Chapter 4
Testing and Achievement

STANDARDS AND ASSESSMENTS

Content standards provide a framework for the knowledge and skills that students are expected to acquire. Performance standards determine how well students should be able to perform relative to the content standards. Assessments provide information regarding the attainment of standards. Elementary, secondary, and special education programs rely on state standards-based assessment systems to evaluate the effectiveness of federal programs. According to federal expectations for Title I (the federal program that helps states and schools meet special education needs), academic standards must be rigorous and exceed minimum competencies. They must be fair, valid, and reliable, and include all students. Assessment results should be reported for schools and districts, and they must include demographic categories (gender, race and ethnicity, English proficiency, disability, migrant status, and low income status). Title I legislation specifies that performance standards must provide information for at least three levels of performance.

NATIONAL ASSESSMENT OF EDUCATIONAL PROGRESS

The National Assessment of Educational Progress (NAEP) has conducted assessments of American students since 1969. The federally funded NAEP is the only regular national survey of educational achievement at the elementary, middle, and high school levels. It is authorized by Congress and administered by the National Center for Education Statistics (NCES). The Augustus F. Hawkins—Robert T. Stafford Elementary and Secondary School Improvement Amendments of 1988 (P.L. 100-290) established the National Assessment Governing Board (NAGB) to formulate policy guidelines for NAEP. The NAGB determines which subjects will be assessed and how they will be assessed.

According to the U.S. Department of Education, the NAEP has two major goals: to discover what American students know and can do in key subject areas and to measure educational progress over long periods of time. Designed as a measure of the nation's educational system, the NAEP is a series of reading, writing, mathematics, science, history, civics, and geography tests. The tests are given periodically to randomly selected samples of youth ages nine (grade four), thirteen (grade eight), and seventeen (grade twelve) attending both public and private schools. Student performance in all grade levels is measured on a proficiency scale of zero to five hundred. This allows a comparison of younger students with older ones, as well as an assessment of progress from year to year. Table 4.1 shows the schedule from 1990 to 2010 for assessing individual subjects and releasing long-term trend data. (See Table 4.1.)

Beginning with the 1990 assessments, the NAGB also developed achievement levels for each subject at each grade level in an effort to measure the match between students' actual achievement and their desired achievement. A panel of teachers, education specialists, and other members of the general public categorized these levels into basic, proficient, and advanced. Basic level was defined as "partial mastery of prerequisite knowledge and skills that are fundamental for proficient work at each grade." Proficient level was defined as "solid academic performance for each grade assessed. Students reaching this level have demonstrated competency over challenging subject matter, including subject-matter knowledge, application of such knowledge to real-world situations, and analytical skills appropriate to the subject matter." The advanced level was defined as "superior performance." Achievement levels provide another way to report assessment results, allowing comparisons between percentages of students who achieve a particular level on one assessment with the percentage who achieve that level the next time that subject is assessed. They are also used to make comparisons between states and the nation.

READING PERFORMANCE

The ability to read is fundamental to virtually all aspects of the education process. If students cannot read well, they usually cannot succeed in other subject areas. Eventually, they may have additional problems in a society that requires increasingly sophisticated job skills.

The NAEP assesses proficiency in three different types of reading: reading for literary experience, reading for information, and reading to perform a task. The first type of reading involves various genres of creative writing, including fiction, poetry, drama, biography, myths, legends, and folktales, and students are asked to explore such literary features as characters, plots, and themes. "Reading for information" assesses students' ability to use magazines, newspapers, textbooks, and essays to gain knowledge about the world. The third type of reading assessment addresses students' capacity to use what they learn from such informational sources as train schedules, instruction manuals, game directions, and maps. This component is not part of fourth-grade assessments. Overall, the NAEP reading assessments examine students' ability to understand and interpret the material they read, to make connections between the material and their own experiences, and to evaluate the structure and content of what they have read.

NAEP reading performance is described in terms of the percentage of students attaining the three achievement levels—basic, proficient, and advanced. A 300 score indicates relative proficiency in understanding complicated literary and informational material.

According to the National Center for Education Statistics, long-term trend data from 1971 to 2004 indicate that the average reading scale score for nine-year-olds increased from 208 in 1971 to 219 in 2004. Reading scores of nine-year-olds went up from 212 in 1999 to 219 in 2004. The reading scores of thirteen-year-olds went up four points between 1971 and 2004, from 255 to 259, but the score stayed the same between 1999 and 2004 (259). For seventeen-year-olds, reading scores were the same in 2004 as in 1971 (285), and showed a three-point decrease after 1999, when the average reading scale score was 288. (See Figure 4.1.)

Prior to 1998, administration procedures for the NAEP reading assessment did not permit the use of accommodations such as extra time or individual (rather than group) administration for special needs students. In 1998 administration procedures were introduced that allowed students with disabilities and limited English proficiency to use such accommodations.

The NAEP defines an accommodation as a change in testing conditions that removes barriers to participation but does not alter what is being tested. Examples of accommodations include extended time to complete the test, testing in small group or one-on-one sessions, use of a scribe to write the student's answers, and reading the instructions aloud. Bilingual test booklets can be used for subjects other than reading assessments.

