Hi, Leila,

With only 30 subjects, and ceiling effects for Likert variables, there is not much wiggle room for finding effects. Associated statistical power limitations would make it tough for complex analysis. Probably it's a good idea to begin with a simple "exploratory" approach. For example, first see if the intervention affected mood. Next see if mood predicted any of the dependent measures.

If you end up using a regression analysis, you may be interested in an analysis which is used to fine-tune a regression model so that it predicts scores near the extreme ends of the scale, rather than scores in the middle of the scale (i.e., which remedies the problem of "regression toward the mean"). Below I give a link to two free-to-view articles which demonstrate how correlation and regression models predict scores near the middle of the response distribution, and how the analysis I mention improves prediction of outlying (high and low) responses.

• https://odajournal.com/2013/09/20/maximizing-the-accuracy-of-multiple-regression-models-using-unioda-regression-away-from-the-mean/
• https://odajournal.com/2013/10/21/assessing-technician-nurse-and-doctor-ratings-as-predictors-of-overall-satisfaction-of-emergency-room-patients-a-maximum-accuracy-multiple-regression-analysis/

I would be remiss not to mention a much more powerful analysis than regression analysis--should nothing else work...

• https://odajournal.com/2016/09/19/novometric-analysis-with-ordered-class-variables-the-optimal-alternative-to-linear-regression-analysis/

Optimal wishes!

Paul

Hi Paul Yarnold,

I appreciate the information you provided me with. I will try the recommended analyses and share any update here.

I just read an article in which "Jonckheere-Terpstra" (an analysis similar to Kruskal-waliis H test) was recommended. Any Idea about that?

Thank you

Leila

Thank you, Leila,

If I may of assistance with respect to conducting maximum-accuracy analysis, please let me know.

Yes, I am familiar with the Jonckheere test. I was planning to do to it -- what I do to all suboptimal tests (i.e., tests which do not explicitly maximize the objective function)... However, there is too much work, and too little time, to accomplish even a fraction of my desires. That test is seldom used -- much more important targets require attention. The same is true for the Kruskal-Wallis approach (both of these statisticians were my father's dissertation advisers at the University of Chicago--ancient history).

Suffice it to say that the FUTURE requires new tools, methods, and ideas. The old tools have failed. Einstein defined insanity as believing that repeating a failed behavior ("doing the same thing over and over again and expecting different results") will lead to success. I recommend that young researchers look forward, not backwards, in time.

For example, the discoveries being made today in quantum mechanics (and quantum computing) are almost magical! It has been demonstrated that it is possible to instantaneously teleport a single electron more than 300 miles ("beam me up, Scotty"--from Star Trek), and that it is possible to move a single electron a measurable distance backwards in time (time travel is possible). QM scientists are also working with much larger molecules. Particles are being redefined in terms of waves! Space-time is being challenged!!

How much do we know? The universe is dominated by (undetected) dark matter, which remains a mystery. Nevertheless, scientists are discovering new, previously unknown, states of matter!

• https://phys.org/news/2020-02-physicists-accidentally-state.html

To me, the unimaginably magical future is much more interesting than the forgettable failed past, in science...

Hi Paul Yarnold ,

Thank you very much for the valuable guidance. your assistance and help would be greatly appreciated.

I used Wilcoxon test to compare before after effects of the intervention on my mood data which was significant.

The next step is doing Maximal accuracy analysis to see if mood predict my dependent variables.

I read your papers which was definitely informative. But, honestly, I do not know how to do the analysis.

Is there any way to do it using SPSS?

Thank you very much.

leila

Hi, Leila,

Great that W worked--promising! ODA has been compared to W:

• https://odajournal.com/2015/12/06/unioda-vs-wilcoxon-rank-sum-test-a-small-sample-paired-experiment/
• https://odajournal.com/2015/12/12/unioda-vs-wilcoxon-rank-sum-test-invariance-over-monotonic-transformations/

It would be prudent to continue this conversation off-line, so as to maintain the confidentiality of your specific design and data. Please check your eMail.

You are welcome, and I thank you for your interest!

Paul