Measuring Behaviour

Question 1

Introduction

What is Psychology?
What is Behaviour?

Answer 1

Measuring behaviour –why important?

• direct benefit to humans
• fundamental understanding of biology
• Knowledge of mechanism can greatly inform understanding of behaviour – and vice versa

The direct observation of behaviour is used for many purposes, ranging from understanding cognition and how it develops to investigations of the current utility of behaviour and how it evolved in the past. The techniques for measuring behaviour are being combined with molecular and physiological approaches, but their use is important in its own right and will always remain a central part of behavioural biology and psychology.

Psychology –how describe discipline?
“Psychology is the scientific study of the mind and how it dictates and influences our behaviour, from communication and memory to thought and emotion.
It’s about understanding what makes people tick and how this understanding can help us address many of the problems and issues in society today.”
BPS -https://www.bps.org.uk/public/what-is-psychology

•Psychology is the scientific study of the mind occurring partly via the study of behaviour.
•Psychologists study human and non-human animal behaviour by observing, measuring and testing, then arriving at conclusions that are rooted in sound scientific methodology.
•Analysing behaviour informs how we think, feel, act and interact individually and in groups.
———————————————————————————-
Behaviour
•is the ultimate phenotype –result of animals’ own individual decision making processes and the expression of emotions (Darwin, 1982; Dawkins, 2004).
•reflects relationships
•reflects motivational state
•is an immediate, accessible, informative and quantifiable measure of welfare
Behaviour can be measured non-invasively (e.g. does not require breaking the skin), and often non-intrusively (the animal not disturbed).

Niko Tinbergen (1963), who pointed out that four distinct types of problem are raised by the study of behaviour biology:
Current - proximate - how does it work?
- ultimate - what is it for?
Historical - proximate - how did it develop?
- ultimate - how did it evolve?

Question 2

Thinking before you measure behaviour

Answer 2

Think before you measure:

1. Level of analyses
2. Choosing the species
3. Choosing where to study (lab, natural environment)
4. Choosing when to observe (time of day)
5. Effects of observer
6. Anthropomorphism
7. Ethical considerations

Moral
•Think about broader context –get too focussed and you miss a lot!

Question 3

In depth look at what you should think about before measuring behaviour;

1. Level of analyses
2. Choosing the species
3. Choosing where to study (lab, natural environment)

Answer 3

1. Level of analyses
   •The question you are interested in dictates level of analyses
   •Details versus high-level patterns (precise 3-d patterns of movement or general activity?)
   •Considerations; Distance and visibility (is human/animal in lab or in the real world, how much are you gonna be able to control this)

A fine-grained analysis is only appropriate for answering some sorts of question, and a full understanding will not necessarily emerge from describing and analysing behaviour at the most detailed level. While a microscope is an invaluable tool, in some circumstances it would be useless – say, for reading a novel. In other words, the cost of gaining detail can be that higher-level patterns, which may be the most important or relevant features, are lost from view. For example, recording the precise three-dimensional pattern of movements for each limb may be desirable for certain purposes, such as analysing the neurophysiological mechanisms underlying a particular locomotor behaviour pattern.

However, higher-level categories such as ‘walk’ or ‘run’ are often more appropriate. On a more practical note, recording large amounts of unnecessary detail may obscure broader issues, simply by presenting you with overwhelming quantities of data to analyse and interpret. The pattern of events you notice is affected by spatial scale: being close to a subject reveals details that would not be seen at a distance; being at a distance may reveal the wider context in which the behaviour is being expressed. Imagine that you are an observer among a group of migrating wildebeest on the Serengeti plain of East Africa. The impressions you would form close up to the animals would be totally different from those obtained from an aeroplane, where you would see the long columns of animals. Similar arguments apply to the time scale of measurements.

2. Choosing the species
•Right for research question (e.g. lifespan)
•Previous research and knowledge
•Sample –enough?
•Visible? (habitat, size, speed)
•Habituated? (response to humans)
•Daily activity? (diurnal/nocturnal)
•If in captivity –financial/practical considerations, breed, dietary requirements.

some questions to think about when choosing a species in pg 14 of measuring behaviour book

