Skip Navigation
Search

Dr. Gary W. Phillips
Acting Commissioner of the National Center for Education Statistics

THE RELEASE OF THE NATIONAL ASSESSMENT OF EDUCATIONAL PROGRESS (NAEP) FOURTH-GRADE READING 2000
April 6, 2001

Good morning. My name is Gary Phillips, Acting Commissioner of the National Center for Education Statistics. This morning I am releasing findings from Fourth-Grade Reading 2000 for the Center. The findings describe the reading achievement of 4th-grade students across the nation. I will provide data on fourth graders' reading performance in 2000 with comparisons to scores from three prior assessments, conducted in 1992, 1994, and 1998. Jumping ahead just a bit, average scores for fourth-graders have shown no improvement over the past eight years.

About 8,000 4th-graders participated in the 2000 assessment-6,000 public school students and 2,000 nonpublic school students. The 4th-graders read complete texts from typical grade-specific sources. They read two types of texts, representing two different purposes for reading-reading for literary experience and reading to gain information. The students then answered multiple-choice questions and open-ended questions that required written answers.

Before I present the results, let me explain how student reading performance is reported. Student reading performance is reported in two ways: 1) average scale scores, and 2) achievement levels. The average scale score reflects the overall reading performance of a particular group or subgroup of students. Results are also reported according to three reading achievement levels established by the National Assessment Governing Board, which establishes policy for NAEP. These achievement levels are intended to describe standards for what students should know and be able to do. The three achievement levels are defined as Basic, Proficient, and Advanced. NAGB believes that all students should be performing at or above the Proficient level.

I also need to make a point about making comparisons with NAEP data. Because NAEP scores are based on samples, there is a margin of error associated with each score. When we make comparisons between scores, or compare the percentages of students at the various achievement levels, we must test the differences to see if they are larger than the margin of error involved-differences that we call statistically significant. NCES only discusses significant differences.

Results

Overall Sample
Average Scale Scores. Overall, reading scale scores for the nation's fourth graders have not changed, with scores of 217 in both 1992 and 2000. However, there has been some movement in scores for certain groups. When we look at scores for high-performing students, those at the 75th and 90th percentiles, we see increases from 1992 to 2000. The 75th percentile score rose from 242 in 1992 to 245 in 2000, while the 90th percentile score rose from 261 in 1992 to 264 in 2000. At the same time, scores for the lowest-performing students, those at the 10th percentile, declined from 1992 to 2000, as the 10th percentile score dropped from 170 to 163.

Achievement Levels. If we look at Achievement Levels, we see a similar pattern. In 2000, 37% of fourth graders were below Basic in reading, while 63% were at or above Basic. Thirty-two percent were at or above Proficient, and 8% were Advanced. The percentages for at or above proficient and Advanced were higher in 2000 than 1992 (29% at or above Proficient and 6% at Advanced), reflecting the improvement in scores for high-performing students.

Race/Ethnicity
Average Scale Scores. If we look at scale scores for racial and ethnic subgroups, we see that only Asian/Pacific Islanders increased from 1992 to 2000. Scores for whites, blacks, Hispanics, and American Indians remained unchanged for this time period. At the same time, scores in 2000 for whites (226) and Asian/Pacific Islanders (232) were higher than scores of blacks (193), Hispanics (197), and American Indians (196). In addition, the gap in performance between whites and blacks and the gap between whites and Hispanics has not changed between 1992 and 2000.

Achievement Levels. We see similar patterns when we look at achievement level results for these subgroups. Only Asian/Pacific Islanders showed any changes, with an increase in the percentage at or above Proficient, from 25% in 1992 to 46% in 2000. We also see that the percentages at or above Proficient for whites (40%) and Asian/Pacific Islanders (46%) were higher than for blacks (12%), Hispanics (16%), and American Indians (17%).

A Changing Population. The data from the 2000 assessment show how the makeup of the nation's fourth-grade population has changed since 1992. The percentage of white fourth-grade students has fallen from 71% of the 4th-grade population in 1992 to 66% in 2000. Similarly, the black fourth-grade population has decreased from 16% to 14%, while the Hispanic population has grown from 9% to 15%. The 2000 population shares for Asians/Pacific Islanders (3%) and American Indians (2%) remain unchanged from 1992.

Gender
Average Scale Scores. Scores for male and female students remained relatively stable over time. However, the gap between the two groups did increase from 1998 to 2000, from 6 scale points in 1998 to 10 scale points in 2000. (1998 scores: 214 and 220; 2000 scores: 212 and 222.) Female students have traditionally scored higher than males in reading.

