Experimental Design of Behaviour Projects

Question 1

Three ways of describing behaviour

Answer 1

Structure, consequences, spatial relation

Question 2

What does it mean to describe a behaviour using structure

Answer 2

Appearance, physical form or temporal patterning (timing) of the behaviour
Described in terms of subjects posture and movements

Question 3

What does it mean to describe a behaviour using consequences

Answer 3

The effects of the subjects behaviour on the environment, on other individuals, on itself

Question 4

What does it mean to describe a behaviour using spatial relation

Answer 4

Relation to features of the environment or to other individuals (landmarks)
Position is important; not so much what subject is doing
e.g. approach or leave

Question 5

Three guidelines to choosing categories (types of behaviours) for projects

Answer 5

1. Enough categories to describe the behaviour in detail
2. Define each category precisely
3. Categories should be independent of one another

Question 6

How do behavioural hypotheses change over time

Answer 6

Tend to initially be broad, narrow down as more is discovered

Question 7

What are the four types of behaviour measures

Answer 7

Latency
Duration
Frequency
Intensity

Question 8

What is the latency measure? E.g.?

Answer 8

Measured in units of time
Time from some specified event to the onset of the first occurrence of the behaviour
e.g. a ball is placed in pen at time “0,” length of time for animal to approach ball is noted

Question 9

What is the frequency measure

Answer 9

Measured in reciprocal units of time
Number of occurrences of the behaviour pattern per unit time
Measures rate of occurrence

Question 10

What type of recording do we use in latency, frequency, duration

Answer 10

Latency = continuous
Frequency = continuous
Duration = time sampling or continuous

Question 11

What is the duration measure? Two types?

Answer 11

Measured in units of time
Length of time for which a single occurrence of the behaviour pattern lasts (bout length)
Total duration or mean duration

Question 12

total duration vs mean duration

Answer 12

Total = total length of time for which all occurrences of a beh lasted (proportion of observation period)
Mean = mean length of a single occurrence of the behaviour pattern (record duration of each, calculate mean)

Question 13

What is the intensity measure? E.g.?

Answer 13

Judgement of the intensity or amplitude of a behaviour pattern
e.g. weight of food eaten, number of prey captured, distance traveled, how loud a call is, how high they jump

Question 14

What type of behaviour is measured by intensity, frequency, duration and latency?

Answer 14

Quantitative behaviours

Question 15

Event vs state

Answer 15

Event = beh patterns of short duration (discrete body movements, vocalization) approximated as points in time
State = beh patterns of long duration (body postures, proximity)

Question 16

What measure do we use to measure events? States?

Answer 16

Events = frequency
States = duration

Question 17

Most important thing to make sure when marking individuals

Answer 17

It does not affect the behaviours you are trying to observe
If you can’t say it doesn’t then don’t use it as marking (could be a confounding factor)

Question 18

Group vs parties

Answer 18

Group = association whose composition is known (know all the animals in the group)
Parties = aggregations whose membership is uncertain

Question 19

What is reliability? vs validity (with regards to behaviour measures)

Answer 19

Reliability = good measures (inter observer reliability = all observers code the same beh as being the same)
Validity = right measures (are you measuring what you think you are measuring)

Question 20

How do we check the quality of our measurements before data collection

Answer 20

Reliability and validity

Question 21

Within observer reliability vs between observer reliability

Answer 21

Within = observer consistency, extent to which single observer obtains consistent result when measuring the same thing on diff occasions
Between = extent to which two or more observers obtain similar results when measuring simultaneously (>90%)

Question 22

What is Kappa coefficient

Answer 22

Want >90% between observer reliability

Question 23

What factors affect reliability

Answer 23

Practice and experience
Frequency of occurrence (fast/rare behaviours missed unless recorded)
Observer fatigue
Adequacy of definitions (drift with passage of time, observer unconsciously “improves” definition

Question 24

What is the pooling fallacy

Answer 24

Treating repeated measures of the same subject as though they were independent of one another
Additional measurements from same subject do not substitute for additional subjects

Question 25

What can affect the independence of the measures

Answer 25

The pooling fallacy
Group effects
Non-independent categories
Confounding factors
Pseudo-replication

Question 26

What is group effect

Answer 26

Behaviour of one animal affect beh of other
Beh may be similar amongst related individuals
Social interactions between members of group can lead to problems of independence
Within-group variation must be = to between-group variation

Question 27

What is wrong with non-independent categories

Answer 27

Important to make sure behavioural measures are independent of each other
Causes problems in interpreting associations btw measures
Sleeping is negatively correlated with being awake

Question 28

What are confounding factors

Answer 28

Difficult to vary one condition without varying something else
Make sure other factors did not affect our results, or randomize

Question 29

What is pseudo-replication

Answer 29

Experiment may be limited in scope because only a restricted conclusion may be drawn from results

Question 30

E.g. of pseudo-replication

Answer 30

Many piglets may respond the same way to one tape recording of vocalizations, but they may of responded differently to another tape recording that is thought to be the same but is subtly different

Question 31

What are ontogeny studies

Answer 31

Cross-sectional research: studying different individuals at each age (disconnected animals, faster/more data)
Longitudinal research: measuring the individuals repeatedly over time (follow one individual, see how life experiences affect beh)

Question 32

What are choice tests

Answer 32

Test differential responsiveness or preference

Question 33

Downside of simultaneous choice tests

Answer 33

Simultaneous presentation of test stimuli may be distracting to subject
Subject may be “trapped” by first choice
May need to include a blank choice