Long-Term Trend Reading Scores of Nine-Year-Olds

Scores of nine-year-olds increased for all ethnic groups and both genders between 1971 and 2004, and the average reading score of nine-year-olds in 2004 was higher than in any previous assessment. The achievement gap between scores of white and African-American nine-year-olds decreased between 1971 and 2004. Between white and Hispanic students at age nine, the achievement gap decreased from 1975 to 1999, but it stayed the same from 1999 to 2004. (See Table 4.2.) On average, females scored higher than males in reading in 2004. Among nine-year-olds, the gender gap between females and males decreased from thirteen points in 1971 to five points in 2004. The score gap between white and African-American students decreased eighteen points from 1971 to 2004 and nine points between 1999 and 2004.

Long-Term Trend Reading Scores of Thirteen-Year-Olds

Scores of thirteen-year-olds increased for all ethnic groups and both genders from 1971 to 1999, but in 2004 the average scores were not statistically different from those in 1999. The achievement gap between scores of white and African-American students decreased from 1971 to 2004, but has stayed about the same since 1999. The gap between white and Hispanic thirteen-year-olds stayed about the same from 1975 to 2004. (See Table 4.3.) Among thirteen-year-olds, there was only a small change in the gender gap between males and females from 1971 to 2004; on average, in 1971 females scored eleven points higher than males in reading, in 1999 females scored twelve points higher, and in 2004 females scored ten points higher than males in reading. The score gap between white and African-American students at age thirteen decreased seventeen points from 1971 to 2004.

Long-Term Trend Reading Scores of Seventeen-Year-Olds

On average, scores of all seventeen-year-olds showed no statistically significant difference from 1971 to 2004. However, scores of African-American students increased from 1971 to 1999. For Hispanic students they increased from 1975 to 1999, and for both groups they stayed about the same from 1999 to 2004. The achievement gap between scores of white and African-American students decreased from 1971 to 1999. They also decreased between white and Hispanic students from 1975 to 1999, but there was no significant difference in scores for either group from 1999 to 2004. (See Table 4.4.) Among seventeen-year-olds, there were only small differences in the gender gap between males and females from 1971 to 2004; on average, in 1971 females scored twelve points higher than males in reading, in 1999 females scored thirteen points higher, and in 2004 females scored fourteen points higher than males in reading. The score gap between white and African-American students at age seventeen decreased twenty-four points from 1971 to 2004.

Early Childhood Factors Associated with Reading Performance

According to the National Center for Education Statistics, children whose family members read to them eventually demonstrate higher reading performance and do better in school. In 2001–02, one-third (33%) of children about nine months of age were read to by a relative on a daily basis; even fewer (27%) were told stories. These numbers compared unfavorably with those for other activities, including nearly three-quarters (74%) who were sung to, 68% who played peek-a-boo, 64% who were taken on errands, and nearly half (47%) who played outside every day. (See Table 4.5; while totals in Table 4.5 reflect percentages for children nine months of age, children were assessed as young as six months and as old as twenty-two months.)

More nine-month-old white children (41%) than African-American (23%), Hispanic (21%), Asian/Pacific Islander (26%), or Native American (18%) children were read to every day. Nearly half (48%) of children whose mothers have a bachelor's degree or higher were read to daily. Poor children (22%) were less likely to be read to than nonpoor children (36%). In terms of other literary activities, three-quarters (75%) of white children were sung to every day, which was not much different from African-American (73%), Hispanic (70%), or Asian/Pacific Islander (71%). About 64% of Native American children were sung to every day. Children ages fourteen to twenty-two months were more likely to be read stories every day than younger children. Nearly half (44%) of children in this age group were read stories on a daily basis. (See Table 4.5.)

Another way to look at the early development of children is by the number of family risk factors. Risk factors include living in a household that is below the poverty level, having a language other than English spoken as the primary language at home, having a mother whose highest education was less than a high school diploma or equivalent, and living in a single-parent household. According to the National Center for Education Statistics, 41% of nine-month-old children who had no family risk factors were read to on a daily basis, while 25% of those with one risk factor and 20% of those with two or more risk factors were read to daily in 2001–02. (See Figure 4.2.)

Homework Habits and Reading Performance

The NAEP asked students background questions about their school and home experiences. One question asked was how much time they spent on homework the day before. Possible responses included that no homework was assigned; that homework was assigned but the student did not do it; less than one hour was spent on homework; one to two hours were spent on homework; or more than two hours were spent on homework. The National Center for Education Statistics then analyzed the responses of students by age, and how they scored on the assessment. Nine-year-old students in 2004 who did not have homework assigned the day before had average reading scores of 217; those who had homework but did not do it had average scores of 204; students who spent less than one hour on homework or one to two hours on homework had average scores of 221; and those who spent more than two hours on homework had average scores of 204.

For thirteen-year-olds, the 2004 average reading scale scores for those who did not have homework assigned the day before was 248; students who had homework but did not do it scored 245; those who spent less than one hour on homework scored 261 on average; students who spent one to two hours on homework the day before scored 268; and those who spent more than two hours on homework achieved an average score of 272.

