Research Methods Flashcards
Scientific method
observation of phenomena, hypothesis formulation and experimentation, must use empirical methods and be objective
Replicability,
must be possible in order to establish validity and reliability
Peer review
research checked by experts before publication
Random sampling
selected without bias, using random number tables or out of a hat, but doesn’t guarantee a representative sample
Opportunity sampling
makes use of available people but is often biased
Volunteer sampling
obtained by adverts/posters, unlikely to be representative
Systematic sampling
choosing every nth person, unbiased but not certainly representative
Stratified sampling
P’s selected in proportion to their frequency in the target population
Reliability
extent to which results are consistent, internal (consistent within itself) external(measure consistent with other measures), to test use test-retest or inter rater
Validity
how a research measures what it claims to measure, internal(does the test measure what it says it does) and external( for far the results can be generalised)
Ethical issues
consent, no harm, debrief, deceit, confidential, ethical committees, suitable observation
Graphs
bar charts(discrete), histograms( continuous), frequency polygons, scatter grams
Measures of central tendency and dispersion
mean, median, mode, range and standard deviation
Type 1 errors
accepts found difference when in fact it’s down to chance
Type 2 errors
rejects difference as insignificant but are wrong as the LOS is too low
Nominal data
frequency of qualitative variables
Ordinal
involves ranking data in order
Interval/ratio data
uses equal measurement intervals such as time or weight
Chi-squared
difference, nominal, independent groups
Mann-whitney
difference, ordinal, independent groups
Wilcoxon matched ranks
difference, ordinal, repeated measures
Spearman’s rho
correlation, ordinal, data is pairs of scores from the same person or event
Content analysis
quantifying qualitative data, data placed in categories and counted
Report writing
prose, past tense, clear and formal, important for replicability