Table 1

Stanford Achievement Series, Ninth Edition (Stanford 9 TA)
Fall 1998-99
Average Scores by Ethnicity

Test Grade All Students Asian Black Hispanic White
Fall 1998 Fall 1998 Fall 1998 Fall 1998 Fall 1998
 
Total Reading 4 64 59 36 39 72
  6 74 70 46 54 80
  9 72 67 49 50 78
 
     Vocabulary 4 58 50 37 35 65
  6 70 64 47 53 76
  9 68 64 49 51 74
 
     Reading Comprehension 4 66 63 37 43 73
  6 74 72 45 55 80
  9 73 68 49 50 79
 
Total Mathematics 4 68 75 37 46 73
  6 79 84 47 59 83
  9 74 80 41 49 79
 
     Problem Solving 4 71 74 41 50 77
  6 83 85 53 65 87
  9 77 80 45 53 82
 
     Procedures 4 63 75 37 46 66
  6 70 79 43 51 74
  9 65 75 35 42 70
 
Language 4 68 73 43 50 73
  6 65 68 40 48 70
  9 63 63 38 40 69
 
Partial/Basic Battery 4 66 68 40 46 72
  6 73 75 47 56 78
  9 70 71 45 49 76
 
Science 4 77 76 45 56 83
  6 72 70 39 51 79
  9 80 79 50 57 86
 
Social Science 4 66 64 38 44 73
  6 67 63 42 48 73
  9 76 73 51 55 81

Cautions in Interpreting School Test Results

Readers are cautioned not to rely too heavily upon test scores as measures of instructional quality, and not to use test scores exclusively to compare schools or school systems. Any interpretation of test results should take into account the following:

1. The Stanford Achievement Test, Ninth Edition, Form TA (Stanford 9) is a nationally-normed test required by the Virginia Department of Education. It was administered to students in grades 4, 6, and 9 in the fall of 1998. Average scores are expressed as national percentile equivalents; these percentiles indicate how students scored relative to the students in the national sample on which the test was normed in 1995. In the spring of 1997, the Stanford 9 was administered to students in grades 3, 5, 8, and 11. It was not administered during the 1997-98 school year. Scores from the fall 1998 administration are not directly comparable with the spring 1997 scores.

2. The tests do not attempt to assess any specific local school curriculum or a more general "national curriculum."

3. The multiple-choice format limits the type of questions that can be asked and the skills that can be covered. For instance, students are not asked to write a sentence or paragraph. These tests do not assess listening or oral communications skills. Nor do they attempt to measure such work-related attributes as responsibility, initiative, and creativity.

4. Changes in scores may reflect only an additional item or two answered correctly or incorrectly. For instance, for an individual, a change in one question from a correct to an incorrect answer may change the score by as much as twenty percentile points.

5. The average scores of students in smaller schools will tend to vary more from year to year than the average scores of students in larger schools.

6. Average scores for schools do not give information about how many students scored at a high level or at a low level. In each school, some students scored at a high level and others scored at a low level.

7. Test scores do not necessarily take into account the unique characteristics of a particular population, such as the proportion of limited English proficient students, transiency rates, boundary changes, etc.

8. A test score is only one indicator of instructional quality. It can help to point out possible weaknesses in instructional programs, but in order to determine whether a weakness actually exists and, if so, the causes of that weakness, other information besides test scores must be used.

8. What the tests in this report assess is limited. When used in conjunction with other information, test results can help improve students' learning. When used in isolation and simplistically, test results can damage effective instruction.

10. Each year's average score for a school represents the results for a completely different group of students.




Last Updated on 1/29/99
By Office of Planning, Testing, and Evaluation