The Program Evaluation Prism Using Statistical Methods to Discover Patterns

by Abbott, Martin Lee

• ISBN13: 9780470579046
• ISBN10: 0470579048
• Edition: 1st
• Format: Paperback
• Copyright: 2010-11-23
• Publisher: Wiley

Summary

This book is a comprehensive treatment of correlation/regression techniques and using SPSS for interpretation of findings. Striking a balance between detailed coverage and approachability, this book provides a thorough treatment of the elements of regression and how they can be used with real research problems in program evaluation.The author begins with a basic introduction to evaluation methodology, and its ability to recognize embedded patterns of meaning in research data. Subsequent chapters explore the statistical tools that can be applied by researchers and evaluators irrespective of the design that was used to generate this data.Topics of coverage include: correlation, single predictor regression, multiple correlation, part and partial correlation, detection of extreme scores, multiple regression, regression with continuous predictors, coding of categorical data, regression with categorical predictors, methods for entering predictors in multiple regression, and interaction in multiple regression.Each chapter is presented in the same comprehensive format: an introduction to the topic, followed by a discussion of its primary elements, illustrations of the data through numerous tables and figures, SPSS procedures for designing the analysis, SPSS output of the analysis , and guidance on how to interpret findings from the analyses. Discover Note and Research Steps sections illustrate how using statistical processes can unveil unobserved patterns and assist readers with identifying such patterns in their own data.Real-world analyses are used throughout the book, utilizing meaningful social issues as a catalyst for teaching statistical procedures, and a related Web site features additional data sets, solutions, and research projects for readers.

Author Biography

Martin Lee Abbott, PhD is Professor of Sociology at Seattle Pacific University, where he also serves as Executive Director of the Washington School Research Center, an independent research and data analysis center funded by the Bill and Melinda Gates Foundation. Dr. Abbott has held positions in both academia and industry, focusing his consulting and reaching in the areas of program evaluation, applied sociology, statistics, and research methods.

