Research methods

3 variables

Independent

Dependant

Extraneous

Independent variable

The thing that is changed throughout the experiment

Dependent variable

What is recorded in the experiment

Extraneous variable

Other factors that are control and kept constant through the experiment

Confounding variable

A variable that isn’t controlled and that could affect the data

Operationalisation of variables

Urned into a measurable form

Lab experiments

IV is manipulated by the researcher in an attempt to change the participants’ performance on the DV. Held in an artificial environment.

Lab experiment strengths

• Extraneous variables can be controlled, therefore cause and effect can be inferred and it can be stated that it is the IV that has an effect on the DV

• Greater control over of variables means the study can be replicated and if the same results are achieved it shows that it is reliable and valid - can be replicated

Lab experiment weaknesses

• Usually artificial task that are conducted in an artificial environment, which might lead to poor ecological validity, meaning that the results from lab experiments might not reflect real life situations

• There is a higher risk of demand characteristics - people might guess the true aim of the experiment if its too artificial and act how they think the experimenter wants them to, meaning it might lose validity

• There is an ethical issue as the IV is manipulated which could cause psychological harm to the participants in some experiments

Field experiments

The IV is manipulated but the study is carried out in the participants natural environment

Field experiment strengths

• Conducted in more natural environments which means the DV will be more realistic , leading to high ecological validity and the results can be generalised to real life situations

• Low risk of demand characteristics - isn’t artificial enough got the participants to guess the aim so results will be more valid

Field experiments weaknesses

• More difficult to control extraneous variables , so the researchers not be able to infer cause and effect and can’t claim that it is the IV that has an effect on the DV

• Lower control of variables means the study will be more difficult to replicate and the researcher won’t be able to see if the same results are achieved and can’t check the reliability or validity of the experiment - difficult to replicate

• Ethical issue is that the IV is manipulated, which could cause psychological harm to participants in some experiments

Quasi experiments

The experimenter doesn’t have control over the IV - instead the IV s a condition that already exists

Done in a controlled environment - evaluate as you would a lab experiment

Done in a realistic environment - evaluate as you would a lab experiment

Quasi experiment evaluation difference

Because the IV isn’t manipulated participants shouldn’t experience psychological harm

Natural experiments

The experimenter does not have control over the IV - instead it is changed by natural occurrences

Natural experiments strengths

• Conducted in a natural environment which means the IV is naturally occurring and the DV is a measurement of natural behaviour - high ecological validity and can therefore be generalised to real life situations

• Lower risk of demand characteristics - environment isn’t artificial so its hard for participants to guess the aim

• Because IV isn’t manipulated participants shouldn’t experience psychological harm

Natural experiments weaknesses

• More difficult to control extraneous variables so researchers might not be able to infer cause and effect and can’t claim that it is the IV that effected the DV

• Lower control over variables means it is difficult to replicate and the researches can’t see if the same results are achieved. Can’t check reliability and validity - difficult to replicate

Ecological validity

Wether the results can be applies to real life situations outside the experiment set up

Yes = high, no = low

Cause and effect

If control of extraneous variables are high, cause and effect can be inferred - IV caused an effect on DV

Control of EV high = yes, low = no

Demand characteristics

Participants guess the aim and act in a way they think the experimenter wants them to

Screw you effect

The participants guess the aim and act in a way to ruin the experiment

Hypothesis

Clear statement of what will happen

What’s included on a hypothesis

Operationalised IV and DV

Directional/ one tailed hypothesis

Predicts the direction of the effect or difference

If previous research shows a direction of the effect

Non- directional/ two tailed hypothesis

Predicts a difference but not in a certain direction

If previous research is contradictory or there is no previous research

Directional/ one tailed structure

IV condition one will have a higher/lower number of ___(DV) than IV two

Non-directional/ two tailed structure

There will be a difference in the DV between IV condition 1 and IV condition 2

Null hypothesis

There is no effect of the IV on the DV

Null hypothesis structure

There will be no difference in the DV between IV condition 1 and IV condition 2

Target population

All the members of the group from which the sample has been taken, who the research is aimed at

Sample

A small group taken from the target population who carries out the experiment

Representative

Be typical of the whole population and fairly represents them

Biased

If a sample isn’t representative

Sample size

• If too small, it won’t be representative and statistical testing might be inaccurate

• If its too big it will be expensive and time consuming - usually about 30 in experiments (15 in each condition)

Random sampling

Occurs when every member of the target population has an equal chance of bing selected

