COMM 2001: October 11 th
*MOCK EXAM FOR THIS COURSE THROUGH PASS*
PASS is at: Monday 10am to 1130am & Tuesday 12:30am-2pm
Email: [email protected]
Expected Values= (Row Total x Column Total)/ Sample Size
Review: Nominal Data
When talking about correlation…
If you get the value of one, it usually means you did something wrong as it is highly unlikely that any one
behavior is perfectly correlated with another. That means that your graph would be perfectly linear and
you could predict the dependent variable from the independent variable.
When looking at Cramers V Value…
If you have a value you need to move the decimal point to the left once and then you can figure out the
percentage. An example would be if you had 0.102 that would mean a low correlation as the percentage
is actually just 10.2%.
Ordinal Level Data
Logic=Proportional Reduction of Error (PRE)
Measure which tell us whether knowing the value of one variable reduces chance of making
an error in predicting the value of the other (the dependent variable)
Correlation value which varies between -1 and +1
Indicates direction and strength
*We need to understand what direction the scale is going.
Strength of Correlation of Variables: (Error knowing Not Knowing