Yr2 Research methods Flashcards
What are descriptive statistics?
summary statistics that identify trends and analyse sets of data
What are the 3 types of descriptive stats?
measures of central tendency
measures of dispersion
graphs and tables
What is an inference?
a reasonable conclusion that can be drawn from information presented
What is inferential stats?
the use of stats tests that tell whether or not the differences/ relationships that have been found are statistically significant or not
helps decide which hypothesis to accept = alternative/ null
What is a statistically significant result?
a statistically significant result is one that is unlikely to have occured by chance = reject the null hypothesis
Is P<0.01 more or less stringent?
more stringent = testing at a harsher probability level
What is probability?
the likelihood or chance that something is the case or will happen
Why is a P=0.05 level of significance used?
it represents a reasonable balancing point between the chances of making a type 1 or type 2 error
What is a Type 1 error?
occurs when a null hypothesis is rejected when it is in fact true = a less stringent level of significance was used
What is a Type 2 error?
when a null hypothesis is accepted when it is in fact false = a more stringent level of significance was used
What factors affect the choice of stats test?
the type of data
the type of research design
is it a parametric test
What are the 3 levels of measurement?
nominal
ordinal
interval
What is nominal data?
the data is in separate categories
What is ordinal data?
data is ordered in some way but there is not equal measurement between score , they are simply numbers
What is interval data?
data is measured uing units of equal measurement
What are related designs?
matched pairs and repeated measures
What is an unrelated design?
independant measures
What is an experiment?
testing for a difference between the samples of data
What is a correlation?
a test that detects relationships in samples of paired data
Why are parametric tests more powerful than non-parametric tests?
they are better at detecting significant differences/relationships = uses means and SD rather than rank-ordring and nominal data
What criteria must be met to use a parametric test?
the level of measurement must be interval
data comes from a population that has a normal distribution
the variances of the 2 samples are similar
How do you identify which stats test was used?
- is it looking for a difference or relationship?
- unrelated or related data?
- nominal, ordinal or interval data?
Name all the stats tests.
chi squared
sign test
mann-whitney
wilcoxon
spearman’s rho
unrelated t test
related t test
pearson’s r test
What is the criteria for a chi squared?
test of difference
unrelated data
nominal data
What is the criteria for a sign test?
test of difference
related data
nominal data
What is the criteria for a Mann-Whitney?
test of difference
unrelated data
ordinal data
What is the criteria for a Wilcoxon?
test of difference
related data
ordinal data
What is the criteria for a Spearman’s rho?
test or correlation
related data
ordinal data
What is the criteria for a unrelated t test?
test of difference
unrelated data
interval data
What is the criteria for a related t test?
test of difference
related data
interval data
What is the criteria for a pearson’s r test?
test of correlation
related data
interval data
How do you find the calculated value of T for a Wilcoxon test?
find the rank of differences - if not done for you
the less frequent sign - add up the ranks
What is reliability?
how consistent the findings from an investigation of measuring device are
What are the 2 ways of assessing reliability?
Test-retest method
inter-(observer/rater/observer) reliability
What type of methods is the test-retest used for?
questionaires
psych tests/ experiments
interviews
What does inter-rater reliability assess the reliability of?
content analysis
How is a test-retest carried out?
assess the same person using the same testing method on 2 separate occassions
2 sets of results are correlated and if the correlation coefficient is at least +0.8 then the results are reliable
What is the method of inter-observer reliability as a way of assessing reliability?
the same event is being observed by 2 observers at the same time but independently of each other
data is correlated, if correlation coefficient is at least +0.8 then it is assumed that the behavioural categories were reliable
How can you improve the reliability of questionaires?
some questions may need to be taken out or rewritten
remove ambiguous/ complex Qs
change open Qs to closed
How can you improve the reliability of interviews?
better training for interviewers
use structured
use the same interviewer
How can you improve the reliability of experiments?
high levels of control = lab expts are better
ensure replication to check for correlations b/w repeats
How can you improve the reliability of observations?
behavioural categories are properly operationalised - they are measurable, do not overlapp and all behaviour is covered on the checklist