E.g. put all of target populations name in a hat and choose at random

Random sampling strengths

• Provides th best chance of an unbiased representative sample of a target population because everyone has an equal chance of being selected

Random sampling weaknesses

• Compiling a list of everyone in the target population is impractical and therefore it is rarely used

Stratified sampling

Involves dividing the target population into important subcategories and then selecting members of these categories in the correct proportions

Stratified sampling strengths

• Should be representative of the target population and the results can therefore be generalised

Stratified sampling weaknesses

• It can be difficult to identify important characteristics in.a target population

• Time consuming to do

Opportunity sampling

Selecting participants that are available at the time

Opportunity sampling strengths

• It is quick and convinient because the participants are easily available

Opportunity sampling weaknesses

• Unrepresentative and often biased as students are often used - students are more educated than other groups

Self selecting (volunteer) sampling

Individuals who have determined their involvement in the study, such as responding to a poster or or newspaper article

Self-selecting (volunteer) sampling strengths

• Continent because once the poster/advert is made it is merely waiting for participants to respond, therefore little effort is required

• As paticipants have volunteered it shows they have give their consent to take part before the study had started therefore it is ethical in terms of informed consent

Self-selecting (volunteer) sampling

• Volunteers are often different to the rest of the population, meaning the sample may e unrepresentative and the finding may np generalise

  • Often more outgoing, helpful, higher intelligence, like psychology

Systematic sampling

A sample is obtained by selecting every nth person, this numerical interval is applied consistently

Systematic sampling strengths

• Unbiased as participants are selected using an objective system, therefore the sample should be representative

Systematic sampling weaknesses

• Not truly random unless you select a number using a random method and start with this person, ten select every nth person

observations

Involve watching and recording behaviour.

There is no IV/DV (this would make it an experiment) ​

However observations are often used within an experiment as a way of assessing a DV. ​

naturalistic observation

• refers to the observation of behaviour in its natural setting. ​

• Researcher makes no attempt to interfere with or influence the behaviour.

controlled observation

• some variables in the environment are controlled by the researcher​

• Reduces the naturalness of the environment ​

• It may be conducted in the lab ​

overt observation

• The person being observed is aware of the observation ​

• Researchers try to be unobtrusive e.g. one way mirrors

covert observation

• The person being observed is unaware of being observed before or during the study ​

• They may be informed afterwards ​

non-participant observation

• Watching/listening to the behaviour of others ​

• No interaction with Ps

participant observation

• Observer is part of the group being observed ​

• Some interaction with the Ps ​

naturalistic observation evaluation

• High realism - behaviour is observed in a natural environment with no attempt to change it so research should have high ecological validity​

• Lack of control over variables, therefore replication to check for validity/reliability is difficult. It also means that conclusions can not be made about patterns of behaviour - so cannot infer cause and effect

controlled observation evaluation

• Lack of realism - some variables have been controlled so people may not act in a ‘normal’ way. Research may therefore lack ecological validity​

• High control over variables, therefore replication to check for validity/reliability is possible. It also means the researcher can make conclusions about patterns of behaviour - so cause and effect can be inferred

covert observation evaluation

• High internal validity -Ps are unaware that they are being observed so unlikely behaviour will be influenced by demand characteristics, social desirability bias and observer effects. ​

• Ethical issues: lack of informed consent (covert observations should therefore only be carried out in public places)

overt observation evaluation

• Low internal validity – when people are aware that their behaviour will be observed it is likely that it will be influenced by demand characteristics, social desirability bias and observer effects ​

• Lack of ethical issues - can get informed consent from Ps who can give agreement to take part​

participant observation evaluation

• Greater insights into behaviour as the observer is part of the group, this means the observer can understand the behaviours more. Because of this research should have high internal validity ​

• It is difficult to be an objective observer if you are part of the group. The observer may even effect the behaviour of the group merely by their presence. This could impact the internal validity

non-participant observation evaluation

• Data lacks richness as it might not be possible to understand the feelings and motivations of the group when you are not involved in the interactions, potentially lowering internal validity. ​

• Lack of direct involvement with the group ensures greater objectivity. It also means the researcher is less likely to impact the behaviour of the group, increasing internal validity. ​

observation general evaluation point

• Observer bias ​

• Observers are not passive, they select what behaviours to include in their checklist and will have pre-existing ideas about those behaviours. ​

• Therefore when conducting observations a researcher may experience expectancy effects​

