EXTRANEOUS VARIABLES: 'nuisance' variables that do not vary systematically with the IV - may be controlled

CONFOUDING VAIRABLES: variables that vary systematically with the IV so we cannot be sure if any observed change is due to the IV or not - must be controlled

DEMAND CHARACTERISTICS: any cue from researcher/research situation that reveals aim of study leading to participant reactivity

INVESTIGATOR EFFECTS: any effect of investigator's behaviour on research outcome

1 of 35


AIM: general expression of what the researcher intends to investigate

HYPOTHESIS: defined and measurable (operationalised) statement of what the researcher believes to be true

DIRECTIONAL - states the nature of the change (increase/decrease etc)

NON-DIRECTIONAL - simply states that there will be a change

EXPERIMENTAL METHOD: the manipulation of the IV to record the effect on the DV

2 of 35


RANDOMISATION: use of chance to control for the effects of bias

STANDARDISATION: using exactly the same formalised procedures for all participants

CONTROL GROUP: a baseline group that helps to establish causation by comparison with the experimental group

SINGLE BLIND: participants don't know the aims of the study to reduce demand characteristics

DOUBLE BLIND: neither the participants nor the researcher knows the aims of the study to reduce demand characteristics + investigator effects

3 of 35


INDEPENDENT GROUPS - one group experience one condition, the other group experience another. Random allocation is required to reduce bias.

no order effects and fewer demand characteristics but participant variables may act as EV/CVs and more participants are needed

REPEATED MEASURES - all participants experience all conditions. Counterbalancing is required to reduce order effects.

fewer participant variables and fewer participants needed but order effects can arise as well as demand characteristics

MATCHED PAIRS - participants are matched based on a participant variable relevent to the experiment and one of each pair experiences one condition.

fewer participant variables and order effects are controlled for but matching is not always perfect and more participants may be needed

4 of 35


LAB EXPERIENT - controlled environment where EVs and CVs are regulated

  • participants go to researcher
  • IV manipulated - effect on DV recorded

control means greater internal validity and easier replication but the level of control may mean a lack of external validity and demand characteristics may cause participant reactivity

FIELD EXPERIMENT - behaviour takes place where it would naturally occur

  • researcher goes to participants
  • IV manipulated - effect on DV recorded

the natural environment means greater external validity and participants may not be aware they're being studied but it is more difficult to control CVs and informed consent may not always be given

5 of 35


NATURAL - IV would have varied even without the researcher's interest

  • DV may be naturally occurring (e.g. exam results) or measured by the experimenter

this may be the only ethical option (e.g. institutionalisation of children) and the study of real-life events means greater external validity but the events may only occur occasionally and participants are often not randomly allocated

QUASI - IV based on pre-existing difference (age, gender) with no manipulation

  • DV may be natrually occurring (e.g. exam results) or measured by the experimenter

often a high degree of control meaning high internal validity and it allows comparisons to be made between participants but participants are not randomly allocated and the lack of manipulation means causation cannot be established

6 of 35


POPULATION: large group of people a researcher is interested in studying

SAMPLE: smaller group of the population that the researcher will use as participants

GENERALISATION: when the sample is representative of the population so the results from the study can be applied to them

BIAS: when groups within a sample are over or under represented

7 of 35


OPPORTUNITY SAMPLING - people who are the most available are selected (e.g. asking people in the street)

quick and convenient method but will inevitably be biased

VOLUNTEER SAMPLING - participants select themselves to take part (e.g. via adverts)

quick and convenient method and participants more likely to engage but likely to be a biased sample (volunteer bias)

RANDOM SAMPLING - every person in the target population has an equal chance of being selected

  • assign all members of population a number and randomly generate numbers

free from researcher bias but representation is not guaranteed

8 of 35


SYSTEMATIC SAMPLING - participants selected using a sampling frame where every nth person is selected from a list of the target population

free from researcher bias but takes time and effort

STRATIFIED SAMPLING - participants selected according to frequency in target population

  • subgroups ('strata') identified, relative proportions calculated, population reflected in samole

designed to be representative but subgroups don't always reflect all the ways in which participants are different

9 of 35


ETHICAL ISSUES: arise when there is a conflict between the rights of the participant and the aims of the research

BPS CODE OF CONDUCT - a quasi-legal document to protect participants based on respect, competence, responsibility and integrity

ETHICS COMMITTEES - weigh up costs and benefits to decide whether or not a study should go ahead

INFORMED CONSENT - participants should make an informed decision on whether or not to take part