3. Choosing where to study
   •Captivity or wild?
   •Field observations –ecologically valid and generate hypotheses. any environment in which individuals can range freely and interact with their own and other species. The observer notices the circumstances in which an activity is performed and those in which it never occurs, thereby obtaining clues as to what the behaviour pattern might be for (its function) and how it is controlled (its proximate causation). it uncovers aspects of behaviour that would not otherwise be known about, providing the raw material from which research questions and hypotheses can be formulated. Moreover, it provides an understanding of how an animal’s behaviour is adapted to the environment in which it normally lives, in the same way that its anatomical or physiological characteristics are adapted.
   •Hypotheses testing in controlled experimentation in captivity
   •Experimental (in which conditions are actively manipulated) or observational

issues with field observations:
- An animal under observation may frequently disappear from view, wrecking the best-laid plans for
systematic recording over a fixed period.
- ensuring randomness in the choice of individuals to be studied within a population may not be easy because some identified individuals are difficult to find.
- Bad weather may make observation impossible.
- The animals may prove to be much more shy than had
been expected and require months or years of habituation before they will allow you near enough to make useful observations.
- The conditions for recording behaviour in the field are rarely ideal and high-quality data are not easy to collect. - In pursuing the possibility of working on an animal in the field, you need to discover whether you must
overcome practical problems such as requiring special permissions from governments or landowners.
- How easy will it be to get to where you want to be and to live when you get there?

A solution to this problem is to conduct controlled field experiments, so that certain aspects of the
environment can be manipulated or controlled. An example of a controlled field study comes from work on egg discrimination in free-ranging weaverbirds in the Gambia, West Africa (Lahti & Lahti, 2002). In order to assess the birds’ abilities to discriminate between their own eggs and those from other parents, the nests were artificially ‘parasitised’ with the eggs of other weaverbirds. To control for disturbance of the nests, eggs were removed from other nests, marked and replaced.

Question 4

Compared to studying captive animals, in the wild often difficult to

Answer 4

•Observe directly and frequently (finding them! obscuring vegetation, data biased by species, activities, time of day, etc)
•Identify individually (history)
•Have experimental control
- Naturally occurring comparisons
- But possibilities of field experimentation
——————————————————————————
West and Packer 2002:
from the picture of lions, realise two of them are not real, they are comparing the colour of the mane and seeing which of these lions the lionesses prefer. lionesses prefer lions with darker manes; sign of good genes.
——————————————————————————–
4. Choosing when to observe
•Consider your research question
•Data collection must be representative
•Pre-determined schedule
•choosing appropriate season and time of day; because the nature of behaviour will change with both.
•Solutions for 24/7 activity budgets (multiple days & individuals)
- 24hr continuous recording (strongest)
- Different time each day to build up 24/7 picture
- Random time sampling (weakest)
- Must avoid time of day confounds

5. Effects of observer
   Hawthorne effect in humans (we change our behaviour when we know we are being observed)/observer-expectancy effect (observers affecting studies).
   Both captive and field studies
   Solutions:
   •Hides to conceal yourself (but limit mobility)
   •One way glass be placed between you and the animals being observed. This technique for concealing the observer relies on the illumination being much brighter on the subject’s side of the screen. (captivity, beware mirror effects)
   •Habituation (beware subtle effects on activity, predation)
   •concealed Video/camera trap and watch the behaviour on a remote screen. The ability of video to capture detail may not be as good as direct observation, but this method has the advantage that a permanent record can be kept, it may capture rapidly occurring events that are easily missed, and the record can be analysed many times.

Question 5

11 different steps involved (all inter-related)

Answer 5

1.Ask a question (Broad or specific)
The particular choice of question (or questions) may be influenced by a variety of factors, including previous
knowledge, interests and observations made in the course of other research and the priorities of the group in which you work. Sometimes the impetus for a study stems from little more than a hunch or from a wish to see what an animal will do next.
——————————————————————————-
2.Make preliminary observations
A period of preliminary observation is generally invaluable in deciding what measurements to make and should be regarded as a crucial part of any study. Jumping straight in and collecting ‘hard data’ from the very beginning is rarely the best way to proceed.
Quantitative recording of behaviour should be preceded by a period of informal observation, aimed at understanding and describing both the subjects and the behaviour you intend to measure.

Preliminary observation is important for two reasons: first, because it provides the raw material for formulating questions and hypotheses; and second, because choosing the right measures and recording methods requires familiarity with the subjects and their behaviour. Preliminary observation is especially important if the problems or animals are new to you. After a period of trial recording sessions, in which
behavioural categories and measurement techniques have been tried out, preliminary data should be analysed. It is at this stage that methods should be modified if necessary.
•What fed?
•When fed?
•Individual identification?
•Visibility?
—————————————————————————–
3.Identify behavioural variables to be measured
Measures must provide best account of what you have observed. Definitions of behavioural categories should be clear, comprehensive and unambiguous. Write down the definitions before starting to collect data.
e.g. What behaviours would indicate aggression?
•Physical contact? Biting, hitting, chasing
•Aggressive displays? (subordination) Piloerection, swaying (shrinking posture)
•Aggressive vocalisations?
•Facial expressions?

