure yields the most reliable predictor of proficiency in the societal language.
• Current exposure in the home yields the optimal predictor of proficiency in the heritage language.
• The greatest level of questionnaire detail did not yield more informative predictors.