PRESUMPTIVE - ask a similar group

PRIOR GENERAL - agree to be deceived

RETROSPECTIVE - debrief and get consent after the study

10 of 35


DECEPTION - deliberately misleading or witholding information from participants (meaning consent is not informed)

  • participants should be debriefed on the aims of the study, details not given during the study (other groups/conditions), what their data will be used for and their right to withold data

PROTECTION FROM HARM - ensuring participants are not at increased risk of harm from their everyday lives

  • constantly reminded of right two withdraw
  • reassured that behaviour was normal during debrief
  • provide counselling if participants distressed etc

PRIVACY/CONFIDENTIALITY - the right to control information about ourselves and have our details protected legally

  • use initials, numbers or false names
  • do not share personal data with other researchers
11 of 35


CORRELATION: illustration of the strength and direction of an association between two co-variables, with one of each plotted on the axes of a scattergram

POSITIVE CORRELATION - co-variables rise or fall together

NEGATIVE CORRELATION - co-variables change oppositely

ZERO CORRELATION - no association

  • only shows associations, not relationships - no cause and effect due to lack of manipulation of IV
  • influence of EVs not controlled so third-variable/intervening variable may be an issue

provide hypotheses for future study and are relatively economical but do not establish causation  and methods of measurement may be unreliable

12 of 35


NATURALISTIC - takes place where behaviour would normally occur

high external validity but low control

vs CONTROLLED - some manipulation of variables, control of EVs

can be replicated but have low external validity

COVERT - participants unaware they're being studied

demand characteristics reduced but ethically questionable

OVERT - participants aware of being studied

more ethically acceptable but demand characteristics may reduce validity

PARTICIPANT - researcher becomes a part of the group to be studied

greater insight but potential loss of objectivity (going native)

NON-PARTICIPANT - researcher remains separate from the group to be studied

more objective but potential loss of insight

13 of 35


OBSERVATION: watching and recording the behaviour of participants to assess the DV

can capture unexpected behaviour (e.g. not shown in self report) but risks observer bias through subjective interpretation

BEHAVIOURAL CATEGORIES - breaking up the target behaviour into observable categories (similar to operationalisation)

make observations more objective but can be ambigious and overlap and not all behaviours may be included

TIME SAMPLING - observations made at regular intervals (e.g. once every 30 seconds)

reduces the number of observations to be made but may be unrepresentative of the behaviour

EVENT SAMPLING - target behaviour recorded every time it occurs

may capture infrequent behaviour but complex behaviour may be oversimplified or unrecorded

14 of 35


QUESTIONNAIRE: pre-set list of questions to which a participant responds, potentially used to assess the DV in an experiment

can be distributed to a lot of people easily and reduces social desirability bias compared to interviews but responses may not always be truthful and responses may be biased

GOOD QUESTIONS - avoid jargon, double-barrelled questions and leading questions

CLOSED QUESTIONS - respondent has limited choices meaning quantitative data is gathered

easier to analyse but options may not reflect participant's true experiences

OPEN QUESTIONS - respondent replies however they which in words meaning qualitative data is gathered

repondents aren't restricted but data is more difficult to analyse

15 of 35


INTERVIEW: face-to-face interaction between interviewer and interviewee where questions are asked in real time

STRUCTURE INTERVIEW - list of pre-determined questions asked in fixed order

easier to replicate but interviewees cannot elaborate

UNSTRUCTURED INTERVIEW - no set questions but a general topic to be discussed with encouraged elaboration from the interviewee

greater flexibility but difficult to replicate so more open to observer bias

SEMI-STRUCTURED INTERVIEWS - list of questions determined in advance but interviewers can ask follow-ups when appropriate

GOOD INTERVIEWS have an established interview schedule (a standardised list of questions to cover), are conducted in a quiet room, establish rapport to relax the participant and remind interviewees that answers will be treated with confidence

16 of 35


QUANTITATIVE DATA - data expressed in numbers

easier to analyse but oversimplifies behaviour

QUALITATIVE DATA - non-numerical, expressed in words

represents complexities of behaviour but is more difficult to analyse

PRIMARY DATA - collected first hand from the participants for the purpose of the investigation

designed for the research but requires time and effort to collect

SECONDARY DATA - collected from the participants by someone other than the person conducting the research i.e. it already exists

inexpensive to collect but quality may be questionable

META-ANALYSIS - secondary data involving combining data from a large number of studies and calculating effect size

increases validity of conclusions but is at risk of publication bias (leaving out negative or non-significant results)

