Inferential testing Flashcards
steps of the sign test
- use plus or minus sign to indicate direction of difference for each p
- calculate observed value by counting up the times the less frequent sign occurs
- use critical value table and decide if hypothesis in one tailed or two tailed
- count how many p’s there are and don’t included omitted results ( THIS IS N)
- find critical value
- see if less, more or = to significance level
how to select and inferential test
3 questions
nature of hypothesis
level of measurement
research design type
Chi Squared
Difference
nominal
independent research deisgn(unrelated)
Mann Whitney
Difference
Ordinal
indepenedet research design( unrelated)
unrelated T test
Difference
Interval
independent research design
Related T test
Difference
interval
repeated designs/matched pairs
Sign test
Difference
Nominal
Repeated/matched pairs
Wilcox
Difference
Ordinal
Independent
Spearman’s rho
Correlational
Ordinal
Repeated/Matched Pairs
Pearson
Correlational
Interval
Repeated/Matched Pairs
Type I error
when a difference/relationship in data is accepted as a real one when in fact it is not.
Type II error
when a difference/relationship in data is rejected but it actually does exist.
How can these errors be reduced?
increasing sample size
the statistical significance
indicates the degree of which a difference/relationship is true
