Tennessee’s education department released preliminary statewide results for the 2013-14 Tennessee Comprehensive Assessment Program, or TCAP, Tuesday. The results spotlight statewide trends and figures. District- and school-level results will come later this month.

A lot rides on the state’s standardized test: TCAP scores are used for everything from students’ end-of-year grades to teacher evaluations to decisions about whether to close or open schools.

Overall, there was a slight improvement in scores this year. But when it comes to test scores, a lot of the interest is in the details: How did minority students fare? What’s happening in eighth grade?

So, how did Tennessee students fare on this year’s test?

Overall, not bad, not great. The numbers looked positive at the high school level. Scores continued to nudge upward on most of the end of course tests, as they have for the past several years. The results from elementary and middle school are more muted, growing less than in previous years. And depending on how you slice up the numbers, they weren’t even positive. For example, the percentage of students scoring proficient increased on 19 of 31 tests overall. But a lot of that can be attributed to tiny improvements in science and social studies scores. If you just focus on math and English, fewer than half of the tests (eight out of 17) showed improvements.

Overall math scores continue to grow faster than reading scores. Improvements in math have outpaced improvements in reading every single year for the last four years for grades 3 through 8, with the number of students scoring proficient in reading falling for the first time this year. Interestingly, math proficiency scores started off eight percentage points lower than reading in 2010, surpassed reading scores in 2013 for the first time and now seem to be moving steadily ahead.

But more and more math students fall to the bottom the longer they are in school: Between grades 3 and 8, the portion of students scoring very low in math grows as students advance through school. More students score below basic each year, from just 7 percent of students in 3rd grade to 24 percent in 8th grade.

Nobody fails social studies. Zero percent of social studies and U.S. History students scored below basic. This could mean that the state has set the failing score so low that guessing on every question still would not cause you to score below basic. Fewer students take the U.S. History test than any other end of course test in high school.

The achievement gap between minorities and white students narrowed a little. Sort of. The difference between the percent of minority and white students who scored proficient on end-of-course tests decreased on every test except for two. The most progress was made on the Algebra I and English II tests, where the gap between minority and white students decreased two and three percentage points respectively. But the data lumps together all black, Hispanic and Native American students, so it is unclear if the gains made were reflected across all minority groups or were predominantly focused on one or two.

The achievement gap for other disadvantaged groups remains wide. The gap between low-income students and their peers did not change significantly between this year and last. Between 20 and 30 percent fewer low-income students met basic standards on TCAP tests all levels. The gap between students with disabilities and their peers actually grew this year in a majority of subjects.

English language learners’ scores are still lower in English. But this year, they’re falling farther behind in math. The gap between English language learners and their peers widened on high school math tests and decreased slightly on English tests. But the overall gap between ELL students and their peers on English tests is still larger than the gap in math, with between 33 and 45 percent fewer English language learners meeting basic standards.

What was missing from today’s data? This data doesn’t include district or school performance information. That will come later. But the state also held back a few other key numbers. We don’t know how many students in each subgroup took which test and why. This data also didn’t give any historical perspective on how students did at the very high or very low end of the testing results. More on this below.

More students tested in high school this year, fewer students tested in elementary and middle. Between 1,000 and 2,000 fewer students tested in grades 3-8 this year than last year. The number of students who took tests in high school increased this year on every test except for Algebra II, and barely increased on English III. Algebra II and English III are the most difficult end of course tests, according to the data released, with fewer than half of test takers scoring proficient on each. This raises questions about whether schools are pushing their students to achieve at their highest levels or whether they are keeping students from taking difficult tests that may lower their test scores.

What do the numbers actually mean? The numbers people are talking about today are mostly the percentage of students who scored proficient or advanced—3 or 4 on a scale of 1 to 4. Students who scored a 3 are said to be at grade level. So when you hear people talk about the percentage of students who passed the test increasing or decreasing, they are saying there are more or fewer students whose scores have either dropped below 3 or risen above 2. But the data released today doesn’t tell us much about how many students moved between 1 and 2, or between 3 and 4. Some districts have been accused of focusing too much on “bubble students“–students who score a 2 or a 3. The worry is that schools are not doing enough for very low-performing students because it would take too much effort to raise their scores, or for very high-performing students because they don’t receive extra credit for top students doing even better.