17 of 35


MEAN - arithmetic average where all scores are added up and the total is divided by the number of scores 

sensitive because it involves all the scores but may be unrepresentative

MEDIAN - middle value when data is placed in ascending order 

unaffected by extreme scores but less sensitive than the mean

MODE - most frequent or common value in nominal data

only possible measure for discrete data but is overly simplified

18 of 35


RANGE - difference between highest to lowest value

easy to calculate but doesn't account for distribution of scores

STANDARD DEVIATION - measure of the average spread aorund the mean where the larger the standard deviation, the larger the spread

more precise than the range but may be misleading by 'hiding' some characteristics of the data where extreme values not revealed

19 of 35


TABLE - raw scores displayed in columns and rows with a summary paragraph beneath to explain results

BAR CHART - discrete data along the x axis with frequency on the y

HISTOGRAM - continuous data with a true zero and bars that touch one another

LINE GRAPH - frequency on on y axis, continues data on the x showing how a value changes over time

SCATTERGRAM - correlational analysis where one dot represents a piece of related continuous data

20 of 35


NORMAL DISTRIBUTION - symmetrical bell-shaped curve with most of population in middle areas and few at the extremes, mean median and mode occupy same mid-point

SKEWED DISTRIBUTION - most people either at upper or lower end of distribution

NEGATIVE SKEW - most of population concentrated at right side with a small mean dragging the tail to the left

POSITIVE SKEW - most of population concentrated at left side with a large mean dragging the tail to the right

21 of 35


SIGNIFICANCE: the difference/association between two sets of data is greater than what would occur by chance, determined through statistical testing

PROBABILITY: numerical measurement of the likelihood that certain events will occur - the level at which the researcher accepts or rejects the null hypothesis

THE SIGN TEST - score from condition B subtracted from condition A to give sign of difference, take less frequent sign as S, if S is equal to or less than the critical value (N - 0 difference) the results are significant

22 of 35


  • all aspects of investigation scrutinised by experts in field
  • should be objective and unknown to researcher
  • allocates research FUNDING
  • VALIDATES quality and relevance of research
  • suggests IMPROVEMENTS and amendments

ensures published research is of good quality but may be used to criticise rival research and can be subject to publication bias where editors only want to publish headline grabbing findings and ground-breaking research may be buried

23 of 35


CASE STUDY: in-depth investigation, description and analysis of single individual, group, institution or event

  • often study unusual events or individals
  • qualitative data gathered through a case history using interviews, observations, questionnaires
  • quantitative data gathered through experimental testing
  • temd to be longitudinal and involve family members, peers

rich, detailed insights that shed light on unusual behaviour, may contribute to our understanding or normal behaviour and generate hypotheses for future study but lack generalisability and often rely on self report

24 of 35


CONTENT ANALYSIS: involves the indirect study of behaviour through communications that we produce (emails, texts, TV) with the aim of summarising communication systematically so conclusions can be drawn


- initial stage

- data analysed via meaningful units (number of times a word/phrase is used)

- produces quantitative data


- analysis of data through implicit and explicit ideas that recur throughout it

- may collect new data after thematic analysis to test validity of themes

circumnavigtes ethical issues and produces both quantitative and qualitative data but risks loss of context and involves subjective analysis

25 of 35


RELIABILITY: consistency - to what extent are the findings the same when retested?

TEST-RETEST - give the same participant the same test on different occassions and correlate the ratings for significance, must be sufficient time to ensure participant doesn't remember previous answers but not so much that the ansfers change

INTER-OBSERVER RELIABILITY - pilot study involving rating of behaviour (application of behavioural categories, correlate findings


QUESTIONNAIRES - test-retest, rewrite ambiguous or open questions

INTERVIEWS - same interviewer each time, avoid leading or ambiguous questions, use structured design

EXPERIMENTS - standardisation to allow replication

OBSERVATIONS - operationalise behavioural categories with no overlap or omission of possible behaviours

26 of 35


VALIDITY: accuracy - to what extent is the effect a result of what we think it is a result of and how far can it be generalised beyond the setting in which it was found?

