Peggy G. Carr, Ph.D.
Acting Commissioner of the National Center for Education Statistics
2015 National Assessment of Educational Progress (NAEP) Grades 4 and 8 Mathematics and Reading
October 28, 2015
Today, I am releasing the results of the 2015 mathematics and reading assessments from the National Assessment of Educational Progress— The Nation's Report Card. The NAEP assessments were administered in early 2015 to 279,000 fourth-graders and 273,000 eighth-graders. The results for the nation include both public and private school students. At the state level, results are for public schools only, for all 50 states, along with the District of Columbia and the Department of Defense Education Activity, which are treated as states for comparison purposes. I will also highlight the main findings from our Trial Urban District Assessment, abbreviated as TUDA.
We report student performance both in terms of scale scores and the percentages of students attaining various achievement levels. For mathematics and reading, scores for students in grades 4 and 8 are reported on a 0-500 scale, and these scales are developed separately for each subject. Achievement levels were developed by the National Assessment Governing Board. They set standards for what students should know and be able to do. For each subject and for each grade, the Governing Board has established standards for Basic, Proficient, and Advanced performance. Students who reach the Proficient level display solid academic performance, demonstrating competency over challenging subject matter.
The assessments included both multiple-choice questions and questions requiring either short or long written answers, which we call "constructed-response" questions.
I will first discuss the overall national results in mathematics and reading, before turning to more detailed results for each subject, in turn, for the nation and the states. Finally, I will describe results for the large urban school districts we assessed. We will only discuss differences that are statistically significant.
Highlights of 2015 NAEP Results
Before describing the results in detail, I want to point out a few main findings from the 2015 assessments. In mathematics, for the first time since the early 1990s, there were declines nationally in scores at both grades 4 and 8 since the last assessment. Average mathematics scores were lower for students across the performance range, from high performers to lower performers.
At the state level, mathematics results were mixed— at grade 4, average scores were higher in 2015 than in 2013 in 3 states and/or jurisdictions. In 30 states, however, the average score in mathematics was lower in at least one of the grades.
At grade 8, national average scores were lower than in 2013 in both mathematics and reading and this was generally true for students across the performance range. The average grade 8 reading score was higher in 1 state; the average mathematics score did not increase in any state. Scores were lower in 26 states in either mathematics or reading, or in both subjects.
Over the longer term, scores were higher in 2015 compared to the initial assessments in the early 1990s, as were the percentages of students at or above Proficient.
The NAEP mathematics assessment measures student performance over five content areas: number properties and operations; measurement; geometry; data analysis, statistics, and probability; and algebra. These content areas were established by the NAEP Mathematics Framework developed by the National Assessment Governing Board.
In fourth grade, the national average score is 240 which is about 2 points lower than in 2013. Going back to 1990, the average score in 2015 was 27 points higher. We can also look at average scores over time for students across the performance range at various percentile levels (10th, 25th, 50th, 75th, and 90th percentiles). Students at all percentiles, except the 90th, scored lower than their peers did in 2013. The national eighth-grade average score of 282 was around 3 points lower than in 2013. Over the long term, the 2015 score was about 20 points higher than in 1990. When we look at the performance of students at the performance distribution, students at all percentiles scored lower in 2015 than their peers in 2013.
Now let us consider the mathematics results in terms of the percentages of students who performed at the three NAEP achievement levels— Basic, Proficient, and Advanced. At both grades, the percentage of students at or above Proficient was lower in 2015 than in 2013, but higher than in 1990.
Next we will discuss mathematics score changes at the national level for racial/ethnic groups and associated score gaps. Compared with 2013, the only score change for racial/ethnic groups was a 2-point decrease for White students at grade 4; scores for White, Black, and Hispanic groups declined at grade 8. Scores in 2015 for Asian/Pacific Islander students were not different from those in 2013 at either grade. Scores for all the groups for which we can report data since 1990 have increased for both grades. The White-Black score gap at grade 4 narrowed from 26 points in 2013 to 24 points in 2015. There were no other changes in the White-Black or White-Hispanic gaps between 2013 and 2015.
By gender, the average score for fourth-grade female students decreased two points and scores for both male and female students decreased at grade 8.
