Hans Fredrik Sunde
@hfsunde.bsky.social
Researcher at the Centre for Fertility and Health (NIPH) in Oslo. Interested in assortative mating, behavioral genetics, and bias in research.
www.hansfredriksunde.com
www.hansfredriksunde.com
I tried making a figure explaining how the sum of variances law can be understood as the diagonal of a bivariate normal, and how a correlation between the two antecedent variables — such as under assortative mating — will result in greater variance and more cases above a threshold.
September 10, 2025 at 12:50 PM
I tried making a figure explaining how the sum of variances law can be understood as the diagonal of a bivariate normal, and how a correlation between the two antecedent variables — such as under assortative mating — will result in greater variance and more cases above a threshold.
Accompanying the paper is an interactive web page with figures and tables showing the prevalence of psychological codes in the ICPC-2 by age, sex, and parental income quartile. Check it out here:
hfsu.shinyapps.io/prevalence_b...
hfsu.shinyapps.io/prevalence_b...
August 4, 2025 at 8:45 PM
Accompanying the paper is an interactive web page with figures and tables showing the prevalence of psychological codes in the ICPC-2 by age, sex, and parental income quartile. Check it out here:
hfsu.shinyapps.io/prevalence_b...
hfsu.shinyapps.io/prevalence_b...
Our new paper is out today! 🎉 In it, we use administrative register data to document how psychiatric disorders are strongly linked to parental income, from childhood far into adulthood. Furthermore, we attempt to separate causation and selection using kinship-based models.
doi.org/10.1111/jcpp...
doi.org/10.1111/jcpp...
August 4, 2025 at 8:45 PM
Our new paper is out today! 🎉 In it, we use administrative register data to document how psychiatric disorders are strongly linked to parental income, from childhood far into adulthood. Furthermore, we attempt to separate causation and selection using kinship-based models.
doi.org/10.1111/jcpp...
doi.org/10.1111/jcpp...
Our new study is just out in Psychological Science! We study cognitive ability at age 18 and mental health 20 years later in 270k Norwegian men. We include different mental disorders, compare education by ability, and run sibling-fixed effects. Check it out here: doi.org/10.1177/0956...
June 28, 2025 at 11:32 AM
Our new study is just out in Psychological Science! We study cognitive ability at age 18 and mental health 20 years later in 270k Norwegian men. We include different mental disorders, compare education by ability, and run sibling-fixed effects. Check it out here: doi.org/10.1177/0956...
Our new paper was just published! It has all the right ingredients:
- Behavioural genetics ✅
- Assortative mating ✅
- Discussions about nature-nurture of social outcomes ✅
- "Things are complicated"-conclusion ✅
- Summary statistics available for reanalysis! ✅
See Eivind's thread for more.
- Behavioural genetics ✅
- Assortative mating ✅
- Discussions about nature-nurture of social outcomes ✅
- "Things are complicated"-conclusion ✅
- Summary statistics available for reanalysis! ✅
See Eivind's thread for more.
June 20, 2025 at 7:41 PM
Our new paper was just published! It has all the right ingredients:
- Behavioural genetics ✅
- Assortative mating ✅
- Discussions about nature-nurture of social outcomes ✅
- "Things are complicated"-conclusion ✅
- Summary statistics available for reanalysis! ✅
See Eivind's thread for more.
- Behavioural genetics ✅
- Assortative mating ✅
- Discussions about nature-nurture of social outcomes ✅
- "Things are complicated"-conclusion ✅
- Summary statistics available for reanalysis! ✅
See Eivind's thread for more.
... This came in my inbox today
June 16, 2025 at 9:54 AM
... This came in my inbox today
I have also taken care to write supplementary notes where I go through the logic more thoroughly than in the main paper, so have a look at that if you are curious!
June 11, 2025 at 9:52 PM
I have also taken care to write supplementary notes where I go through the logic more thoroughly than in the main paper, so have a look at that if you are curious!
After accounting for indirect assortment, we find that genetic transmission only accounted for ~62% of the parent-offspring correlation. The rest was attributable to direct (~16%) and especially passive (~23%) environmental transmission.
June 11, 2025 at 9:52 PM
After accounting for indirect assortment, we find that genetic transmission only accounted for ~62% of the parent-offspring correlation. The rest was attributable to direct (~16%) and especially passive (~23%) environmental transmission.
When we apply the iAM-COTS model, we estimate education’s sorting factor (panel d) to be only 29% heritable, whereas 60% (!) could be attributed to sibling-shared environmental factors and gene-environment correlations.
(Panel c shows results for observed education).
(Panel c shows results for observed education).
June 11, 2025 at 9:52 PM
When we apply the iAM-COTS model, we estimate education’s sorting factor (panel d) to be only 29% heritable, whereas 60% (!) could be attributed to sibling-shared environmental factors and gene-environment correlations.
(Panel c shows results for observed education).
(Panel c shows results for observed education).
We apply these models to extended families of twins and full siblings in Norwegian registry data. We find that MZ-twins-in-law correlations are about the same as for partners (r~.46), while DZ-twins-in-law and regular siblings-in-law correlations are lower but still quite high (r~.29).
June 11, 2025 at 9:52 PM
We apply these models to extended families of twins and full siblings in Norwegian registry data. We find that MZ-twins-in-law correlations are about the same as for partners (r~.46), while DZ-twins-in-law and regular siblings-in-law correlations are lower but still quite high (r~.29).
We present two models: one single-generation model (the iAM-ACE model) and one multi-generation model (the iAM-COTS model) that uses partners of twins and their children to investigate indirect assortment and its consequences for intergenerational transmission.
June 11, 2025 at 9:52 PM
We present two models: one single-generation model (the iAM-ACE model) and one multi-generation model (the iAM-COTS model) that uses partners of twins and their children to investigate indirect assortment and its consequences for intergenerational transmission.
