1. The degree to which the following assumptions are met determine the degree to which findings from the tests can be generalized from the sample to the population: (a) significance level of the test is based on probability theory, (b) data are assumed to come from a normally distributed population, (c) the appropriate variables are included in the test, and (d) individuals participating in the research project should be selected through probability sampling.
2. A correlation is a simple description of the degree to which two variables are related.
3. Causation cannot necessarily be established with correlation.
4. A correlation coefficient must be interpreted for its direction and its strength or magnitude.
5. Researchers rely on r² to describe the amount of variance shared between the two variables.
6. Regression is an extension of correlation; however, multiple regression can test for the influence of multiple independent or predictor variables on the dependent or criterion variable.
7. Regression is particularly well suited for communication research as it tests the relationship among naturally occurring variables.
8. R² provides information about the amount of variance of the dependent variable explained by the independent variables separately or in common.
9. The beta weight, or ß, provides information about the direction and strength of influence for each independent variable.