We also looked at data for students in various types of locations, those with disabilities, and English language learners. Average scores for several groups declined compared with 2013, with the exception of students in cities and suburbs, students with disabilities, and English Language Learners in grade 4. At grade 8, scores for students in cities and English language learners did not change. Some groups were not reported separately in 1990, so we cannot report changes from the earliest assessment for those groups.
Now let us consider the mathematics results for the states and jurisdictions. As I noted in my introduction, the state results did not always follow the national pattern. Average scores among public school students in the states at grade 4 ranged from 231 to 251, compared to the national average for public school students of 240. Grade 4 mathematics scores increased for several states and jurisdictions— these were the Department of Defense Education Activity, the District of Columbia, and Mississippi. Sixteen states had declines in average scores and there were no significant changes in scores from 2013 to 2015 in 33 states.
At grade 8, state scores ranged from 263 to 297, compared to the national public average of 281. Between 2013 and 2015, average scores declined in 22 states, while in 30 states and jurisdictions the scores were not significantly different. There were no score increases.
The NAEP assessment framework treats reading as a dynamic cognitive process that involves understanding written text, developing and interpreting meaning, and using meaning as appropriate to the type of text, purpose, and situation. The assessment uses both literary and informational texts. Literary texts can include fiction, literary nonfiction, and poetry. Informational texts include exposition, argumentation and persuasive texts, as well as procedural texts and documents. The assessment measures three "cognitive targets"— three mental processes that underlie reading comprehension. Students are asked to "Locate and Recall," to "Integrate and Interpret," and to "Critique and Evaluate."
The average score for the nation's fourth-graders in 2015 was 223, which was higher than in all previous years except 2013. The average score in 2015 was 6 points higher than the initial assessment in 1992. The average scores for students at the various performance levels did not differ from those in 2013 for any level of performance shown. For example, high-performing students, those at the 90th percentile, had an average score of 266, not significantly different from 2013. Similarly, students performing at the 10th percentile had a score of 174, the same as in 2013.
At grade 8, the 2015 score of 265 was almost 2 points lower than in 2013, but the same as in 2011; the 2015 score at grade 8 was 5 points higher than in 1992. When examining changes at the across the range of performance, the pattern at grade 8 was different from that at 4th grade. Eighth-grade scores declined at every percentile, reflecting the decrease in the national average score.
The percentages of fourth-grade students at all achievement levels in 2015 were not different from those in 2013. Compared to 1992, the percentages at or above Proficient and at Advanced were higher in 2015, and the percentage below Basic was lower. There was a decline at grade 8 in the percentage of students at or above Proficient, from 36 to 34 percent compared to 2013.
The percentage at or above Proficient was higher than in 1992, and the percentage below Basic was lower.
At grade 4 there were no changes in performance compared to 2013 by race/ethnicity or gender. However, scores for White, Black, and Hispanic students at grade 8 declined, as well as those for both male and female students. Compared to 1992, scores increased at both grades for all groups for which separate results were available in 1992. The White-Black and White-Hispanic gaps in reading did not change from 2013 to 2015 at either grade. These gaps ranged from 21 to 26 points.
Results for students in various types of locations follow a similar pattern— no change at grade 4 from 2013, but declines for all but students in city schools at grade 8. Scores for students with disabilities, and those without disabilities, improved at grade 4, but declined at grade 8, again following the general pattern. English language learner scores did not change at either grade from 2013 to 2015.
The average reading score for public school students in 2015 was not different from the 2013 score at grade 4, but was two points lower at grade 8. There was no significant change in scores for Catholic school students at either grade over this period.
As with mathematics, results across the states and jurisdictions are varied. At grade 4, state scores ranged from 207 to 235, compared to the national public school average of 221. Reading score increased in 13 states and decreased in 2 states (Minnesota and Maryland) compared to 2013. There were no changes in grade 4 average scores in 37 states or jurisdictions. At grade 8, average scores ranged from 248 to 277, compared to the national public school average of 264. Scores students in one state, West Virginia, increased since 2013, and declined in 8 states. There were no changes in 43 states and jurisdictions.