How shall we measure aggression?
Number of occurrences? (Can calculate rate/hr)
Duration, Intensity,
Type (Contact and non-contact aggression)
Separate out aspects?
•Threat displays and submissive gestures
———————————————————————————
How to Describe Behaviour:
Behaviour can be described in a number of different ways,

Structure–appearance, physical form or temporal patterning of the behaviour (i.e. the subject’s posture or movements (“press switch down using index finger” is a structural description))

Consequence–the effects of the subject’s behaviour on the environment, on other individuals or on itself (“turn on light”)

4.Choose recording methods - MAIN THEME
•Cannot record everything! because any stream of behaviour could, in principle, be described in an enormous number of different ways.
•Decide best way, practice and assess reliability and validity of categories
•measure inter-and intra-observer reliability at the
beginning and end of data collection
•Drop unreliable and irrelevant measures, and tweak/redefine during pilot

Collect and analyse the data
•use same measurement procedures throughout
•Decide how much data your collecting in advance (Power calculations)
•Decide how to analyse before you collect the data!
Interesting evolutionary comparisons with humans
- use appropriate statistical tools for analysing the data

Question 6

11 different steps involved (all inter-related)

Answer 6

6. Formulate precise hypotheses
   •A hypothesis is a proposed explanation for a phenomenon
   •A clear scientific hypothesis invites a direct test, but remember that hypotheses may be tested by observing natural variation in a population as well as by performing experiments
   - Formulating hypotheses is a creative process, requiring imagination as well as some knowledge of the issues involved.
   - think about the weaknesses of their proposals as a guide to formulating a new hypothesis.

7.Make predictions from the hypotheses
A clear hypothesis should, by a process of straightforward reasoning, give rise to one or more specific predictions that can be tested empirically. The more specific the predictions are, the easier it
usually is to distinguish empirically between competing hypotheses, and thereby to reduce the number of different ways in which the results could be explained. A failed prediction may not necessarily mean that the hypothesis was wrong; the method of testing it may have been at fault
e.g.
Hypothesis: Desirability and distribution of food will impact aggression levels: direct prediction:
1.Preferred food will lead to more aggression than less preferred.
2.Defendable food (clumped) more aggression than when scattered?

A clear hypothesis invites a direct test

8.Design the tests
Even if steps 2 to 4 have already have been passed through, it may be necessary to consider them again. The variables that are to be measured should then be chosen so as to provide the best test of the different predictions made by competing hypotheses. Experiments can be greatly improved by careful design and thoughtful use of control groups. Good design allows fewer subjects to be used overall when several treatments are combined in a single experiment (see Chapter 8). Moreover, if the treatments combine to influence the outcome, such interactions between
them will be uncovered. The numbers of subjects used must, however, not be so small that the study will reveal no clear conclusions. Advice from colleagues expert in the design of experiments is almost always helpful.
e.g.
- Preferred food will lead to more aggression than less preferred. Examples of food?
Preference tests?
Compare banana/fruit vs veg?

• Defendable food (clumped) more aggression than when non-defendable (scattered)?
  Compare across feeding methods (in one location, vs scatter feed).
  What food characteristics?
  Identical food, size etc

9.Run tests on hypotheses (i.e. collect data/do the experiment)
Use the same measurement procedures throughout and try, if possible, to collect data ‘blind’ so that you do not unconsciously select data that fit your hypotheses. Stop collecting data when you have reached a pre-determined threshold that enables you to provide clear and reliable answers to your questions. Some inexperienced observers will stop when they get a statistically significant result, forgetting that with small sample sizes flukes are more likely to arise. On the other hand, do not go on collecting data simply because it is possible to do so. Once collected, make
sure that your data are properly labelled (checksheets), dated and include crucial information about the conditions and by whom the data were collected.

10.Analyse results
- prepare data in spreadsheets so that they are easily inspected and made available for subsequent statistical
analysis
- Combine categories and guard against some common mistakes when treating data points as though they were independent.
Beware non-independence!
- Employ the appropriate statistical tools, both for presenting and exploring the data, and for testing the hypotheses.
- Carry out exploratory data analysis to obtain the maximum amount of information from the data and to discover unexpected results that generate new questions.
- Do not sacrifice clarity for complexity by using unnecessarily complicated statistics.
- Use confirmatory analysis to test hypotheses.
- Distinguish between testing existing hypotheses and generating new ones (see Chapter 9).

