research methods Flashcards
Aim
A general statement of what the researcher intends to investigate. This is the purpose of the study.
E.g. An investigation into how colour of text affects recall of words.
Hypothesis
A clear, precise and testable statement that states the relationship between the variables to be investigated.
e.g. words will take longer to recall if they are presented in the same colour as the meaning of the word.
Variables
A factor that can change in an investigation.
Independent Variable (IV)
An aspect of the experimental situation that is manipulated by the researcher or changes naturally so the effect on the DV can be measured.
Dependent Variable (DV)
The variable that is measured by the researcher. Any result/change on the DV should be caused by the change in the IV.
directional hypothesis
states the kind of difference or relationship between the IV and the DV. It is sometimes called one-tailed hypothesis.
non-directional hypothesis
simply predicts that there will be a difference between conditions (but does not predict which direction it will go). It is sometimes called two-tailed hypothesis.
Extraneous variables
Any variable apart from the IV which can effect the DV if not controlled. However they can be maintained or eliminated.
Confounding Variables
A variable apart from the IV which can effect the DV. However it can’t be controlled.
Demand Characteristics
Any cue from the researcher or research situation that can be interpreted by participants as revealing the purpose of the investigation leading them to changing their behaviour.
Experimenter Effects
Where the experimenter changes a persons views usually sub-consciously through body language.
How can we control extraneous and confounding variables?
- Standardisation
-All participants should be subject to the same experimental condition (i.e. environment, time etc).
-Standardised instructions (scripted/recorded). - Randomisation
-Using chance in order to control for the effects of bias in an experiment.
-E.g. Randomly allocating questions to participants.
Conditions could be randomly allocated.
Participants
people who take part in research.
Population
the group of people from whom the sample is drawn.
Bias
when certain groups are under- or over-represented in a sample (not representative).
Random Sampling
Every member of the target population has an equal chance of being chosen.
1.Compile a list of all target population.
2.Assign each name a number.
3.Select sample using a random number generator (e.g. computer or draw from a hat).
advantages of random sampling
+No researcher bias.
+Confounding variables should be distributed evenly – improves validity.
disadvantages of random sampling
-Difficult to get a complete list of target population.
-Time consuming.
-Participants may refuse to take part.
-May randomly draw a non-representative sample.
Systematic Sampling
Every nth member of the target population is selected (e.g. every 5th person on the register).
1.Create a list of the target population in an order (e.g. alphabetical) – this is the sampling frame.
2.Take a sample from this list.
advantages of systematic sampling
+Objective - avoids researcher bias. Researcher has no influence once system is chosen.
disadvantages of systematic sampling
-Could still draw a non-representative sample.
-Time consuming and costly.
-Participants may refuse to take part.
Stratified Sampling
Composition of the sample reflects proportions of certain subgroups in the target population.
1.Identify different subgroups in the population.
2.Work out proportion of each group.
3.Participants in each sub group are selected randomly in the same proportions as the target population.
advantages of systematic sampling
+Avoids researcher bias.
+More representative of the whole population so findings are more generalizable.
disadvantages of systematic sampling
-Stratification is never perfect. Complete representation of population is not possible.
Opportunity Sampling
Sample from people who are available and willing when the study is carried out (e.g. psychology undergraduates).
advantages of opportunity sampling
+Quick
+Convenient
+Less costly
disadvantages of opportunity sampling
-Unrepresentative of the target population – cannot generalise.
-Researcher controls selection so may avoid certain people – researcher bias.
Volunteer Sampling
Self-selected sampling – participants become part of the study when asked or in response to an advert.
advantages of volunteer sampling
+Convenient.
+Less time consuming.
+No researcher bias.
disadvantages of volunteer sampling
-Often unrepresentative. Volunteers may have similar profile (e.g. people with spare time)
– volunteer bias.
name three experimental designs
independent groups, matched pairs and repeated measures
Independent Groups
-Participants only take part in one condition.
-Requires a separate group for each condition (e.g. control & experimental requires 2 groups.
-Results for each group are then compared, usually by comparing mean results.
advantages of independent groups
+Avoids order effects.
