Phew! Just came out of the Preconference Yoga Research 101: The Basics of Yoga Research for Therapists, Instructors, and New Investigators. It was a very fast, intensive session with four speakers individual and sitting on a panel, plus questions. This could only have happened with Sat Bir running this program, it's safe to say. I'll have to write this in parts because in 20 minutes we are to be back in class for a live-video feed with Dr. Ornish - don't want to miss that!!
Lorenzo Cohen spent a quick 25 minutes on his topic of Documentation, Evaluation, and Measurement. (of course this is all aimed at Yoga research). He told us that the most important thing with developing your research is to use standardized measures. If you want to measure back pain you need to find an instrument that already exists for assessing that type of outcome and not just developing something for yourself. You will need to do literature searches to identify these measures. If it’s not in the literature, then it isn’t valid, isn’t standardized, and don’t go out an invent something because you're probably going to be wasting lots of time (and money).
He talked about how some treatment effects will be obvious in terms of improvement. Flexibility, for example, is expected and easy to measure. It's easy to assess; clients can easily assess it themselves.
But, getting at more subtle aspects of effects of yoga, he says, is the real challenge and opportunity and that we haven't been able to quantify these more subtle things.
Other quick points he made, having so little time to present, were:
Longer follow-up intervals are better because we'd like to study longterm effects. We need to know if people who are in a study are continuing to do the yoga and if the effects are the same as previously noticed.
Timing is also important , and other presenters hit on this too. Exactly when do you ask people the followup question? Recommends up to a week after and then certain intervals after the formal instruction is over.
Sample size: larger samples can detect smaller treatment effects. There is a bias in the literature in favor of larger sample size, more on this when I get to Cohen's presentation.
Be aware that there is a difference between what might be clinically significant versus statistically significant. P.05 (statistical significance) is usually considered necessary to get a study published. But people often forget what is clinically significant. What is a clinically meaningful difference between groups? This all needs to be written into the discussion section of the study.
Cohen emphasized that the more demographic characteristics that can be known about the subjects the better off you will be. In addition to the demographics, also find out the medical characteristics - and not by relying on the patients. Cohen related how most patients in a particular study were unable to report at what stage their cancer was in. Get access to medical records if possible. Other covariates that influence the outcomes are such things as knowledge of medications, or how could a job be influencing the outcomes for a person in a low back pain study, for example.
Yes, this is a crash course on Yoga research. Back soon.