Sampling decisions are made whenever it is not possible or practical for the observer to continuously observe and record all of the behaviour of an individual or a group, and must therefore settle for a partial record (Altmann, 1974). It is important to understand that data that results from any sampling method can only answer certain classes of questions as the data generated is only a representation of the real data. Some behaviours will be missed due to the fact that everything that occurs in real time is not captured when sampling. The first decision that the observer must make is to choose the sampling rules (which subject to watch and when). There are four different sampling rules described:

Ad libitum sampling: no systematic constraints are placed on what is recorded or when (Martin and Bateson, 2007). The observer records whatever they can see and that they think is relevant at the given time (opportunistic observations). It is informal, non-systematic, and often used in field notes. Since the observer cannot keep track of everything that is going on, the results of these observations will always be biased by the behaviours, individuals, or situations that most attract the observer’s attention (Lehner, 1996). It is useful in the early stages of a project, in the generation of an ethogram or if the events that the observer is interested in are rare but relevant (Mills and Nankervis, 1999).

Examples:

-The observer is interested in the behaviour of a group of rabbits and in a 30 minutes observation period, records whatever he can see and that he thinks is useful (avoiding interpretation of what has been observed).

At 10.00 the observer starts to records:

10.05 Rabbit A sniffs the ground

10.06 Rabbit C drinks

10.07 Rabbit A moves to Rabbit B

10.08-10.15 Rabbit A grazes close to Rabbit B

10.10 Rabbit D thumps the ground with its hind legs

10.12 Rabbit C rears

10.15 Rabbit B moves away from Rabbit A

10.22 Rabbit A follows Rabbit B

Focal animal sampling: This technique is useful if the observer is interested in what a particular individual is doing over a given time (Mills and Nankervis, 1999), and can be used to record states or events. The observer focuses their attention on a specific subject for a specified period of time and records all instances of behaviour relevant to the study. Usually a set of behavioural categories (e.g. an ethogram) would be established prior to behaviour recording to assist this (Martin and Bateson, 2007). If the observer wants to study different subjects of the same social group, the choice of the focal subject should be made before the beginning of the observation period and the sequence in which individuals are watched should be varied systematically. This is done because at the same time of the day different subjects could be involved in the same activity

For example, at every recording session the observer watches Cat A from 9 to 9.30am and Cat B from 10 to 10.30am. The meal time is at 9am. She records eating behaviour for Subject A, but not for Subject B. From the records Cat A eats more often than Cat B, however more accurate information would be obtained if the observation timings were alternated each successive day.

Focal sampling is the most satisfactory approach to studying groups, because the observer focuses his attention on a single individual, so he can provide accurate data on frequencies and durations of studied behaviours (Lehner, 1996).

Examples:

-The observer is interested in the social behaviour of dogs that live in the same household. He observes dog A for 1 hour and records all the interactions with other subjects. The observer can score if dog A approaches or is approached by other subjects and if he has a high or low body posture, for example. Then he can observe dog B for 1 hour, and then dog C and record the same data.

-The observer is interested in play behaviour of kittens and therefore observes Kitten A for 15 minutes and records all the play behaviour displayed. Then the observer focuses on Kitten B, and then on Kitten C for the same amount of time.

-The observer is interested in mother bird-young behaviour in the nest. They focus initially on nest A and record how often the mother bird feeds the young and how often the young beg for food and then the observer moves to focus their observations on next B..

Scan sampling: a whole group of subjects is rapidly scanned, at regular intervals and the behaviour of each individual is recorded (Martin and Bateson, 2007). A single scan may take anything from a few seconds to several minutes, depending on the size of the group and the amount of information recorded for each individual. This technique is useful for recording the behaviour of more subjects in less time but at the cost of some detail (Mills and Nankervis, 1999). It is a method best suited for recording behavioural states, since events are likely to be missed. Sampling interval is determined by the number of individuals being scanned and the number of behaviours being recorded (Lehner, 1996).

Examples:

-The observer is interested in sheep grazing behaviour and scans the sheep at 5 minute intervals, noting down how many sheep are grazing, how many have their heads up looking around, and how many are lying down.

-The observer wants to know where her cats spend their time in the house. She scans the house at 15 minute intervals and records where each cat is.

-The observer is interested in the behaviour in a group of frogs. He scans at 10 minute intervals and record how many subjects are in the water and how many are on the ground.

Behaviour sampling (or all occurrence sampling): the observer watches the whole group of subjects and records each occurrence of a particular behaviour, describing the context in which it occurs in as much detail as is required (Martin and Bateson, 2007). This is used if you are interested in a specific behaviour rather than the overall activity of an individual or group (Mills and Nankervis, 1999). The behaviour under study should be obvious to the observer, and not so frequent that recording becomes impossible. This technique is mainly used for recording rare but significant types of behaviour where it is important to record each occurrence.

Examples:

-The observer is interested in fighting behaviour in a specific group of cats. He can record each fight, subjects involved and the context (e.g. the presence of resources).

-The observer is interested in rearing behaviour in a group of horses and therefore records each time a horse rears.

-In an experimental situation the observer is interested in studying how rats approach a novel object and thus records each time a rat touches the object.

Sampling method	Most useful for States or events	Recommended uses
Ad libitum sampling	States and events	Generation of an ethogram, records of rare but significant behaviours
Focal animal sampling	States and events	Percentage of time, duration, nearest neighbour relationships
Scan sampling	States	Synchrony in displaying a behaviour, percentage of time
Behaviour sampling	Events	Records of rare but significant behaviours

Task: The following videos illustrate several cats engaged in various behaviours. Utilise these videos to practice different sampling methods. You can also consider whether the behaviours demonstrated are states or events and consider different types of measures (e.g. frequency, duration)