Exploring International PISA 2018 achievement gaps between 15-year-olds whose primary home language did and did not match the test language

• In Event: Poster Session 2
  In Poster Session: Exploring International PISA 2018 achievement gaps between 15-year-olds whose primary home language did and did not match the test language

Proposal

The Program for International Student Assessment (PISA) is an international assessment of 15-year-olds’ achievement in reading, mathematics, and science literacy, with the most recent data collection held in 2018. This proposed poster investigates achievement gaps between students whose primary home language did and did not match the language of the PISA assessment. Identifying gaps in achievement based on primary home language can help inform educators and policymakers on whether students whose primary home language differs from the test language need additional support.
Analyses focus on five countries (the United States, Germany, Belgium, Sweden, and New Zealand) with similar percentages of students speaking primary home languages that do not match the test language (14 – 18 percent).
Across select PISA 2018 participants, how did achievement vary by subject and primary home language?
For all five countries, students whose primary home language did not match the test language had lower average reading and science achievement than students whose primary home language matched the test language. In the United States, this difference was 29 points for reading (511 v. 482) and 34 points for science (509 v. 475). Similarly, in four out of five countries, students whose primary home language did not match the test language had lower average math achievement than students whose primary home language matched the test language (no significant difference was observed in New Zealand).
Across select PISA 2018 participants, what achievement gaps exist between students whose primary home language did and did not match the language of the test? How do these gaps differ by subject?
Risk ratio analysis (relative risk) was used to compare these two student groups’ likelihoods of scoring below the PISA median (500). For reading and science, in all five countries, students whose primary home language did not match the test language were more likely to score below the PISA median than other students. In reading, U.S. students whose primary home language did not match the test language were 1.27 times more likely than their counterparts to score below the median. The risk ratios in reading for the other four countries ranged from 1.33 (New Zealand) to 1.81 (Sweden). In science, U.S. students who spoke a different primary home language were 1.30 times more likely to score below the median than their counterparts. The risk ratios in science for the other four countries ranged from 1.27 (New Zealand) to 1.89 (Germany).
For mathematics, in all countries of interest except New Zealand, students whose primary home language did not match the test language were more likely to score below the median than other students. U.S. students whose primary home language did not match the test language were 1.20 times more likely than their counterparts to score below the median. Risk ratios in mathematics for other countries with significant differences ranged from 1.66 (Belgium) to 1.78 (Germany).

Authors

• Amy Burton

• Jason Solinsky, American Institutes for Research