Statistical Analyses of Repeated Measures Designs: Problems and Some Solutions

When repeated measures are obtained on the same subjects, interobservation dependencies frequently are generated. The major ways in which these dependencies can arise are illustrated and it is shown how these dependencies may invalidate the use of analysis of variance (ANOVA) and its extension, tren...

Full description

Saved in:
Bibliographic Details
Main Author Rosmann, Michael R
Format Report
LanguageEnglish
Published 1973
Subjects
Analysis of Variance
Correlation
Goodness of Fit
Hypothesis Testing
Mathematical Models
Measurement Techniques
Nonparametric Statistics
Observation
Repeated Measures Design
Research Design
Statistical Analysis
Statistical Significance
Time Series Analysis
Transformations (Mathematics)
Trend Analysis
Online AccessGet more information

Cover

More Information
Summary:When repeated measures are obtained on the same subjects, interobservation dependencies frequently are generated. The major ways in which these dependencies can arise are illustrated and it is shown how these dependencies may invalidate the use of analysis of variance (ANOVA) and its extension, trend analysis, as methods of evaluating the data of such experiments. Several alternative procedures which involve ANOVA and trend analysis and which may avert interobservation dependencies are reviewed. The use of time series analysis with repeated measures experiments is also examined. Finally the advantages and limitations of graphic display only of data and directional correlation are discussed briefly. (Author)