Among seventeen-year-olds in 2004, students who did not have homework the day before scored 270, on average, in reading. Those who had homework assigned but did not do it scored 279. Students who spent less than one hour on homework the day before scored 287, and those who spent one to two hours on homework scored 295. The average reading scale scores of students who spent more than two hours on homework the day before were 304 in 2004.

Daily Reading and Reading Performance

Students at all three ages were asked about the number of pages per day they read in school and for homework. According to the National Center for Education Statistics, nine-year-olds who reported reading five or fewer pages per day had average reading scale scores of 211 in 2004. The score was 220 for students who read six to ten pages per day, 222 for students who read eleven to fifteen pages per day, 223 for students who read sixteen to twenty pages per day, and 222 for students who read more than twenty pages per day. (See Figure 4.3.)

Among thirteen-year-olds, the average reading scale score in 2004 for students who read five or fewer pages per day was 249. For students who read six to ten pages per day, the score was 260. It was 262 for students who reported reading eleven to fifteen pages or sixteen to twenty pages per day, and 263 for students who read more than twenty pages per day. (See Figure 4.3.)

In 2004 seventeen-year-olds who read five or fewer pages per day had average reading scale scores of 268. The scores were 282 for students who read six to ten pages per day, 287 for those who read eleven to fifteen pages per day, 293 for students who reported reading sixteen to twenty pages per day, and 297 for students who read more than twenty pages per day. (See Figure 4.3.)

Students were asked how often they read for fun. According to the National Center for Education Statistics, among nine-year-olds, those who read for fun almost every day had average reading scale scores of 220 in 2004. Those who reported reading for fun once or twice per week scored 224 on average. The scores were 216 for students who read for fun once or twice per month, 209 for those who read for fun a few times per year, and 203 for students who reported that they never or hardly ever read for fun. (See Figure 4.4.)

The average reading scale score was 271 for thirteen-year-olds who reported reading for fun almost every day. The score was 261 for students who read for fun once or twice per week, 256 for those who read for fun once or twice per month, and 236 for students who never or hardly ever read for fun. Too few students responded that they read for fun only a few times per year for a reliable estimate of the average reading scale score to be reported for this group. (See Figure 4.4.)

For seventeen-year-olds who reported reading for fun almost every day the average reading scale score was 305 in 2004. It was 288 for students who reported reading for fun once or twice a week, 287 for those who read for fun once or twice a month, 272 for those who read for fun a few times per year, and 268 for students who never or hardly ever read for fun. (See Figure 4.4.)

MATHEMATICS PERFORMANCE

Since 1973 the NAEP has assessed the mathematics performance of nine-, thirteen-, and seventeen-year-olds. The 2005 NAEP mathematics framework specifies five content areas to be assessed: number properties and operations; measurement; geometry; data analysis and probability; and algebra. According to the National Center for Education Statistics, on the 2003 mathematics assessment, 39% of all grade eight students scored at the basic level, 23% at the proficient level, and 5% at the advanced level. The remaining students were categorized as below basic level.

Long-term trend data for 1973 to 2004 from the National Center for Education Statistics indicate that the average mathematics scale scores for nine-year-olds increased from 219 in 1973 to 241 in 2004. Between 1999 and 2004 mathematics scores of nine-year-olds went up nine points from 232 to 241. The mathematics scores of thirteen-year-olds rose fifteen points between 1973 and 2004, from 266 to 281, with an increase of five points from 1999 to 2004 (276 to 281). For seventeen-year-olds, mathematics scores were not much different in 2004 (307) from the 1973 level (304). However, the 2004 scores represented a slight decline from 1999, when scores for seventeen-year-olds had reached an all-time high of 308. (See Figure 4.5.)

Since 1996 schools have been encouraged to make testing accommodations on the mathematics NAEP for students with disabilities or limited English proficiency. According to the National Center for Education Statistics, average mathematics scores increased for grade four students from 213 in 1990 to 235 in 2003. In 1996 students in grade four who had testing accommodations scored the same as students who did not have testing accommodations (224). In 2000 grade four students with accommodations scored slightly lower (226) on average than students who did not have accommodations (228). Among students in grade eight, scores increased from 263 in 1990 to 278 in 2003. In 1996 students with accommodations scored 270, while those without accommodations scored 272. There was a two-point difference between the scores again in 2000, when students with accommodations scored 273, and those without accommodations scored 275. Since there were no significant differences between scores of students with and without accommodations, after 2000 the data for all students was reported in only a single category, with accommodations. (See Figure 4.6.)

Long-Term Trend Mathematics Scores of Nine-Year-Olds

According to the National Center for Education Statistics, scores of nine-year-olds increased for all ethnic groups and both genders from 1973 to 1999 and from 1999 to 2004. The achievement gap between mathematics scores of white and African-American nine-year-olds narrowed between 1973 and 1999, but it stayed about the same from 1999 to 2004. Between white and Hispanic students at age nine, the achievement gap in mathematics stayed about the same from 1975 to 1999, but it decreased from 1999 to 2004. (See Table 4.2.) Although there has been a small score gap since 1973 and a three-point gap in the 2004 mathematics scores of male and female nine-year-olds, it was not considered statistically significant. The score gap between white and African-American students decreased twelve points from 1973 to 2004.

