AQA Psychology - Year 2 Research Methods

Reliability

Refers to how consistent a measuring device is - and this includes psychological tests or observations which assess behaviour

Ways of assessing reliability

Test - retest : Same test/questionnaire given to the same person after a while

2 sets of results are positively correlated at 0.8 or above

Inter observer/rater: 2 or more observers use some coding checklist on same pp's/materials

2 sets of results are positively correlated at 0.8 above

Improving reliability of observations

Behavioural categories operationalised, measurable and self-evident without overlap

Improving reliability of experiments

Standardisation of procedures will minimise extraneous variables

Improving reliability of questionnaires

Don't use ambiguous, complex or subjective questions

Improving reliability of interviews

Don't use ambiguous and leading questions

Use the same interviewer

Don't ask two questions in one

Empiricism

the belief that accurate knowledge can be acquired through observation

Replicable

pertaining to a study whose results have been obtained again when the study was repeated

Objectivity

treating facts without influence from personal feelings or prejudices

Falsifiability

a feature of a scientific theory, in which it is possible to collect data that will prove the theory wrong

case studies

studies that involve extensive, in-depth interviews with a particular individual or small group of individuals over time

content analysis

A research technique that enables the indirect study of behaviour by examining communications that people produce, for example, in texts, emails, TV, film and other media.

Thematic analysis

An inductive and qualitative approach to analysis that involves identifying implicit or explicit ideas within the data. Themes will often emerge once the data has been coded.

Evaluation of case studies

+lots of data

+use many other research methods

-time consuming

-expensive

-cannot determine cause & effect

-findings may be subjective

Evaluation of content analysis

+No ethical issues, as no people are directly involved

+High external validity

+Can produce both qualitative and quantitative data

-Coding of materials is subjective so may be biased

how to conduct a content analysis

choose sampling method, how the data should be recorded, who is going to collect the data, how the material should be categorised or coded

Validity

the extent to which a test measures or predicts what it is supposed to

internal validity

the degree to which the effects observed in an experiment are due to the independent variable and not extraneous factors

external validity

extent to which we can generalize findings to real-world settings

ecological validity

The extent to which a study is realistic or representative of real life.

Temporal validity

the degree to which the results can be generalized across time

Ways of assessing validity

Face validity: whether a test appears to be testing the correct thing on the face of it

Concurrent validity: whether the results match up with current well-established tests

Improving validity of experiments

- Use a control group, so you know IV is causing change in DV

- Standardise procedures, to minimise participant reactivity and investigator effects

- Use single-blind and double-blind design in order to remove demand characteristics and investigator effects

Improving validity of questionnaires

- Use a lie scale and anonymity to reduce social desirability

Improving validity of observations

- Findings may be more authentic in covert observations.

- Behavioural categories may be too broad

Triangulation

the use of multiple research methods as a way of producing more reliable empirical data than are available from any single method

Choosing a statistical test

Simon Cowell wants more singers receiving unanimous praise

Sign Test

Chi-squared

Wilcoxon

Mann-whitney u

Spearman's Rho

Related t-test

Unrelated t-test

Pearson's r

Nominal Data

Data of categories only. Data cannot be arranged in an ordering scheme. (Gender, Race, Religion)

ordinal data

data exists in categories that are ordered but differences cannot be determined or they are meaningless. (Example: 1st, 2nd, 3rd)

Interval Data

Interval data are based on numeric scales in which we know the order and the exact difference between the values. Using units of measures such as cm, g

Probability

A measure of likelihood that a particular event will occur where 0 indicates statistical impossibility and 1 statistical certainty.

Significance

A statistical term that tells us how sure we are that a difference or correlation exists. A 'significant' result means that the researcher can reject the null hypothesis.

Critical value

when testing a hypothesis, the numerical boundary or cut-off point between acceptance and rejection of the null hypothesis

Type 1 error (false positive)

The incorrect rejection of a true null hypothesis

Type 11 Error (False Negative)

The failure to reject a false null hypothesis

level of significance

The usual level of significance is 0.05 (5%) or in a medical study it is 0.01 (1%). This means that the probability that the observed effect was down to chance is 5%.

Paradigms

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

Paradigm shift

The result of a scientific revolution: a significant change in the dominant unifying theory within a scientific discipline.