Why a Child’s IQ Score Can Vary Between Tests and Retests
A child’s IQ score is an estimate of performance on a particular set of cognitive tasks—not a permanent reading of intelligence. Scores can differ across tests and retests because of measurement error, test design, development, health, stress, familiarity, and the conditions in which testing occurs. Standardized IQ tests are typically scaled to an average of 100 with a standard deviation of 15, so even a professionally administered score should be interpreted alongside its confidence interval and the child’s broader history. Understanding those limits helps families and schools distinguish meaningful change from ordinary variation.
IQ Score Variability: What a Child’s Score Measures
An IQ score is a norm-referenced estimate of selected cognitive abilities compared with those of children in the same age group. The American Psychological Association describes intelligence testing as the assessment of abilities such as reasoning, problem-solving, and learning; no single score captures every form of ability or predicts a child’s future on its own. The score is therefore best understood as a snapshot produced by a particular test, at a particular time, under particular conditions.
The score, the scale, and the confidence interval
Most contemporary IQ tests use a standard-score scale with a mean of 100 and a standard deviation of 15. On that scale, about 68% of the reference population falls between 85 and 115, assuming a normal distribution. A reported score is not perfectly precise: test manuals provide confidence intervals that reflect measurement uncertainty. For example, a score of 100 may be accompanied by an interval such as 95–105, depending on the instrument, the child’s age, and the confidence level selected. The exact interval should come from the test report rather than be assumed.
The National Association of School Psychologists emphasizes that test results must be interpreted in context and alongside other sources of information. A small numerical difference between two scores may therefore be less informative than the confidence intervals, the pattern across subtests, and evidence from classroom performance or daily functioning.
Full-scale IQ and specific cognitive abilities
A full-scale IQ (FSIQ) combines performance across several areas, which commonly include verbal reasoning, visual-spatial or fluid reasoning, working memory, and processing speed. These component scores are related but not interchangeable. A child may reason well on untimed puzzles yet work slowly on timed tasks, or understand spoken explanations better than written ones. The combined score can conceal such differences, especially when a child’s abilities are uneven.
This is why a report should be read as a profile rather than as a single label. The test edition, subtests administered, language of testing, and the child’s age at testing all affect what the numbers mean. When comparing two assessments, families should check whether the same test or comparable instruments were used and whether the evaluator considered differences in norms and scoring.
Sources of Variation in a Child’s IQ Score
Score variation has several overlapping causes. Some reflect the ordinary limits of measurement; others reflect genuine changes in a child’s skills, development, or circumstances. A difference between scores is not automatically evidence that a child’s underlying ability has changed.
Measurement error and regression to the mean
Every test score contains measurement error: the difference between the observed result and the child’s unknowable “true” score on the measured abilities. Test reliability describes how consistently an instrument distinguishes among people under specified conditions; it does not guarantee that one child will receive precisely the same score every time. The National Research Council’s review of psychological testing explains why reliability and standard error matter when interpreting individual results.
Regression to the mean is also relevant. If a child’s first result is unusually high or low partly because of chance factors, a later score is statistically more likely to be closer to the child’s typical level. This does not mean the first test was invalid or that the second one is necessarily more accurate. It means that extreme scores should be interpreted cautiously, particularly when decisions depend on small differences.
Development, age, and schooling
Children’s cognitive skills develop unevenly. Language, attention, working memory, problem-solving, and test-taking strategies may change at different rates, while schooling provides experience with vocabulary, numbers, symbols, and structured tasks. Because IQ scores are generally compared with same-age norms, a child can learn new skills without a large change in relative standing; conversely, a child’s relative position can shift if development or educational opportunity differs from that of peers.
A retest also takes place at a different age and may use updated age norms. Comparing raw scores, standard scores, or subtest scores without accounting for those differences can lead to misleading conclusions. The evaluator should explain which scores are comparable and whether observed changes exceed the test’s expected measurement uncertainty.
Health, mood, attention, and testing conditions
Performance can be affected by sleep, illness, hunger, pain, anxiety, frustration, medication effects, attention, and rapport with the examiner. Distractions, interruptions, unfamiliar surroundings, or a poorly matched language of assessment can also interfere with a child’s ability to show what they know. These influences do not make a test useless, but they should be recorded and considered when interpreting the result.
