The choice to employ snowball sampling in this study is driven by the unique and sensitive nature of studying smoking habits. It is challenging to identify and explore through conventional sampling methods. Snowball sampling, by relying on existing social networks and participant referrals, provides a valuable approach to reach individuals engaged in smoking within this particular context. Is there any way to change it to probability sampling technique?
I think "useless" is too strong a word, but your statistical analyses would have to be regarded as "provisional" or "exploratory" rather than actual probabilities.
What could matter most is what previous research has done in this area. In particular, if there is widespread agreement that random sampling is impractical for this population, then you will be providing "useful" information even if it does have limitations.
It will not be seen as useless. Snowball sampling is a recognised method of sampling. However it may just take you a little longer to obtain respondents because it will be based on referrals. However note that it can also be seen as purposeful sampling as you are only targeting respondents who smoke. So to me, its like a mixture/hybrid of snowball and purposeful sampling.
With a small population/sample, statistical power will be a concern as well as the possibility of not being able to run certain tests.
Useless is too strong word to use. However, to strengthen your research it may be good to answer the questions which may arise due to selecting snowballing as a method of sample selection. Some questions you may answer to are:
1. Have earlier researches on similar topic/respondent group mentioned random sampling as a major limitation to get the desired sample?
2. Are time and financial constraints reasons for opting snowballing instead of random sampling?
3. For acceptable statistical power, can desired sample size be achieved through snowballing technique?
4. Can stand for external and internal validity?
There is a relation between the sampl and the original research community, in addition the sampl must be representative to the research community
Snowball sampling will not render your quantitative research "..useless.." but will limit results for generalisability.
The word useless seems weighted. It will not be rendered useless. I think you may need to look at your sample/population size for you to get more statistically significant data.
I hope this helps!!!
No, your quantitative research would not be rendered useless, if you employ the Snowball Sampling Technique in a study where your subject of study is rare to find. You need a non-probabilistic sampling technique like that of snowball sampling to locate the research respondents to get answers to the research questions.
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology
- Research Papers