8.2 Interpreting Score Reports: Battery Scores, Warning Symbols, APR vs. GPR & Profiles
Key Takeaways
Level 8 battery raw scores are read out of fixed totals: 54 Verbal, 50 Quantitative and 52 Nonverbal items.
Riverside's warning symbols are ‡ (inconsistent responses), • (chance-level score) and ^ (ten or more items omitted with at least 80 percent of attempted items correct).
Profile letters come from overlapping confidence bands (A, B, C) or from SAS scores 24 or more points apart (E).
Students who are young for their grade have higher age percentile ranks than grade percentile ranks, and older students the reverse.
Riverside advises against routinely grouping students by score level or profile and recommends using the full profile rather than the composite alone.
8.2 Interpreting Score Reports: Battery Scores, Warning Symbols, APR vs. GPR & Profiles
Section 1.2 explains what each CogAT score means. This section is about reading an actual Level 8 report: which columns appear, what the warning symbols mean, why age and grade percentile ranks can disagree, how the ability profile is printed and how Riverside suggests using it in the classroom. Districts choose which optional scores appear, so not every report shows every column described here.
What a Level 8 Report Shows
Riverside's Score Interpretation Guide walks through the List of Student Scores and the Individual Profile Narrative. For each battery and the composite, a student's line can include:
| Column | Meaning | Always Shown? |
|---|---|---|
| No. of Items / No. Att | Items on the battery and items the student attempted | Yes |
| RS (raw score) | Number of items answered correctly | Yes |
| USS (universal scale score) | Developmental scale used as the entry point to the norms tables | Yes on most reports |
| SAS (standard age score) | Age-normed score, mean 100, SD 16 | Yes, if selected |
| APR and AS | Age percentile rank and age stanine | Yes |
| GPR and GS | Grade percentile rank and grade stanine | Optional |
| LPR | Local percentile rank, comparing the student with others tested locally | Optional |
| Composite or partial composite | Overall composite, or VQ, VN or QN if the school chose one | Yes (one type) |
| APR graph with confidence bands | A bar for each battery with a shaded band showing measurement error | Yes |
| Ability profile | For example, 9B (V−) | Yes, on Complete Battery reports |
At Level 8 the battery denominators are fixed: 54 Verbal, 50 Quantitative and 52 Nonverbal items (156 in total). A raw score is therefore read as, for example, 41 of 54. Raw scores are not compared across batteries, because each battery has its own items and its own conversion to scale scores.
Warning Symbols: Check These Before Using Any Score
Riverside tells schools to scan the report for warning indicators before making any decision. A score marked with a warning should not be used until the cause has been investigated.
| Symbol or Signal | What It Means | What Riverside Advises |
|---|---|---|
| ‡ before a raw score, often with a very wide confidence band | An inconsistent response pattern, such as missing many easy items but answering many hard ones, or very different results on the three subtests of a battery | Do not make decisions from that battery score or from the composite that includes it |
| • before a raw score | A targeted (chance) score: a raw score low enough to be obtained by guessing or random marking | Do not base instructional decisions on it; many targeted scores in one class suggest a booklet, level or scoring mix-up |
| ^ | Many items omitted: the student skipped ten or more items but answered at least 80 percent of attempted items correctly | Treat the scores as minimum estimates of the student's ability |
| A difference of 10 or more between No. of Items and No. Att | The student did not answer ten or more items on that battery | For one student, investigate; for a whole Level 5/6–8 class, suspect an administration irregularity and do not use that battery or the composite |
For a second grader, the ^ symbol is the one families most often need to understand. A slow, careful child who leaves items blank can earn a score that understates ability, which is one reason the guide recommends answering every item (see section 8.1).
Confidence Bands and the Profile Letter
The graph beside each student's scores shows the APR for each battery with a shaded confidence band. The band shows the range in which the student's score would likely fall on another occasion. Riverside assigns the profile letter from these bands:
- A: all three bands overlap.
- B: two bands overlap and the third is clearly above or below them.
- C: one battery is a relative strength and another a relative weakness.
- E: at least two SAS scores differ by 24 or more points.
A family looking at the graph can often see the letter before reading it. Section 1.2 gives the national shares of each pattern at Levels 5/6–8 (about 50 percent A, 26 percent B, 12 percent C and 9 percent E).
Why APR and GPR Can Disagree
Age percentile ranks compare the student with students of the same age. When Riverside scores the test, ages are grouped in one-month intervals; hand scoring with the norms booklet uses three-month intervals. Grade percentile ranks compare the student with all students in the same grade, using fall, midyear or spring grade norms.
