So you have a numericĀ outcome variable. Question: What is/are your predictor variable(s)?
NO PREDICTORS – I’M TESTING A
CLAIM ABOUT A SINGLE POPULATION
Example: You want to determine if college students do not get an average of eight hours of sleep. In this case, you have no predictor variables and are testing a specific claim (mean hours of sleep equals eight hours) for a single population (college students).
Example: You want to see if there is a relationship between the number of hours students study and their grades on a final exam. So you want to test if a single numeric predictor (hours spent studying) explains a numeric response variable (grade on exam).
I HAVE ONE CATEGORICAL PREDICTOR
Example: You want to test if the cost of rent differs across neighborhoods in your city. Here, you want to test if a categorical predictor (neighborhood) explains a numeric response variable (rent cost).
I HAVE MORE THAN ONE PREDICTOR
Example: You want to figure out which demographic characteristics (gender, ethnicity, age, income) can explain differences in salaries of employees of a state agency. Here, you want to see the effect of multiple predictors (all the demographic variables) on a numeric response variable (salary).