Table of Contents

Preface	p. xi
Acknowledgments	p. xiii
Introduction	p. 1
Initial Considerations	p. 4
Book Plan	p. 5
Real Examples	p. 6
Using Statistical Programs	p. 7
The Evaluator's Journey	p. 7
The Elements of Evaluation	p. 9
The Nature of Evaluation	p. 9
Evaluation Concerns	p. 10
Evaluation Standards	p. 12
Methods used in Evaluation	p. 12
The Evaluator's Tools	p. 13
Evaluation Hurdles	p. 14
Quantification	p. 14
Resistance to Quantification	p. 15
The Nature of Quantification	p. 15
Qualitative Methods	p. 17
Specialization	p. 18
Statistical Issues	p. 19
Certainty versus Probability	p. 19
Statistical Significance	p. 20
Effect Sizes	p. 20
Can We Achieve Certainty?	p. 21
Dispelling the Mystique of Statistics	p. 22
Research Literacy	p. 23
The Discovery Questions	p. 24
School Characteristics and Student Learning	p. 24
Worker Participation	p. 25
The Impact of Technology on the Classroom	p. 27
Classroom Observation Data	p. 28
Discovery Learning	p. 29
Terms and Concepts	p. 29
Using SPSS“	p. 31
General Features	p. 32
Management Functions	p. 34
Reading and Importing Data	p. 34
Sort	p. 34
Split File	p. 35
Transform/Compute (Creating Indexes)	p. 36
Merge	p. 37
Analysis Functions	p. 41
Graphing Functions	p. 41
Correlation	p. 47
The Nature of Correlation	p. 47
Prediction	p. 49
Correlation Is Not Causation	p. 50
Pearson's r	p. 51
Strength and Direction	p. 51
A Note on the Nature of the Data	p. 54
Interpreting Pearson's r	p. 56
Testing the Statistical Significance of a Correlation	p. 57
The "Practical Significance" of r: Effect Sizes	p. 60
An Evaluation Example of Correlation: The Impact of Technology on Teaching and Learning	p. 63
Influences on Correlation	p. 71
Restricted Range	p. 71
Extreme (Outlier) Scores	p. 73
Other Kinds of Correlation	p. 73
A Research Example of Spearman's rho Correlation	p. 75
Nonlinear Correlation	p. 78
"Extending" Correlation to Include Additional Variables	p. 78
Correlation-Detail for the Curious	p. 78
Computing Pearson's r	p. 78
Assumptions of Correlation	p. 79
Nonlinear Correlation	p. 81
Discovery Learning	p. 81
Terms and Concepts	p. 83
Real World Lab-Correlation	p. 84
Description of the Data	p. 84
Evaluation Questions	p. 86
Regression	p. 87
The Regression Line-Line of "Best Fit"	p. 88
The Regression Formula	p. 91
Standard Error of Estimate	p. 93
Confidence Interval	p. 94
Residuals	p. 94
Regression Example with Achievement Data	p. 98
The Results of the Analysis	p. 101
The Graph of the Results	p. 105
Standard Error of the Estimate	p. 106
The Confidence Interval	p. 108
Detail-For the Curious	p. 110
Assumptions of Regression	p. 110
Fixed and Random Effects Modeling	p. 111
Nonlinear Correlation	p. 111
Calculating the Standard Error of the Estimate	p. 116
Discovery Learning	p. 117
Terms and Concepts	p. 118
Real World Lab-Bivariate Regression	p. 120
Cleaning the Data-Detecting Outliers	p. 121
Univariate Extreme Scores	p. 122
Multivariate Extreme Scores	p. 124
Distance Statistics	p. 128
Influence Statistics	p. 131
Discovery Learning	p. 134
Terms and Concepts	p. 135
Real World Lab-Extreme Scores	p. 136
Multiple Correlation	p. 137
Introduction	p. 137
Control Variables	p. 139
Mediator Variables	p. 140
Using Multiple Correlation to Control Variables: Partial and Semipartial Correlation	p. 142
Partial Correlation	p. 143
Semipartial (Part) Correlation	p. 147
Discovery Learning	p. 149
Terms and Concepts	p. 149
Real World Lab-Partial and Semipartial Correlation	p. 151
Multiple Regression	p. 153
Multiple Regression with Two Predictor Variables	p. 154
Uses of Multiple Regression	p. 154
Multiple Regression Outcomes	p. 155
Omnibus Findings for the Overall Model	p. 155
Individual Predictors	p. 157
Additional SPSS“ Results	p. 158
Multiple Regression: How to Enter Predictors	p. 165
Stepwise Regression and Other Methods	p. 167
Assumptions of Multiple Regression	p. 171
Multicollinearity	p. 171
Cleaning the Database	p. 174
Multiple Regression with More Than Two Predictor Variables: Research Examples	p. 174
Predicting the Impact of School Variables on Teaching and Learning: The TAGLIT Data	p. 174
Omnibus Findings	p. 177
Results of Individual Predictors	p. 177
The "Larger Model" of School Achievement	p. 178
Discovery Learning	p. 180
Terms and Concepts	p. 181
Real World Lab-Multiple Regression	p. 182
Coding-Using Multiple Regression with Categorical Variables	p. 185
Nature of Dummy Variables	p. 185
One Categorical Variable with Two Groups	p. 186
Creating Dummy Variables	p. 193
Creating Subvariables in SPSS“	p. 194
One Categorical Variable with More Than Two Groups	p. 199
A Hypothetical Example	p. 200
An Example from the School Database	p. 202
Discovery Learning	p. 203
Detail for the Curious-False Dichotomies	p. 203
Terms and Concepts	p. 204
Real World Lab-Dummy Coding	p. 205
Interaction	p. 207
Interactions with Continuous Variables	p. 209
Interaction with Categorical Variables	p. 216
Discovery Learning	p. 221
Terms and Concepts	p. 222
Real World Lat-Interaction	p. 223
Discovery Learning Through Correlation and Regression	p. 225
Overall Discovery Notes	p. 225
Findings from the Data	p. 226
Student Academic Achievement	p. 226
Workplace Participation	p. 227
Impact of Technology on Student Learning	p. 228
Advanced Statistical Techniques	p. 229
Hierarchical Linear Modeling	p. 229
Structural Equation Modeling and Path Analysis	p. 230
Other Regression Procedures	p. 231
Practical Application Analyses	p. 233
Real World Lab-Correlation	p. 233
Real World Lab-Bivariate Regression	p. 235
Real World Lab-Extreme Scores	p. 239
Real World Lab-Partial and Semi-Partial Correlation	p. 245
Real World Lab-Multiple Regression	p. 247
Real World Lab-Dummy Coding	p. 249
Real World Lab-Interaction	p. 252
Appendix	p. 259
References	p. 305
Index	p. 307
Table of Contents provided by Ingram. All Rights Reserved.