Scores are calculated in three stages
BrainTypeIQ scores are built in three stages: response data → task scores → composite scores.
| Stage | Processing | Output |
|---|---|---|
| Step 1 | Calculate raw scores from responses | Raw scores for each task |
| Step 2 | Convert raw scores into task scores | IQ-scale scores for 9 tasks, with mean 100 and SD 15 |
| Step 3 | Calculate composite scores from task scores | 5 domains, GAI, CPI, and FSIQ |
Correct and incorrect responses are not turned directly into overall IQ. A raw score is calculated for each task, converted to a common scale, and then related tasks are combined into composite scores.
Step 1: from responses to raw scores
For each task, raw scores are calculated first from response data. A raw score is not simply the number of correct answers. Which items were answered correctly can change how it is reflected, because item difficulty and task structure are taken into account.
For tasks that measure processing speed, such as Symbol Search, the raw score is the number of correct responses minus the number of incorrect responses. In speeded tasks, guessing can create correct answers by chance, so subtracting errors reduces the effect of guessed scoring.
At this stage, scores cannot yet be directly compared across tasks. Vocabulary, Matrix Reasoning, and Symbol Search differ in number of items, difficulty, time limits, and response format, so raw scores alone do not place the tasks on a shared scale.
Step 2: convert raw scores into task scores
Next, raw scores are converted to a common scale so that tasks can be compared.
Because each task has a different difficulty structure, the same number of correct answers can have a different meaning. BrainTypeIQ converts scores while considering the difficulty of each task and places them on an IQ scale with mean 100 and standard deviation 15. This makes it easier to place different task scores side by side.
It also considers how cognitive abilities can appear differently by age. For example, processing speed is more easily affected by age, while crystallized intelligence tends to be more preserved. Because the same raw score can mean different things across age bands, that difference is adjusted.
For age and cognitive abilities, see intelligence and aging.
Step 3: from task scores to composite scores
Once task scores are aligned, related tasks are grouped to calculate composite scores.
- 5 domain scores - Gc, Gf, Gv, Gwm, and Gs
- GAI - calculated from tasks related to Gc, Gf, and Gv
- CPI - calculated from tasks related to Gwm and Gs
- FSIQ - calculated from all tasks
Composite scores are not simple averages of task scores. A statistical adjustment is applied to account for correlations among tasks.
Tasks that measure similar abilities tend to move together. If they were simply added, the same information could be counted more than once. The adjustment compresses overlap so that composite scores are easier to read as overall level or domain-level patterns. This way of building composite scores is also used in major intelligence tests.
Scores should be read as a profile, not as isolated numbers
Overall IQ is useful for seeing the general level. However, differences across domains are averaged into it.
Even with the same overall IQ, the meaning changes depending on whether Gc is higher, Gf is higher, or load is more likely to increase in Gwm or Gs. For that reason, BrainTypeIQ is designed to be read through overall IQ together with the five domains and task scores.
Rather than treating one number as an exact label, it is more useful to see which conditions make ability easier to show and which conditions make load more likely.