I am conducting a qualitative study on the perception of Physiotherapy among patients with Low Back Pain!
There is confusion about the necessity of sample size calculation in the qualitative study. Some argued that it is not mandatory but others suggest calculating based on some formula.
Dear respected researchers, what do you think? Please share your thought and pieces of evidence.
Thanks in advance for your kind help.
If you are not going to do inferential statistics it is not important to calculate a sample size. See a good statistics book. David Booth
I believe that research is becoming more hybrid meaning deterministic and Stochastic input.
More deterministic includes quantitative data and reporting more stochastic - qualitative less reporting..
Both above answers are good however I always believe in why did you ask the question in the first place. You state that this is for your study on perception of lower back pain, therefore a VAS scale i presume maybe used? Sample sizing offers the reader an opportunity to see whether it would be applicable to their wider population, it adds validity and reliability to your study. Therefore, who will benefit from your study and will they be able to drawer conclusions that will be applicable to a wider population, if its purely an exploratory study for yourself, then this would be fine not to sample size.
Lee Johnson Thank you, professor. This study will not use any scales. This qualitative study takes in-depth individual interviews of participants with chronic LBP using a semi-structured questionnaire. Results will be helpful when discussing the promotion of physiotherapy among patients with LBP in a low and middle-income country, for example, Bangladesh.
I am also conducting a qualitative study using semi-structured interviews, and I do not believe that sample size calculation is mandatory. The sample size in this type of study is usually not defined a priori. Most of the time, it is based on "data saturation". This means that when the interviews do not provide new information relevant to your research question, you can assume that your sample size is sufficient. But an idea of sample size would be between 10 and 15. However it depends on your interview guide and your research question.
Qualitative studies do not attempt to generalize results to a larger population and to make inferential statistics, unlike quantitative studies. Hope this helps.
Part of what we are talking about is Deterministic vs. Stochastic Research.
Large cohorts for quantitative study eliminates variables so that algorithms can be determined. They eliminate outliers to improve results.
This results in studies that are less applicable and valuable in the real world.
For example, children cannot take a vaccine or medication because they were not including in cohorts.
Woman over 40 cannot use our device because the cohort only accepted women under 40.
I am not a statistician but from a common-sense approach I assume that a mandatory sample size suggests that it should be calculated, in which case it becomes in some senses a quantitative survey with quantitative conclusions
I'm assuming that to calculate a minimum sample size is more about applying statistics conventions. One needs to avoid disraeli's aphorism. " 'Lies, damned lies, and statistics' is a phrase describing the persuasive power of statistics to bolster weak arguments" as Wikipedia profanes!!
Surely common-sense tells us that a larger study size is better than a small one but surely the important question is whether the questions are useful
Just a world-weary thought :)
Like prof has said, it is necessary to have the size of your sample calculated which helps to give the reader the leverage of agreeing with your eventual conclusions. For example, if your low back patients are only about 300 and you have worked with only 20, your conclusions will be weak because it wont be a good representative of the number of hospital patients upon whose opinion you are drawing your conclusions on.