• = see what they expect to see

event sampling

Record how many times a certain behaviour (event) occurs in a set amount of time

time sampling

You record what behaviours are occurring at different time intervals

recording observations

• Operationalised observer checklist:​

• A checklist of specific categories that can be ticked each time a behaviour occurs​

• This will produce quantitative data​

• If you are using different observers, they must have clear descriptions of the behaviours they are observing otherwise observers may have different interpretations of events​

• Observers should be trained in use of the checklist perhaps using video footage ​

• This will help ensure inter-observer reliability (all observers agree on number of behaviours observed- findings are consistent)​

pilot study observations

• A small scale trial run of your study carried out before your main study​

• This is to check that there are no problems with the procedures before you spend lots of time and money carrying it out​

• To check the behaviours in the list are appropriate and see if any behavioural categories are missing​

• To ensure that observers are clear about the behaviours that they are identifying so that tallies are accurate - behaviours in the checklist should be operationalised and without overlap. ​

• The length of the observation can be reviewed

• To check the inter-observer reliability - you may also need to check the training of observers to ensure inter-observer reliability is high. ​

• This can be done by getting the observers to use the checklist with video footage. ​

• If inter-observer reliability is low more training is needed/the checklist needs to be reviewed

primary data strengths

The researcher has control over the data this is because the data collection is designed to fit the aim of the study

primary data weaknesses

Designing, conducting the study and analysing results is very time consuming

secondary data

Information that was collected for a purpose other than the current one. The researcher could use previous data that they have collected or data collected by another researcher eg use of government statistics, medical reports, teachers' observations. Correlations and meta-analyses are likely to use secondary data.

secondary data strengths

Simpler, cheaper and quicker to access someone else's data

Such data will already have been analysed so it will be known whether data is significant

secondary data weaknesses

Data may not exactly fit the needs of the study

meta-analysis

This is a method for collecting data from many different studies that have investigated the same behaviour/topic. This means we can look at overall trends or patterns of behaviour (also known as effect size). Eg in attachment we will look at a meta-analysis which analysed children's attachment types all over the world. The overall effect was that most children are securely attached.

evaluation of meta analysis

Reviewing the results from a group of studies rather than just from one study can increase the validity of the conclusions because they are based on a wider sample of Ps. However, the research designs in the different studies may vary considerably which means that studies are not truly comparable. Putting them together to calculate the effect size may not be appropriate so conclusions may not actually be valid.

qualitative data

data in non-numerical form- e.g. transcript of an interview with an alcoholic explaining why he drinks. It allows a detailed response and can be used to express feelings. However it is open to bias and difficult to analyse.

quantitative data

behaviour measured in the form of numbers e.g. 56% of people like onions! It is easy to analyse. It is objective & less open to bias.

advantages of qualitative data

Gains access to thoughts and feelings that may not be assessed using quantitative methods with closed questions.

Provides rich details of how people behave because participants are given a free range to express themselves.

reverse for quantitative

weaknesses of qualitative data

More difficult to detect patterns and draw conclusions because of the large amount of data usually collected.

Subjective analysis can be affected by personal expectations and beliefs.

reverse for quantitive function

correlation

measure the relationship between 2 co-variables.

can be positive or negative

positive correlation

if both variables increase

+1

negative correlation

as one variable increases the other decreases

-1

no correlation

no relationship between the variables

0

correlation coefficient

between -1 and +1

strengths of correlations

• Measuring the strength of relationships: Correlational analyses provide valuable information on the strength of the relationship between variables

• Value when doing exploratory research: Can be used to explore relationships in complex situations and can suggest ideas for further research.

• More ethical than experiments because there is no manipulation of an IV, instead the researcher just takes two measurements from a participant.

weaknesses of correlations

• Causality: It is impossible to establish cause and effect using correlational analyses. The technique only measures relationships not the effect of an IV on the DV. For example, the finding that there is a strong relationship between the symptoms of schizophrenia and high levels of the neurotransmitter dopamine. However, we cannot say that high levels of dopamine cause schizophrenia, it could be that having schizophrenia causes changes in dopamine levels.

• Spurious relationships- can detect meaningless patterns. E.g. Snedecor (1956) reported a correlation of -0.98 between the production of iron in USA and the birth rate in Britain. This ability to detect nonsense relationships is a severe limitation.

