Self-Help Portal

Selection standards should meet three criteria: 

• fairly identify academic talent among students who differ in ethnicity, economic circumstances, and familiarity with the English language
• apply to students whether they have even or uneven ability profiles
• produce the best possible match between a student’s cognitive resources and the cognitive demands of the talent development program

Age- versus Grade-Normed Scores

Interpretations of scores on ability tests and judgments about academic talent generally rely on age norms, which offer a different perspective on development from grade-normed scores on achievement tests. On CogAT, age norms are used for the following score types:

• standard age scores 
• national age percentile ranks 
• local age percentile ranks

For most students, national age percentile ranks (PRs) and grade PRs do not differ greatly, so program selection based on national age PRs is an equitable criterion. When national age PRs and grade PRs do differ, it is often because the student is much older or younger than other students in the same grade. In these cases, students’ scores may warrant special consideration.

Students who are older than their grade peers will have lower age PRs than grade PRs because their scores are being compared with those of students of the same age, some of whom are in higher grades and have had more schooling. Therefore, these older students are less likely to qualify for programs for the academically talented if you use CogAT standard age scores or national age PRs rather than national grade PRs to make that decision. An older student with very high grade PRs may need a greater academic challenge even if his or her age PRs do not meet the criteria set for your program for academically talented students.

Composite and Partial Composite Scores

Requiring a high composite score for all three batteries eliminates many of the most-able students. The composite most accurately summarizes the abilities of students who have an A profile (all CogAT scores are at approximately the same level) and least accurately summarizes the scores of students with an E profile (two CogAT scores show an extremely large difference). Therefore, when identifying gifted students, the profile of scores on all three batteries provides a better indicator of cognitive development than the overall composite score alone.

Although the overall composite must be used with care, the quantitative-nonverbal (QN) partial composite score can be quite helpful for talent identification.

The QN partial composite is useful for these reasons:

• It better predicts achievement in mathematics and some technical domains than either the Quantitative Battery or Nonverbal Battery alone.
• It allows for meaningful score interpretation for English language learner (ELL) students since none of the items on either the quantitative or nonverbal tests require language.

The verbal reasoning score, on the other hand, is the best predictor of success in a wide range of achievement domains that rely heavily on language. 

Considerations for Different CogAT Score Patterns 

It is important to understand the issues that arise for students with different score patterns before reviewing recommendations for using CogAT scores. The following are explanations and examples of issues regarding two types of ability profiles:

• even patterns (A profiles)
• uneven patterns (B or C profiles)

A Profiles 

When schools set a cut score for the standard age scores or percentile ranks that will determine eligibility for programs for academically talented students, issues arise regarding selecting students with even score patterns (A profiles). If the standard is based on national norms, standard age scores or age PRs will be the primary criteria. 

If eligibility is based on local norms, local percentile ranks (LPRs) will be the primary criteria. In this case, you could still use the pattern of scores summarized in the ability profile, but the median age stanine is not as helpful since it is based on national norms. For example, a student with a profile of 7A may have LPRs that would map onto local stanine scores of 8 or 9. In either case, the A profile would still hold.

For ease of discussion, national norms are used in the examples that follow.

The usefulness of relatively flat or A profiles depends on the kinds of programming options available. Consider the following examples:

• Whole-grade acceleration requires advanced reasoning and achievement in multiple domains. Students with strong scores on all three CogAT batteries are more likely to succeed than students with uneven profiles, especially if the relative weakness is in verbal or quantitative reasoning. 
• Single-subject acceleration or enrichment activities in a particular domain are more likely to require more domain-specific reasoning abilities. For example, a student with strong scores on the QN partial composite and excellent mathematics achievement may be a strong candidate for acceleration in mathematics. The student with strong scores on all three batteries would also be a strong candidate for single-subject acceleration. 

As the preceding examples illustrate, the problem with academic enrichment programs requiring students to have A profiles and high composite scores lies not in the students admitted to such programs but in the highly able students who are not admitted.

B and C Profiles 

Approximately 60 percent of students who obtain a median age stanine of 9 have significantly uneven score patterns (B or C profiles). These high-scoring students are much more likely to have a relative weakness on one of the three batteries than a relative strength. Most surprising is that these high-scoring students are much more likely than others to show an extreme relative weakness. At Levels 9–17/18, only 3.2 percent of average-ability students show an extreme weakness. However, among the most-able students at these levels, that percentage is 16.2, or five times more likely. In fact, as many of these very able students with a median stanine of 9 show a significant or extreme weakness as show a flat profile. This underscores the importance of measuring reasoning abilities in multiple domains rather than in a single domain, especially for students who have extremely high or low scores. 

When students have very high scores on two batteries, their composite scores can also be very high, that is, above the 90th percentile. Nevertheless, these composite scores can be misleading because they sometimes conceal a weakness in a cognitive area that is essential for high-level achievement in a particular program of study. 

Talent Identification Case Studies

When identifying academic talent among students who are native speakers of English, emphasis should be placed on the verbal and quantitative abstract reasoning skills that are fundamentally involved in almost all higher-level academic learning tasks. The problems that arise in using the composite score for such students can be illustrated by case studies based on the four students whose CogAT scores are shown in the table on the next page. 

CogAT Scores for Students with Uneven Score Patterns

These case studies represent frequently occurring uneven score patterns in which all scores are above average. Although these students have similar composite scores, they do not all have the same probability of attaining very high levels of achievement in a program for academically talented students.

Student 1:  Strengths in V and Q, Weakness in N

Student 2:  Strength in Q and N, Weakness in V

Student 3: Above-Average on Q and V, Very High Score on N

Student 4: Extreme Relative Weakness in V

The composite score is especially misleading when there is an extreme relative weakness.

Who Qualifies for a Talent Development Program in the B and C Profile Groups?

Because approximately 60 percent of students with stanines of 9 on at least two batteries have uneven ability profiles, using the composite score for such students can lead to inappropriate identification or placement of individuals in talent development programs. We strongly recommend that the overall composite not be used to determine admission to such programs. Students are better selected for specific educational programs based on the match between their particular cognitive strengths and the demands of the program. An extreme relative weakness suggests that the student might be twice-exceptional.