Part Tabela Part 1: Some True/False Statements (just enter Tor F for each statement). Q1-4 Samples must be Large (ie, a large number of observations, so collect as much sample data as you can) Random (.e., randomly drawn from the population of interest) Similar to your population of interest in all respects. Similar to your population of interest in terms of the issues you are studying Q5-8 In looking at p-values A low p-value is inherently good (1.e. the lower the p-value, the better) A p value is only useful when you are using sample data to estimate a population A high p-value means you probably had a bad sample A high p-value means you probably had a good sample Q9-15 Now some general statements from the final part of semester For categorical variables, categories should be overlapping To say that two variables are related means that changes in one variable correspond to changes in the other variable. P-values can be applied to t-tests, regression, and crosstabs, but not to correlations. An R2 of 4.25 is desirable Correlations range from negative -1 to positive 1 When you interpret Crosstab results to determine whether two variables are related to each other, it helps to look at just frequency counts rather than frequency percent (%) calculated from frequency counts Regression coefficients do not have any units they are unit free)