Can Braining-Training Programs Raise IQ?

Dr. Russell T. Warne

Jun 17, 2025

LearningRx clients see an average 15-point increase in IQ and an average age equivalence increase of 3.4 years in as little as 12 weeks. (LearningRx, 2018, p. 139)

. . . significant increases in general intelligence, of 28 points on average, can be produced by undertaking online ... skills training. (Roche, 2016)

Attempts to increase intelligence are not limited to social interventions. Since the early 2000s, there has been a desire among psychologists, educators, and the public to improve intelligence by training people how to think better and solve problems. These programs take many different forms, but with interactive technology becoming cheaper and more available (especially through mobile devices) several user-directed “brain-training” programs have become popular. The theory behind these programs is that training people to use the cognitive skills that intelligence test items require will result in improved problem solving and, therefore, intelligence (e.g., Cassidy, Roche, Colbert, Stewart, & Grey, 2016).

Brain Training: The What and the Why

The theory behind these brain-training programs is plausible because everyone uses these thinking skills as they answer intelligence test items – and smarter people are better at using those skills. If the mastery of thinking skills causes people to earn higher scores on intelligence tests, then training these skills may raise intelligence.

One important skill for solving intelligence test items is working memory, which is a temporary mental store of information. By storing information temporarily, a person can easily recall it and use it for conscious problem solving (Baddeley, 1992). A simple math problem can show how most people use their working memory:

Most people approach this problem by solving the portion in the numerator first: 20 – 4 = 16. They then divide 16 by 2 to produce a final answer: 16 ÷ 2 = 8. Completing this process requires that a person remember the number 16 (the solution from the first step) in order to begin the second step. That number 16 is held in working memory for conscious use. Many multi-step mental tasks require some use of working memory, including backwards digit span, analogies, and matrix reasoning problems.

Although everyone has a working memory, some people are able to hold more information in their working memory or are able to use their working memory more efficiently (Baddeley, 1992; G. A. Miller, 1956). So, it should be unsurprising that people with a larger or better-functioning working memory score higher on intelligence tests (Brydges, Reid, Fox, & Anderson, 2012; Kyllonen & Cristal, 1990). As a result, some people have hypothesized that improving people’s working memory will produce higher intelligence (Melby-Lervåg, Redick, & Hulme, 2016).

Additionally, it is reasonable to expect programs that train people’s thinking to increase intelligence because of the most successful known method of raising intelligence: education. It is known that requiring someone to stay in school longer raises IQ (Ritchie & Tucker-Drob, 2018), and this is most likely because schooling teaches students how to think, solve problems, and use information (Ceci, 1991). Thus, a targeted intervention designed to teach these very skills could raise intelligence without requiring extensive time in school.

Evaluating Brain Training

Unquestionably, brain-training games improve scores on the tasks that users engage in (Protzko, 2017b; Simons et al., 2016; Stojanoski, Lyons, Pearce, & Owen, 2018). This should be unsurprising because it means that training people on a task makes them better at that task. It is also clear that brain-training programs produce improvements in similar tasks – a phenomenon called near transfer (Melby-Lervåg et al., 2016; Sala & Gobet, 2017). Again, this is not surprising because near transfer happens frequently in life: teaching a person how to dance the tango will give them skills that help them learn how to dance the foxtrot better.

Brain-training programs, though, rely on one of two processes to raise intelligence: the first is called far transfer, which occurs when the training on a task improves skills on dissimilar tasks. An example of far transfer would be if teaching someone how to dance a tango made them better at computer programming. The second process by which brain training could work is if it develops general skills or traits (like intelligence) that apply to many tasks.

The evidence is clear that brain-training interventions do not result in either far transfer or general training (Protzko, 2017b; Redick, 2019; Sala & Gobet, 2017, 2019; Simons et al., 2016; Stojanoski et al., 2018). For example, the authors of a thorough review of 87 studies on working memory training found that, “there is no good evidence that working memory training improves intelligence test scores or other measures of ‘real-world’ cognitive skills” (Melby-Lervåg et al., 2016, p. 512). Studies that show large gains in IQ from brain-training programs are generally poorly designed; the better designed the study is, the less likely it is to show any far transfer or general skill training (Redick, 2019; Sala & Gobet, 2017, 2019).

