Infant abuse in rhesus monkeys

Issue 12, June 1998

A researcher* carried out a study of maternal infant abuse in rhesus macaque monkeys (Macaca mulatta) in a research centre in USA. [This species was used in the famous study of maternal deprivation carried out by Harlow in the 1950s.] Four groups of monkeys were studied, each consisted of 2 - 5 adult males and 25 - 30 adult females with their offspring. The researcher began ad libitum sampling until he observed one of the following pieces of behaviour which indicated abuse: dragging, crushing, throwing or sitting/stepping on their infant. As soon as one piece of abusive behaviour was seen the researcher switched to focal sampling.

Each parent-infant pair (10 in all), i.e. abusive mother and its abused infant, were observed for 30 minutes four times each week, randomly between 0800 and 1900 hours and continued until the infant was 12 weeks old. The observations were made ftom a platform that allowed a clear, uninterrupted view of the area. The researcher provided definitions of each of the patterns of abuse: for example, dragging - the mother drags her infant by its tail or leg while walking or running. Some of the data are illustrated in Figure 1. He also made observations on a group of mothers (10 in all) which were never observed engaging in any abusive behaviour, these mothers served as controls.

Figure 1. Percentage of occurrence of different abuse patterns.

1. What is focal sampling? (1)
2. Why did the researcher switch from ad libitum to focal sampling?(1)
3. Why are controls used in scientific studies of animals? (2)
4. Why do researchers need to define behaviours carefully? (1)
5. Outline one possible weakness in describing a piece of abusive behaviour as "stepping/sitting on". (1)
6. Which behaviour is the modal abuse pattern? (1)
7. The researcher used an unpaired (unrelated) t test to determine if there was a significant difference in the age of the two groups of mothers, i.e. abusive and non- abusive mothers. Why was an unpaired (unrelated) t test used rather than a paired (related) t test?(2)
8. The t test is often described as being a robust test. What does the term `robust' mean in this context? (1)
9. The researcher stated in his report that " probabilities less than or equal to 0.05 were considered significant". Why is it important to make such a statement before undertaking the statistical analysis of the data? (2)

*Maestripieri, D. 1998. Parenting styles of abusive mothers in group-living rhesus macaques. Animal Behaviour, 55, 1 - 11.

[We are grateful to Academic Press for permission to reproduce Figure 1 a (page 5) from the article by Maestripieri, D. in Animal Behaviour, 1998.]