|
Andrew M. Ball MS MBA PT -> Re: Proposal for study of Chiropractic (December 4, 2001 11:20:00 AM)
|
Wisecracker,
Very good questions. You're right to be concerned with issues of study validity. Keep in mind, however, that I don't believe the state of the existing knowledge base to yet lend itself to meaningful quantitative study. I propose a ethnographic study using semi-structured interviews of patients (including discussion of perceptions of traditional healthcare, chiropractic care, lifestyle, and PMH over the past 5 years), triangulation of PMH data with PCP records.
Each participant interview would last 60 minutes or so, be recorded, transcribed, and fed into a NUD*IST N4 database (that's like SPSS for non-numerical data). The sampling would be a purposeful sample of chiropractic patients, and interviews would continue until the data is saturated (e.g NUD*IST detects no new themes with additional interviews). Sample randomization is a non-issue because the intent of study is not to generalize a cause-and-effect relationship, but rather to describe the culture and experiences of chiropractic patients, and compare that to a purposeful sample of generally demographically "matched" non-chiropractic patients.
A power analysis, is therefore unnecessary because no parametric statistics are going to be run, and no generalization (at least not in the same sense as quantitative research) is going to take place.
Your question about "How will the questionnaire questions be defined per population? ie, males have tendencies toward downplaying symptoms, while females may pay more attention to symptoms. Also, different populations may view symptoms completely different manners, etc." is a great one, and gets into the qualitative validity issues of REFLEXIVITY, POSITIONING, and DATA TRIANGULATION.
The goal of this study will be to describe two different cultures and experiences, that of chiropractic patients, and that of non-chiropractic patients. This study will not attempt to test yet another hypothesis of a cause-and-effect relationship of known variables, but rather will begin with the idea that not all independent variables are known, and that they must be exposed before any meaningful quantitative analysis can occur.
A common complaint of quantitative researchers is that qualitative results cannot be generalized to larger populations. Not only is this assumption not entirely true, it misses the point, of qualitative research such as this. Qualitative research is not as concerned with generalization to larger populations as it is to different samples, settings, and situations. In other words, the experiences of chiropractic patients may be similar to the experiences of massage therapy patients if similarities in patient characteristics can be established . . . paving the way for future quantitative research in this area.
There is therefore no need to control and measure the effect of unknown variables through traditional tools of random sampling, precision measurement, and sophisticated statistical analysis because this study will make no attempts to test a hypothetical cause-and-effect relationship.
As previously stated, the closest approximation of threats to the validity of qualitative research such as this, surround the issues of reflexivity, positioning, and triangulation of data.
Reflexivity refers to the influence that the researcher's background, bias, and line of questioning will have upon the participant during interviews. This not a limitation of qualitative research per se, rather a situational variable of which the qualitative researcher must be aware. Specifically, some people who go to chiropractors have had poor prior experiences with PT's, and vice versa. Being a physical therapist, this reality will undoubtedly affect any participant interview that I conduct.
The only truly limiting factor of this type of research is that of position. Positioning (or more to the point, positioning mismatch) is a problem when the gender, class, ethnicity, sexual orientation, age, experience, and/or disability, of the researcher and participant are different. Upon write-up, detailed description of the purposeful sample will include relevant demographic information, taking care to point out differences in gender, class, ethnicity, age, experience, and education of study participants, and the examiner.
The reliability and validity of the data will be enhanced through data triangulation, including systematic recording of interviews, and a logical consistency in interpretation, as described by DePoy and Gitlin. Attempts will be made to further enhance reliability and validity of data through a team approach to the coding of data, using several other researchers (PT's and DC's alike) to code the transcribed data before entering the NUD*IST N4 database.
In summary, qualitative research completes the circle of research by aiming to, "describe the experiences of people in particular settings and to understand their perspectives. Its purpose is also to develop hypothesis, concepts and theory." It is guided by three underlying assumptions. First, human behaviors extend beyond the observable and incorporate subjective meanings, values, and perceptions that are difficult, (if not impossible), to quantify. Second, actions and ideas can only truly be understood from within the physical, economic, and socio-cultural contexts in which they exist. Third, people (including researchers using any methodology) interpret realities differently due to past experiences. There are therefore no objective truths . . . but the culture and experiences of a group (chiropractic patients) can be described and compared to that of another group (non-chiropractic patients). It's then up to the researcher to test, by subsequent quantitative study, the hypotheses generated from the qualitative study . . . or more to the point --- pave the way for someone else to do that.
Drew
P.S. NUD*IST is a Trademark of QSR, and stands for Non-numerical Unstructured Data* Indexing, Searching, and Theorizing. I thought I'd clear that up for a few RehabEdge members who got excited about purchasing a "NUD*IST in a box." LOL!
|
|
|
|