Module Nine    Part II: Experimental and Quasi-Experimental Research  

before going on, please complete the student evaluation.

Correlation 
Click here to access the information on how to do a correlation 

Linear Regression  
 

We will not do calculations for the linear regression. It will be enough for you to estimate this visually. Take the scatter plot you have created for the Number of Employees Supervised and the Stress Experienced by Supervisors.  

Can you approximate the regression line?  

Draw it through the points minimizing the distance between the line and each point while making sure the line is absolutely straight. You are not connecting the dots.  

This line is called the regression line or line of best fit. The formula is the same as for a straight line where "a" is the y-intercept (where the line crosses the y-axis) and "b" is the slope (rise over run). 
 

The proportion of variance explained is the correlation squared. If we square the correlation we obtained with the data from the Number of Employees Supervised and the Stress Experienced by Supervisors, we will get the proportion of variance in the dependent variable, Stress Experienced by Supervisors, accounted for by all the predictor variables taken together, Number of Employees. The proportion of variance explained in this example is 77%. 

Create the scatter plot for Class Size and Math Test. Approximate the regression line. Calculate the proportion of variance explained. 

T-test
Click here to access  the information on how to do a T-test

ANOVA 

We will not be calculating ANOVA's. The procedure is similar to the T-test and interpretations are made in comparable ways. The ANOVA is the most appropriate method when the research question is similar to those answered by the T-test, but more than two groups are being compared or a greater number of independent variables are being included and subjects can not be matched on all variables. When more than one dependent variable is included in the study, a MANOVA is ususally the most appropriate method. All of these analysis procedures analyze the variation among means of different groups. 

Chi-Square
Click here to access  the information on how to do a Chi-Square