11.Consider alternative interpretations
Do not draw conclusions beyond evidence, but do try to formulate a list of questions and ideas suggested by
the data that could form the basis of future research. Be prepared to consider a range of alternatives. Try to come up with the best explanation that would, if correct, fit with background knowledge. The aim is to unify and offer coherence where it might have been lacking before.

Question 7

4.Choose recording methods - MAIN THEME

Answer 7

•Behaviour is a continuous stream of movements and events (difficult to capture behaviour into distinct behaviours and categories)
•Need to divide into discrete units or categories
•Suitable categories are dependent upon the research questions.
–Example gorilla feeding
———————————————————————————
Choosing categories –general guidelines:
•Enough categories to describe behaviours in sufficient detail to answer the questions and, preferably, to provide some additional background information.
•Precisely defined and summarise as much relevant information about the behaviour
•Usually independent of each other (i.e. not different ways of measuring same thing); that is, two or more categories should not be merely different ways of measuring the same thing.
•Be homogeneous – means that acts included in the same category should share same properties (e.g. displacement activity).
———————————————————————————
When choosing categories it can sometimes be helpful to have descriptions of the main types of behaviour pattern that typify the species. Often when we are trying to decide what behaviours we go to the Ethograms.

Ethograms
•A catalogue of descriptions of discrete species-typical behaviour patterns that form the basic behavioural repertoire of the species.
•Can also have vocal ethograms
•Useful for learning about a species and promotes valid comparison across studies
Problems -
•Unavailable for many species (require in depth studies)
•Not all individuals behave in a species-typical way

Question 8

3. Identify behavioural variables to be measured

4. Choose recording methods

Answer 8

Defining behavioural categories
•Each category clearly, comprehensively and unambiguously defined, using easily understood criteria.
•These criteria should unambiguously distinguish it from other categories.
•Decided during informal observations and pilot data collection
•Write these down before you start.

A detailed and complete definition of each category and the associated recording method should be written down before the data used in the final analysis are collected. Two types of definitions are used:

• Operational definitions specify the physical operations that are required by the researcher to make the
  measurement; these are most commonly used when measuring the consequences of behaviour.
• Ostensive definitions involve giving an example of the case to which the category applies using diagrams and
  written descriptions, for example when an individual plays with an object. These definitions are most often used in the direct observation of behaviour. An ostensive definition of an action should enable another observer to recognise the same pattern of behaviour.

Recording behaviours - Types of measures
- Latency is time from onset of event to specified behaviour. need unit of measurement (sec, min, hrs, yrs etc.)
- Frequency is total no of occurrences by total observation time. need unit of time.
- Duration; need unit of measurement AND over time period. Often use proportion or percent.–Total duration is 6 min in one hour (=0.1, or 10%) – Mean duration –Need to know total time to interpret behaviour
- Intensity is the amplitude of behaviour as represented by height – unlike latency, frequency and duration, has no universal definition –depends upon behaviour. Sometimes presence or absence of components of act (aggression)
—————————————————————————
The different levels of measurement:
Nominal - If observations are simply assigned to mutually exclusive, qualitative classes or categories, such as male/female or active sleep/quiet sleep/awake, then the variable is measured on a nominal (or categorical) scale. If only two outcomes are possible (e.g. yes/no, or male/female) the data are said to be binary.
Ordinal - If the observations can also be arranged along a scale according to some common property then the
variable is measured on an ordinal (or ranking) scale. The number assigned to each measurement is its rank. For example, if in a given period of time individual A played more than B who played more than C, then A would be ranked highest on this measure, B next and C lowest.
Interval - If, in addition, scores can be placed on a scale such that the distance between two points on the scale is meaningful – i.e. the difference between two scores can be quantified – the variable is measured on an interval scale. The zero point and unit of measurement are arbitrary for an interval scale. A temperature measured in degrees Celsius is measured on an interval scale.
Ratio - The highest level of measurement is attained when the scale has all the properties of an interval scale but also has a true zero point. This is referred to as a ratio scale since, unlike an interval scale, the ratio of any two measurements is independent of the unit of measurement. Mass, length and time are measured on ratio scales. For example, the ratio of two periods of time is the same whether they are measured in seconds or days. True frequencies, durations and latencies are measured on ratio scales.
——————————————————————————–
Research Q -What is the best way to feed the chimpanzees to ensure equal food intake?
Design –compare scattered and clumped food