+Reduces demand characteristics (from guessing the aim).
disadvantages of independent groups
-Needs lots of participants – costly.
-Differences between groups (participant variables) may affect the result.
(Random allocation can help to overcome this.)
Order Effects
When the order of the conditions in an experiment has an effect on participant behaviour.
Repeated Measures
-Participants take part in all conditions.
-Everyone experiences the control and the experimental condition.
-Results for the two conditions are then compared.
advantages of repeated measures
+Avoids participant variables as everyone does all conditions.
+Fewer people needed.
disadvantage of repeated measures
-Order effects more likely. Requires counterbalancing.
-Demand characteristics more likely as participants are more likely to guess the aim.
Counterbalancing
An attempt to control order effects in a repeated measures design
e.g. ABBA where when group does the experimental condition first where as the other does the control condition.
Matched Pairs
-Participants are matched in each condition for any characteristics that may affect performance. E.g. age, gender, IQ.
-Results are compared between members of each pair.
advantages of matched pairs
+Reduces participant variables.
+Reduces order effects and demand characteristics.
disadvantages of matched pairs
-Very time-consuming (costly)
-Impossible to match pairs exactly (even for twins). May be unexpected confounding variables.
Lab Experiment
an experiment carried out in a controlled setting.Participants know they are being studied.
Advantages of a Lab Experiment
+ Can establish cause and effect
+ Few if any extraneous variables
+Easy to replicate
+High internal validity
Disadvantages of a Lab Experiment
- Lacks ecological / external validity
- Demand characteristics can occur
- Behaviour in a lab is often different
- Experimenter effects can occur
Field Experiment
an experiment conducted outside of the lab but the IV is still manipulated. Participants usually unaware that they are being studied.
Advantages of a Field Experiment
+ More ecologically/externally valid
+ Fewer demand characteristics
+Replication can occur to some extent
+ Fewer experimenter effects
Disadvantages of a Field Experiment
- Chance of extraneous variables
- More time consuming
- Ethical issues (informed consent)
- Need a skilled researcher
Natural Experiment
an experiment in which the experimenter has not manipulated the independent variable directly.
Advantages of a Natural Experiment
+ No demand characteristics
+ No researcher effects
+ Fewer ethical issues
+ Allows participants who wouldn’t normally be tested to take part.
Disadvantages of a Natural Experiment
- Lack of control (extraneous variables)
- Short term behaviour may be displayed
- No random allocation can create confounding variables
- Harder to replicate
Quasi Experiment
studies that are almost experiments. The IV doesn’t vary- its a set condition that exists.
advantage of quasi experiment
+Carried out in controlled environment – shares strengths of lab experiments
disadvantage of quasi experiment
-Cannot randomly allocate participants to a condition, like natural experiment - may be confounding variables.
pilot studies
-a small scale trial run of a study, completed before the actual research
-can check methodology to highlight what isn’t working
-you can then change things without it affecting your actual study
-results of the pilot study are irrelevant
why do people use pilot studies?
-check if the IV is manipulated correctly
-check if the measure of the DV is correct
-check if the measure is appropriate and if people understand it
-avoid wasting time and money
BPS
British Psychological Society
Deception
Information is withheld from participants: they misled about the purpose of the study and what will happen during it.
Right to Withdraw
Participants should be told this at the start of the research. No attempt should be made to encourage them to remain.
Protection from harm
Participants should not be put through anything they wouldn’t normally be expected to.
Debrief
Researchers should discuss the aims of the research with the participants making sure they know how they’ve contributed to meeting the aims
Privacy
Participants’ right to privacy must be maintained.
We should only observe people where they would expect to be observed by others in public places.
Confidentiality
Information about our participants is protected by the Data Protection Act.
Participants must not be identifiable in published research.
Participants are given numbers or referred to by code or their initials.
Dealing With Informed Consent
-Participants issued with a consent letter or form detailing all relevant information that might affect their decision to participate.
-If under 16 parents need to sign
What if getting consent would ruin your study…?
-Presumptive consent
-Prior general consent
-Retrospective consent
Single Blind Procedure
-Participant does not know which experimental condition they are in.
