choosing a statistical test - A-level Flashcards
statistical test definition
used to determine is a significant difference or correlation exists, and consequently if the null hypothesis should be accepted or rejected
sign test defintion
statistical test used for a difference in scores between two related items. data should be at least nominal
levels of measurement definition
quantitative data can be classified into types or levels of measurement such as nominal, ordinal or interval
spearman’s rho definition
a test for a correlation when data is at least ordinal level
person’s r definition
a parametric test for a correlation when data is at interval level
wilcoxon definition
test for a difference between two sets of scores. data should be at least ordinal and use a related design
mann- whitney definition
test for a difference between two sets or scores. data should be at least ordinal and use an unrelated design
related t-test definition
parametric test for a difference between two sets of scores. data must be interval level with a related design such as repeated measures or matched pairs
unrelated t-test
parametric test for a difference between two sets of scores. data must be interval level with an unrelated design
chi-squared definition
test for an associated between two variables or conditions. data should be nominal level using and unrelated design
what do the results of a stats test show
is the null hypothesis should be accepted or rejected and therefore if the difference is statistically significant
what factors are needed when deciding to use a statistical test
-if it is a difference or correlation
-in the case of difference, what experimental design is being used
-level of measurement
how can you tell if it is a difference or correlation
from the wording of the hypothesis
what designs are related designs
repeated measures and matched pairs
what designs are unrelated designs
independent groups
how many conditions do repeated measures participants do
all conditions
how many conditions do matched pairs do
one condition but are matched on a variable making them related
how many conditions do independent groups do
one condition
conditions for chi-squared
-test of difference
-unrelated design
- nominal data
conditions for mann-whitney
-test of difference
-unrelated design
-ordinal data
conditions for unrelated t-test
-test of difference
-unrelated design
-interval data
conditions for sign test
-test of difference
-related design
-nominal data
conditions for wilcoxon
-test of difference
-related design
-ordinal data
conditions for related t-test
-test of difference
-related design
-interval data
conditions for chi-squared
-correlation
-nominal data
conditions for spearman’s rho
-correlation
-ordinal data
conditions for pearson’s r
-correlation
-interval data
what test?
-test of difference
-unrelated design
- nominal data
chi-squared
what test?
-test of difference
-unrelated design
-ordinal data
mann-whitney
what test?
-test of difference
-unrelated design
-interval data
unrelated t-test
what test?
-test of difference
-related design
-nominal data
sign test
what test?
-test of difference
-related design
-ordinal data
wilcoxon
what test?
-test of difference
-related design
-interval data
related t-test
what test?
-correlation
-nominal data
chi-squared
what test?
-correlation
-ordinal data
spearman’s rho
what are the 3 levels of measurement
-nominal
-ordinal
-interval
how is nominal data represented
in the form of categories
is nominal data discrete
yes as only that one item can appear in one of the categories, for example if asked to name favourite football player the vote only appears in one category
how is ordinal data represented
rating scales
what is ordinal data
does not have equal intervals between each unit. lacks precision as is based on subjective opinion not an objective measure such as what is 4 and 8 on rating scale will mean different things to different people
how is ordinal data used in an IQ test
questions are derived from view of what constitutes intelligence rather than any universal measurement
how are questionnaires and psychological test using ordinal data
do not measure something real as do not measure physical entities and measure psychological constructs
why is ordinal data sometimes referred to as ‘unsafe’ data
lacks precision, due to unsafe nature is not used in statistical testing and instead raw scores are converted to ranks and the scores are not used in the calculation
what is interval data
based on numerical scales that include units of equal, precise and defined size
why is interval data ‘better’ than ordinal data
detail is preserved
what sort of things is interval data used for
take measurements with things such as a stopwatch or thermometer that produce measurements based on accepted units of measurements
precision of interval data
most precise and sophisticated form of data in psychology and is necessary for parametric tests
