choosing statistical test

Q1

does the research seek to compare two or more groups OR try to understand how they relate to each other

compare is what kind of design

cause and effect

terms that indicate relate

associate , correlate, predict

Q2 compare

compare: are there multiple IV

yes: complex comparison

no: basic comparison

Q2 relate

are there more than two total variables

yes: complex relational ( predictive model)

no: basic relational

Q3

what type of data are we working with

q3 compare

only interested in the DV

q3 relational

interested in the data of ALL the variables not just one of them

parametric test indicate

stronger stat tests

parametric assumptions

• normality

• homogenity of variance

• independence assumption

what if there is a violation of one or more assumtpions

we treat the interval or ration data as ordinal

q4 only applies to

comparison( caus and effect)

are between subjects or within subjects stronger

between subjects design

longitudinal

taking data from the same people over and over across a long period of time and then comparing those measures

cross over

mix of between and within subjects design

matched

• considered WITHIN SUBJECTS DESIGN

• matching people randomly based on specific characteristics to ensure comparability between groups.