Long-Term Trend Mathematics Scores of Thirteen-Year-Olds

The National Center for Education Statistics reports that the scores of thirteen-year-olds increased for all ethnic groups and both genders from 1973 to 1999, as well as from 1999 to 2004. The achievement gap between mathematics scores of white and African-American students decreased from 1973 to 1999 but stayed about the same between 1999 and 2004. The gap between white and Hispanic thirteen-year-olds also decreased from 1975 to 1999 and stayed about the same from 1999 to 2004. (See Table 4.3.) Among thirteen-year-olds, there was only a small change in the gender gap from 1973 to 2004; on average, in 1973 females scored two points higher than males in mathematics; in 1999 males scored two points higher; and in 2004 males scored four points higher than females in mathematics. The mathematics score gap between white and African-American students at age thirteen narrowed nineteen points between 1973 and 2004.

Long-Term Trend Mathematics Scores of Seventeen-Year-Olds

According to the National Center for Education Statistics, on average, mathematics scores of seventeen-year-olds showed no statistically significant difference from 1973 to 2004. However, scores of female, white, and African-American students increased between 1973 and 1999, and scores of Hispanic students improved from 1975 to 1999. Between 1999 and 2004 there were no statistically significant changes in scores for any group. The achievement gap between mathematics scores of white and African-American students decreased from 1973 to 1999, but it stayed about the same from 1999 to 2004. The gap also decreased between white and Hispanic students from 1975 to 1999, but it showed no significant difference from 1999 to 2004. (See Table 4.4.) Among seventeen-year-olds, the mathematics gender gap between males and females decreased five points from 1973 to 1999; on average, in 1973 males scored eight points higher than females in mathematics, while in 1999 males scored three points higher. No change was seen in the mathematics gender gap between 1999 and 2004. The score gap between white and African-American students at age seventeen decreased twelve points from 1973 to 2004.

Mathematics Instruction and Math Performance

Students taking the mathematics NAEP were asked several background questions related to course taking, homework, television watching, and computer use. Each of these factors was analyzed to see how it related to performance in mathematics.

At age thirteen, students were asked what kind of mathematics class they were in that year. According to the National Center for Education Statistics, thirteen-year-olds who were in regular mathematics had average scale scores of 269 in 2004. Those in pre-algebra scored 284, and those in algebra scored 296. Students who were taking another mathematics course scored 279. The sample of students who reported that they were not taking mathematics was too small to provide a reliable estimate of an average score. (See Figure 4.7.)

At age seventeen, students were asked whether they were currently taking, or had ever taken, any course in a list of mathematics courses. Students who were taking or had taken pre-algebra (or less) as their highest-level mathematics course scored 270 on average in 2004. According to the National Center for Education Statistics, those who were taking or had taken first-year algebra as their highest course scored 282, and those for whom geometry was the highest mathematics course scored 296. Second-year algebra students scored 310, while students who had taken or were taking calculus scored 336. The sample of students who reported that they had taken another course, such as trigonometry or general mathematics, was too small to provide a reliable estimate of an average score. (See Figure 4.8.)

Students age seventeen were asked about the frequency with which they did mathematics homework. According to the National Center for Education Statistics, students who reported doing mathematics homework often had average scale scores of 312 in 2004. Among those reporting that they did mathematics homework sometimes, the score was 296, and students who never did math homework scored 289 on average. (See Figure 4.9.)

At ages thirteen and seventeen, students were asked about their use of computers for mathematics work. According to the National Center for Education Statistics, thirteen-year-olds who reported receiving computer instruction had average scale scores of 283 in 2004, while those who did not receive computer instruction scored 280. Students age thirteen who had access to a computer in school for mathematics scored 283, and those who did not have access in school scored 282. Those who used a computer to solve mathematical problems scored 283, while those who did not use a computer scored 278 on average in 2004. (See Figure 4.10.)

According to the National Center for Education Statistics, seventeen-year-olds who reported receiving computer instruction had average mathematics scale scores of 307, which was the same as for those who did not receive computer instruction in 2004. Students at age seventeen who had access to a computer in school for mathematics scored 308, and those who did not have access in school scored 303. Those who used a computer to solve mathematical problems scored 309, while those who did not use a computer scored 303 on average in 2004. (See Figure 4.10.)

Television Viewing Habits and Mathematics Scores

Students at all three ages (nine, thirteen, and seventeen) were asked how much television they watched every day. According to the National Center for Education Statistics, among nine-year-olds, students who reported watching zero to two hours of television per day had average mathematics scale scores of 244 in 2004. Those who watched three to five hours of television daily scored 245, while those who reported watching six or more hours of television every day had scores of 229 on average. (See Figure 4.11.)

According to the National Center for Education Statistics, thirteen-year-olds who reported watching zero to two hours of television per day had average mathematics scale scores of 288 in 2004. Those who watched three to five hours of television on a daily basis scored 279, while students who reported watching six or more hours of television every day had scores of 264 on average. (See Figure 4.11.)

