On Intelligence and Geniuses (Part 2): Why IQ Matters — And Why We Measure It

Whatever It Measures, It Matters

Melancolia I Albrecht Dürer

In my first article (On Intelligence and Geniuses (Part 1): What Is Intelligence and How Do We Measure It?), I elaborated in detail what, in psychological research, we understand by the concept of intelligence. We reached the following conclusion:

Intelligence describes a colloquial term for mental performance. It is not directly graspable and scientifically difficult to capture. The general factor ‘g’ (general intelligence) is a specific statistical construct that empirically approaches the concept of intelligence. It describes the fact that, on average, a positive correlation exists between different tasks that measure mental performance. It is thus the ‘engine’ that ‘drives’ performance in various mental tasks. The IQ is the measurement value that a standardised test provides. It offers an estimate of the individual level of ‘g’ in comparison to a norm group.

We now know what we mean by intelligence in the psychometric‐scientific sense, and we ask: what is it all for? It seems we remain attached to Boring’s definition and may even be going in circles:

“Intelligence is what the intelligence test measures”

So why is this concept so relevant?

Quite simply: even though intelligence, strictly speaking, is only quantified via performance on specific tests, the outcome is far from meaningless. What we measure (IQ) is associated with many outcomes one might not expect, and is among the best predictors we have for variable human performance. Some examples:

• Criminality (−): Jovanović et al. (2012)

• Religiosity (−): Zuckerman et al. (2013)

• Life expectancy (+): Sanchez‑Izquierdo et al. (2023)

Later we will discuss:

• School performance (+)

• Educational attainment (+)

• Academic success (+)

• Career success (+)

• Poverty (−)

• Socio‑economic status (+)

Thus we can firmly conclude: we may not know precisely what intelligence is, and may define it in relation to test performance, but that very performance has enormous predictive validity for many outcomes. In other words, even a sceptic would have to concede: whatever we are measuring, it matters.

I could end the article here, since the evidence is clear. Yet I wish to delve deeper into some of the most important factors: school performance, academic success, career success, and socio‑economic status.

School Performance and Academic Success

There is a well‑established relationship between intelligence test scores and school performance. In fact, “highly aggregated measures of general intelligence (…) are the best single predictors of school achievement” (Helmke & Weinert, 1997). Intelligence has even been called “the capacity for higher education” (Asendorpf, 2004). This finding is among the most robust in empirical psychology. It makes logical sense: students with higher IQs learn faster and grasp difficult content more swiftly (Helmke & Schrader, 2007). Indeed, IQ is the best predictor of school performance. Nevertheless, the strength of the relationship between IQ and specific subjects varies.

A meta‑analysis by Roth et al. (2015) aggregated 240 independent studies with 105,185 participants. The following correlations (r) emerged:

• Mathematics & Science (r = 0.49)

• Languages (r = 0.44)

• Social Sciences (r = 0.43)

• Art & Music (r = 0.31)

Similarly, or even more clearly, with final examination performance:

• Overall performance: r ≈ 0.81

• Mathematics: r ≈ 0.77

• English (native tongue): r ≈ 0.67

• Geography: r ≈ 0.65

• Foreign languages (French & German): r ≈ 0.64, r ≈ 0.61

• History: r ≈ 0.63

• Music: r ≈ 0.54

• Biology: r ≈ 0.51

• Art & Design: r ≈ 0.40

An interesting differentiation emerged in intelligence across different school years (Roth et al., 2015). Overall, intelligence correlated more strongly with performance in later grades (r ≈ 0.54–0.58) than in earlier ones (r ≈ 0.45). Additionally, the correlation declined over time, that is, intelligence still significantly predicts school performance, but less so than it did prior to around 1983.

In summary, intelligence is by far the best predictor of school achievement across nearly all subjects. Other factors, such as motivation and self‑efficacy, are likely responsible for unexplained variance. It should also be emphasised that personality traits like self‑efficacy and motivation are positively associated with school performance, and they are themselves influenced by intelligence, which affects performance!

When school performance is measured via a standardised achievement test (rather than by report‑card grades), the correlation with intelligence becomes very high. Indeed, it has been said that “internationally and nationally the different scales of school‑achievement studies across age levels empirically measure almost the same constructs as intelligence tests” (Rindermann, 2006).

The situation is similar for academic success at university. Entrance tests such as the SAT measure general intelligence to a very high degree (r ≈ 0.70–0.80), and scores on such entrance tests are highly correlated with first‑year GPA (r ≈ 0.53). Correlations between intelligence test scores and specific university subjects vary (Sackett et al., 2009). In particular, STEM fields typically have students with higher average intelligence test performance. For example, Kuncel et al. (1997) found a difference of about 0.92 SD in SAT scores between engineering and education majors.