-Limits demand characteristics
Double Blind Procedure
-Participant and researcher do not know which experimental condition the participant is exposed to (another researcher oversees the investigation).
-Limits demand characteristics and removes researcher bias
Observational Techniques:Naturalistic
-Watching/recording behaviour in the setting where it would normally take place.
-All aspects of the environment are free to vary.
evaluation of naturalistic
+More external validity.
+Findings can be generalised to real life.
-Replication is difficult
Observational Techniques:Controlled
-Watching/recording behaviour within a structured environment where variables are managed.
-Control over variables and also extraneous variables.
evaluation of controlled
+Replication easy due to control
-Lacks ecological validity
-Cannot be generalised to real-life situation
Observational Techniques:Covert
-Participants’ behaviour is watched and recorded without their knowledge or consent.
evaluation of covert
+Removed participant reactivity
+Increased validity
-Ethical Issues
Observational Techniques:Overt
-Participants’ behaviour is watched and recorded with their knowledge or consent.
evaluation of overt
+Ethically acceptable
-Could increase participant reactivity.
Observational Techniques:Participant
-Researcher becomes a member of the group whose behaviour they are observing.
evaluation of participant
+Researchers experience the whole situation – gives them insight.
-Lose objectivity
Observational Techniques:Non-participant
-Researcher remains outside of the group whose behaviour they are observing.
evaluation of non participant
+Maintain objective in observations
-May lose insight due to being ‘on the outside.’
Recording Observations:Unstructured
Write down everything the observer sees.
Provides rich detail… often too much.
evaluation of unstructured
-Produce qualitative data, more difficult to record and analyse.
+richer data
-higher risk of observer bias.
Recording Observations:Structured
Target behaviours for a main focus.
Allows the researcher to quantify their observations using a pre-determined list of behaviours and sampling methods.
evaluation of structured
+Use of behavioural categories make the recording easier and more systematic.
+Data more likely to be numerical - analysis is more straightforward.
Sampling Methods:Event sampling +evaluation
counting the number of times a particular behaviour occurs.
+Useful when the target behaviour or event happens quite infrequently and could be missed if time sampling was used.
sampling methods:time sampling +evaluation
recording behaviour with a pre-established time frame.
-Effective in reducing the number of observations that have to be made.
what are Questionnaires
-Set of written questions used to assess a person’s thoughts and/or experiences.
-May be used to assess the DV.
what are open questions ?
-Provide qualitative data- hard to analyse but provides rich data.
-Allow people to give opinions and feelings.
what are closed questions?
-Fixed choice (yes/no, tick boxes)
-Provides quantitative data- easy to analyse.
strengths of questionnaires
+Cost effective
+Large amounts of data
+Easy to distribute to large numbers of people
+Easy to analyse data (normally)
weaknesses of questionnaires
-Demand characteristics
-Response biases
different type of questionnaire designs
-Likert
-rating
-fixed questions
interviews: structured
Pre-determined questions asked in a fixed order.
interviews:semi-structured
Likely to experience in everyday
life, list of questions that have been decided but also free to expand (job interview)
interviews: unstructured
Works like a conversation, no set questions, interviewee encouraged to expand on questions.
evaluation of structured interviews
+easy to replicate due to standardised format
+reduces differences between interviewers
-not possible to debate topic
evaluation of unstructured interviews
+gives more insight due to flexibility
-analysis of data is hard (qualitative)
-social desirability
-drawing conclusions may be difficult
what does a correlation illustrate?
strengths and direction of association between two or more co-variables.
Positive Correlation
As one co-variable increases, the other co-variable also increases.
Negative Correlation
As one co-variable increases, the other co-variable decreases.
Testing the strength of a correlation
calculate a correlation coefficient to show how strong the association is between two co-variables.
what does a correlation co efficient do?
enables us to see the direction of the correlation and how strong the correlation is.
what are the strengths of correlations?
+Useful as a preliminary tool to assess the strength of a relationship.
+May suggest ideas for later research
+Quick and economical to carry out
what are the weaknesses of correlations?
-Cannot demonstrate cause and effect
-Third variable problem (intervening variable)
-Can be misinterpreted or misused to peoples advantage