According to the National Center for Education Statistics, seventeen-year-olds who reported watching zero to two hours of television per day had average mathematics scale scores of 313 in 2004. Those who watched three to five hours of television daily scored 300, while students who reported watching six or more hours of television every day had scores of 286 on average. (See Figure 4.11.)

SCIENCE PERFORMANCE

Since 1977 the NAEP has assessed the science performance of students at age nine, thirteen, and seventeen. According to the NAEP Web site (http://nces.ed.gov/nationsreportcard/science/whatmeasure.asp), the science assessment addresses students' knowledge of facts in areas of earth, physical, and life sciences and measures their ability to "use the tools, procedures, and reasoning processes of science to develop an increased understanding of the natural world." The most recent results that have been released reflect the assessments conducted in 2000.

In general, the 2000 NAEP assessments indicated that science performance has remained stable since 1996. High-performing eighth-grade students increased their science scale scores between 1996 and 2000, while those of middle-performing twelfth-graders decreased. NAEP science assessments were again conducted in 2005, with results to be made public in 2006.

According to the National Center for Education Statistics, average scores on the NAEP science assessment were not significantly different overall in 2000 from 1996 levels. However, Native American students in all three grades and twelfth-grade white students had lower scores in 2000 than they did in 1996. Eighth-grade male students' average science score was higher in 2000 than it was in 1996. Twelfth-grade males and females scored lower in 2000, on average, than they had in 1996.

WRITING PERFORMANCE

In 2002 NAEP administered a writing assessment to students in grades four, eight, and twelve. Students were asked to write narrative, informative, and persuasive essays based on a variety of materials, including photographs, cartoons, and poems. They were asked to draft, revise, and edit their work, which was measured at three achievement levels: basic, proficient, and advanced. The basic level means partial mastery of the knowledge and skills that are fundamental for proficient work at a given grade. The proficient level represents solid academic performance; students reaching this level demonstrate competency in challenging subject matter. The advanced level indicates mastery of both the basic and proficient levels and represents superior performance. The next writing assessment is scheduled for 2007. (See Table 4.1.)

Writing scores at each grade level ranged from 0 to 300. Basic level achievement scores in grade four ranged from 115 to 175; proficient level ranged from 176 to 224; and advanced level scores ranged from 225 to 300. Basic level achievement scores in grade eight ranged from 114 to 172; proficient level scores from 173 to 223; and advanced level scores from 224 to 300. Basic level achievement scores in grade twelve ranged from 122 to 177; proficient level scores from 178 to 229; and advanced level scores from 230 to 300. Performance is scaled separately, so average scale scores cannot be compared across grades.

Comparisons between 1998 and 2002

According to the National Center for Education Statistics, all the 2002 writing results for grade four, and some of the eighth and twelfth grade results, were significantly different from the 1998 assessment. The proportion of grade four students who scored below the basic level was 16% in 1998, while in 2002 it had dropped to 14%. (See Table 4.6.) Another 61% of fourth-graders scored at the basic level of writing achievement in 1998, while 58% were at this level of achievement in 2002. Proficient students increased from 22% in 1998 to 26% in 2002, and advanced students increased from 1% in 1998 to 2% in 2002. Among eighth-grade students, 58% scored at the basic level in 1998; this decreased to 54% in 2002. Students writing at the proficient level increased from 25% in 1998 to 29% in 2002, and advanced students increased from 1% in 1998 to 2% in 2002. The proportion of grade twelve students at the basic level decreased from 57% in 1998 to 51% in 2002. Students in grade twelve scoring at the advanced level increased from 1% in 1998 to 2% in 2002. The only statistically significant change that showed decreased writing proficiency was between the proportion of seniors scoring below basic in 1998 (22%) and in 2002 (26%).

Results by Race/Ethnicity

According to the National Center for Education Statistics, fourth-grade Asian/Pacific Islander students had the highest average writing scores of all racial and ethnic groups in 2002 at 167, followed by white students (161), Hispanic students (141), and African-American students (140). At grade eight, white and Asian/Pacific Islander students both scored an average of 161 on the assessment, followed by Hispanic students (137) and African-American students (135). White twelfth-graders scored highest among racial and ethnic groups at that level in 2002 (154), followed by Asian/Pacific Islander students (151), Hispanic students (136), and African-American students (130). The National Center for Education Statistics indicated that average scores were not available for Native American/Alaska Native students because the sample size was too small to ensure the accuracy of results.

Results by Gender

At all three grade levels, females had higher average writing scores than males in 2002. Fourth-grade females scored an average 163, compared to 146 for males. Eighth-grade girls scored 164, compared to 143 for boys, and twelfth-grade girls scored 160, compared to 136 for boys. Average writing scale scores increased five points for fourth-grade girls between 1998 and 2002 and four points for fourth-grade boys. Writing scores for eighth-graders also showed improvement, with girls' scores rising four points between 1998 and 2002 and boys' scores increasing an average of three points. However, average writing scale scores did not show improvement among twelfth-graders between 1998 and 2002. The average score for female students edged up slightly from 159 to 160, but the scores for male students declined by four points from 140 to 136.

