SlowcoachCan IQ increase for a person with borderline intellectual functioning?
To answer directly, an IQ score measured by a test can rise or fall. But a change of a few points alone does not tell us that someone can do more in daily life or at work.[1][2]
IQ scores can change. However, changes in scores, changes in actual functioning and life outcomes need to be examined at different levels.
Missing this distinction can lead to two extremes: believing that “IQ is fixed for life, so effort is pointless,” or, conversely, that “specific training will bring everyone into the normal range.” The answer suggested by research lies between the two.
1. Start with a fictional workplace scene
Suppose an employee receives the following instructions while closing a café.
“Count the milk in stock. If there are 3 or fewer, write it on the order sheet, and record the refrigerator temperature.”
In the first week, they missed the second condition and finished in 22 minutes after their supervisor re-explained the instructions three times. After using a task card, they completed all three steps at the same level of difficulty, asked for help once before making a mistake, and finished in 15 minutes.
This is a clear change in work functioning. But without a standardized retest, we cannot say that their IQ increased. Nor can we yet say that their life outcomes improved without several weeks of attendance and work records.
This is why we need to ask three separate questions about the same person at the same time.
- Did the test score change?
- Was less support needed in actual work?
- Did that change continue for several weeks or more and help with employment and daily life?
2. What did research on adolescent IQ show?
A 2011 study retested 33 healthy adolescents after an interval of approximately 3.5 years. Individual changes in full-scale IQ ranged from -18 to +21 points, and changes in verbal and non-verbal IQ were associated with structural changes in related brain regions.[1]
This study matters because it shows that saying “IQ scores never change” is an oversimplification.[1]
However, we should not jump from this finding to the conclusion that training can raise IQ.
The participants were healthy adolescents, not adults with borderline intellectual functioning. The study did not test the effects of a particular educational program, app or job-coaching approach.[1]
It is also not contradictory to say that someone’s relative position among their peers is fairly stable while their individual absolute score can change. To describe what changed, we need to distinguish their position relative to other people from their performance relative to their own past performance.[5]
3. Can a change in environment lead to long-term change?
In a study that randomly assigned children who had experienced severe early institutional care to foster care or care as usual, the foster-care group’s IQ at age 18 was, on average, 9 points higher than that of the care-as-usual group.[3]
This is strong evidence that the environment can affect long-term cognitive development. It is not merely a claim that “it depends on your mindset”; the study changed the actual caregiving environment and followed the children over a long period.[3]
However, the participants were children who had experienced extreme early deprivation. The result cannot be applied directly to everyone with borderline intellectual functioning or to job coaching for adults.[3]
A 2024 review of interventions for borderline intellectual functioning also examined 10 studies in its final review and outlined the potential of early intervention, treatment of co-occurring conditions and structured programs. At the same time, it concluded that more research on education, treatment and prognosis was needed.[2]
What we can currently say, therefore, is that changes in scores or functioning have been reported under particular conditions. To establish which outcomes change when a particular intervention is provided to particular people in a particular amount, we need to examine the population, environment and measures separately.[2][3]
4. What should we check when a retest score rises?
Taking the same test again can produce a practice effect: scores may increase as the person becomes familiar with the question format. A WAIS-R retest study also reported that practice effects should be considered when interpreting score differences over short intervals.[4]
Tests themselves also involve measurement error. Rather than treating a single score as one exact point, we should interpret it alongside a confidence interval reflecting that test’s measurement error.[5]
A gain of a few points on a retest does not, on its own, prove that an intervention worked or establish that someone is ready for employment.[4][5]
When examining a score difference, it is safer to check the following with a professional.
- Which test and subtests were used?
- How long was the interval between the two tests?
- Were health, fatigue, anxiety, concentration and testing conditions similar?
- Did the difference exceed that test’s range of measurement error?
- Was the change sustained afterward, rather than occurring only once?
There is no universal number of points above which a difference counts as “real change” across every test. Professional interpretation appropriate to the type and purpose of the test is needed.[5]
5. What becomes visible when we separate change into three levels?
I find it practical to separate the phrase “things improved” into the following three levels.
First, test scores.
Examine whether standardized IQ and subtest scores changed. Record the test name, date, conditions, scores and confidence intervals together.[5]
Second, actual functioning.
Look at which of a task’s total steps the person completed independently, the repeated explanations and prompts needed, corrections after errors, appropriate requests for help, and whether the learned approach was applied in other situations.
Third, life outcomes.
Look at how long attendance and work were sustained, whether the range of roles the person could take on expanded, their own fatigue and satisfaction, and whether the support time required from the manager decreased.
Change at one level does not automatically demonstrate change at the next.
For example, person A’s IQ might rise by 5 points on a retest while the instructions needed at work remain unchanged. Person B’s IQ score might stay the same while the steps they complete independently increase from 4 to 8 and the prompts they need decrease from 6 to 2.
The progress that employment support should not overlook is not one score, but a reduction in the support needed and an expansion in what the person can actually do.
Even if person B remains employed for a long time, however, we cannot immediately attribute that to coaching. We also need to record task difficulty, working hours, how instructions were given, managerial support and changes in work to understand what made a difference.
6. What if you recorded just one task from today?
The recording method below is a suggestion for observation to avoid missing changes in actual work; it is not a validated, standardized adaptive behavior scale.
Choose one task from today and fill in the following line.
Date / Task name / Total steps: __ / Steps completed independently: __ / Repeated explanations: __ / Requested help before an error: O·X (yes/no) / Corrections: __ / Completion time: __ minutes
Do not judge a person from one record. Repeat work of similar difficulty and examine the following.
- Does the number of steps completed independently continue to increase?
- Are direct instructions being reduced to choices or prompts?
- Are repeated errors decreasing, with the person making corrections independently?
- When stuck, does the person request help rather than hide the problem?
- Can they use the same approach in other tasks or locations?
This record does not measure IQ by proxy. Instead, it makes concrete changes in actual functioning visible that scores alone cannot show.
7. What does Slowcoach need to test?
There is currently no evidence to claim that Slowcoach improves the measures above.
The hypothesis we want to test is that breaking work into small steps and providing explanations and phrases for requesting help when needed may increase the number of steps performed independently and help people continue working.
To examine this, we need to distinguish repeated records of the same task, observations of actual users, assessments by job coaches and long-term employment outcomes. An impression that things improved after using the app is not enough to claim that the product’s effects have been demonstrated.
Hope does not lie in promising a particular IQ increase. It lies in accurately recording what someone can do today and expanding that range while adjusting the support they need.
An earlier article explains why borderline intellectual functioning is difficult to assess from a single score.
Is an IQ of 70-85 the criterion for borderline intellectual functioning?