Trial Urban District Assessment (TUDA)
As I mentioned earlier, the NAEP program produces results for 21 large urban school districts through the Trial Urban District Assessment (TUDA) program. The TUDA program began in 2003 with 10 districts including the six original TUDAs from 2002 and has expanded to 21 districts that participated across the country in 2015.The program is a collaboration among NCES, the National Assessment Governing Board, and the Council of the Great City Schools. Participation in the program by the school districts is voluntary. Participating districts must fulfill certain criteria for inclusion in the program, such as population size (250,000 or more), and having enough students to provide sufficient samples for assessments in three subjects at one time. In addition, at least half the students must be nonwhite, and half must be eligible for the National School Lunch Program. The number of districts included also depends on available funding. Participation in the NAEP program affords these urban districts the opportunity for their educational performance to be measured on a common yardstick with other districts, large city students in general, the states in which they are located, and the nation. Thus, we compare the performance of TUDA districts to public school students in the nation and large cities. Large cities are a peer group for comparison, because they have a similar population size and demographic makeup as the TUDA districts.
Unlike the national and state results, there was not as strong of a pattern in any subject or grade, and the results were more varied. Scores for large cities, a peer comparison group for TUDA districts, were not significantly different from 2013 across all the subjects and grades. Large cities and seven TUDA districts had larger score gains than the nation in at least one subject or grade in comparison to 2003. There were score gains in 6 districts compared to 2013, and 10 districts had score losses in one or more subjects and grades in comparison to 2013.
When we look at the summary of the changes from 2013 to 2015, three districts had score gains in more than one subject and grade- Chicago, District of Columbia Public Schools, and Miami-Dade. There were three districts that saw significant decreases in their overall average scale scores in more than one grade or subject- Albuquerque, Baltimore City, and Hillsborough County, FL. Five districts had no change in their overall average scale scores in either grade or subject.
Mathematics Scores for Large Cities and TUDAs
At grade 4, the scores for large cities public schools were unchanged from 2013. Over time, there have been greater gains for large city schools than the nation, which have helped narrow the gap between large cities and the nation since 2003. The gap narrowed from 10 points in 2003 to 6 points in 2015.
We can also explore score gains since 2003 and see that large cities have made greater gains than that nation— scores for large cities increased by 10 points and scores for the nation (public schools) increased by 6 points. Over this same period, there are several TUDA districts that are outpacing the nation in score gains— Houston, Atlanta, Boston, Chicago and the District of Columbia. And three of these districts, Boston, Chicago and the District of Columbia, had greater gains than large cities. Allow me to clarify what we mean by the "District of Columbia" — you may remember that the District of Columbia was also one of the jurisdictions that improved in the state-level data. This refers to the District of Columbia as a state-level jurisdiction which includes both charter and non-charter public schools. As a TUDA, the District of Columbia includes public schools only that are not charter schools only; charter schools are not included in the DC TUDA results.
We can also see how the percentage of students at or above Proficient have increased since 2003. In large cities at grade 4, the percentages of students at all achievement levels in 2015 were not different from those in 2013. Across the national public schools, the percentage of students at Proficient declined by two percentage points and the percentage of students scoring below Basic rose by one percentage point. Looking back to 2003, both large cities and the nation had statistically significant score changes across most of the achievement levels. Large cities reported increases in the percentages of students at Advanced and Proficient, while there was a decrease in the percentage of students performing below Basic. The percentage of students rated as Basic was statistically unchanged. The national results had increases in the percentage of students scoring at the Advanced and Proficient levels, and decreases in the percentage of students at Basic and below Basic.
Now let us explore the TUDA results in more details. In grade 4 mathematics, average scores were higher in 2015 than in 2013 for three districts— Dallas, the District of Columbia, and Miami-Dade. Continuing with overall TUDA results for grade 4, 10 districts did not change significantly and 7 districts experienced declines in mathematics performance from 2013.
Fourth grade mathematics score changes by racial/ethnic groups varied across the districts. The districts that experienced score decreases in grade 4 mathematics, showed declines for some student groups as well. There was one exception— scores for Hispanic students in Miami-Dade, increased.
While the nation saw a decline in scores in the average mathematics score at grade 8 from 2013 to 2015, the average score for large city eighth-graders was unchanged. As we saw in fourth-grade, large city schools have made greater gains compared to the nation over time, since 2003. The gap has narrowed from 14 points in 2003 to 8 points in 2015.