GEOGRAPHY PERFORMANCE

In 2001 NAEP administered a geography assessment to students in grades four, eight, and twelve, measuring performance at three achievement levels: basic, proficient, and advanced. The Geography NAEP measures students' knowledge, understanding, and ability to apply knowledge about "space and place"; "environment and society"; and "spatial dynamics and connections." For example, students must know where the world's largest tropical rain forest is located, understand why it is located near the equator, and apply that knowledge in support of the conclusion that tropical rain forests promote wide species variation.

Comparisons between 1994 and 2001 Scores

According to the National Center for Education Statistics, at grade four the percentage of students performing at the basic or proficient levels in geography increased from 70% in 1994 to 74% in 2001. The percentage of eighth-grade students performing at or above the basic level increased from 71% in 1994 to 74% in 2001. The percentage of twelfth-graders achieving at least a basic level of proficiency rose slightly from 70% in 1994 to 71% in 2001.

At all three grade levels, white students achieved the highest average scores in 2001: white fourth-graders averaged 222, eighth-graders averaged 273, and twelfth-graders averaged 291. Asian/Pacific Islander students had the second-highest averages at fourth grade (212) and eighth grade (266), but were edged out by Native American students in twelfth grade (286 and 288). Hispanic students had higher average scores than African-American students in all three grades. However, the average scores of African-American fourth-graders improved dramatically from 168 in 1994 to 181 in 2001 and almost closed the gap with Hispanic fourth-graders, who averaged 184.

In 2001 male students at all three grade levels had higher average scale scores than females, but there were no significant changes in the scores of males and females from 1994 to 2001. The next geography assessment is scheduled for 2010. (See Table 4.1.)

U.S. HISTORY PERFORMANCE

In 2001 NAEP administered a U.S. history assessment to students in grades four, eight, and twelve, measuring performance at three achievement levels: basic, proficient, and advanced. According to the NAEP Web site (http://nces.ed.gov/nationsreportcard/ushistory/whatmeasure.asp), the test focused on several themes in American history, including:

1. "Change and continuity in American democracy: ideas, institutions, practices, and controversies;
2. The gathering and interactions of peoples, cultures, and ideas;
3. Economic and technological changes and their relation to society, ideas, and the environment; and
4. The changing role of America in the world."

The NAEP U.S. history assessments evaluated students' knowledge and understanding of historical events, concepts, and movements and measured their ability to analyze and interpret historical issues, patterns, and relationships.

Comparisons between 1994 and 2001 Scores

According to the National Center for Educational Statistics, in grade four the percentage of students performing at the basic level or above increased from 64% in 1994 to 67% in 2001, but there were no significant differences in the percentages performing at the proficient or advanced levels. In grade eight, the percentage of students who performed at the basic level remained steady at 48% between 1994 and 2001, but increases were noted in those achieving the proficient level (from 13% in 1994 to 15% in 2001) and the advanced level (from 1% to 2%). In grade twelve there were no significant changes in performance at any level of proficiency. More than half of twelfth-graders (57%) failed to achieve even the basic level of proficiency. About one-third of twelfth-graders scored in the basic range in 2001, 10% reached the proficient ranking, and 1% were advanced.

Average Scores by Race/Ethnicity

At all three grade levels, white students had higher scale scores than African-American, Hispanic, and Native American students. Asian/Pacific Islanders ranked second behind white students in both the fourth and eighth grades but had the highest average scores among twelfth graders. African-American students achieved significantly higher average scores in 2001 (188) than they had in 1994 (177). The average for white fourth-grade students also showed improvement from 215 in 1994 to 220 in 2001. Eighth-grade white students scored a higher average in 2001 (271) than they had in 1994 (267), and twelfth-grade Hispanic students increased their average scores significantly in 2001 (274) from 1994 (267).

Results by Gender

Male fourth graders improved significantly from 203 in 1994 to 209 in 2001, matching the average reached by female fourth-grade students in 2001. Male students in eighth grade averaged 264 in 2001, which was three points above the average for female students in the same grade. Among twelfth-graders in 2001, male students surpassed the female average, with male students measuring 288 to the female students' 286. The next U.S. history assessment will be in 2006. (See Table 4.1.)

Computer Use and U.S. History Performance

Students and teachers answered questions for the 2001 NAEP U.S. history assessment related to the ways in which they use computers at school in history and social studies classes. According to the National Center for Education Statistics, about one-quarter of students in grade four and one-third of students in grades eight and twelve reported that they use computers for social studies at least once every few weeks. At all three grades, students who used computers every day in class had lower average scale scores than students who used computers less frequently.

INTERNATIONAL COMPARISONS

In 2003 the Trends in International Mathematics and Science Study (TIMSS, formerly known as the Third International Mathematics and Science Study) compared the mathematics and science performance of fourth-graders in twenty-five countries and eighth-graders in forty-five countries around the world. This comparison with students in other countries allows the United States to monitor its progress toward the goal of being first in the world in mathematics and science achievement.

