I'm in need of some assistance. And always find this forum helpful. I'm out of my depth at minute and under pressure to finish my proposal for cut off date. I have most done, accept some of the methodological aspects.
Ok, so, a RCT, single blind. The experiential group will test the intervention + TAU and compare it against the prominent evidence based TAU currently in use. The first null hypothesis is no significant difference in outcome effect size, and its alternative. the second null hypothesis is no difference in drop out rate and its alternative, extant literature suggests the experiential intervention cuts drop out rates in half as well as improving overall effect size.
The experiential treatment involves eliciting feedback from research subjects as a routine outcome measures, two differentiated psychometric instruments are administered at the beginning and end of each individual session (mode) and delivered weekly (dosage). The instruments can be delivered manually, or by iPad, the iPad is linked to a computer based system
The current computer based outcome system that stores and analyses the data for clinical use, provides information on cutoff scores, compares intervention to normative data from up to 500,000 cases on the system, and the overall effect size as established through over 40 years of data (meta-analysis/RCT). It can 'predict' the trajectory of treatment based on these previous data sets and current research (hope I'm making sense here?) and highlight risk of drop out (hypothesis 1) and contributes to overall effect size in outcome (hypothesis 2). Just to say, even if people drop out, the system will still hace results of there outcomes measures to point of dropout
Given that this system can track these issues, my first question is, will this system be enough for analyzing my data? will I also need to use a statistical software package? if so, for the purpose of writing up my proposal, what should I need to focus on explaining when writing the data analysis part? correlation, cronbach etc. Again, I have a very basic understanding of analyzing quantitative data (I'm more qualitative)
Regarding the sample size, I am seeking to establish clinical significants in a naturalistic setting, so is there a recommended sample size? I may be able to find the (N=) population for those seeking such interventions yearly through national prevalence stats, however, right now I'm unaware of these. What sample size would be considered acceptable for generalization in this case?
Regarding outcome measures. It is envisaged that the RCT will be multi-center. Research subjects will be presenting to centers to address an issue, we will call construct A and associated/related issues, constructs B. Generally, research would focus on primary issue construct A, with many measures available to assess this. However, the current intervention will be focused on contructs B, so should I also measure construct A with a pre-post outcome measure? Also, which is primary and secondary measures in this case?
Final question, for now is there an established method for randomizing control groups? is there any variables I should be considering? Just to mention, the people delivering the experiential intervention will be receiving a short 2 day training to assist them with the intervention. I assume therefore, that the control group should be provided with TAU by a different group of practitioners?
Anything else I'm missing?
sorry about long post, hope that I am clear in the information and my questions?
Any help is greatly appreciated guys/gals