Psychology- Research Methods

research methods

HideShow resource information

Laboratory Experiment:

Researcher directly manipualates IV to see its effect on the DV

Takes place in highly controlled conditions


  • High levels of control of Iv and Evs
  • Replicable
  • Can infer cause and effect


  • Artificiality can lead to low validity
  • Strong chance of investigator and participant effects

Ethical issues:

It is sometimes impossible to gain fully informed consent because of potential demand characteristics

1 of 23

Field Experiment:

Researcher directly manipulates IV to see its effect on the DV
Takes place in a natural setting


  • Higher ecological validity ( than in laboratory)
  • Reduction in participant effects (demand characeristics)
  • Can infer cause and effect


  • Less control over Evs
  • Less control over participant sample
  • Difficult to replicate exactly
  • Can be more time consuming to set up and carry out

Ethical issues: Consent, deception and right to withdraw are all important issues as participants often do not realise they are in an experiment

2 of 23

Natural Experiment:

Researcher takes advantage of a naturally occuring event to see the effect on the DV, the IV is not directly manipulated


  • Useful where it would be impossible/unethical to manipulate the IV or undertake the investigation in a lab setting
  • High levels of ecological validity


  • Lack of control of Evs- leads to low validity
  • Less possible to infer cause and effect confidently
  • Less control of participant sample
  • Difficult to replicate or generalise as the natural event is usually a one-off

Ethical issues:

Consent, right to withdraw and confidentiality are all important issues

3 of 23

Correlational Analysis:

  • Technique for analysing data by measuring the strenght of the relationship between two variables. The analysis will show onr of 3 things:
    • a positive correlation
    • a negative correlation
    • no correlation
  • Strengths:
    • Can establish the strength of a relationship between two variables and measure it precisely
    • Allows researchers to investigate things that could not be manipulated experimentally for ethical or practical reasons
  • Weaknesses:
    • Cannot infer cause and effect
    • Can only measure linear relationships ( clear positive or negative correlations)- does not detect curvlinear relationships

Ethical issues: Consent, confidentiality and right to withdraw are issues as participants are often unaware their data is being used

4 of 23

Naturalistic Observation:

Researcher obserbs participants in their own environment

No deliberate manipulation of variables


High ecological validity

Participants usually behave more naturally

Useful preliminary research tool- can suggest hypotheses for further research


No control over Evs

Ethical issuesL

Privacy, confidentiality and consent are important issues

5 of 23

Controlled Observation:

Researcher observes participants in a controlled environment

Often involves manipulation of variables e.g. in Ainsworth's Strange Situation


Higher levels of control over EVs


Participants usually know they are being observed- might not behave naturally - lowers validity

Ethical issues:

Informed consent, confidentiality and right to withdraw are issues

6 of 23


A set of questions used to collect data from a large sample of participants
Can be given face to face, by post, phone or internet or simply left in public places

  • Strengths:
    • Can reach a large sample of people relatively quickly and cheaply
    • Can collect large amounts of data
    • Time efficient as researcher does not have to be present
    • Reduces investigator effects
    • Data can be easily analysed ( if quantitive)
    • Replicable
  • Weaknesses:
    • Social desirability/ lack of honesty leads to low validity
    • Postal surveys have low response rate- reduces representativeness of sample
    • Questions may be ambiguous - researcher is not there to explain
    • Questions (if closed) limit the depth of response
    • Quesitons (if open) can be difficult to analyse

Ethical issues: privacy, protection from harm, confidentiality and informed consent

7 of 23


Researcher asks participants questions directly face to face. Can be structured or unstructured

  • Strengths:
    • Structured: Data analysis straight forward
    • Less risk of investigator effects
    • Less training needed for interviewers
    • Interviewer available to clarify any ambiguity
    • Unstructured: Researcher can follow up issues raised
    • Interviewee can expand on answers so provide new insights
    • More informal- interviewer can be more sensitive when asking for personal info
  • Weaknesses:
    • Structured: Interviewer cannot follow up interesting answers
    • Formal situation may inhibit honest/ full answers
    • Unstructured: Interviewer effects
    • Social desirability
    • High levels of training needed for interviewers
    • Time-consuming and expensive
    • Can be difficult to analyse qualitive data