According to the National Center for Educational Statistics, at grade eight U.S. mathematics scores increased from 1995 to 2005, but there was no significant difference in average scores between 1999 and 2003, so the gain was made between 1995 and 1999. The TIMSS indicated that in 2003 U.S. students in grades four and eight scored above the international average in mathematics. On average, grade four students scored higher than students in thirteen countries and lower than students in eleven countries. The average score of U.S. students in grade eight in mathematics was higher than students in thirty countries, but students in thirteen countries outperformed U.S. eighth graders. (See Table 4.7.)

The National Center for Education Statistics reports that in 2003 students in the United States in grades four and eight scored, on average, above the international average in science. At grade four, U.S. students outperformed students in sixteen countries, while students in three countries scored higher than U.S. students. At grade eight, students in the United States outperformed students in thirty-six countries, while students in eight countries scored higher than U.S. eighth graders. (See Table 4.8.)

CARNEGIE UNITS AND HIGH SCHOOL COURSE AVAILABILITY

In response to the recommendations of the National Education Goals Panel (see Chapter 5), many state legislatures, local school boards, and departments of education have attempted to strengthen high school graduation requirements. While state-mandated standards cannot necessarily measure activities in the classroom, they tend to indicate a state's desire to improve its schools.

The District of Columbia and most states have established minimum Carnegie units (one unit equals an academic year course of two semesters) required for high school graduation. According to the Council of Chief State School Officers (CCSSO), as of 2002, thirty-eight states required four credits in English (see Figure 4.12), twenty-five states required 2.5 or more credits in mathematics (see Figure 4.13), twenty-two states required 2.5 or more credits in science (see Figure 4.14), and thirty-six states required 2.5 or more credits in social studies. (See Figure 4.15.)

Table 4.9 shows the average Carnegie units earned by high school graduates in various subject fields in 2000. According to the National Center for Education Statistics, the average number of Carnegie units earned by public high school graduates in 2000 was just over twenty-six. Students earned an average of 4.4 credits in English, 3.8 in history/social studies, 3.6 in mathematics, 3.2 in science, 2 in foreign language, 2 in art, 4.2 in vocational education, and 0.8 in computer science.

In "The 36th Annual Phi Delta Kappa/Gallup Poll of the Public's Attitude toward the Public Schools" (Lowell C. Rose and Alec M. Gallup, Phi Delta Kappan, September 2004), those polled were asked whether they favored or opposed requiring high school students to take four years of English, mathematics, and science to graduate from high school in their community. More than three-quarters (78%) favored this requirement. Twenty percent opposed requiring four years of English, mathematics, and science for high school graduation. (See Table 4.10.)

According to the National Center for Education Statistics, since 1982 the proportion of students taking

advanced coursework in mathematics, English, science, and foreign language has increased. However, advanced courses are not offered in all schools. In 2000 schools in a central city, schools in the Northeast and Southeast, and schools with twelfth-grade enrollment of 450 or more were more likely to offer advanced courses. More than two-fifths (43%) of large schools (twelfth-grade enrollment of 450 or more), 31% of schools in the Northeast, and nearly a third (32%) of schools in central cities offered at least four advanced courses each in mathematics, English, science, and foreign language. (See Figure 4.16.)

GRADUATION REQUIREMENTS AND HIGH SCHOOL EXIT EXAMS

According to Key State Education Policies on PK—12 Education: 2004 (Council of Chief State School Officers, Washington, DC, 2004), seventeen states reported requiring specific mathematics courses, and twenty-three states reported requiring specific science courses for high school graduates in 2004. Algebra, required by seventeen states, was the most common mathematics course required, and biology, required by fourteen states, was the most common science course. Health was required in twenty-eight states in 2004; thirty-three states required

physical education; and foreign language was required in fourteen states. In addition, eleven states required speech or communications classes, twenty-one states required government, and twenty-four states required state, national, or international history.

In the early 2000s school accountability became a major issue of school reform. Many states mandated what children should learn in each grade, developed assessments to measure student achievement, designed school report cards, rated their schools and publicly identified failing schools, assisted low-performing schools with additional funding, and even closed or took over failing schools. Included in the various accountability measures is the high school exit examination.

According to the Council of Chief State School Officers, twenty-nine states were either using high school exit exams or were in the process of developing such tests in 2002. Forty-two percent of the states required the exam for graduation, 20% were developing an exam, and 38% of the states did not require a high school exit exam. (See Figure 4.17.) The CCSSO reported in Key State Education Policies on PK—12 Education: 2004 that 60% of states would require an exit examination for the class of 2009.

Of the states that require an exit exam prior to graduation, a few differentiate the diplomas according to whether the test was passed or not. Most states initially administer the exit exam in tenth or eleventh grade. This allows time for remediation or other interventions to be provided for students who fail the test the first time. All states with high school exit exams allow students to take the test multiple times.

In 2004, according to the National Center for Education Statistics, nine states tested mathematics, English, science, and social studies; one state tested mathematics, English, and science; ten states tested mathematics and English; one state tested mathematics, English, and computer skills; and thirty-one states did not require a high school exit exam. In 2004 there were five states in the process of phasing in an exam. (See Figure 4.18.)