Achievement Levels. We see a similar pattern when we look at Achievement Level results, with a greater percentage of females at or above Proficient (36%) than males (27%). In addition, the percentage of females at or above Proficient increased from 32% in 1992 to 36% in 2000.

Public/Nonpublic
Average Scale Scores. There was no measurable change over time for public or nonpublic schools. 2000 Scores for nonpublic school students (234) remained higher than scores for public school students (215).

School and Home Factors Related to Reading Performance
The Fourth-Grade Reading 2000 assessment identified several encouraging patterns of behavior:
  • A higher percentage of the 4th graders reported reading at least 11 pages a day in school and for homework in 2000 as compared to both 1992 and 1994 (60% as compared to 56 % and 54% respectively). Students who read 11 or more pages a day had higher scores than those who did not, both in 2000 and in prior assessments.
  • The percentage of students who reported that they were never assigned homework declined, from 16% in 1992 and 13% in 1994 to 10% in 2000, while the percentage who reported spending half an hour a day on homework increased. Students who said they were never assigned homework had lower scores than those who reported spending either a half hour or an hour a day on their homework. Seventy-two percent of the students said they spend one half to one hour a day on homework, and these students scored the highest.
  • The percentage of students who reported watching no more than one hour of television per day increased in 2000 as compared to both 1992 and 1994-25% as compared to 19% for both prior assessments. These students had higher scores than those who watched four or more hours a day.
  • Eighty-one percent of students said they write long answers to questions on tests or assignments that involved reading at least once a month. These students score higher than their peers who engage in this activity less frequently.
  • Three quarters of students said they read for fun at least weekly, and 43% said they read for fun every day. Students who read for fun every day scored highest. At the same time, the percentage of fourth-graders who said that they never or hardly ever read for fun on their own time increased from 12% in 1994 to 14% in 2000. These students have lower reading scores than other students.
  • Seventy-six percent of students said they talk about reading with family or friends at least monthly. However, the percentage who said that they never or hardly ever talk about reading with family or friends also increased, from 21% in 1994 to 24% in 2000. These students have lower reading scores than other students.
  • The percentage of fourth-graders who said they had four types of reading materials in their homes (newspapers, magazines, books, and encyclopedias) declined, from 38% in 1994 to 34% in 2000. These students have higher reading scores than other students. However, 68% of students say they have at least three types of reading in the home, and these students have higher scores than those students who say they have two types of reading or less at home.
Accommodated/Non-Accommodated
NCES assessed two separate samples in the 2000 reading assessment. One assessment allowed accommodations for students with disabilities and limited English proficiency (LEP) students who needed them and one did not. NCES used two separate samples to obtain an understanding of what effect accommodations had on performance and so that we would have a sample whose assessment results could be compared with results from previous assessments, which did not allow accommodations. The "no accommodations" sample included scores from 7,914 students and the "with accommodations" sample included scores from 8,074 students. All 2000 data presented so far has been based on the "no accommodations" sample.

Fifteen percent of the 4th-graders who were selected for the assessment without accommodations were identified as either students with disabilities or LEP students. About half of these students, amounting to 7 percent of all the students in the sample, were excluded from the assessment. Seventeen percent of the students who were selected for the assessment with accommodations were identified as either students with disabilities or LEP students. About a third of these students, amounting to 6 percent of all the students in the sample, were excluded, because they could not be accommodated in the NAEP reading assessment.

Average Scale Scores. The group that was assessed with accommodations had a lower average score, 215 compared to 217 for the group that was assessed without accommodations. Hispanics assessed in the "with accommodations" sample had a lower average score (190) than those who were assessed without accommodations (197). "Accommodated/non-accommodated" comparisons for other subgroups showed no differences. American Indians in the "accommodated" sample had a higher score than Hispanics in the "accommodated" sample (201 as compared to 190).

One possible reason for the decline in Hispanic scores for the "with accommodations" sample is that NCES treated the reading assessment as an "English language reading assessment." Accommodations such as translations and use of foreign language dictionaries were not allowed, and this may have affected the performance of LEP students.

Achievement Levels. "Accommodated/non-accommodated" comparisons showed no differences in the percentage at or above Proficient for any groups or subgroups.

Additional Information

Fourth-Grade Reading 2000, the full report, is available on the web, along with additional information, including sample questions and student answers and instructional information taken from surveys of fourth-grade teachers and elementary school principals in connection with the assessment.