In chapter two, the authors talked about the different methods that can be used to conduct research. First, the researcher must determine the different dependent and independent variables. The authors also provided two different diagrams: The Table of Statistical Tests, in which the researchers can look up the type of methods to use based on the number of dependent and independent variables used. The second diagram is of a Decision-Making Tree, which is constructed on four types of research questions such as relationship between variables, group differences, membership, and structure. In addition, there are three types of tests that researchers can use to explore the relationship between variables. These tests includes Bivariate correlation and regression, multiple regression, and path analysis.
Similarly, the first test, Bivariate correlation looks at the relationship among the two variables by examining the Pearson correlation coefficient. Contrary to bivariate correlation, bivariate regression utilized the independent variable to calculate or assume the results of the dependent variable. Thus, there exist one independent and one dependent variables in bivariate correlation or regression. Likewise, multiple regression categorizes the best group of independent variables in order to accurately predict the dependent variable. Thus, this method normally consists of multiple independent variables. In addition, multiple regression consists of two or more independent variables, while having only one dependent variable. Furthermore, path analysis examines both direct and indirect relationship among variables in dependent and independent variables through the use of multiple regression. This method consists of two or more independent variables and one or more dependent variables. Thus, this method is used to examine the strength of the relationship from the proposed hypothesis.
Next, the authors discussed about group differences. One of the purpose for testing group differences is to see if there is a relationship between the independent and dependent variables that the researcher wants to test. Thus, a common test that is used to examine this difference is the t test. In t tests, two groups are being compared where the independent variable have two categories and the dependent variable is quantitative. Furthermore, one way analysis of variance (ANOVA) can be used to also test group differences among two or more variables while also examining the differences between and within the groups. Thus, in one way ANOVA, there exist one independent variables with two or more categories and one quantitative dependent variable. Another example of group comparison can be found through the use of one way analysis of covariance (ANCOVA). In this method, two or more groups are used to compare the mean of the dependent variable, but another variable is being controlled for in order to get more accurate results and reduce other variable influences whether from pretreatment or environmental effects. Therefore, in ANCOVA, there is one independent variable with two categories, one dependent variable, and one or more covariant (controlled variable). Likewise, one way multivariate analysis of variance (MANOVA) can also be used to determine within group differences. In MANOVA, two or more dependent variables can be examined at one time to see if there’s a correlation among the two different dependent variables. Thus, in one way MANOVA, there exist one independent variable with two or more categories and two or more quantitative dependent variables.
Another test that can be used to examine the differences between group is multivariate analysis of covariance (MANCOVA). In MANCOVA, similar to ANCOVA, there is one independent variable with two or more categories, two or more quantitative dependent variables, and one or more covariate. Moreover, factorial multivariate analysis of variance (factorial MANOVA) can be used when there are more than one independent variables. In factorial MANOVA, there are two or more categorical independent variables and two or more quantitative dependent variables. The last test to compare group differences is the factorial multivariate analysis of covariance. In this test, one or more dependent needs to be adjusted for. Thus, there exist two or more categorical independent variables, two or more quantitative dependent variables, and one or more covariate. Contrarily, in prediction of group membership, discriminant analysis is used because it examines one dependent variables and predict various independent variables. This type of test has two or more quantitative independent variables and one dependent variable with two or more categories in it.