Proponents of high school exit exams believe that standardized tests are the best way to ensure high standards and accountability. They maintain that tests can communicate what is expected of students and teachers and assess whether progress is being made. Supporters believe that if tests are aligned to a rigorous curriculum, they are the best chance that low-performing students have to get the education they need and to narrow the minority achievement gap.

Those who oppose these tests point out that a single test is not an accurate measure of a student's performance. Opponents claim that these tests put poor and minority students at a disadvantage. A further criticism of "high-stakes" tests is that they push teachers to "teach to" the test, taking too much time away from classroom practices that support true learning.

"The 36th Annual Phi Delta Kappa/Gallup Poll of the Public's Attitude toward the Public Schools" asked members of the American public whether they favored or opposed using a single standardized test in the public schools in their community to determine whether a student should receive a high school diploma. Results were mixed: 51% favored standardized exit exams, while 47% opposed them. (See Table 4.11.)

SCORES ON COLLEGE ENTRANCE TESTS

Students trying to enter most colleges and universities in the United States generally take either the SAT (formerly called the Scholastic Assessment Test) or the American College Test (ACT) as part of their admission requirements. The SAT is the primary admissions test for twenty-two states, mostly in the East and on the West Coast. The ACT is more popular in twenty-eight states in the Midwest, South, and West, where a large percentage of students attend public colleges and universities. Most colleges will accept either the SAT or the ACT. In addition, some schools require three SAT II subject tests.

These two college entrance tests are standardized three-hour tests intended as an assessment of readiness for college. The SAT measures students' mathematical and verbal reasoning abilities. The ACT is curriculum-based and tests four areas: English, mathematics, reading comprehension, and science reasoning. Students who elect to take these tests usually plan to continue their education beyond high school; therefore, these tests do not profile all high school students.

Performance on the SAT is measured on a scale of 200 to 800. According to the College Board, the mean SAT scores for 2003 were 508 for the verbal section and 518 for the mathematics section. The verbal score was well below the 1972 level (530), while the mathematics score was higher than in 1972, when it was 509. The ACT results are measured on a scale of one to thirty-six. The 2004 average composite ACT score was 20.9.

According to the College Board, about 1.4 million college-bound seniors took the SAT in 2004; the American College Testing Program reports that 1.1 million students took the ACT. The number of students taking both the SAT and the ACT has grown steadily, especially in the early 2000s. In general, the more students taking the tests, the lower the scores will be, because a greater number of test-takers will likely include students who are less academically accomplished.

Trends in Scores

Verbal SAT test scores dropped in the 1970s, 1980s, and early 1990s, and math scores dropped during the 1970s and early 1980s. Observers attributed the decline to the rise in the number of students from lower scholastic achievement levels taking the tests since the 1970s. While that may explain the initial drop, a major part of the decrease also resulted from a decline in performance among the kinds of students who had previously done well in these tests. Although verbal scores remained virtually unchanged from 1995 to 2004, mathematics scores increased slightly starting in the 1990s, leading some officials to be cautiously optimistic. From 2003 to 2004, verbal SAT scores increased one point, and mathematics SAT scores dropped one point.

Gender of Test-Takers

In 2004 females accounted for the majority (53%) of students taking the SAT. Females have historically scored lower than males on college entrance examinations. The SAT was found to over-predict male college grades and under-predict female grades, so in 1995 the SAT was revised, partly to reduce these gender-related differences.

In 2004 female students'mean verbal SAT score was 504, compared to 512 for males, and female test-takers' mean mathematics score was 501, compared to 537 for males. Among ACT-takers, females' 2004 average composite score was 20.9, compared to 21 for males. (See Table 4.12.)

Race and Ethnicity of Test-Takers

According to the College Board, in 2004 white students accounted for nearly two-thirds (63%) of those taking the SAT, although the proportion of minority test-takers has risen steadily from 13% in 1973 to about 37% in 2004. African-American students made up 12%; Asian-American students 10%; Hispanic students 10%; and Native American students about 1%. Four percent of test-takers classified themselves as "other" in the race/ethnicity category.

Gains have been made since 1991, but overall SAT scores for minorities (with the exception of Asian-American students) still lagged behind the scores of white students. White students averaged 528 on the verbal section and 531 on the mathematics section in 2004, an increase of eight points and twelve points, respectively, since 1994. African-American students scored 430 on the verbal component and 427 on the mathematics. These scores were two points and six points higher, respectively, than those of 1994. So while scores for African-Americans have increased, the gap between African-Americans and whites has nevertheless widened. In 2004 the average scores of Hispanic, Latino, Mexican-American, and Puerto Rican students ranged from 451 to 461 on the verbal section and from 452 to 465 on the

mathematics section, up from the 1994 span of scores, which were 444 to 460 for verbal and 442 to 464 for mathematics.

According to the American College Testing program, for those taking the ACT in 2004 the mean composite score of white students was 21.8. For Asian/Pacific Islander students it was 21.9, while scores for Hispanics (18.4 to 18.8), Native American/Alaska Indians (18.8), and African-Americans (17.1) were lower. (See Table 4.13.)