A practical example is a child assessed while recovering from illness who has difficulty sustaining attention on timed tasks. A later assessment under better conditions might show stronger processing-speed performance even if the child’s broader reasoning abilities have not suddenly changed. Evaluators should distinguish such situational explanations from persistent patterns documented across settings.
Practice effects and changes in the test
Practice effects are score changes associated with prior exposure to similar tasks, instructions, or test items. Familiarity may make a child more comfortable with the format, though it does not necessarily mean the underlying ability has increased. The size of any practice effect depends on the test, the time between assessments, and whether the same or alternate forms were used. Test publishers provide guidance on retest intervals and interpretation; clinicians should follow those recommendations rather than assume that repeating a test is neutral.
Tests also change across editions. Updated instruments may use new items, revised norms, different subtests, or altered scoring procedures. A score from one edition is not automatically equivalent to a score from another. In addition, older norms can become less representative over time as populations and educational experiences change. The Flynn effect—the long-observed rise in average scores on some IQ tests across generations—illustrates why test norms are periodically updated, though its size and causes vary by place, period, and test.
How to Interpret Differences Between Tests and Retests
Interpreting a score difference means asking whether it is statistically reliable, educationally meaningful, and consistent with evidence beyond the test. A change that appears noticeable on paper may fall within expected measurement error; a modest score change may still matter if it accompanies a clear, persistent change in learning or daily functioning.
Compare like with like
Before comparing results, check the test name and edition, the child’s age at each assessment, the domains measured, the language and accommodations used, and the time between tests. Also note whether the reports provide confidence intervals, standard errors, or guidance on statistically significant differences. A change in FSIQ may partly reflect a change in the mix of subtests, rather than a uniform shift across all cognitive abilities.
Look at the profile and real-world evidence
A useful interpretation considers the child’s subtest pattern, achievement testing, teacher and caregiver observations, developmental history, language background, and response to instruction. For example, a child with strong reasoning and weaker working memory may need written directions, short task segments, or reduced memory load even if the overall IQ score is average. These supports should be based on documented needs, not on one number alone.
In schools, assessments may contribute to eligibility or planning, but IQ scores do not by themselves establish a diagnosis or determine every service. Evaluation rules differ by jurisdiction and purpose. Families can ask the examiner how the score relates to the referral question, which limitations apply, and what additional evidence is needed before making decisions.
When a retest may be useful
Retesting can be useful when there is a clear clinical or educational question—for example, when an earlier assessment was incomplete, circumstances substantially affected performance, or updated information is needed for planning. It is less useful when requested solely to obtain a higher score or repeated without an appropriate interval, because practice effects and normal score variation can complicate interpretation.
- Ask whether the assessment was conducted under suitable conditions and whether any health, language, attention, or emotional factors affected performance.
- Request an explanation of confidence intervals and whether the difference between scores is statistically meaningful.
- Compare subtest patterns and functional evidence, not only the full-scale score.
- Discuss the purpose and timing of any retest with a qualified psychologist or school evaluation team.
Conclusion: A Score Is an Estimate, Not a Fixed Label
A child’s IQ score can vary because of measurement error, development, health and mood, test conditions, practice effects, and changes in instruments or norms. Standard scores provide useful information when interpreted with their confidence intervals, component abilities, and the child’s real-world learning and functioning. Parents and educators should avoid treating small score differences as proof of a lasting change in intelligence. For further guidance, ask the psychologist who administered the test to explain the report and consult the child’s school or healthcare team about decisions that depend on the results.
Sources: American Psychological Association, Intelligence, https://www.apa.org/topics/intelligence; National Association of School Psychologists, Assessment and Evaluation, https://www.nasponline.org/standards-and-certification/nasp-practice-model/assessment-and-evaluation; National Research Council, The Nature of Psychological Measurement, https://nap.nationalacademies.org/catalog/2110/the-nature-of-psychological-measurement; American Psychological Association, Standards for Educational and Psychological Testing, https://www.apa.org/science/programs/testing/standards; Nisbett et al., Intelligence: New Findings and Theoretical Developments, American Psychologist, https://doi.org/10.1037/a0026699.