The study is a Residential nature-based intervention programme to be conducted for 5 days focussing on strengthening the environmental behaviour of the young adults (18 to 25 age group). The study involves a Pre-Post control group design in which the participants to treatment groups (control and experimental group) will be assigned based on their pre-test scores (of a standard questionnaire). Moreover, matching of the participants to the two groups will be done based on the gender of the participants as well.
The participants shall be assessed at three intervals of time; pre-test prior to the programme and an immediate post-test and a follow-up test after a month of the programme completion.
The aim of the study is to assess the effectiveness of the nature-based intervention programme on the environmental behaviour of young adults.
In this regard, I am unable to find relevant literature to calculate the minimum sample size required for the study. To be precise, the study area falls under the field of "Educational Psychology". On a high note, conducting a pilot study to determine the sample size, with fewer resources and funds also becomes not feasible.
Considering all these, it would highly be appreciable if I get help in deciding the minimum samples required for this residential intervention programme with proper literature cited.
ANY OTHER SUGGESTIONS ARE ALSO WELCOME WITH REGARD TO THE STUDY DESIGN AND METHODOLOGY.
Howdy,
It should be possible to do a power analysis if you have estimates from previous research. Depending on the specific statistical method you plan to use to test your hypotheses you might be able to use G*power https://www.psychologie.hhu.de/arbeitsgruppen/allgemeine-psychologie-und-arbeitspsychologie/gpower (which has a lot of tutorials online and on Youtube). Alternatively, if you are familiar with R, there are lots of R functions for power analyses (you could search something like "*name of statistical test* power analysis R". If you are more comfortable in Excel, there are options for that too (e.g., monte carlo simulation - https://www.riskamp.com/how-to)
As Daniel P. Moriarity said - G*Power and R are very good options to estimate how many participants you need. It is good not only for dissertation, but also in any kind of research.
In longitudinal study you should also remember to recruit more people in first measurement than you need basing on estimations to have at the end to do proper analysis. For example, I have conducted longitudinal study with short- and medium-term post-tests (day, two weeks, and two months after experiment) few months ago. Our plan was to have ~ 215 participants who will take part in all measurements, so we've recruited ~ 400 people at the beginning of the study. Finally, we had planned number of participants.
Drop-out rate depends on many conditions (e.g. length of intervals, financial gratification, subject of the study, how time-consuming the study is for participants). So, you should consider how many people can leave your study between the first and the last measurement.
Good luck!
A good read for you can be
Hickey, G. L., Grant, S. W., Dunning, J., & Siepe, M. (2018). Statistical primer: sample size and power calculations-why, when and how?. European journal of cardio-thoracic surgery : official journal of the European Association for Cardio-thoracic Surgery, 54(1), 4–9. https://doi.org/10.1093/ejcts/ezy169
If you find formulas are difficult for you you, you can also check the book titled Multivariate Data Analysis (7th Edition) 7th by Joseph F. Hair Jr, William C. Black, Barry J. Babin, Rolph E. Anderson for rules of thumb on sample size.
Hi. Determining a minimum sample size is subject to to the type of the research you are using as well as the population you are working on. You can use scientific procedure to estimate the sample.
Friends' description is very comprehensive and complete, but I suggest using NCS-passast software
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size