Week 4 Discussion Capstone 3
Name:
Instructor:
Date:
Appropriate test statistics
There were two programs, 1) Children anxiety and 2) underlying condition, according to the database formed. The variables for each program had ordinal measurement according since they represent categorical ordered levels. The total outcome of each instrument will also be ordinal since it is the summation of ordinal variables. Comparing the pre-test and post-test variables for children anxiety group for example, requires a non – parametric test statistics. In this case, Wilcoxon Signed Rank test was the most suitable test statistics which is the counterpart of the parametric paired samples t-test (Shams & Hesamian, 2021). This test statistics is used when qualitative ordinal characteristics of the participants is measured in different periods with the aim of determining whether a treatment had an impact on the subject.
Outcome that will be evaluated
The Wilcoxon Signed Rank test output (Z and p-value) will be evaluated in order to determine whether the differences were statistically significant. The pre and post scores will be compared in this study.
The variables that will be used are; SCAS_Pre_Total and SCAS_Pst_Total which measures children anxiety before and after treatment. Another pair of variables that can be used is ASI_Pre_Total and ASI_Pst_Total for the addiction severity index in the underlying condition program.
Comparison of different groups
A scenario when two groups are compared, say the treatment and the control group, Mann-Whitney U test can be used instead (Karadimitriou et al., 2018). This is the equivalent non-parametric test to the independent samples t-test. This helps in comparing groups from ordinal measurement.
Reference
Shams, M., & Hesamian, G. (2021). Wilcoxon Signed-Rank Test based on Fuzzy Random Variables and Fuzzy Hypotheses. Journal of Decisions and Operations Research.
Karadimitriou, S. M., Marshall, E., & Knox, C. (2018). Mann-Whitney U test. Sheffield: Sheffield Hallam University.
