How Science Self-Corrects Its Way to Truth
A journey through genetic research on cognitive performance and the scientific process of refinement
When a team of distinguished scientists published a groundbreaking study in 2014 claiming to have identified specific genes linked to human intelligence, the scientific community took notice. Their work, which connected cognitive performance to four particular genes, represented a potential breakthrough in understanding the biological basis of thinking and learning. But science, unlike dogma, progresses not through unchallenged declarations but through relentless verification. Just a few months after their initial publication, the researchers did something remarkable—they issued a correction. Not for fraudulent data or methodological flaws, but for misspelled gene names and incomplete analysis. This story isn't about scientific failure; it's about science working exactly as it should, refining its understanding one correction at a time.
Initial identification of genes potentially linked to cognitive performance
Refinement of findings through transparent error correction
In their 2015 correction notice published in Proceedings of the National Academy of Sciences, the researchers clarified two important errors from their original paper 9 :
To the non-scientist, these might seem like trivial typographical errors. But in genetics, precision in naming is crucial because these designations point to specific biological entities with known functions:
The correction ensured that future researchers would accurately investigate the right genetic pathways, maintaining the integrity of the scientific record.
Scientific progress depends not just on discovery but on the meticulous correction and refinement of those discoveries. The Rietveld correction exemplifies how science self-corrects to enhance accuracy and reliability.
The researchers faced a significant challenge: studying the genetics of cognitive performance directly requires massive samples of people who have undergone extensive cognitive testing, which is expensive and time-consuming. Their innovative solution? The proxy-phenotype method—a two-stage workaround that leverages more easily obtainable data 1 5 .
Think of it this way: If you wanted to study the genetics of basketball ability but couldn't test thousands of people on court skills, you might instead look at height data, which correlates with basketball performance. Similarly, educational attainment (years of schooling) serves as a proxy for cognitive performance since the two are moderately correlated 5 .
Using correlated traits to enhance discovery power
The team conducted a genome-wide association study (GWAS) of educational attainment in a massive sample of 106,736 individuals, identifying 69 education-associated genetic variants 1
Using independent samples (24,189 people), they tested these education-linked variants for association with actual cognitive performance 1
This approach allowed them to focus their cognitive testing resources on the most promising genetic candidates rather than searching blindly through the entire genome.
The researchers' methodology followed a systematic approach 1 :
Analyzing genetic data from over 100,000 people
Identifying 69 SNPs linked to educational attainment
Testing SNPs in 24,189 individuals
Investigating biological pathways
The results offered intriguing insights into the genetic architecture of cognition:
After rigorous statistical correction for multiple testing, three SNPs showed significant association with cognitive performance 1
Each variant had an extremely small effect—approximately 0.3 points on the standard IQ scale per copy 5
| Genetic Variant | Effect on Cognitive Performance | Significance |
|---|---|---|
| rs1487441 | ~0.3 IQ points per copy | Survived multiple testing correction |
| rs7923609 | ~0.3 IQ points per copy | Survived multiple testing correction |
| rs2721173 | ~0.3 IQ points per copy | Survived multiple testing correction |
| Gene Name | Biological Function |
|---|---|
| KCNMA1 | Potassium channel function |
| NRXN1 | Neuronal connection and communication |
| POU3F2 | Nervous system development |
| SCRT | Neuronal development |
Modern genetic research relies on sophisticated tools and approaches that enable scientists to detect incredibly small signals among vast amounts of biological data.
| Research Tool | Function | Application in This Study |
|---|---|---|
| Genome-Wide Association Studies (GWAS) | Scans the entire genome for variants associated with traits | Identified education-associated variants in large samples 1 |
| Polygenic Scoring | Combines effects of many genetic variants into a single measure | Created composite scores predicting cognitive outcomes 1 2 |
| Bioinformatics Analysis | Uses computational tools to interpret biological data | Identified biological pathways connecting the significant genes 1 |
| Proxy-Phenotype Method | Uses correlated traits to enhance discovery power | Leveraged educational attainment to boost discovery of cognitive performance variants 5 |
| Linkage Disequilibrium Analysis | Measures how genetic variants are correlated | Determined that one significant SNP was related to previously identified variants 9 |
Large sample sizes are crucial in genetic studies to detect variants with small effects. The proxy-phenotype method allowed researchers to leverage larger datasets for initial discovery.
The two-stage proxy-phenotype approach represented a significant methodological advance in cognitive genetics, allowing researchers to:
The most important takeaway from this research is that cognitive performance is influenced by numerous genetic factors, each with minuscule effects. As one of the researchers noted, "The effect of these variants was extremely small," accounting for just tiny fractions of IQ points per variant 5 . This polygenic nature explains why no single "intelligence gene" exists and why cognitive ability varies continuously across populations.
These findings offer several crucial insights:
Many genes with small effects
Convergence on synaptic plasticity
Both factors crucial for cognition
Self-correction enhances reliability
Subsequent research has dramatically advanced this field. By 2017, a large-scale meta-analysis identified 70 independent genomic loci associated with general cognitive ability by combining even larger samples and advanced statistical methods 8 . By 2023, studies were expanding to diverse populations, including rural South African communities, helping to create a more inclusive understanding of cognitive genetics across human populations 6 .
The story of the Rietveld correction reminds us that science is not a collection of settled facts but a process of successive approximation toward truth. The initial publication represented an important step forward in understanding the genetic architecture of cognitive performance. The subsequent correction refined that understanding with greater precision and transparency.
What makes this story compelling isn't just the discovery of genes associated with how we think, but the demonstration of how science thinks—critically, rigorously, and correctively. As the researchers noted in their FAQ about this study, cognitive performance is influenced by both genetic and environmental factors, and traits that are genetically influenced may still be quite malleable 5 . Just as eyewear can correct vision regardless of genetic predisposition, educational interventions and environmental enrichment can potentially enhance cognitive outcomes.
In the end, this research illuminates not only the biological basis of cognition but also the human capacity for intellectual honesty—the willingness to correct, refine, and improve our understanding, one gene, one letter, one discovery at a time.