Okay, enough theory. Why do partners have similar education? Can we freely estimate the degree of genetic and social homogamy? It turns out that we can! How? In the paper, we present one way, namely the use of partners of twins (i.e., twins-in-law)!
(pic from first presentation of paper, May '24)
(pic from first presentation of paper, May '24)
June 11, 2025 at 9:52 PM
Okay, enough theory. Why do partners have similar education? Can we freely estimate the degree of genetic and social homogamy? It turns out that we can! How? In the paper, we present one way, namely the use of partners of twins (i.e., twins-in-law)!
(pic from first presentation of paper, May '24)
(pic from first presentation of paper, May '24)
We propose defining genetic and social homogamy as the degree to which genetic or social factors influencing the trait are also factors that partners assort upon—that is, their association with the trait’s sorting factor.
June 11, 2025 at 9:52 PM
We propose defining genetic and social homogamy as the degree to which genetic or social factors influencing the trait are also factors that partners assort upon—that is, their association with the trait’s sorting factor.
A general framework of assortative mating would separate the induced correlations in causes from the importance of those causes. Our theoretical framework accomplished this by conceptually separating the "focal phenotype" (the observed trait, like education) from its "sorting factor" (S).
June 11, 2025 at 9:52 PM
A general framework of assortative mating would separate the induced correlations in causes from the importance of those causes. Our theoretical framework accomplished this by conceptually separating the "focal phenotype" (the observed trait, like education) from its "sorting factor" (S).
Assortative mating thus induces cross-partner correlations between the causes of the trait in question. Because this includes induced genetic similarity, there are genetic consequences of assortative mating. These are proportional to the genotypic partner correlation.
June 11, 2025 at 9:52 PM
Assortative mating thus induces cross-partner correlations between the causes of the trait in question. Because this includes induced genetic similarity, there are genetic consequences of assortative mating. These are proportional to the genotypic partner correlation.
Compare this with other processes leading to partner similarity, where partners see a shared change in trait value as the result of a shared, identical cause, or when one individual’s trait value is a direct consequence of the other individual’s trait value.
June 11, 2025 at 9:52 PM
Compare this with other processes leading to partner similarity, where partners see a shared change in trait value as the result of a shared, identical cause, or when one individual’s trait value is a direct consequence of the other individual’s trait value.
First, it is important to note that “assortative mating” is not synonymous with “partner similarity”. Instead, assortment is only one of several processes that can lead to partner similarity, the others being convergence, stratification, and inbreeding.
June 11, 2025 at 9:52 PM
First, it is important to note that “assortative mating” is not synonymous with “partner similarity”. Instead, assortment is only one of several processes that can lead to partner similarity, the others being convergence, stratification, and inbreeding.
Crucially, the answer to the question “why do offspring resemble their parents” hinges on why partners are similar. The wildly different final sentences in these two abstracts illustrate the point:
June 11, 2025 at 9:52 PM
Crucially, the answer to the question “why do offspring resemble their parents” hinges on why partners are similar. The wildly different final sentences in these two abstracts illustrate the point:
We know this because direct assortment comes with very clear expectations that can be tested in various ways, all of which fail for education. For example, partners and siblings-in-law are too genetically similar.
June 11, 2025 at 9:52 PM
We know this because direct assortment comes with very clear expectations that can be tested in various ways, all of which fail for education. For example, partners and siblings-in-law are too genetically similar.
Last week, our new paper on indirect assortative mating was published.🍾 Let’s take a closer look at what this means, why it matters, and what we found (🧵/32):
www.nature.com/articles/s41...
www.nature.com/articles/s41...
June 11, 2025 at 9:52 PM
Last week, our new paper on indirect assortative mating was published.🍾 Let’s take a closer look at what this means, why it matters, and what we found (🧵/32):
www.nature.com/articles/s41...
www.nature.com/articles/s41...
I will write a detailed thread next week. If something is confusing until then, I highly recommend the supplementary notes, where I go through the logic more slowly and in greater depth.
June 6, 2025 at 6:01 PM
I will write a detailed thread next week. If something is confusing until then, I highly recommend the supplementary notes, where I go through the logic more slowly and in greater depth.
Results imply that partners are strongly assorting (r=.68) on education-associated trait(s) with large shared-environmental effects (i.e. Social Homogamy). Accounting for this in intergenerational models reveals previously hidden or underestimated environmental effects.
June 6, 2025 at 6:01 PM
Results imply that partners are strongly assorting (r=.68) on education-associated trait(s) with large shared-environmental effects (i.e. Social Homogamy). Accounting for this in intergenerational models reveals previously hidden or underestimated environmental effects.
Our paper on indirect assortative mating is now out in @natcomms.nature.com! In it, we provide refined definitions of terms used to explain partner similarity, develop statistical models, and find evidence of surprisingly high social homogamy for education.
Link: doi.org/10.1038/s414...
Link: doi.org/10.1038/s414...
June 6, 2025 at 6:01 PM
Our paper on indirect assortative mating is now out in @natcomms.nature.com! In it, we provide refined definitions of terms used to explain partner similarity, develop statistical models, and find evidence of surprisingly high social homogamy for education.
Link: doi.org/10.1038/s414...
Link: doi.org/10.1038/s414...
Do you remember how the results compare with this from MoBa (Norway)?
May 25, 2025 at 6:35 AM
Do you remember how the results compare with this from MoBa (Norway)?
Regardless of cause, it is worth reiterating that the association between socioeconomic background and psychiatric disorders is large and prevails far into adulthood.
October 24, 2024 at 1:25 PM
Regardless of cause, it is worth reiterating that the association between socioeconomic background and psychiatric disorders is large and prevails far into adulthood.