BRENDGEN ET AL. (2005)
THE CONTEMPORARY STUDY: AGGRESSION IN 6-YEAR-OLD TWINS

This study was carried out by Mara Brendgen and colleagues at the University of Montreal in Canada. They tested a large sample of Canadian twins using questionnaires given to their classmates and teachers. Brendgen is particularly interested in the connection between “social aggression” (taunting, malicious gossip) and the physical aggression studied by other psychologists.

This study is significant for students in other ways:

• It shows how scientific research proceeds, because previous research has focussed on physical aggression’s links to biology and learning, but social aggression is a new area for research.
• It illustrates features of the Biological Approach, since it uses the twin study technique to isolate genetic and situational variables – nature and nurture.
  
• It illustrates the power and shortcomings of the self report method, because the researchers were using reports of aggression, rather than observing it themselves.
  
• It shows the growing importance of evolutionary psychology – the idea that aggressive behaviour may be passed on genetically because it has survival value

TWIN STUDIES

A twin study investigates nature and nurture by comparing identical (MZ) and non-identical (DZ) twins.

• Monozygotic (MZ) twins are genetically identical. They split from the same egg shortly after conception. They share 100% of their genes and (because of this) will always be the same sex. They probably look identical too.
  
• Dizygotic (DZ) twins are non-identical. They developed from separate eggs, just like any pair of siblings, but at the same time. They are just like any brother or sister, except that they are born at the same time. Same-sex DZ twins may share up to 50% of their genes and may look identical (but don’t have to). DZ twins may be mixed-sex and don’t have to be any more similar than any other set of brothers and sisters.

A twin study looks for a concordance between sets of MZ twins compared to sets of DZ twins in a certain behaviour. If MZ twins share the behaviour more than DZ twins, then the behaviour probably has a genetic basis. We can be sure of this because both sets of twins share the same home life and upbringing: it’s only their genetic similarity that sets them apart from each other.

SOCIAL AGGRESSION & EVOLUTION

Not all aggression involves punching an inflatable clown! As well as physical aggression, there is social aggression. This has two components:

• Indirect aggression, which is covert (hidden), such as spreading malicious gossip
• Relational aggression, which is overt (in the open) but non-physical, such as breaking off a friendship, pulling faces or “bitchiness”

Some researchers link this to evolution. Aggressive behaviour helps an animal protect itself and prosper, but only up to a point. If it is too aggressive, it may be rejected by the group or even attacked.

Therefore, social aggression has more survival value than physical aggression: it has the same benefits without the risks. Gordon Ingram (2014) shows that young children show more physical aggression than social aggression, but, as they grow into adolescence, this reverses and social aggression (gossiping, rumour-spreading) dominates. Brendgen thinks this is because young children don’t have the verbal or social skills to practise social aggression, but they acquire these once they start school.

If this analysis is correct, it leads to several research questions:

• You would expect physically aggressive children to grow into socially aggressive adults. At the age of 6, this transformation should be detectable. Is it?
• If there is a genetic basis for both physical aggression, you would expect it to be shared by identical (MZ) twins who share 100% of their genes, less shared by non-identical (DZ) twins and least shared by non-identical opposite-sex twins who have the fewest genes in common. Is this the case?

PHENOTYPES

Brendgen uses the word “phenotype” in her study and it’s worth considering this concept.

• Our “genotype” is our complete genetic makeup. Identical (MZ) twins share a genotype but DZ twins will only be partially similar in their genotype. Genotype is the influence of nature, not nurture.
• Our “phenotype” is the combination of our genes and our environment – so nature plus nurture. Even MZ twins will have some differences in their environment – they may have different hobbies, different friends, even go to different schools – so they will only be partially similar in their phenotype. However, their genotype similarity means their phenotypes will still be more similar, overall, than for DZ twins.

Brendgen and her colleagues try to separate out each child’s phenotype into three compoents: (1) genes (the genotype), (2) shared environment (home life, parenting), and (3) non-shared environment (friendships, hobbies, school differences).

BRENDGEN ET AL.'S STUDY
APRC

Aim

To find out if there is a difference between physical and social aggression in 6-year-old school children by surveying their teachers and classmates. In particular, to see if physically aggressive children are also socially aggressive and to investigate whether the link is down to genetics or social situation by comparing MZ and DZ twins.

There are other studies into the origins of aggression. Bandura's Bobo Doll studies look at how children learn aggression from the environment and Raine et al. look at how aggression is linked to brain deficits (but may have environmental explanations too)

IV

As a twin study, this looks for a correlation between (1) the aggression scores for identical (MZ) twins and (2) the aggression scores for non-identical (DZ) twins.

It also examines the correlation between (3) teacher ratings for the children’s aggression and (4) peer ratings from classmates for the children’s aggression.

As a natural experiment, it looks at the differences between (5) MZ and DZ twins and (6) girls and boys

DV

Teacher ratings for social and physical aggression were calculated out of 6 each; Peer ratings were taken from classmates who were asked to identify classmates from photographs who fitted various descriptions.

