For Descriptive statistics, we usually find out mean, median and mode as well as standard errors and standard deviation.
Almost in research articles, I can see mean with standard deviation only. Say for example, daily fish catch data: 27, 28, 29, 30, 31, 32 and 33. Number of fish catch per week 30 ± 0.8.
For some data, the standard deviation value is maximum or close to mean value. How to determine, whether it is good/reliable data?
If data is good/reliable can not be judged from the data. You must check if the experiment is well designed and thoroughly performed.
Statistics of standard deviation (SD) is meant for consistency/normalcy of the data. As far as large datasets are concerned, SD frlm mean value tends to be smaller compared to SD small dataset. This is why, larger datasets are preferred.
The sample standard deviation (SD) itself is a random variable (a sample statistic), so it has uncertainty. This is known as the "uncertainty of the uncertainty" (UOU) by the GUM. The UOU decreases with increasing the sample size. Therefore, large datasets are always preferred whenever possible.
- Badges
- Science topic