• Measurement of non-linear relationships: Non- linear relationships cannot be measured by correlation. e.g. the graph below shows the relationship between level of attention and time of day. You can see that over the early morning there is an increase in attention but as lunchtime approaches the attention decreases. The positive and negative relationships cancel each other out when calculating a correlation coefficient and no correlation is shown even though the graph shows a clear relationship.

the difference between experiments and correlations

• Experiments investigate the difference between 2 or more conditions and therefore looks at how the IV affects the DV

• Correlations look at the relationship between 2 co-variables (no IV or DV)

• In lab experiments it is possible to infer cause and effect; it is NOT possible to infer cause and effect in a correlation

questionnaires overview

A questionnaire survey involves asking participants questions about, for example, their attitudes, behaviours or intentions. The use of a questionnaire allows the researcher to gain information from large numbers of people relatively quickly and efficiently. The aim is to obtain information from a specified population, usually by administering the questionnaire to a sample of this population. Questionnaire surveys may be carried out face to face, by post, telephone or the Internet.

questionnaires - closed questions

These are questions where the researcher determines the range of possible answers (participants often reply by ticking boxes or circling the appropriate answers, or rating on a scale). These questions are best used when factual information is required. They produce quantitative data that is easy to analyse but may lack realism due to the forced choices available.

questionnaires - open questions

These are questions where the researcher does not restrict the range of answers (e.g. what are your views on the death penalty?). These produce a greater depth of qualitative data but the data is much more difficult to analyse.

strengths if quedtionnaires

• Speed and cost - a large amount of data can be collected from a large number of participants quickly and cheaply compared to interviews. This means that larger samples can be obtained, if there is a larger sample the results from the study are more likely to be representative of the target population, therefore the results should be generalisable.

• Range of data - both qualitative and quantitative data can be obtained. A benefit of collecting qualitative data is that open questions provides rich detail about the topic. Whereas a benefit of collecting quantitative data is that data from closed questions is easy to analyse using statistics, meaning that the psychologist is able to compare answers from different groups.

weaknesses of questionnaires

• Untruthful answers - there is no guarantee that respondents will answer truthfully. This is due to Social desirability bias, participants may lie as they want their answers to be seen in the best light. This would result in low internal validity.

• Researcher effects - if a researcher administers the questionnaire personally then participants may be influenced by such things as the researcher's gender, age or ethnicity. Even unintentional smiles or frowns may have an effect. This would result in low internal validity.

• Difficultly with controls - it is hard to ensure that data is collected under controlled conditions when the questionnaire is not completed face to face.

• Different interpretations of questions - different participants may interpret the questions differently e.g. sometimes' may mean different things to different people. In an interview the participant would be more likely to ask for clarification if they didn't understand a question or were unsure how to respond.

• questions to avoid in a questionnaire

• leading questions

• ambiguity

• emotive questions

• jargon/technical terms

• double-barrelled questions

• negatives

• impossible questions

• questions to include in a questionnaire

• filler questions

• east questions

• lie detection questione

• tendency to answer yes

structured interviews

A pre-determined set of questions are asked in a fixed order. This is called an 'interview schedule' and it should be standardised so that each interviewer uses it the same way. Structured interviews are like a questionnaire except they are conducted face to face (or over the phone).

unstructured interviews

These interviews are far less rigid - just the topic for the interview is decided in advance. There are no set questions and it is more like a conversation. The interviewee is encouraged to expand on answers and is prompted by the interviewer.

designing interviews

Begin the interview with easy/neutral questions to make the participant feel relaxed and comfortable, this should establish good rapport. Avoid the same types of questions as with designing questionnaires e.g. leading questions

Recording the interview:

Note taking may be done to record participant's responses. But this is likely to interfere with interviewer's listening skills and it may make the respondent feel anxious about what is being written. Ps may also feel undervalued if they say something and the interviewer doesn't write it down. Because of this it is better to use audio or video recording, then use the recordings to write a transcript of the interview.

strengths of structured interviews

• The subject is unlikely to deviate from the desired topic as the questions are pre-set.

• It requires less training of the interviewer as they will be reading from a list of pre-set questions

• As questions are the same it is easy to replicate.

weaknesses if structured interviews

But the interviewer must follow the pre-determined structure so cannot follow new lines of enquiry which emerge from the Ps responses.

weaknesses of unstructured interviews

• The subject is like to deviate from the desired topic as the questions aren’t pre-set.

• It requires more training of the interviewer as they won’t be reading from a list of pre-set questions

• As questions aren’t the same it isn’t easy to replicate.

strengths of unstructured interviews

But the interviewer don’t have to follow the pre-determined structure so can follow new lines of enquiry which emerge from the Ps responses.