Latency (is time from onset of event (the moment the food is placed) to specified behaviour (the first time they put an item of food in the mouth)
•to put first item of food in mouth from food delivery –(predict quicker in scattered vs clumped)
•first outbreak of aggression from initial food delivery –(predict quicker in clumped vs scattered)

Frequency (is total no of occurrences by total observation time)
•Average number of times chimpanzees put food in mouth (predict more similar across individuals in scattered vs clumped, ieSD or SE lower in scattered than clumped)
•Total frequency of aggressive interactions (predict less in scattered vs clumped)

Duration(time spent in seconds by total observation time)
•Time spent foraging and feeding (predict more similar across individuals in scattered vs clumped, ieSD or SE lower in scattered than clumped)
Or
•Total time spent engaged in aggressive interactions (predict less in scattered vs clumped)

Intensity (is the amplitude of behaviour) Tricky!:
Intensity of aggressive interactions (contact vs non-contact aggression (predict more contact in clumped vs scattered)

How we are going to plot the data?
mean/median latency (in seconds) to put food in mouth after food delivery in two food distribution conditions (n=14 chimpanzees) with minimum and maximum values noted (or SE’s).

expect the Mean/Median latency in seconds to be lower in the scattered food condition, and higher in the clumped food condition.

Question 9

(continuation)

Answer 9

How to cope with out of sight (Oos)
•Stop the clock and start again when the animal is back in sight (so, e.g. 20 min in total, but may watch for 30 min, but Oos for 10)
•Set minimum time required and adjust (e.g. need minimum of 15 min/20 min session)

Please calculate
You observed 5 aggressive events in a 20 minute observation session, but for 5 minutes the chimpanzee was out of sight.
What was the rate per hour?
5 in 15 (60/15 = 4), therefore 5 events x 4 = 20/hr•

You counted 12 aggressive events in a 15 minute observation session but for 5 minutes out of sight.
What was the rate per hour?
12 in 10 (60/10 = 6), therefore 12 events x 6 = 72/hr

Question 10

Behavioural events vs behavioural states, and identifying individuals

Have to decide here which animals to watch and what to record from them.

recording methods and rules

Answer 10

Recording behaviours
Events and states
•Sampling methods: Having decided on the questions, and which behaviours to collect to be able to answer these questions, you have to decide which animals to watch and how.
•When choosing the type of measure to describe a behaviour pattern, it is helpful to distinguish between two fundamental types of behaviour pattern which lie at opposite ends of a continuum. There are two fundamental types of behaviour pattern, which lie at opposite ends of the continuum:
- Events are behaviour patterns of relatively short duration, such as discrete body movements or vocalisations, which can be approximated as points in time. The salient feature of events is their frequency of occurrence. For example, the number of times a dog barks in one minute would be a measure of the frequency of a behavioural event.
- States are behaviour patterns of relatively long duration, such as prolonged activities, body postures or proximity. The salient feature of states is their duration (mean or total duration, or the proportion of time spent performing the activity). For example, the total time a dog spends asleep over a 24-h period would be a measure of the total duration of a state. (Note that the term ‘state’ is also used in the behavioural literature to refer to a motivational state, such as hunger or thirst, so it is important not to confuse the two.)
- The onset or termination of a behavioural state can itself be scored as an event and measured in terms of its frequency.
———————————————————————————
Identifying individuals:
Focusing on the behaviour of an individual in a group
is virtually impossible without a way of distinguishing reliably between one individual and another. Moreover, when differences in behaviour of known individuals are recorded, the resulting data are likely to be much more
informative. Only by identifying and watching individuals does it become clear that all individuals in a species do not behave in the same ‘species-typical’ way.
•Vital for many studies
•Individual differences of interest
•Marking raises issues and does not last long under field conditions (welfare, changing own or others’ behaviour)

example; research revealed that coloured plastic leg bands placed on male zebra finches affected how attractive they were to members of the opposite sex. Female zebra finches preferred males wearing red leg bands over unbranded males, while males preferred females with black leg bands. Both males and females tended to avoid members of the opposite sex wearing green or blue leg bands (Burley, 2006). These findings clearly show that for zebra finches, and probably many other species, methods conventionally used to identify them can have a significant effect on
behaviour.
—————————————————————————
Recognising Individuals:
In some species, individuals have naturally distinctive markings.
Capuchin and Squirrel Monkeys
Capuchin - differences in faces
Squirrel - much more difficult to tell apart, so have collars on them to tell them apart.