A main advantage of the present study is it assessed behavior not only by teachers but also by peers, who are rarely employed as a reporting source in twin studies - Mara Brendgen

The teacher overall scores for each child take the form of interval/ratio level data, but the children’s responses are frequencies for each child that take the form of nominal level data.

Sample

234 pairs of twins, taken from the Quebec Newborn Twin Study. This was a longitudinal study that was already going on and Brendgen used the data from it when the children were age 6.

There were 44 sets of identical (MZ) male twins, 50 sets of identical (MZ) female twins, 41 non-identical (DZ) males, 32 non-identical (DZ) females and 67 sets of non-identical (DZ) mixed-gender twins.

Because the sample was taken from another study, this should be considered a cluster sample.

Procedure

The children had been entered into the Quebec Newborn Twin Study when they were born. They were assigned to MZ or DZ based on physical resemblance; 123 pairs of twins were DNA-tested and this backed up the assignment to MZ or DZ 94% of the time, which was considered reliable enough.

By age 6, 88 pairs of twins had dropped out of the study, but Brendgen obtained data on the remaining 234 twin pairs, got written consent from the parents and approached their schools. Because this was Quebec, some of the schools were English-speaking and some were French-speaking, so they surveys had to be translated and the researchers had to speak both languages.

The teachers’ questionnaires asked them to rate each child on a 3-point scale (0 = never, 1 = sometimes, 2 = often) on these 6 statements:

• tries to make others dislike a child
• says bad things or spreads nasty rumors about another child
• becomes friends with another child for revenge
• gets into fights
• physically attacks others
• hits, bites or kicks others

The scores for social aggression and physical aggression were added together to produce two overall scores.

For the peer ratings, the children were given a simpler task. They were shown photos of their classmates and asked to circle the photos of 3 children who best fit these 4 descriptions:

• tells others not to play with a child
• tells mean secrets about another child
• gets into fights
• hits, bites or kicks others

Results

Brendgen carried out a Chi Squared inferential test, comparing the teacher- and the peer-ratings of the identical (MZ) and non-identical (DZ) twins for physical and social aggression. The differences were not significant. This suggests the two sets of ratings can be combined together.

Brendgen also looked at the correlation between how the teacher rated each child's aggression and how the child's friends rated it. A strong correlation implies the two views were in agreement, making them more valid.

There was also a “moderate” correlation between teacher and peer ratings for physical aggression (r=0.25) and for social aggression (r=0.33).

However, the teachers did rate boys as more physically aggressive and the girls as more socially aggressive; peer ratings from classmates showed the boys to be both more physically and more socially aggressive.

Brendgen also compared the correlations between teachers and friends for the MZ twins and the DZ twins. In the correlation between MZ and DZ twins, the MZ correlations for physical aggression were twice as high as same-sex DZ correlations.  For teacher-ratings, this was r=0.79, a strong correlation. This suggests the MZ twins had strikingly similar behaviour, whether it was aggressive or non-aggressive.

This suggests there was little disagreement about the MZ twins' physical aggression, but a lot more about the DZ twins. Perhaps this was because the DZ twins were more influenced by their surroundings.

The MZ twins' correlations for social aggression were similar to the DZ twins' correlations. This suggests social aggression is less linked to genetics, because both types of twins seemed to be equally affected by their surroundings, unlike with physical aggression.

Conclusions

Brendgen concludes that about 50-60% of physical aggression can be linked to genes, since it was shared by MZ twins but not so much by DZ twins. For social aggression, genes only seem to account for about 20%.

Brendgen also considers “non-shared environment” as an influence, which would be outside influences other than home life (such as friends, school, etc). This seems to account for the other 40% of physical aggression and 60% of social aggression; Brendgen links about 20% of social aggression to “shared environment”, which is probably parenting strategies. Brendgen points out that, if parents control children by withholding love and affection, the children learn to control peers by withholding friendship.

So physical aggression is mostly nature (genes), but social aggression is mostly nurture (environment).

The overlap between social and physical aggression was only moderate. Brendgen proposes that genes might give children a general predisposition towards aggression, but this only becomes social aggression if they have an environment that encourages it.

a generalized, largely genetically driven individual disposition for aggressive behaviour may shift from physical to social aggression as children mature - Mara Brendgen

EVALUATING BRENDGEN ET AL. AO3
GRAVE

Generalisability

Brendgen uses a large sample (234 twin pairs), so unusual children (anomalies) with very high or low levels of aggression out to be “averaged out” by the size of the data. This makes the sample representative.

However, it’s worth noticing that 88 twin pairs disappeared from the Quebec Newborn Twin Study before the age-6 testing point. If these children included some of the more disturbed children or those with the most chaotic backgrounds, the study might not be so representative after all. This tendency for people to drop out of studies is called sample attrition.