Career Success

Let us now turn to the relationship between intelligence (as measured by IQ) and success in one’s career.

First, it must be noted that average IQ differs greatly across occupational groups. More cognitively demanding professions have individuals with higher average IQs. Intelligence differences explain more variance in managerial job performance (with r² ≈ 0.45) than in simpler roles like clerical or driving jobs (r² ≈ 0.20). Thus, intelligence tests are especially predictive for complex jobs (Salgado et al., 2003).

What about actual career success? Although job‑specific selection tools can sometimes better predict performance in a given role, a meta‑analysis by Sackett et al. (2022) found that general intelligence is actually the best job‑unspecific predictor of career success.

Popular culture often emphasises personality traits such as ambition, self‑efficacy, and extraversion instead. But reality shows first that traits like self‑efficacy often positively correlate with intelligence, because higher intelligence leads to better performance and thus boosts self‑efficacy, motivation, and related traits, and second that these traits carry far less predictive power for job performance than intelligence itself (Damian et al., 2015, Wiese & Burk 2024)

Thus intelligence is startlingly relevant in everyday life, immensely important for personnel selection.

Therefore we should not dismiss IQ (as a test outcome) as irrelevant to real life. A fitting analogy: consider a soldier. Recruits are often assessed via sit‑ups, pull‑ups, or endurance tests. These abilities are not directly relevant to combat or defence, yet they have predictive power regarding a soldier’s broader fitness and capacity. Similarly, IQ tests should be viewed in the light of job performance and intellectual capabilities.

Influence of Socio‑Economic Status (SES)

A central finding in intelligence research is that gifted individuals are disproportionately represented in higher social strata and average IQ rises with increasing socio‑economic status (SES). We therefore need to ask: how much influence does IQ have on later SES and financial success, and how much do one’s existing SES and environment (e.g. family background) contribute? We seek the net effect of IQ, meaning how much of the association is solely attributable to IQ itself. This approach is known as a “communalities analysis,” in which the predictive power of IQ and SES are disentangled. To compute the pure net effect of IQ, SES is held statistically constant. In simple terms, one examines within the same SES stratum how much IQ influences later success.

What is remarkable: an analysis by Sackett et al. (2009) found that IQ’s influence on later success remains even when SES is held constant, whereas SES’s effect disappears when IQ is controlled for. This has been supported by further analyses (see Sorjonen et al., 2012 or Damian et al., 2015, Wiese & Burk 20). Concretely, this means intelligence has a stronger effect on occupational status than a person’s socio‑economic background.

I confess that, when I first encountered these findings, I found them somewhat shocking. I do not wish to delve into the political implications here, however I believe they should not be neglected. But the facts remain: IQ has genuine predictive power and is relevant.

Does IQ Matter at the National Level?

We have thus established that IQ, or rather, quantified intelligence, is indeed associated with later academic and career success and is a relevant predictor. But what about at the level of nations? Rindermann (2018) estimated a national g‑factor (derived among other things from standardised achievement tests), conceptually comparable to individual-level g. This national-level g correlated with GDP per capita (0.60), economic freedom (0.52), and economic growth (0.44). Longitudinal analyses also showed that rising prosperity has a small effect on cognitive ability, while rising cognitive ability significantly boosts prosperity. Interestingly, the intelligence level of the top 5 % of the population proved decisive for national development.

However, such correlations must be interpreted with caution. That's because cross-national research faces major methodological challenges, including the comparability and representativeness of samples, differences in test familiarity, and test administration conditions. Many studies have been criticised for poor sample quality and limited representativeness. More rigorous studies, are rather rare and methodologically demanding. Moreover, although the g-factor appears psychometrically universal, it may require culturally specific interpretations, as no text is completely culturally free. Thus, while cognitive ability seems relevant to national development, we must see these results as embedded in historical, educational, and structural contexts.

Significance of the Findings

One might dismiss these figures, as you may have noticed, most correlations lie in the range of 0.3–0.6. A perfect correlation (1.0) would indicate perfect linear predictability. But let me illustrate how significant such correlations are in another field: in medicine, after stroke or myocardial infarction, patients are advised to take acetylsalicylic acid (aspirin) to reduce the risk of recurrence by inhibiting platelet aggregation. The correlation between aspirin use and reduced mortality risk is only about r ≈ 0.04! Or, pardon the example, consider Viagra, often lauded as a “miracle drug” for erectile dysfunction and broadly satisfying to users; its effect size corresponds only to about r ≈ 0.40 (see Meyer et al., 2001). You thus see: correlations in the 0.3–0.6 range, as we observe in intelligence research, are scientifically and practically highly significant, and comparatively rare in research.