For example, zebras’ stripes are like human
fingerprints, no two individuals being identical. . Identifying animals by naturally occurring features can be difficult and requires patience and practice, but it is the best approach in terms of minimising suffering and disruption.
When experienced observers are convinced that they can recognise individuals without recourse to written
records, some verifiable demonstration of their ability is advisable. One technique is to photograph the animals as they are simultaneously identified by the observer. The tester then removes from each photograph additional environmental cues that might be helpful in identification, and records the identity of each individual. Days or weeks later, the tester presents the observer with the pictures in random order and asks the observer to name each individual.
———————————————————————————
Individual differences:
Caterpillars are faced with very different requirements from those of the adult butterflies and moths of the same species. Female mammals caring for their young have very different priorities from those male mammals that play no part in parental care. Individuals of the same age and sex may exhibit qualitatively different modes of behaviour because they differ genetically or because they or their parents have developed under different environmental conditions (Jablonka & Lamb, 2005).
Functional and evolutionary analysis suggests that
much variation in behaviour within members of the same species of the same age and sex has been maintained by Darwinian evolution. Two or more types of reproductive tactic such as holding a harem or sneaking copulations may co-exist within populations of a given species.

Assessing individual distinctiveness:
Observers’ ratings of animals can also provide useful information about subtle aspects of an individual subject’s style that is not easily obtained in other ways. Rating methods of animals’ personalities can capture the overall pattern of an individual’s behaviour that remains elusive when discrete events are measured. The overall pattern involves behaviour occurring in a wide variety of conditions and it may also take into account what happens in complex social interactions.

For example Wemelsfelder, Hunter, Mendl et al. (2001) gave nine observers freedom to choose their own descriptive terms when watching the behaviour of pigs. The independent ratings showed significant agreement in the terms applied to each pig such as ‘confident’, ‘nervous’, ‘calm’ or ‘excitable’. One of the most
successful and reliable distinctions applied to personality differences in animals of a variety of different species is between individuals that are bold and those that are timid (Carere & Eens, 2005).

Defining a group:
In practice, the rules for defining groups are usually implicit: groups can often be defined intuitively by assessing how the animals are distributed in space and observing the relative distances between individuals. Those below a certain distance are regarded as being within a group, and those above it as outside.

For example, more cryptic species – such as lemurs –
that tend not to emit loud vocalisations and occupy forest environments, will have a smaller detection distance than larger more vocal species occupying open environments, such as baboons. This informal approach can cause difficulties if the groups are not tightly clustered or if individuals spend only some of their time together. It is important therefore to make the rules for defining a group as explicit as possible, and to use carefully chosen criteria when deciding on the critical distances and minimum time spent with others that define membership of a group.

Summary
In many studies individuals must be identified. Such identification is particularly important when individuals differ greatly from each other, either because of genetic (or epigenetic) differences or because their early environments induced substantial differences in phenotypes. Such behavioural variation in animals may be likened to personality differences in humans. In some studies the social group must be defined.
——————————————————————————-
Recording Methods:
Behavioural Sampling rule - who or what is watched and when.
- Ad libitum sampling; recall what seems to be interesting at the time that seems relevant. the most problematic; no systematic constraints, observations will be biased towards those behaviour patterns and
individuals which happen to be most conspicuous. but useful for preliminary observations
- focal sampling; focusing on one individual or object for a specified amount of time and how individuals interact with that object, under continuous observation. often best way when have individual identification, and want detailed records of events. choice of focal determined in advance. perfect if many observers. can be particularly difficult under field conditions because the focal individual may leave the area and disappear completely
- scan sampling; a whole group of subjects is rapidly scanned, or ‘censused’, at regular intervals and the
behaviour of each individual at that instant is recorded. excellent for only behavioural states and when many individuals and only one observer. Scan sampling results may be biased because some individuals or some behaviour patterns are more conspicuous
than others. For example, de Ruiter (1986) used it to study the activity of two groups of wild capuchin monkeys. The technique enabled him to obtain
data that were evenly representative across all individuals, time of day and season, allowing various behavioural and ecological comparisons to be drawn between the two groups. Such a broad spread of data would not have been practicable had focal animal sampling been used.
- behaviour sampling; observer watches whole group and records each occurrence of each type of behaviour and what individuals are involved. good for rare but significant types of behaviour (e.g. fights, copulations).
More than one technique may be employed simultaneously.

