If I am measuring, say, the height of individuals in a population, and wish to state results like mean, standard deviation, etc., then I know that my sample size will affect the margin of error of these results. I want to know how to calculate the appropriate precision I should use to state my results, based on the precision of the individual measurements. To use an extreme example, if each of my measurements is only to the nearest meter, it doesn't appear to make sense to state the mean to the nearest millimeter.
have a look at to this;
Article TO DETERMINE SKEWNESS, MEAN AND DEVIATION WITH A NEW APPROAC...
I would report the mean to one decimal place better than my measurements. So if I measure weight to the nearest gram, I will report the mean to the nearest tenth. I would report the standard deviation to the nearest hundredth, or two decimal places better than the measurements. The reason is that this will reduce the round-off error if someone were to try using my results in another analysis.
Sometimes I will increase reported accuracy. Sometimes by accident, sometimes to keep everything similar. Maybe I measure total body weight to the nearest gram, but the weight of the lungs to the nearest tenth of a gram. I would then report mean lung weight to the nearest hundredth of a gram, and it might make the table look nicer to also report total body weight to the nearest hundredth of a gram.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size