I am planning to do a study with the help of 45 farmers who are going to produce hybrid chilly seed for the first time, and I want to measure in what way semi-automated greenhouse will affect on the productivity of hybrid chilly seed. I am planning to allow 43 farmers out of 45 to use semi-automatic greenhouse while, 2 farmers without the semi - automatic feature in their greenhouses. Structure, type, size as well as conditions of each and every greenhouse is similar to each other. What type of design will mostly suit for this study???
I suggest you to have more balanced groups if you want to perform a statistical comparison of the control versus treatment group. In the current design, you only have 2 farmers in your control group (non semi-automatic) which is not even enough to perform non-parametrical statistical tests.
You should randomize if you can, which will probably involve control and treatment groups of the same size. You should check the literature on experiments to see if you have enough statistical power with just 45 units.
See, e.g.,
https://www.povertyactionlab.org/research-resources?view=toc
This reference should be also useful:
https://www.povertyactionlab.org/sites/default/files/research-resources/2018.03.21-Rules-of-Thumb-for-Sample-Size-and-Power.pdf
I understood that the study will be conducted in a Green house with only one hybrid chilly seed. The number of control farmers (2) is not satisfactory. I think you can use a factorial complete block design. Advisable to check more possible designs.
Thank you!
- Badges
- Science topic
- Similar topics
- Agriculture
- Horticulture
- Seeds