Hong Qiu1, Guo-qin Jin, Ru-feng Jin, Wei-kang Zhao. 1. Department of Preventive Medicine and Medical Statistics, Shanghai University of Traditional Chinese Medicine, Shanghai 201203, China.
Abstract
OBJECTIVE: To introduce the method of analyzing repeated data measured by water maze with SPSS 11.0, and offer a reference statistical method to clinical and basic medicine researchers who take the design of repeated measures. METHODS: Using repeated measures and multivariate analysis of variance (ANOVA) process of the general linear model in SPSS and giving comparison among different groups and different measure time pairwise. RESULTS: Firstly, Mauchly's test of sphericity should be used to judge whether there were relations among the repeatedly measured data. If any (P<or=0.05), multivariate ANOVA should be taken next, or Greenhouse-Geisser corrected results should be taken. Treated effect could be evaluated by estimating between-subject variance. Repeated measurement effect or its interactive effect with treated group could be evaluated by estimating within-subject variance. The method of Bonferroni should be used to do pairwise comparisons of the repeatedly measured data in different measurement time of each treated group. With multivariate ANOVA, data in different treated group of each measurement time could be compared pairwise. CONCLUSION: The repeated measures process of the general linear model is suitable for variance analysis of repeatedly measured data. SPSS statistical package is available to fulfil this process.
OBJECTIVE: To introduce the method of analyzing repeated data measured by water maze with SPSS 11.0, and offer a reference statistical method to clinical and basic medicine researchers who take the design of repeated measures. METHODS: Using repeated measures and multivariate analysis of variance (ANOVA) process of the general linear model in SPSS and giving comparison among different groups and different measure time pairwise. RESULTS: Firstly, Mauchly's test of sphericity should be used to judge whether there were relations among the repeatedly measured data. If any (P<or=0.05), multivariate ANOVA should be taken next, or Greenhouse-Geisser corrected results should be taken. Treated effect could be evaluated by estimating between-subject variance. Repeated measurement effect or its interactive effect with treated group could be evaluated by estimating within-subject variance. The method of Bonferroni should be used to do pairwise comparisons of the repeatedly measured data in different measurement time of each treated group. With multivariate ANOVA, data in different treated group of each measurement time could be compared pairwise. CONCLUSION: The repeated measures process of the general linear model is suitable for variance analysis of repeatedly measured data. SPSS statistical package is available to fulfil this process.