Categorizing Participants and chi-square Flashcards
a priori hypothesis
Based on prior knowledge, expectations, theory
Nominal data and distributions
Nominal data doesn’t distribute like other types of variables, so we use nonparametric stats
Observed frequency
the number of people/responses in a sample that falls within a category of a variable (if 9 people favor vanilla ice cream in your survey about favorite flavors, then 9 is your observed frequency for that variable)
Expected frequency
the number of people/responses expected to fall in a category of a variable- based either on a prior hypotheses or random responses
Random responses
Equal expected frequencies
Unequal expected frequency
Expected frequency isn’t equal across levels of variable.
Use published research and government reports to develop expected frequencies
Degrees of freedom
How many values are free to vary. Chi square used degrees of freedom by subtracting it from levels of variable.
one way chi square or goodness of fit
one-way means we only have one variable; used when you have one sample spread across levels of one nominal variable.
Cohen’s w
Calculated to find effect size.
chi square value/sample size, then take square root
What are the levels of effect size?
.1- small, .3- medium, .5- large
two-way chi square or test of independence
Asks if there is a relationship between variables (as opposed to asking of observed frequencies were different from expected frequencies)
used when we have a 2x2 or greater table (more than one sample)
Degrees of freedom for two-way
(# of levels for v1-1) x (# of levels for v2-1)
Cramers V
effect size for a two-way (contingency coefficient)
What is an insignificant result from a one-way chi square?
Expected and observed are similar
What is an insignificant result from a two-way chi square?
No relationship between categories