Average = Individual?

A key concept to grasp is that all these values are statistical. They describe an average trend observed in large samples. That does not imply that specific individuals cannot deviate from the pattern. People exist who are particularly intelligent yet have low career success, or vice versa. Such cases are certainly possible and not excluded.

Conclusion

Although IQ tests measure a theoretical construct that approximates the concept of intelligence, the test results, which might at first seem abstract or irrelevant, are extremely relevant to performance across various life domains. We must concede that many portrayals of IQ and its role fundamentally understate its significance. Perhaps this is because the reality leads to uncomfortable societal implications: on average, more intelligent individuals simply tend to be more successful than less intelligent ones. I do not intend to explore those political questions here. The fact remains: IQ has genuine predictive power and is clearly relevant.

In the next article in this series, I will turn to a particularly interesting topic in intelligence research, in my view, that is, the hereditary and biological foundations of intelligence, and gender differences in intelligence. Stay tuned, if you’re interested.

References

Asendorpf, J. B. (2024). Persönlichkeitsbereiche. In Persönlichkeitspsychologie für Bachelor (pp. 85–151). Berlin, Heidelberg: Springer Berlin Heidelberg.

Damian, R. I., Su, R., Shanahan, M., Trautwein, U., & Roberts, B. W. (2015). Can personality traits and intelligence compensate for background disadvantage? Predicting status attainment in adulthood. Journal of Personality and Social Psychology, 109(3), 473.

Deary, I. J., Strand, S., Smith, P., & Fernandes, C. (2007). Intelligence and educational achievement. Intelligence, 35(1), 13–21.

Jovanovic, D., Novakovic, M., Salamadić, A., Petrovic, N., & Maric, S. (2012). Analysis of the relation between intelligence and criminal behavior. Journal of Health Sciences, 2(3), 224–231.

Kuncel, N. R., Campbell, J. P., & Ones, D. S. (1998). Validity of the Graduate Record Examination: Estimated or tacitly known?

Meyer, G. J., & Archer, R. P. (2001). The hard science of Rorschach research: What do we know and where do we go? Psychological Assessment, 13(4), 486.

Rindermann, H. (2018). Cognitive capitalism: Human capital and the wellbeing of nations. Cambridge University Press.

Roth, B., Becker, N., Romeyke, S., Schäfer, S., Domnick, F., & Spinath, F. M. (2015). Intelligence and school grades: A meta-analysis. Intelligence, 53, 118–137.

Sackett, P. R., Kuncel, N. R., Arneson, J. J., Cooper, S. R., & Waters, S. D. (2009). Does socioeconomic status explain the relationship between admissions tests and post-secondary academic performance? Psychological Bulletin, 135(1), 1.

Sackett, P. R., Zhang, C., Berry, C. M., & Lievens, F. (2022). Revisiting meta-analytic estimates of validity in personnel selection: Addressing systematic overcorrection for restriction of range. Journal of Applied Psychology, 107(11), 2040.

Salgado, J. F., Anderson, N., Moscoso, S., Bertua, C., De Fruyt, F., & Rolland, J. P. (2003). A meta-analytic study of general mental ability validity for different occupations in the European community. Journal of Applied Psychology, 88(6), 1068.

Sanchez-Izquierdo, M., Fernandez-Ballesteros, R., Valeriano-Lorenzo, E. L., & Botella, J. (2023). Intelligence and life expectancy in late adulthood: A meta-analysis. Intelligence, 98, 101738.

Schrader, F. W., & Helmke, A. (2008). Determinanten der Schulleistung. In Lehrer-Schüler-Interaktion: Inhaltsfelder, Forschungsperspektiven und methodische Zugänge (pp. 285–302). Wiesbaden: VS Verlag für Sozialwissenschaften.

Sorjonen, K., Hemmingsson, T., Lundin, A., Falkstedt, D., & Melin, B. (2012). Intelligence, socioeconomic background, emotional capacity, and level of education as predictors of attained socioeconomic position in a cohort of Swedish men. Intelligence, 40(3), 269–277.

Weinert, F. E., & Helmke, A. (Eds.). (1997). Entwicklung im Grundschulalter. Beltz, Psychologie-Verlag-Union.

Wiese, B. S., & Burk, C. L. (2024). Karriere: Berufliche Entwicklungsprozesse verstehen und gestalten.

Zuckerman, M., Silberman, J., & Hall, J. A. (2013). The relation between intelligence and religiosity: A meta-analysis and some proposed explanations. Personality and Social Psychology Review, 17(4), 325–354.