Observations Flashcards
What are observations?
Observations often form the starting point for any research in Psychology; we need to observe that people forget before we can study forgetting, we might watch children at play before embarking on research into the purpose of their play.
When using the observational method psychologists simply observe in real life situations or in the laboratory without manipulating an independent variable. THIS CAN BE EXTREMELY USEFUL WHEN IT IS NOT POSSIBLE TO MANIPULATE VARIABLES DUE TO ETHICAL OR PRACTICAL PROBLEMS OR SIMPLY AS A STARTING POINT FOR RESEARCH. Observations usually involve the categorisation or rating of behaviour in precise ways. The emphasis is placed on how people or non-human animals behave in specific situations; no attempt is made to influence the behaviour being investigated in any way.
Observations can differ in several important ways, depending on:
- The setting in which it is carried out (naturalistic vs. controlled)
- The role of the researcher (participant or non participant)
- The awareness of the participants involved (covert or overt observation)
- The amount of structure that is Imposed (structured or unstructured)
- Time or Event Sampling
- Inter-observer reliability
The study setting
- Observational studies are often carried out in a natural setting such as school, canteen or social area. Studies that focus on people’s naturally occurring behaviour are often referred to as naturalistic observations.
- Observational research can also be conducted in a laboratory which allows the researcher more control over the environmental in which the research is carried out. For example, Labs can be designed to resemble playrooms, once children have become completely accustomed to such a room, their behaviour may be observed through a one way mirror. These are mainly referred to as controlled observations.
The Role of the Researcher
A distinction is often made between participant observation, in which the observer actually joins the group of people being studied and non-participant observation, in which the observer remains external to those being observed. In participant observations the researcher may be most interested in trying to understand the meaning of the behaviours they are observing, in which case they typically produce detailed verbal qualitative descriptions of the behaviour they observe. Observers also have to decide whether to remain covert or overt. In overt observations the participants are made aware of the research that is taking place, therefore enabling it to be more ethical, or in covert observations they remain unaware that the research is taking place.
The structure of an observation
-The amount of structure in an observation will differ. To conduct an observation it is necessary to decide how the data will be recorded.
Unstructured observation
Unstructured observations can be conducted with the researcher simply writing down everything he or she observes and analysing it later by looking for patterns in the information that has been recorded. This is very useful if you are studying something for the first time and mainly produces qualitative data. This type of observation may lead you to generate hypotheses that could be investigated experimentally or suggest more structured observations that could be conducted. One problem with unstructured observations is that the behaviour recorded will often be those which are most visible or eye-catching to the observer but these may not necessarily be the most important or relevant behaviours.
Structured observations
The observer using a coding system/scheme in order to count the number of times specific behaviours occur. Behavioural categories are needed so that observer can count the number of times that each behavioural category occurs during a given period of time. This kind of approach produces quantitative data that can be analysed statistically. In order to record systematically, behaviours to be observed must be operationalised defined and measurable. Operationalisation involves breaking down the behaviour being studied into a set of components.
Strengths of controlled observations
-High levels of reliability
This is because of the high levels of control over the environment meaning that the observations can be replicated and findings confirmed or challenged.
Weaknesses of controlled observations
-Likely to have lower ecological validity
This is because they are potentially more artificial due to controlled nature and so may not generalise to real life behaviour.
Strengths of naturalistic observations
-High in ecological validity
This is because they are conducted in real life setting observing natural behaviour so more generalizable to real life behaviour
Weaknesses of naturalistic observations
-More potential for confounding variables to effect the results
This is because of the lack of controls over the situation participants are observed in.
Strengths of participant observations
-Allows observer to gather more detailed accounts due to flexibility.
This is because the observer is involved in the events they are observing so the researchers are more likely to get accurate data.
Weaknesses of participant observations
-May be a potential for observer bias as it is harder to remain objective.
This is because the observer is involved in what they are observing and so may find it difficult to remain detached.
Strengths of non-participant observations
-Reduces the effect of the observer on the participants behaviour.