The researchers themselves complain that their sample was too small. They looked at 409 classrooms, but argue future studies that replicate their procedure should use larger samples.

The study only looked at 6-year-olds. In some ways, this is a representative age group because they are going through the shift from physical aggression towards social aggression, but still show signs of both. In other ways it may not be representative, because some children will be slow developers and others will be fast developers. For example, more mature children may have more non-shared environmental influences (they do more things away from home).

Reliability

Brendgen uses established questionnaires to measure aggression. These can easily be replicated, making the study reliable.

Two researchers visited each classroom, suggesting the study has inter-rater reliability.

There was a strong correlation between teacher- and peer- ratings, suggesting the scores are reliable.

The language differences might make the study less reliable. Questions translated into another language might have slightly different meanings or become confusing. Brendgen’s original questionnaires were in French, meaning the English translations might be unreliable.

The allocation of zygocity (MZ and DZ categories) was based largely on their appearance and wasn’t 100% reliable. In particular, it is possible there were DZ twins in the MZ condition.

Application

If social aggression is strongly linked to environment, it must be possible to reduce the worst effects of verbal bullying, gossiping and “trolling” in social media by children, by using early intervention. Classes and workshops might help children learn less aggressive ways of interacting.

If friends and family are a big influence on social aggression, it suggests that educating parents into better ways of handling their children and being better role models might prevent the children from being socially aggressive with their own friends.

Validity

Twin studies are a valid way of studying nature versus nurture. Since MZ twins share 100% of their genotype, but DZ twins no more than 50%, but both share the same homelife, if MZ twins have similar behaviours where DZ twins do not, this is likely to be due to the genotype.

However, these conclusions are never certain. Correlations like Brendgren’s do not prove causation. Something else could be affecting the MZ twins. For example, MZ twins tend to be physically identical and get mistaken for one another. It might be that one twin gets stereotyped based on the other twin’s behaviour (if one twin is naughty, both get a bad reputation). In Brendgen’s study, teacher- and peer- ratings might have been influenced by stereotypes, with both children being given the same rating regardless of their behaviour.

As a natural experiment, this study cannot show cause-and-effect. For example, the twins might influence each other’s behaviour, with one leading the other astray.

The study avoids taking a reductionist view of human behaviour: it looks at genetics but also at environmental factors. It links in with the findings about aggression by Bandura (who claims it is learned) and Raine et al. (who suggest it is due to brain structure).

Ethics

The parents of the twins agreed for their children to be in the study, so presumptive consent was given by a responsible adult. The teachers also agreed and this is presumptive consent too.

Nonetheless, the study does get children to look at pictures of their classmates and judge them. This might have a bad impact on friendships, especially if the children told each other afterwards who they had selected as the “hitter and biter” or the “tale-bearer”. This could lead to hurt feelings or worse – revenge! This goes against the social responsibility of ethical research and might create risk for the children who participated.

Despite this, understanding the causes and development of social aggression is for "the common good" and this research might maximise benefit for all sorts of school children, not just twins or Canadians.

EXEMPLAR ESSAY
An 8-mark essay on the contemporary study

Evaluate the contemporary study from biological psychology. (8 marks)

• A 8-mark “evaluate” question awards 4 marks for AO1 (Describe) and 4 marks for AO3 (Evaluate). Remember to include a conclusion in order to get the top band (7-8 marks).

Description
Brendgen et al. investigated 234 pairs of twins taken from the Quebec Newborn Twin Study at the age of 6. They got permission from the parents to study the children in school.
The researchers took two sets of measures. Teacher ratings came from a questionnaire with questions like how often the child gets into fights. This produced a score for physical and social aggression.
Peer-ratings came from classmates who looked at photographs of other pupils and circled the ones who they thought were most likely to tell mean secrets or hit and bite. This grouped children according to whether they were socially or physically aggressive.
Brendgen found that physical aggression was genetic 60% of the time, but social aggression was only genetic 20% of the time. Environment is a much more important factor than genetics for social aggression.
 
Evaluation
Brendgen carried out a very reliable study, using a survey method that can be replicated easily. The photo method is a standard procedure that works well with small children.
This is valid because it is a twin study. Since the only difference between the DZ twins and the MZ twins is their genotype, differences in behaviour must have a genetic cause.
On the other hand, correlations do not show causation. It may be that the twins of aggressive children get stereotyped as aggressive by classmates or teachers who get them muddled up.
There may be ethical problems with getting children to vote for the child in their class mostly likely to tell mean secrets. It might break up friendships and lead to bullying.

Conclusion
Brendgen et al. complain that their sample is too small for the results to be generalised. Since this is the first study to investigate social aggression in this way, it should be replicated with a larger sample.

• Notice that for a 8-mark answer you don’t have to include everything Brendgen et al. did. I haven’t mentioned the sample attrition or problems with translation. I haven’t mentioned the results in terms of physical and social aggression or the number of children in the study of each type. But I have tried to make the two halves – Description and Evaluation – evenly balanced.