Reviewing TABE Scores – TABE 9&10

When testing with TABE^® in an educational setting, the raw score (or number correct) is useful only to indicate which items were missed and to obtain a scale score for a particular test performance. This number correct is translated into a scale score; specific items missed can be used to build a personalized instruction plan. Given that the number of test items can vary with different test batteries and test levels, the raw score is of little use in comparing test performances.

From every raw score on every assessment, TABE^® derives a Scale Score (SS) to compare the test performance to the norm. The SS is the most useful score associated with a testing because it provides the starting point and the basis for all other norm-referenced scores that can be derived. It is a three-digit number from zero to 999 and should be the score used for statistical manipulations of score data. The SS is an equal-interval scale, which means that a difference in two scores – 27 scale score points for example – means the same thing at different points on the scale. The purpose of the scale score is to provide an equal-interval interpretation of an examinee’s score that can be comparable across all levels and tests in TABE within a content area. For example, an examinee’s SS on Form 9 could be used to predict performance on Form 10 of the same test. The SS is also the score that should be used to measure educational gain over a phase of instruction. The scale score, therefore, is the most informative score provided by the TABE and should be recorded for every test taker. 

From the SS, examiners can obtain a Grade Equivalent (GE) score as well, which provides a measure of achievement relative to the structure of a typical school education (k-12). While the GE score does not have comparable meaning in non-graded programs like most adult basic education classes, they can be used to (loosely) organize instructional groups and provide a useful reference for the adult student. The GE ranges from 0-12.9, which represents each 10-month school year in traditional elementary and secondary education.

The Grade Equivalency score (GE) should be interpreted with caution. First, GE should not be used to compare achievement across tests. Second, they are not an equal-interval scale. Therefore, they cannot be added, subtracted, or averaged. Finally, the meaning of the GE is often misinterpreted. For example, a student can score a GE of 7.3 on a TABE^® Level M Mathematics Computation assessment. This indicates the student performs at Grade 7.3 on a test which covers content generally taught in fourth and fifth grade (the TABE M content grade level ranges from 4.0-5.9). One could expect the GE score to go down if the examinee is retested with Level D (where the content grade level ranges from 6.0 to 8.9) – because a student who can perform at the 7th grade level on 5th grade content can’t perform at the same level on 7th grade content.

This example demonstrates that it is important to know exactly what the test objectives are when considering a GE score, or any norm-referenced score data, because the scores mean nothing outside the context of the assessment’s content. When moving students up a test level by their GE score, examiners should ensure that the student is testing at the top of the GE range on the correct level pretest. For example, if a student has been pretested with Level M, he or she should have scored at the top of the range for that content grade level (around 9.9 GE). Examiners should reinforce this with other indications of learning progress, such as quizzes and classroom performance. A 6.5 GE Score on the Level M pretest, however, is usually not a high enough pretest score to consider post-testing with Level D, unless the Locator Test indicates that enough progress has been made. As always, the best reference for determining an examinee’s appropriate test level is the Locator Test.

The following chart illustrates the relationship between test level, the content grade level range, and the GE score range:

There are three other scores that are available, and less often used. The Percentile score is helpful in describing test performance generally. The best way to think of the percentile is that it is a ranking of all test-takers. An examinee that scores an 82 percentile performed better than 82 percent of all test- takers in the norm group. The stanine (STAndard NINE) score is a one-digit score from one to nine, related to percentile rank. The last score the TABE^® provides is the Normal Curve Equivalent (NCE), which is like the percentile rank but on an equal-interval scale, and could be used to compare performance in different content areas. All of these test scores are found in the TABE® Norms Book. To make the most of TABE testing, administrators and instructors should be familiar with the Norms Book.