We presume that for a RBD analysis any number of treatments can be analyzed. However, there is limit on minimum as well as maximum number of treatments beyond which its precision is lost.
My query is focused on a typical situation where we have just 3 treatments but 4 replications under a typical RBD experiment. According to that error degree of freedom is just 6. Does such an analysis is statistically valid and viable? Kindly elaborate.
- As per your details, it seems like you are quite aware of the minimum error degree of freedom to be maintained while designing the treatments (that is 12). But this does not always a requirement if your study reports good findings and your manuscript is well composed. We have published an article in one of the MDPI journals (Nitrogen) where our error degree of freedom was 10 (six treatments and three replications). You can see it here 10.3390/nitrogen2030025. In your case the number of treatments is low, and the error degree of freedom too. I had a similar query to my superior while composing my manuscript to submit to a journal, and he said that usually, it is ok if the error degree of freedom is not 12, but it is good that your work is repeated. That means if your current work (4 treatments and 4 replications- EDF- 6 is repeated for three years then 6*2 ((df of the year: 3-1=2)= 12) is done for two or more seasons (in case of field trial) then there is the possibility of acceptance. However it all depends on the perception of reviewers and editors regarding the error degree of fredom. Sometimes we think too much, so I suggest that, if your study reports novel findings, you compose the manuscript and submit it to the journal.
In case you are worried about academic issues, it is better to consult with your advisors.
I hope this helps you.
Dear Krishna
Here error degree of freedom is 3(for replications) x 2 (for treatments) is just 6 (3*2=6). So, give ur opinion on this issue of low df for errors.
If you would have only 3 treatments, you need at least 8 replications.
Kulvir Singh if it is only three treatments and not going to repeat the trial then there may be problems for acceptance.
- Badges
- Science topic
- Similar topics
- Agricultural Science
- Soil Science
- Soil Analysis