INTERNAL VALIDITY: the extent to which the observed effects are due to the indepdnent variable or some other factor (demand characteirstics etc)

EXTERNAL VALIDITY: the extent to which the observed effects can be related to more situations other than the one in which they were found

ECOLOGICAL VALIDITY - generalising findings from one setting to another, possible when studies have high mundane realism

TEMPORAL VALIDITY - generalising findings from one time period to another

27 of 35


FACE VALIDITY: where a test, scale or measure appears to measure what it was intended to measure - determined through 'eyeballing' or examination by expert

CONCURRENT VALIDITY: the extent to which the findings of a measure match that of a pre-established and recognised measure


EXPERIMENTAL RESEARCH - control group determines if changes are due to the independent variable, standardisation to minimise participant reactivity and investigator effects, blinding to reduce demand characteristics

QUESTIONNAIRES - incorporate a lie scale to measure honest and control for social desirability bias, assure participants that responses will be anonymous

OBSERVATIONS - covert to reduce participant reactivity, operationalise behavioural categories

QUALITATIVE METHODS - interpretive validity (the extent to which the researcher's interpretation matches that of participants) achieved through direct quotes and coherence, triangulation (use of number of sources of evidence) to support findings

28 of 35


STATISTICAL TESTS: determine whether a significant difference or correlation exists and thus if the null hypothesis should be accepted or rejected



  • RELATED is repeated measures or matched pairs
  • UNRELATED is independent groups


  • NOMINAL is data presented in discrete categories, mode as measure of central tendency
  • ORDINAL is ordered data on a subjective scale with no set intervals between each unit, median as measure of central tendency and range as measure of dispersion
  • INTERVAL is data on numerical scales of equal, precisely defined size, mean as measure of central tendency and standard deviation as measure of dispersion
29 of 35


  • more powerful and robust than other tests
  • can detect significance within data sets that other tests will not


data should be drawn from population with NORMAL DISTRIBUTION

should be HOMOGENEITY OF VARIANCE where set of scores have similar dispersion

30 of 35


PROBABILITY: measure of the liklihood that a particular event will occur - 0 represents impossibility, 1 represents certainty

SIGNIFICANCE: tells us how sure we are that a difference or correlation exists where it is greater than what would occur by chance

- we use a more stringest significance level when there may be human cost

- when there is a considerable difference between the critical and calculated value we may retest with lower levels of significance

NULL HYPOTHESIS: states there will be no difference between the conditions

CRITICAL VALUE: numerical boundary between acceptance and rejection of the null hypothesis

31 of 35


TYPE I ERRORS occur when the null hypothesis is rejected when it should have been accepted because the level of significance is too lenient

TYPE II ERRORS occur when the null hypothesis is accepted when it should have been rejected because the level of significance is too stringent

32 of 35


ABSTRACT - short summary containing the aims, hypotheses, method/procedure, results and conclusions (can be read by psychologists to identify investigations worthy of future study)

INTRODUCTION - review of general area of investigation, relevant theories/concepts/studies, gradually becomes more specific before presenting aims/hypotheses

METHOD - provide sufficient detail to allow other researchers to replicate, design and justification, size and demographic of sample, sampling method, target population, apparatus/materials, procedure (verbatim report of everything said to participants), ethical issues and how they were adressed

RESULTS - summary of the key findings using descriptive statistics, inferential statistics, thematic analysis of qualitative data, note that raw data appears in appendix

DISCUSSION - verbal summary of the findings in the context of the introduction, discussion of limitations and how they may be adressed, consideration of wider implications of research

REFERENCING - details of source material uses

33 of 35


PARADIGM: set of shared assumptions and agreed methods within a scientific discipline

  • different approaches etc mean there is a lack of paradigms in psychology 

PARADIGM SHIFT: results from a scientific revolution where there is a significant change in the dominant unifying theory within a scientific discipline due to too much contradictory evidence to ignore

THEORY: set of general laws or principles that can explain particular events or behaviours

THEORY CONSTRUCTION: fathering evidence via direct observation using the empirical method

HYPOTHESIS TESTING: the ability to make clear and precise predictions on the basis of the theory that can be scientifically tested

DEDUCTION: deriving new hypotheses from existing theory

34 of 35


FALSIFIABILITY: the idea that a theory cannot be considered scientific unless it admits the possibility of being proved untrue

  • theories that survive most falsification attempts are the strongest
  • many psychological concepts are too vague to be tested

REPLICABILITY: the extent to which scientific procedures and findings can be repeated by other researchers

  • a theory must be repeatable across a number of contexts and circumstances - increases external validity

OBJECTIVITY: minimising all sources of personal bias so as not to distort the research process

EMPIRICAL METHOD: approaches based on gathering of evidence through direct observation and experience

  • e.g. through controlled experiment and observation
35 of 35


No comments have yet been made

Similar Psychology resources:

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