Research methods Flashcards
Nominal
Basic type of data, data into separate categories, counting frequency which behaviour occur in a category
Ordinal
Data is ordered in some way, it is meaningfully rank-able, 1st 2nd 3rd
Interval
Public units of measurements, interval between ranks are regular and has no true zero
e.g. temperature, g (mass)
Ratio
Similar to interval data, has a true zero point
e.g. weight, height, mph
IM/Nominal
Chi-squared
IM/Ordinal or higher
Mann-Whitney
RM/Nominal
Sign Test
RM/Ordinal or higher
Wilcoxon
Correlation
Spearman’s Rank
IM/Interval, Ratio (Parametric)
Unrelated T-test
RM/Interval, Ratio (Parametric)
Related T-test
Correlation (Parametric)
Pearsons
IV
Something which is being manipulated by the experimenter but can also naturally occur
DV
Something which is measured
Primary data
Getting your own data, answering the research questions
e.g. questionnaires, observations
+unbiased, controlled and up to date data
Secondary data
Pre-existing data, surveys, documents , articles
+cheaper, quicker, ethical
External reliability
Aspects that measure acts same way every time
Test-re-test reliability
Has measure done twice, should give similar responses, it can gain high correlation co-efficient
Inter rater reliability
Two or more raters consistently assessing or interpreting data from the PS
Internal reliability
Consistency measure within itself, should be same throughout
Internal validity
Truly measuring what it claims to
e.g. IV causes changes in the DV
Face validity
Whether on the surface or face it measures what it claims to
Construct validity
Whether your measure is consistently tapping into a construct which you think will exist
Criterion validity
Whether measure you are using is backed up by another measure of sae phenomena
Concurrent validity
Two measure which are agreeing with each other
External validity
Findings can be generalised outside the study
Population validity
Whether results from the sample can be generalised to target population
Ecological validity
Measure and results are generalisable to real world - measuring real life
Null hypothesis
No difference, no correlation between variables
Alternate hypothesis (one tailed)
Directional, specific
e.g. longer the hours revising, the more better the results
Alternate (two tailed)
There will be a difference between the correlations
Structured interviews
Structured questions, designed to give a specific answer - often closed questions