This is because the observer does not interact with participants with participants so has no effect on behaviour.
Weaknesses of non-participant observations
-The observer may miss vital information
This is because the behaviour may be shown while the observer is making notes
Strengths of covert observations
-Reduces possibility for demand characteristics
This is because participants were unaware of the observation taking place therefore reducing potential for demand characteristics.
Weaknesses of covert observations
-Difficult to remain hidden
This is because the observer will still have to take notes and so the p may spot the researcher looking suspicious.
Strengths of overt observations
–Removes ethical issues such as invasion of privacy
This is because the p know they are being observed
Weaknesses of overt observations
-Potential for demand characteristics
As participants know they are being observed so may change behaviour to suit what they think the observer wants.
How to design an observation?
Once you have decided what behaviour you are aiming to observe and what type of observation you will conduct you also need to make decisions about precisely how you will gather data
What do you observe?
Behavioural categories are the operationally defined units of events used in a structure observation to break a continuous stream of activity into separate recordable events. They must be observable actions rather than inferred states,
for example, laughing and smiling could be observed but ‘being happy’ could not.
A coding frame is a system for differentiating behaviours to be recorded in an observation, adding detail to the behavioural categories. It may use numbers or abbreviations to represent things such as the severity or duration of a behaviour.
For example, in recording observations of children’s aggression, a researcher might have the behavioural categories of ‘hitting’, ‘biting’ and ‘shouting’. The coding frame could summarise these to H, B and S. When each one is recorded, further details might be scored on a scale indicating where the victim was hit (face, arms, legs) and how loudly the perpetrator shouted (on a scale of 1 to 3)
What is a key issue to consider in the use of coding schemes?
A key issue to consider in the use of coding schemes is if there is more than one person observing the same thing or different observers watching different individuals it is important that they agree on what constitutes certain behaviour. This can be achieved through comparing the observations of each researcher initially and checking if they match. If they do this is known as inter-rater or in this case inter-observer reliability.
To do this, firstly the observers would all need to watch the same individuals but record the participant’s behaviour independently. The results from each observer are then compared, using a correlation. If the observers are seen to agree and a positive correlation is established then inter-observer reliability can be said to be achieved.
When do you observe?
event sampling
A final consideration is for the observer to decide how often or how long they will observe for. One option is to use event sampling, in this kind of technique you list all the behaviours you might see and observe continuously ticking every time you see them.
Biting//// Kicking// Smacking/ Shouting//////
When do you observe?
time sampling
Alternatively an observer can opt for time sampling which is when the researcher observes behaviour at specific times during an event, rather than observing the whole event
If you were observing children at play over a 5 minute period, you may only look and record what they are doing every 20 seconds and you would record what you see at that point in time, then look away.
When do you observe?
time event sampling
If you were observing children at play over a 10 minute period, you may only look and record what they are doing every minute for 30 seconds you would record what you see for the entire 30 seconds (like event sampling for 30 seconds), so over 10 minutes you would observe 10, 30 second events.
Strengths of time sampling
-Easy to replicate as there is a specific structure, so other researchers can re-use the same time units - enabling reliability to be tested
Weaknesses of time sampling
Potential to miss behaviour that is occurring outside of the time when the observation takes place - so lacking in validity as there could be important behaviours ignored.
Strengths of event sampling
Allows observer to make note whenever the behaviour of interest occurs, which increases validity as you can be sure you are measuring all relevant behaviour.
Weaknesses of event sampling
Observer may miss some behaviours if lots of things occur simultaneously, as you could be making notes as the behaviour happens - so lacking in validity as there could be important behaviours ignored
Strengths of time event sampling
Easy to replicate as there is a specific structure, so other researchers can re-use the same time units - enabling reliability to be tested.
Weaknesses of time event sampling
Potential to miss behaviours as observation period is not continuous, or could miss if lots of behaviours occur simultaneously - so lacking in validity as there could be important behaviours ignored.
