Why boys and girls may score differently on state tests

It found that, on average, boys outperform girls on state tests that rely more heavily on multiple-choice questions — a common way to measure math knowledge. Meanwhile, girls outperform boys on exams that rely more heavily on “constructed-response” items, which require students to write their own answers, whether it’s a sentence or an essay.

The study, led by researchers at Stanford University, offers important insights for education leaders working to bolster student achievement and narrow longstanding gaps in scores by gender. The findings also call into question whether school districts should be using high-stakes state exams to place students in certain courses.

Scholars examined data from three sources: the EDFacts Database, which is part of the National Center for Education Statistics; the National Assessment of Educational Progress (NAEP); and the Northwest Evaluation Association’s Measures of Academic Progress (MAP).

They analyzed the test scores of about 8 million students in grades 4 and 8 in math, reading and language arts across 47 states. They also considered whether tests relied more heavily on multiple-choice or open-ended questions.

Journalists should note that the researchers focused on test scores from the 2008-09 school year. Test content and formats may have changed as some states have adopted new standards, including Common Core State Standards.

Here are the main takeaways:

• In each subject area and grade level, girls, on average, outperform boys on state tests when a larger proportion of the total score is determined by students’ answers to constructed-response questions.
• Across subject areas and grade levels, tests favor boys when more of the score is based on how students answer multiple-choice questions.
• Differences in question format explain about 25 percent of the variation in gender achievement gaps among states.