A second-grade class can span more than a year of ages. Riverside's guide explains the result: when a student is very young for the grade, the age scores are higher than the grade scores; when a student is much older than the typical student in the grade, the grade scores are higher. For a typical-age student the two are nearly identical.
| Student | APR | GPR | Why |
|---|---|---|---|
| Young for grade (a late-summer birthday) | 96 | 91 | Compared with same-age children, the score is stronger than it looks against older classmates |
| Typical age | 90 | 90 | Age and grade groups are almost the same |
| Old for grade | 85 | 90 | Grade peers are younger on average |
Districts decide which norms count for eligibility. Many use age-based scores (SAS or APR) because they control for age, and some use local percentile ranks, which Riverside describes as useful when a school's population differs markedly from the national sample.
A Worked Sample Report
The following hypothetical report is for a second grader aged 7 years 8 months. The numbers are consistent with Riverside's scales (an SAS of 118 is the 87th percentile, 130 the 97th and 140 the 99th).
| Battery | Items | Attempted | Raw Score | SAS | Age Stanine | APR | GPR |
|---|---|---|---|---|---|---|---|
| Verbal | 54 | 54 | 43 | 118 | 7 | 87 | 85 |
| Quantitative | 50 | 50 | 48 | 140 | 9 | 99 | 99 |
| Nonverbal | 52 | 52 | 47 | 130 | 9 | 97 | 96 |
| Composite (VQN) | 156 | 156 | 138 | 130 | 9 | 97 | 96 |
Reading it step by step:
- Warnings: none. All items were attempted.
- Median age stanine: the stanines are 7, 9 and 9, so the median is 9.
- Pattern: Quantitative and Nonverbal are close together; Verbal is clearly lower, but the largest SAS gap is 140 − 118 = 22, which is under 24. That points to a B profile with a relative weakness in Verbal: 9B (V−). Had the Verbal SAS been 116 or lower, the 24-point rule would make it an E profile.
- Meaning: Verbal reasoning at the 87th percentile is above average nationally. The minus sign marks a weakness relative to this child's own very high quantitative and nonverbal scores, not a deficit.
Using the Profile in Class
Riverside's guide treats the ability profile, not the composite, as the key to adapting instruction, and its online Interactive Ability Profile Interpretation System (riversideinsights.com/apps/cogat) returns suggestions for any profile code. The guide's main principles:
- Build on strength. Present new material through the domain where the student reasons best. A student with N+ benefits from diagrams, models and drawings.
- Shore up weakness without making it the whole program. For V−, pre-teach key vocabulary with pictures and talk through new ideas; for Q−, use concrete objects and number stories before symbols.
- Group for diversity. Riverside states that CogAT results should not be used to routinely group students by score level or by profile; students learn from classmates whose approaches differ. It also notes that highly able students can benefit from advanced or accelerated classes with similar peers.
- Use the number, then the letter. For A profiles the median stanine carries most of the information; for B, C and E profiles the pattern matters more.
Questions to Ask the School
- Which scores does the district use for eligibility: SAS, APR, LPR, a single battery, the composite or a partial composite?
- Were any of my child's scores flagged with ‡, • or ^?
- Was the test given on paper, online with a proctor, or online with audio?
- Which form (7 or 8) was used, and is retesting with the other form allowed?
- What other information (achievement scores, teacher observations, work samples) will the committee consider?
A second grader with an August birthday has an Age Percentile Rank of 96 and a Grade Percentile Rank of 91. What explains the difference?
The child answered more Verbal items than Nonverbal items correctly
The child is young for second grade, so the age-based comparison favors the child
The child took a paper booklet rather than a computer-adaptive test
A scoring error occurred, because APR and GPR must always match
A Level 8 report shows the ^ symbol beside a child's Verbal score. What does Riverside say it means?
The child was tested off level with a harder test than usual
Ten or more items were skipped, but most of the attempted items were right
The child's answers were inconsistent across the three subtests
The raw score is low enough to come from random guessing alone
A second grader's SAS scores are Verbal 116, Quantitative 140 and Nonverbal 130, with age stanines 7, 9 and 9. Which ability profile fits?
7A, because the median of the three scores is the lowest one
9E (V−), because 140 and 116 differ by 24 points
9B (V−), because only one battery is lower than the others
9C (Q+ V−), because one battery is high and one is low
Sections you finish are checked off in the contents.