Ethical issues: Privacy, protection from harm, confidentiality, informed consent, right to withdraw

8 of 23

Case Study:

An in depth study of an individual or group of people


  • Provides rich data
  • High levels of ecological validity
  • Can suggest new hypotheses for further research
  • Can investigate topics that would be impractical/unethical to investigate experimentally


  • Difficult to replicate
  • Difficult to generalise results
  • Possibility of researcher bias

Ethical issues: Informed consent, invasion of privacy, right to withdraw and confidentiality are all issues

9 of 23



The experimental/ alternative hypothesis predicts the direction or the difference/correlation

e.g. participants will recall more short words than long words in a seial recall task


do not predict the direction or any difference/ correlation

e.g. there will be a difference in the number of words recalled in a serial recall task depending on whether the words are long or short

10 of 23

Experimental Design:

Repeated measures design: same participants in each condition

  • Strenghs: holds individual constant so controls for them, needs fewer participants
  • Weaknesses: order effects (e.g.fatigue,boredom), increased chance of demand characteristics, cannot use same set of stimulus materials in both conditions

Independent groups design: different participants in each condition ( randomly allocated to avoid any bias)

  • Strengths: no order effects, reduced liklihood of demand characteristics, can use same set of stimulus material in both conditions
  • Weaknesses: individual differences, more participants needed

Matched pairs design: participants matched on key participant variables (e.g gender)

  • Strengths: no order effects, reduces effects of individual differences, can use same set of stimulus material in both conditions
  • Weaknesses: difficult to decide on criteria for matching, difficult to match exactly, more participants required
11 of 23

Designing a Naturalistic Observation:

  • How will you identify appropriate behavioural categories?
  • Do you need a pilot study?
    • Useful to establish behavioural categories and iron out problems
  • How will you avoid observer bias?
    • Have clear behavioural categories, more than one observer and compare their ratings to ensure inter-rater reliability, keep the observers blind to the hypothesis
  • How will you address the ethical issues?
    • E.g. Invasion of privacy, informed consent, right to withdraw
  • How will you record the data?
    • Tally charts, structured data collection grids, video recording etc
12 of 23

Designing a Questionnaire:

  • Number and layout of questions
    • Need enough to get sufficient data but not too many that it becomes too long and tedious for participants to fill in. Does the question order make sense? Is there a problem of response set?
  • Ethical issues
    • Need to make sure that participants are aware of right to withdraw etc.
  • Question wording
    • Are the questions ambiguous? do they contain unclear jargon? are they leading questions? do they contain emotive language? are they personal/intrusive?
  • Qualitive or quantitive
    • Closed or open questions?
  • Sampling
    • Deciding on target population;finding a representitve sample;deciding on method of finding participants e.g. advertising etc.
13 of 23

Designing an Interview:

  • How to analyse qualitive data
  • How to record data
    • Written notes, tape recorder
  • Sampling
    • Deciding on target population;finding a representitve sample;deciding on method of finding participants e.g. advertising etc.
  • Structure
    • Structured or unstructured?
  • Ethical issues
    • Need to make sure that participants are aware of right to withdraw etc.
  • Question wording
    • Are the questions ambiguous? do they contain unclear jargon? are they leading questions? do they contain emotive language? are they personal/intrusive?
14 of 23

Types of Extraneous Variables:

  • Participant variables: intelligence, age, gender, personality etc
    • Choosing an appropriate design- repeated measures design or matched pairs avoids individual differences
    • In independent groups design, randomly assigning participants to conditions helps to avoid bias
  • Demand characteristics
    • Cues in the environment that help participants work out what the research hypothesis is which can make them alter their behaviour
    • Single blind technique, i.e. making sure they dont know the hypothesis and dont know what condition they're in ( difficult in repeated measures)
  • Experimenter effects
    • Characteristics of the investigator that might affect participant responses
    • At an unconscious level, the investigator might behave in such a way as to influence the outcome of the study in favour of their own predictions
    • double-blind technique i.e neither participant or investigator know hypothesis or condition participants are in.Involves research assistant.
  • Situational variables
    • Temperature, time, lighting etc.
    • Standardisation i.e making sure conditions are the same for all participants as much as possible
15 of 23