Scan sampling can be used in addition to focal sampling during the same observation session. The behaviour of a focal individual may be recorded in detail, but at fixed intervals (say, every 10 or 20 min) the whole group is scan sampled for a single category, such as the predominant activity or proximity to another individual.
Behaviour sampling is often used in conjunction with focal or scan sampling and is subject to the same source of bias as scan sampling, since conspicuous occurrences are more likely to be seen. Indeed, behaviour sampling is sometimes referred to as
‘conspicuous behaviour recording’.
——————————————————————————-
Recording Rule - how behaviour is recorded
•Continuous sampling (CR) - record all behaviours during the observation period; gives a real time record of exactly what the subject is doing. Events and/or states, but usually need computer for states!
•Time sampling - record the behaviour at regular pre-determined sampling intervals = instantaneous point sampling, where you choose a time interval (e.g. 30 secs) and on the dot of every 30 secs you record the behaviour (e.g. foraging). Gives an estimate of the proportion of time spent in that activity. Time sampling only for behavioural states not events. Less information is preserved and an exact record of
the behaviour is not necessarily obtained.

Time sampling in its turn can be sub-divided into two principal types:

• You can combine instantaneous point sampling (IS) with continuous recording of all occurrences of events (to give a frequency of events).
• Another method to use with time sampling is one/zero sampling. Observer records only once if specific behavioural category occurs at any time during the time interval (frequency and duration ignored). One zero is appropriate or both events and states.

Choosing the sample interval:
Division of an observation session into sample intervals (successive short units of time) for time sampling.
The size of the sample interval used in time sampling will depend on how many categories are being recorded, as well as the nature of the behaviour. The shorter the sample interval, the more accurate a time-sampled record will be. However, the shorter the sample interval, the more difficult it is to record reliably several categories of behaviour at once – especially if the activity is complicated or occurs rapidly

The disadvantages and advantages of time sampling:

Time sampling methods are not perfect. Neither instantaneous sampling nor one-zero sampling gives accurate estimates of frequency or duration unless the sample interval is short relative to the average duration of the behaviour pattern (see Fig. 5.5) – although analytical techniques are available for deriving unbiased estimates of frequency and duration from time-sampled measures. Moreover, time sampling is not generally suitable for recording sequences of
behaviour unless the sample interval is very short. This is because with one-zero sampling two or more behaviour patterns can occur within the same sample period, while instantaneous sampling can miss changes in behaviour that occur between sample points.
MORE ON PG36
——————————————————————————–
Summary
Sampling may be done by focusing on a given individual, scanning a group of individuals or focusing on a behaviour pattern of particular interest. Recording may be done continuously, which is ideal but may be impractical or unreliable. Alternatively, recording may be done by time sampling. Two methods of time sampling involve measuring what is happening at particular moments (instantaneous sampling) or recording whether or not a given event has occurred within a given period (one-zero). Both techniques are easier to use than continuous recording but, without taking certain precautions, they may introduce inaccuracies.

Question 11

Time interval
Disadvantages and advantages of time sampling
Recording mediums
Automation

Answer 11

Time interval:
•It is important to choose an appropriate time interval for the behavioural categories you wish to record. If the time interval is too short (such as 5 seconds) you can’t record accurately, if it is too long, then you miss lots of behaviour.
———————————————————————————Disadvantages and advantages of time sampling:
Disadvantages
•Provides an estimate only (not true frequencies or durations)
•Not normally suitable for sequences (e.g. reciprocity)

Advantages
•Instantaneous sampling gives a reasonable estimate
•Larger number of categories can be recorded and simpler/less demanding than continuous recording
——————————————————————————–
Recording mediums:
Checksheets
Advantage - Simple, cheap, flexible and powerful, can provide reliability and validity
Disadvantage - Two-stage transfer to computer, recording precise durations can be difficult, harder for the results to be transcribed (should be analysed on computer)

Verbal descriptions (written or audio)
Advantage - Great for informal pilot data and recording rare events, especially if systematic, necessary to record complex behaviour
Disadvantage - Exceedingly time consuming to code.

Video recordings
Advantage - Exact visual (and perhaps audio) record. Can slow down. Can be analysed repeatedly and in different ways. Can do intra-observer reliability.
Disadvantage - Exceedingly time consuming (and dull!) to code. Restricted view.

