Chi Square

The chi-square test can be used as a test for ________

independence

What does a chi-square test do

It uses a contingency table to compare two categorical variables

It tests whether there is a difference in proportions between the two sets of data

For what kind of data are chi square tests used

The chi-square test can be used when both variables are nominal or when one variable is nominal and the other one is ordinal

What are H_O and H_A

H_O: The two variables are independent (no association)

H_A: The two variables are NOT independent (association)

Difference between observed and expected test statistic

Observed: refers to the count observed in a cell (what the data say)

Expected: refers to the count that would be expected if Ho is true.

Calculate the expected count

Meaning of a significant p value

P-value < 0.05 = Significant p-value and therefore evidence against the null hypothesis of no association

Give steps for finding the standard deviation

1. Find mean of values

2. Minus mean from each value

3. Square new values

4. Find mean of new values

5. Square root new mean

standard deviation measures _________

spread

Central tendency uses what metrics

mean, median & mode

What is a p value

A p value is the probability that you got the correlation between these 2 data sets (e.g. people who smoke & people who have cancer) completely by chance. What’s the probability that smoking actually didn’t have an effect on cancer but you have these figures anyway? The Ho assumes that there is no relationship between the 2 variables. If p<0.05 Ho is wrong. If p>0.05, there isn’t enough evidence to declare Ho false. Ha assumes there is a correlation between the 2 variables

The median is a robust measure of

central tendency in the presence of outliers.

The mean is a measure of central tendency that is sensitive/not sensitive to outliers

The mean is a measure of central tendency that is sensitive to outliers

What are the preferred measures of central tendency and variability for skewed distributions.

The median and interquartile range