Quantitative research methods

one-tailed hypothesis

directional

two-tailed hypothesis

non-directional

true experiment

controlled experiment where there is manipulation of the IV

between subjects

independent groups = ppts allocated to a group randomly and perform in 1 level

within subjects

repeated measures = each ppt performs all levels and acts as their own control

matched pairs = two ppts are matched on all variables to 1 level each

natural independent groups

researcher selects pre-existing groups

correlational design

no experiment, measuring two variables and looking at the relationship between them

independent variable

manipulated

dependent variable

affected by change and measured

nominal data?

numbers represent categories

no numerical hierarchy

ordinal data?

scores can be ranked and have a logical sequence

intervals are not all equal

interval data?

equal intervals

numbers behave like numbers

there is NO absolute zero

eg IQ and psychological constructs

ratio data?

equal intervals

numbers behave like numbers

there IS an absolute zero - there can be none of it

descriptive statistics for nominal data

categories and numbers of ppts in each

descriptive statistics for ordinal data

median and range

descriptive statistics for interval/ratio data

mean and standard deviation

descriptive statistics for correlation

scattergram and regression line

parametric data?

interval/ratio

non-parametric data?

ordinal

inferential tests for independent groups

parametric = independent t-test

non-parametric = mann-whitney

inferential tests for repeated measures/matched pairs

parametric = paired t-test

non-parametric = wilcoxon

inferential tests for correlation

parametric = Pearson’s

non-parametric = spearman’s

null hypothesis?

no difference between the samples

if p < 0.05

there IS a significant difference between the samples = reject the null hypothesis

if p > 0.05

there is NO significant difference between the samples = accept the null hypothesis.