One of the ways in which genetic research is carried out is through a technique called "family studies."  In such studies, researchers look at the incidence of a behaviour over a number of generations.  Usually, this is limited to three generations in most populations.

Weissman et al (2005) carried out a longitudinal family study with a sample of 161 grandchildren and their parents and grandparents to study the potential genetic nature of Major Depressive Disorder.  The study took place over a twenty year period, looking at families at high and low risk for depression. The original sample of depressed patients (now, the grandparents) was selected from an outpatient clinic with a specialization in the treatment of mood disorders.  The non-depressed participants were selected from the same local community. The original sample of parents and children were interviewed four times during this period.  The children are now adults and have children of their own - allowing for study of the third generation.

Data was collected from clinicians, blind to past diagnosis of depression or to data collected in previous  interviews. In order to establish credibility, researcher triangulation was used.  Children were evaluated by two experienced clinicians - with one being a child psychiatrist and the other a psychologist. The inter-rater reliability of their diagnoses were 0.82 for MDD, 0.65 for anxiety disorders and 0.94 for alcohol dependency. 

The researchers found high rates of psychiatric disorders in the grandchildren with two generations of major depression. By 12-years-old, 59.2% of the grandchildren were already showing signs of a psychiatric disorder - most commonly anxiety disorders.  Children had an increased risk of any disorder if depression was observed in both the grandparents and the parents, compared to children where their parents were not depressed. In addition, the severity of a parent's depression was correlated with an increased rate of a mood disorder in the children.

On the other hand, if a parent was depressed but there was no history of depression in the grandparents, there was no significant effect of parental depression on the grandchildren.

• The study is longitudinal, increasing the reliability of the data.
• The association between parental MDD and child diagnosis is moderated by grandparental MDD status.  The amount of time that a child spent with a health grandparent may be a confounding variable in the study.
• The use of researcher triangulation increases the credibility of the findings.
• A sample of 161 children is a large sample; however, more research would need to be carried out to confirm the reliability of the findings.
• Although family (kinship) studies indicate a potential genetic link to behaviour, there is no actual genotype studied.