Similar to grade 4, we see that large cities have made greater gains (12 points) than the nation (5 points) from 2003 to 2015 in grade 8 mathematics. There are seven districts making greater overall score gains than the national average for public schools since 2003, and four of those districts made greater gains than the large city average— Los Angeles, Boston, Chicago and Atlanta.
In grade 8 mathematics, the percentage of students who scored in the Proficient range declined by two percentage points across the nation, and there was a three percentage point increase of students below Basic. Large cities did not have significant changes in any of the achievement levels compared to 2013. Since 2003, the nation had significant increases in the percentage of students scoring at the Proficient and Advanced levels and corresponding decreases in the percentage of students at Basic and below Basic. The achievement level changes in large cites have the same pattern since 2003.
Turning to more detailed results for the TUDA districts, Grade 8 mathematics scores increased from 2013 in one district, Chicago. Scores in sixteen districts did not change significantly and scores declined in three districts (Dallas, Hillsborough County, and Houston). Score changes by racial/ethnic groups varied across the districts. In Chicago scores for White and Hispanic students improved. The districts that experienced score decreases in grade 8 mathematics, also showed declines for some student groups as well. Dallas, Hillsborough County (FL), and Houston, which all had overall declines, saw Hispanic students' performance decline as well.
Reading Scores for Large Cities and TUDAs
Now let us move onto the reading results for TUDAs. Again, we will first look at the performance of large city schools with the nation's public schools. At grade 4 in 2015, the reading scores for both the nation and large city schools were unchanged from 2013. However, the gap between large cities and the nation (public) has narrowed from 15 points to 8 points from 2003 to 2015.
Large city schools have been making greater gains since 2003 compared to the nation, closing the gap in performance between the two. Five districts made gains greater than the nation since 2003 and two of those districts, Chicago and DC, experienced gains greater than large cities.
Another way to measure the progress since 2003 is to look at the achievement levels. In fourth grade reading, both the nation and large cities have had increases in the percentages of students at the Advanced, Proficient, and Basic levels in grade 4 reading. Since 2013, neither the nation nor large cities showed significant changes in the percentage of students in NAEP's achievement levels.
At grade 4, four districts experienced increases in average reading performance from 2013 to 2015— Boston, Chicago, Cleveland and the District of Columbia. Fifteen districts did not change significantly, and 1 district, Baltimore City, declined.
Of the five districts with score changes in grade 4 reading between 2013 and 2015, the only score changes by race/ethnicity in these districts were a decrease in for White student scores in Baltimore City, and increases for Black students in Boston and the District of Columbia.
Now let us look at eighth grade reading results. Large city schools made greater gains since 2003 compared to the nation, closing the gap in performance between the two from 12 points in 2003 to 7 points in 2015. During this same period, four districts made greater gains than the nation (public) and only one district, Los Angeles, made greater gains than the large cities.
At grade 8, the achievement level story is slightly different than grade 4 reading. Both the nation and large cities have had increases in the percentage of students scoring at Proficient since 2003. However, since 2013, the nation had a decrease in the percentage of students at or above Proficient, while large cities have not had statistically significant changes at this level.
In grade 8 reading, the national scores declined by two points since 2013, a statistically significant change, while large cites had no statistically significant change over the two-year period. In grade 8, Miami-Dade improved in reading by 6 points since 2013, there was no change in 16 districts, and 3 districts declined, Albuquerque, Baltimore City, and Hillsborough County.
In the four districts with score changes in eighth grade reading between 2013 and 2015, both Black and Hispanic students improved in Miami-Dade; White and Hispanic students saw declines in Albuquerque, Black students' performance declined in Baltimore City, and Hispanic students' performance declined in Hillsborough County.
This concludes my summary of the 2015 NAEP results in mathematics and reading. We have much more information on our website through our interactive Report Card. Detailed data are easily available through the NAEP Data Explorer, and released NAEP assessment questions are available on the NAEP Questions Tool. These and other tools are available via our website at https://nces.ed.gov/nationsreportcard/about/naeptools.aspx. I would like to thank the schools and students that participated in the 2015 assessments for their time and effort.