research methods: correlation Flashcards
(15 cards)
what is correlational research?
measurement of the extent to which 2 pairs of related values on 2 variables tend to change together
(the extent to which two variables are related)
why are correlations described in terms of co variables?
both variables change and are measured, neither is set or controlled by the researcher
what is a positive correlation?
both variables increase together
- eg height and weight
what is a negative correlation?
one variable increases as the other decreases
- eg height and temp decrease
what is the correlation coefficient?
strength of the correlation
- less than 0 = neg
- more than 0 = pos
0-0.3 = weak
0.3-0.7 = moderate
0.7-1 = strong
strengths of correlations
- investigate otherwise unethical situations
- control for ppt variables as both sets of data provided by same person
- lead to new research (establishes which areas have relationships so need to be researched more to find a cause and effect relationship)
- replicable method eg self report
weaknesses of correlations
- identify link not cause
- lack validity
- self report is subjective to the individual, and uses questionnaires so may be an element of social desirability.
- uses secondary data
supporting study for correlation
Li et al
- used fMRI scans on chronic heroin use and activity in PCC
- unaffected by social desirability / DCs so valid
- however self reporting heroin use
What methods obtain correlational data?
- observations
- questionnaires
- twin/adoption studies
- experiments
Analysis of correlational data
- ranking two sets of data
- using a scatter gram
- spearman’s rho
Explain how ranking data would analyse a correlation
- rank two sets of data and compare
- if high ranks match with high ranks- then there is a positive correlation
- if high ranks match with low ranks- then there is a negative correlation
Explain who using a scatter gram would analyse correlational data
- plot the co-variables on a scattergram
- draw a line of best fit to establish the correlation coefficient from the gradient
Explain how there is high reliability in correlational data
- easily replicable methods used to gather quantitative data
- statistical tests such as a spearman’s rho can easily be replicated to test for consistency within results.
- (also increases validity as the statistical tests can be used to analyse results- objective- reducing misinterpretation or researcher bias)
Explain how there is high validity
- statistical analysis- no misinterpretation
- controlling for ppt variables as both covariables will e completed by the same person therefore individual, differences will not affect the correlation.
Explain how correlational data is ethical
- allows us to investigate otherwise unethical situations if experimented on such as environmentally causing aggression to produce social aggression