Hi Korey,
Generally, a pilot study is conducted in order to get the estimates needed to do a proper sample size calculation. More specifically, when there is no available evidence on what the effect of the intervention is, then a researcher will conduct a small pilot in order to get an estimate of the means and std devs for each of the groups under study (assuming a continuous outcome variable). These values are then plugged into a sample size formula to get an idea of the sample size needed for a full-fledged study.
So the general wisdom for a pilot is to get as many subjects into the study as your budget/resources will allow, and perform the pilot as you would the broader study (so as not to introduce biases that you'll have to worry about later).
I hope this helps
Ariel
Think it really just depends on what exactly you are trying to manipulate. Ideally, you would have a sample size large enough to rule out confounding variables and gain a better picture of your dependent variable(s). If you are a bit more specific I can probably help get a ballpark on a sample size without doing any type of g-power analysis.
Dear Korey, the searched sample size, is it of the pilot sample? Or, is it of the future study sample?
Hi Korey,
Generally, a pilot study is conducted in order to get the estimates needed to do a proper sample size calculation. More specifically, when there is no available evidence on what the effect of the intervention is, then a researcher will conduct a small pilot in order to get an estimate of the means and std devs for each of the groups under study (assuming a continuous outcome variable). These values are then plugged into a sample size formula to get an idea of the sample size needed for a full-fledged study.
So the general wisdom for a pilot is to get as many subjects into the study as your budget/resources will allow, and perform the pilot as you would the broader study (so as not to introduce biases that you'll have to worry about later).
I hope this helps
Ariel
Hello Korey,
I had only 5 participants for my pilot study, it was more about understanding questions, time needed to fill in questionnaires, and also asked about general opinions.
Hope it helps a bit.
A pilot study is used to get an estimate of the variability needed to do a proper sample size calculation. I recommend the use of the pilot study with 10 participants at least.
Korey, as others have suggested, one important purpose of the pilot study is to furnish values you'll need to do sample size estimation for the full study. But what those values are depends on what type of analysis you'll be doing in the full study (e.g., t-test versus 2x2 table with odds ratio vs some kind of regression model, etc). So I think it would help if you gave a bit more information about the nature of the study, and what type of analysis you'll be using in the full study. HTH.
In addition to what others have already stated:
Does the pilot study reduce the maximum sample size for the main study? This could happen if you have a fixed budget, or if all work must be completed in 2 years. Either way, performing a pilot study reduces the resources available for the full study. In this case you will have to decide where the benefit from doing a pilot study exceeds the loss from the main study.
Also keep in mind that pilot studies tend to be smaller versions of the main research project. So I am now trying to get an accurate estimate of the variance using 10 replicates to plan a sample size for a project with 100 replicates. We know that smaller sample sizes increase the uncertainty about an estimate. We know that with 10 replicates you will have no ability to estimate the error in your estimated variance. Of course, the tactic is to assume that everything is normally distributed in the Pilot study. One gathers a small quantity of data (3+ replicates), and uses the results to plan a larger study. With luck, the pilot study was close enough.
In some cases, the pilot study is more useful as a dry run for the main study. Planning the time it takes to do certain tasks, making sure all the equipment is functioning, training people, and so forth.
Finally, is the pilot study intended to be publishable?
Hi. How can I explain results that are statistically significant but not reach the magnitude pre established at the sample size calculation?
Thank you everyone for your answers. It turns out that the study I originally wanted to do became a survey study of about 80 participants.
Pilot study
Saunders et al., (2007) state that prior to using the questionnaire to collect data it should be pilot tested. Saunders et al., (2007) point out the purpose of the pilot test is to refine the questionnaire so that the respondents will have no problems in answering the questions and also there will be no problems in recording the data
Fink (2003b) as cited in Saunders et al., (2007) state that the minimum number for a pilot study is 10.
Reference
Saunders, M.N., 2007. Research methods for business students, 5/e. Pearson Education India.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size