“The Secular Trend of Intelligence Test Scores in the Present Century: The Danish Experience”, 2021-03-01 (; similar):
Changes in mean intelligence test scores were minimal in Denmark in 2006–132019 [2006–4201014ya: 111.5, 111.1, 110.8, 110.7, 110.6, | 2011–82019: 109.1, 109.2, 109.0, 109.1, 109.3, 109.2, 109.1, 108.7, 108.8].
A change in the format of the intelligence test resulted in a sudden drop in scores.
Neither changes in parental age, dysgenics, or immigration can explain the findings.
Changes in sample composition may conceal a true decline in intelligence test scores.
The present register-based study investigated the secular trend of intelligence test scores during the period from 2006 through 2019 in a Danish population-representative sample, as well as whether the observed trend could be explained by changes in parental age, dysgenics, and immigration or changes in the format of the intelligence test and sample characteristics.
The study population consisted of all Danish men appearing before a draft board during the study period (n = 400,288). Intelligence test scores were obtained by the use of Børge Priens Prøve, typically at age 19. For each of the included draft board cohorts, the intelligence test score mean and standard deviation were estimated.
The results showed that changes in mean intelligence test scores were minimal during the study period. A slight decline was observed from 2006–4201014ya. Furthermore, there was a drop of 1.5 IQ points from 2010–201113ya, which coincided with the change in the format of the intelligence test from paper-and-pencil to computer-based, but there was essentially no change after 2011. Neither changes in parental age, dysgenics, or immigration seem to have influenced the observations. However, changes in sample composition may conceal a true decline in intelligence test scores given that a larger proportion of individuals with low intelligence seems to be exempted from testing.
In conclusion, the study findings suggest no systematic change in intelligence test scores during the last decade, but due to changes in sample composition, it cannot be excluded that there has been a negative secular trend.
[Keywords: intelligence, secular trend, Flynn effect, Denmark]
…A slight decline in mean IQ score was observed 2006–4201014ya, which can be seen as a continuation of the decline previously reported between 1998 and 2004 (2008)…We have had the opportunity to rescale the mean intelligence test scores from the Danish draft board examinations reported by 2008 against our baseline year 1960 to compare them with our observations. The rescaled mean IQ scores are as follows: 1988: 111.0 (SD: 13.0); 1998: 112.4 (SD: 12.7); 2003–4: 111.1 (SD: 12.8). As can be seen, there was an increase 1988–10199826ya followed by a small decline 1998–5200321ya–4. The mean IQ score in 1998 remains the highest recorded using Danish draft board data, whereas the mean IQ score in 2003–4 is comparable with our mean intelligence test score in 2006. As such, there has been a decline of 1.8 IQ points during the period from 1998 through 2010 followed by a drop of 1.5 IQ points which is probably due to the change in the format of the intelligence test and virtually no change from 2011 through 2019. However, the variance has declined statistically-significantly throughout the study period, corresponding to a decline of 0.15 SD per year (p < 0.001). A previous study has suggested that the negative secular trend observed in developmental test performances may be rooted in declining performances of the top percentiles (2018), leading to declining variances. If this is also true in our study where the proportion of individuals with low test intelligence scores who were exempted from testing has increased over time, this might explain our observation of no change in mean intelligence test scores, but a declining variance.