CONTENTS

Table of Contents
Introduction

1. Evaluation and the Program Planning Process
2. Data Analysis, or "What Do You Mean by That?"
3. Special Challenges in Evaluating Prevention Programs

Supplements
Resources

Evaluation for the Unevaluated:
Program Evaluation 102

Inferential Statistics

Another way to see if your program worked is to look at relationships between program activities and outcomes. For example, Jack could look at the relationship between homework assistance and grades to see if Cool After School made a difference. Another name for this type of relationship is correlation.

Two types of correlations are possible:

1. Negative correlation, also known as inverse correlation. The relationship between two variables is such that as one variable's values tend to increase, the other variable's values tend to decrease. In Jack's case, he would want to see a negative correlation between program participation and drug use. As participation increases, drug use should decrease.
2. Positive correlation. The relationship between two variables is such that as one variable's values tend to increase, the other variable's values also tend to increase. Jack would be looking for a positive association between homework assistance and grades. Students receiving more homework assistance should have improved grades.

Many formulas exist to show correlations such as Pearson's correlation coefficient, or Pearson's rho.

Other statistics used to show results are Chi-square and t-tests. Evaluation 201 will discuss statistics in more depth, using examples. Remember, though, that there's more to evaluation than statistics. Jack heard wonderful things about his program and observed positive changes. This kind of evidence can be valuable as well.

There is value in other methods of research, especially for small programs such as Jack's. Interviews, observations, and similar methods are forms of qualitative research. This can be helpful with small programs in which more objective measures (such as grades) may not yield statistically significant results. See User-Friendly Handbook for Mixed Method Evaluations for more information.