Ethical Guideline Categories:

Consent- aims should be clear. parental consent needed for under 16s

Deception- information shouldnt be withheld, participants shouldnt be misled

Debriefing- should be given full explanation after the study has been completed

Confidentiality- data must be kept annonymous /  clearly inform before taking part

Protection of participants- from physical or emotional harm.

Colleagues- if researcher sees them being unethical, they should raise concern

Giving advice- proceed with caution. offer advice within own area of expertise/ refer

Observational research- protection from harm, privacy, not be observed in situations whre they would not normally be watched

Right to withdraw- can withdraw at any time, even at the end-data is then destroyed

16 of 23

Dealing with Ethical Issues:

  • Deception: participants may be misled about the nature of the research. It prevents them from making an informed decision about taking part
  • It might make people distrustful of psychologists.
  • - debriefing and retrospective informed consent deal with it
  • Informed consent: If participants are not given all the facts before agreeing to take part in a study, they may find themselves taking part against their wishes.
  • It might make people distrustful of psychologists.
  • -prior general consent, presumptive consent, for children-parental consent or those in loco parentis (e.g. headteacher) deal with it
  • Protection from harm: Participants have a right to be protected from any physical or emotional harm. They should leave the study in the same state as they entered. Any harm could have long lasting effects
  • -reminding of right to withdraw, terminating research causing distress,debriefing,offering advice/support deals with it
17 of 23

Random Sampling:

Sample in which every member of the target population has an equal chance of being picked

Every member of the target population is identified and then a random sampling technique is applied


Most likely to be a representative sample of the target population ( cant guarantee it)

Difficult to obtain because must have access to every member of target population- usually only possible when the target population is quite small

If anyone drops out or refuses to take part, the sample becomes less random

18 of 23

Opportunity Sampling:

Sample that consists of people readily available to the researcher

The researcher approaches people who happen to be available and asks them if they would be willing to take place in research


High chance that sample will not be very representitive of the target population

Sometimes people feel obliged to take part even if they do not really want to

most popular technique because it is so convenient

19 of 23

Volunteer Sampling:

Sample where participants are self-selected

The researcher advertises for volunteers on posters or in magazines/newspapers


A certain type of person tends to volunteer ans this means that there is a high chance the sample will not be representative

Useful way of finding participants for highly specific types of research

20 of 23


Bar charts: Used for nominal data

  • consists of a set of vertical bars with space between them
  • each bar represents a different category

Histograms: consists of vertical bars of equal width

  • bars are continuous so there is no space between them
  • used for ordinal or interval data
  • frequency is represented by the area of the bar

Frequency polygon: used as an alternative to histograms

  • useful when showing two sets of data on the same graph

Scattergram: Used for showing relationship of two variables, that is for showing correlation

21 of 23

Measures of Central Tendency:

  • the mean- add all of the scores and divide by total number of scores
    • easy to calculate
    • only measures of central tendency for nominal data
    • unaffected by extreme scores
    • Tells us nothing about other scores in the set and may not be typical (central)
    • Limited usefulness if there is more than 1 modal score
    • Not useful for small sets of scores
  • the median- rank the scores and take the middle value
    • can be used on ordinal or interval data
    • unaffected by extreme scores
    • not much use for small data sets
    • can be unrepresentative of data if scores clustered at high and low levels
  • the mode- the most frequent occurring value
    • most sensitive measure taking all scores into account
    • can be distorted by a single extreme value in a set
    • can only be used for data that are at least interval
22 of 23

Measures of Dispersion:


Quick and easy to calculate

takes account of all the scores

Standard deviation (SD):

can be easily distorted by extreme values

more difficult to calculate compared to range

should only be calculated on data measured on an interval scale

23 of 23


No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Research methods and techniques resources »