Synopses & Reviews
SALT provides a step-by-step approach to the most useful statistical analyses for language test developers and researchers based on the programs IBM SPSS, Winsteps and Facets. Each chapter focuses on one particular type of analysis, for example, analysing how items in a test are performing or investigating the relationship between two variables. Each chapter begins with an introduction as to why this particular analysis is important for the language tester and then provides explanations about the terms and concepts which the reader will meet in the chapter.
The method for carrying out each analysis is then described in a systematic manner guiding the reader through the procedure for that particular analysis. The main aspects of the output files are then investigated and the results explained. Eleven of the book's appendices provide further opportunities for the reader to repeat many of the statistical procedures on different data sets. Questions guide the reader in checking their understanding of the statistical procedures and output tables.
Synopsis
Provides a step-by-step approach to the most useful statistical analyses for language test developers and researchers using IBM SPSS, Winsteps and Facets. It contains clearly-worked out examples for each analysis with detailed explanations.
Synopsis
SALT provides a step-by-step approach to the most useful statistical analyses for language test developers and researchers based on the programs IBM SPSS, Winsteps and Facets. Each chapter focuses on one particular type of analysis, for example, analysing how items in a test are performing or investigating the relationship between two variables. Each chapter begins with an introduction as to why this particular analysis is important for the language tester and then provides explanations about the terms and concepts which the reader will meet in the chapter.
The method for carrying out each analysis is then described in a systematic manner guiding the reader through the procedure for that particular analysis. The main aspects of the output files are then investigated and the results explained. Eleven of the book's appendices provide further opportunities for the reader to repeat many of the statistical procedures on different data sets. Questions guide the reader in checking their understanding of the statistical procedures and output tables.
About the Author
RITA GREEN is an international consultant and trainer as well as an EALTA (European Association for Language Testing and Assessment) Expert Member. Her current work involves being the trainer on projects in Austria, Germany, Spain and Thailand, all of which involve statistical work. Rita has directed courses in statistical analysis for over 15 years and taught on the statistical component of the distance MA course at Lancaster University.
Table of Contents
Foreword; J. Charles Alderson
1. Data Entry
2. Checking and Correcting Data Files
3. Item Analysis
4. Descriptive Statistics
5. Analysing Test Taker Feedback
6. Comparing Performance I: Means, Scatterplots and Correlations
7. Comparing Performance II: Parametric and Non-Parametric Analyses
8. Comparing Performance III: ANOVA
9. Factor Analysis
10. Creating a Control File and Convergence Table
11. Analysing the Convergence Table and Variable Map
12. Item and Person Statistics
13. Distracter Analysis
14. Creating and Running a Specifications File
15. Analysing the Iteration Report and Vertical Ruler
16. Raters and Item Measurement Reports
Appendices