Automatic recording devices
Advantage - Automatic! (e.g. actimeter).
Disadvantage - Still essential to watch.

Event recorders
A computer event recorder is simply a portable or hand-held computer or personal digital assistant (PDA) with the appropriate software that enables you to record behavioural observations directly on to the machine. For most applications all that is needed is an ordinary computer or PDA and suitable event-recording software for encoding and interpreting the data inputs. With the right software, virtually any device will do.
Advantage - can record quicker, download to computer, accurate multiple durations, larger quantities of data can be stored
Disadvantage - not be readily adaptable as checksheets, possible to collect too much info, offer a degree of sophistication that may be unnecessary for some applications.

Automation:
using RFID technology to follow the movement of an animal around an enclosure without being there.

Question 12

Anthropomorphism

the attribution of human characteristics or behaviour to a god, animal, or object.

Answer 12

•Beware interpretation in relation to your emotions and intentions
•Occam’s razor –simplest way first
•Deception
•Sensory abilities and observer expectancy
——————————————————————————–
Humans readily interpret the behaviour of other species in terms of their own emotions and intentions. Observations of animals often leave the strong impression that the animals know what they are doing. However, subsequent analysis frequently reveals that seemingly complex and purposive behaviour can be produced by simple mechanisms that do not involve conscious awareness or intentions. Using human emotions and intentions as explanations for animals’ actions can impede further attempts to understand the behaviour.
Nonetheless, slavish obedience to such a maxim tends to sterilise imagination and, although the possibility of anthropomorphism must be acknowledged, an over-emphasis on its dangers can constrain research. If you never think of an animal as though it were a human you are liable to miss much of the richness and complexity of its behaviour. If an animal is invariably thought of as a piece of machinery then some of its most interesting attributes may be overlooked.
A preferable approach is to muster every possible type of mental aid when generating ideas and hypotheses, but to use the full rigour of analytical thought when testing them. Attributing human sensations, emotions and intentions to an animal so that you can do more imaginative science does not mean that, when your efforts are crowned with success, you have proved that it feels and thinks like a human. The distinction between the heuristic value of such projection and its
truth value needs to be explicit.
Another fundamental point is that other species occupy different perceptual worlds from humans – that is, their
sensory abilities may be radically different from our own. Many rodents communicate using ultrasonic vocalisations, some insects and birds can detect ultraviolet light, some snakes can detect prey using infrared sensors and many species have highly developed powers of olfaction.
Humans occupy a perceptual world that is dominated by colour vision, but this is not true for many other species. Thus, an animal may be oblivious to a visual stimulus that seems obvious to a human or, conversely, may respond to a stimulus that a human cannot detect or does not imagine is important to the animal.

Question 13

Ethical considerations

Answer 13

• insensitive to the condition and welfare of their animals
• if some form of experimental manipulation is involved or if the lives of the animals are disrupted.
• abide by the spirit (as well as the letter) of any legislation governing animal research
• should strive to minimise the number of animals used in research and the amount of suffering caused to each animal. can sometimes be reduced by more careful choice of research design, measurement techniques and statistical analysis

three difficult types of judgement to be made (Bateson, 2005a):
- how worthwhile is a proposed piece of research in terms of its likely contributions to scientific understanding?
- how likely is the research to bring benefits – for example to human and veterinary medicine, the economy or the environment?
- how much suffering to animals is likely to result
from the research?

These points are brought together in the ‘decision cube’ shown in Fig. 2.1. Limited animal suffering is treated as acceptable only when both the importance of the research and the probability of medical benefit are assessed as being high. Moreover, certain levels of animal suffering would generally be unacceptable regardless of the quality of the research or its probable benefit.
Research of high importance involving little or no animal suffering would be approved even if the work had no obvious potential benefit to humans. This recognises the need to understand phenomena that have no immediate and obvious benefit for humans. Such understanding is seen as a ‘good’ in itself. It
is often impossible to predict how understanding biology will help to advance medicine in the future. For instance, ornithologists studying bird migration played an important role in understanding the transmission of avian flu.

Summary
In planning research you should think carefully about the level at which you want to work so that you measure appropriate detail but not too much of it. You must take into account whether or not you have chosen the right subjects, where you will work on them and when. You must consider how much your presence will affect them and be careful about how much you project human intentions and emotions on to them. You must consider their particular perceptual abilities and you should be very sensitive to the ethical issues raised by working on animals and the different issues raised by working on humans. You should also consider any pressures arising from the way in which you are funded.