An example of low-quality research suggesting that far transfer or general training can occur is in an article by Cassidy et al. (2016) on the Strengthening Mental Abilities with Relational Training (SMART) brain-training program in which children completed 55 modules to teach them how to identify the relationships between pairs of items in order to understand the relationship between items that were not paired. For example:

RIH is the opposite of MOJ

MOJ is the same as WIK

The child would then be taught how to discern that RIH is the opposite of WIK. More difficult items like this can be created by increasing the number of components and adding superfluous information. Cassidy et al. (2016) found that children who used the SMART program for 8.6 weeks had IQ scores that were an average of 23.3 points higher on the British version of the WISC-IV. Another study in the same article showed that 6.2 weeks of training on the same program raised children’s scores on an educational aptitude test by approximately 30 points. However, these studies were extremely poorly designed. They had small sample sizes (15 in the first study and 30 in the second study), no control group, no follow-up to determine if fadeout occurred, and no attempt to discern whether the gains in IQ were related to improvements in thinking outside the testing situation (e.g., in the form of school performance).

Moreover, the authors’ own data provides evidence against their claims. In both studies, there was a negative correlation between improvement on the SMART program task and the improvement of IQ scores. This indicates that the children who improved the most on the relationship task had the smallest gains in IQ – and vice versa – which is the exact opposite result that one would expect if the SMART program really did raise IQ. So why did IQ scores increase? It is not entirely clear, but one possibility is that the rules that the relationship task taught are also the same rules that are used to solve some intelligence test questions. As a result, the children were well trained to look for patterns in carefully constructed test questions, which caused inflated IQ scores. This shows that it is possible to have gains in IQ without gains in g.

Gains in IQ without gains in g are not unique to the SMART brain-training program. The same result probably occurred in the Milwaukee Project (Jensen, 1989, 1998; see Chapter 15 of this book). It is also probably why matrix tests saw some of the strongest Flynn effect gains (see Chapter 14) of any intelligence test formats: solving matrix items requires mastering a limited number of rules. Once a population learns how to identify the patterns in these rules, scores on matrix items increase greatly (Armstrong & Woodley, 2014). Indeed, the Flynn effect may mostly be a result of improvements of narrow cognitive abilities (in Stratum I and Stratum II of the Cattell–Horn–Carroll model, described in the Introduction) that lead to higher IQ scores without contributing to increases in g.

An extreme example of increasing IQ without increasing g or intelligence occurs with cheating. If an examinee memorizes the answer to intelligence test questions, then their IQ score will increase greatly – but that doesn’t make the examinee smarter. Goode (2002) reported a high-profile example of cheating where a child’s mother gained access to the test’s manual (which included the answer key) and trained her 6-year-old child to give correct answers to the questions. The child obtained an IQ of 298!3 But there is no evidence that he actually became smarter or better at problem solving by memorizing the test answers.

I Fought the Law and the Law Won: Lumosity’s Legal Troubles

The evidence that far transfer occurs or that brain training produces real-world benefits is so thin that one company even got into legal trouble in the United States because of unsubstantiated claims about their brain-training program. Lumos Labs, creator of the Lumosity program, claimed that engaging in their brain-training games could improve school or work performance, improve the symptoms of mental health diagnoses (e.g., attention deficit/hyperactivity disorder, post-traumatic stress disorder), and reverse mental decline in elderly people (Fair, 2016; Simons et al., 2016). Some of their advertising claims promised to increase intelligence. The company’s unproven claims about the effectiveness of brain training resulted in a $2 million fine and over 13,000 customers receiving refunds. The evaluation from the Federal Trade Commission (FTC) about the effectiveness of Lumosity was not flattering for the company:

