Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit

Biological Data

Herds of the African savannah elephant (Loxodonta africana) can contain up to 100 individuals, consisting of one or more adult females and their offspring. Males leave the herd after puberty and spend their time either alone or in smaller all-male groups.

In this study, the researchers were interested in aggressive behaviours of male elephants towards non-conspecifics (either other animals or vehicles). Specifically, they wanted to investigate if there was a difference in aggression between younger (adolescent) and older (adult) males (variable ‘Age’ in the dataset), and if the presence of other younger or older males in the same group would impact aggression (variable ‘NumOver25YOPresent’ indicates how many other individuals over 25 years old were present, while ‘NumUpTo25YOPresent’ indicates how many other individuals of up to 25 years old were present).

The authors observed the elephants for ten minutes after they arrived at watering holes in a nature reserve in Botswana (they only included elephants arriving in groups and always randomly chose a single elephant out of the group to observe). They then scored whether the elephant did or did not exhibit any aggressive behaviours towards non-conspecifics in the 10-minute observation interval (variable ‘AgressionToNonElephant’ in the dataset, where 0 indicates that the elephant did not exhibit any aggressive behaviours, and 1 indicates that they did). They often had more than one observation for the different individual elephants that they observed (the different individual elephants are indicated by the variable ‘ID’ in the dataset).

1) Read the data, check if everything was read properly, and remember to code variables as factors if necessary.

2) Construct a model to predict how age, the number of other elephants up to 25 years old that were present and the number of other elephants over 25 years old that were present affect aggressive behaviours. Include the appropriate random effects structure. Don’t include interactions.

3) Also run a model that includes the interaction between age and the number of elephants over 25 years old that were present. Which model is better? Keep the best model to continue.

4) Determine whether it is necessary to account for overdispersion (again, keep the best model). For the purpose of this exercise, there is no need to do any other model diagnosis checks.

5) Visualize the model outcomes.

6) What do you conclude?