Let’s set the record straight. Playing Lumosity’s games might make you better at those games, the FTC says, but that doesn’t necessarily mean it will sharpen your memory or brain power in the real-world ... If you remember nothing else, remember this: You can be skeptical of any app, product, or service that says it can improve your memory or brain power quickly and easily. (Jhaveri, 2016, paragraphs 7, 9)

I endorse the FTC’s evaluation about the effectiveness of brain-training games in raising intelligence. Healthy skepticism about other claims of large benefits from brain-training programs is also probably warranted (Sala & Gobet, 2019). These companies’ main goal is to make money, and they are happy to let the press tout studies that show that their product raises IQ or cognitive functioning, but disconfirming research rarely gets the same level of attention (Detterman, 2014).

So Close, But Yet So Far: Why Far Transfer Does Not Occur

The failure of brain-training programs to raise intelligence is disappointing, but it is a consequence of the way that cognitive abilities are related to each other. As stated in the Introduction, the leading theories of cognitive abilities are the Cattell–Horn–Carroll model and the bifactor model. Both of these models assume that g is a cause of people’s performance on specific tasks in Stratum I – not vice versa. This is shown in Figures I.5 and I.6 with the arrows that point from g to Stratum II abilities and from Stratum II to Stratum I (in the Cattell–Horn–Carroll model) or from g to Stratum I abilities (in the bifactor model). If either of these models is correct and g really does influence people’s performance in specific tasks, then training people on those tasks will not improve g because those tasks have no causal influence on people’s levels of g (Protzko, 2017b). Although it was not the intent of people who are trying to increase IQ, the failure of brain-training programs to increase g has provided evidence that (a) g is largely independent of explicit training and (b) theories of the relationship between g and performance on specific tasks in the bifactor or Cattell–Horn–Carroll models are correct (Protzko, 2017b; Sala & Gobet, 2019).

But Don’t Lose Hope

This is the third chapter in a row where I wrote that many popular interventions to raise intelligence have disappointing results. These can be depressing chapters to read (and write!). Yes, large gains in IQ are possible, but these are generally only for people in extremely disadvantageous environments and a level of deprivation that is not common in wealthy countries. Interventions designed for people who already live in positive environments – preschool, brain-training programs, etc. – consistently show minuscule or temporary results. To substantially raise intelligence permanently apparently requires major, long-term life changes, such as adoption, requiring additional schooling, or a change from an extremely deprived environment to a better one.

It is not my intention to make readers feel fatalistic about interventions. These results make it tempting to say that it is impossible to raise IQ for people who already live in positive environments. But there is reason to hope: the fact that most interventions have failed to produce intelligence gains does not mean that such improvements are impossible. It merely indicates that, on the basis of current knowledge and technology, nobody knows how to raise intelligence for people who already live in beneficial environments (Lee, 2010). Some scholars who study intelligence are optimistic about the possibility of increasing intelligence (e.g., Haier, 2017a), and no one can say that every possible intervention has been tried. Psychologists, neuroscientists, or educators may one day propose a targeted intervention that does successfully raise intelligence.

However, it is valuable to be realistic about what to expect from interventions. The public should be highly skeptical of claims that a temporary training program based on current technology and science can permanently raise intelligence by 5 points or more. Such claims are not plausible, given the realities of fadeout, the results of brain-training programs, and the massive changes in environment that are needed to produce a 5-point IQ gain (e.g., from adoption). Fads like the Mozart effect (discussed in Chapter 15) can be discarded and ignored. Given the disappointing history of efforts to raise intelligence, the default assumption for the effects of any new intervention should be that it does not work until proven otherwise in well-designed, replicated studies.

From Chapter 16 of "In the Know: Debunking 35 Myths About Human Intelligence" by Dr. Russell Warne (2020)

We hope you found this information useful. For further questions, please join our Discord server to ask a Riot IQ team member or email us at support@riotiq.com. If you are interested in IQ and Intelligence, we co-moderate a related subreddit forum and have started a Youtube channel. Please feel free to join us.