Hi,
I would like to know how to calculate the Specific Growth Rate (SGR) including the standard deviation. The formula for SGR is: (ln(final weight)-ln(initial weight))*100/duration(d)
Let's assume we have
initial weight: 100 ± 5
final weight: 120 ± 7
duration: 10 days
Can anyone tell me how to get the SGR with the standard deviation (the value after ± )?
Hello Stefan,
If you have the actual data set, just compute the SGR scores for each case, then compute the SD of that set of values.
If you only have summary statistics, you'll need to have the SD of each of the three scores, as well as the correlations among the scores (since, the variance of a difference score is the sum of the two variances minus twice the covariance).
Good luck with your work.
Thank you David! That was helpful already.
Next thing would be the Feed Conversion Ratio (FCR) which is: (feed intake/liveweight gain)
Now I have the weight gains of the animals individually, but only the total amount of feed intake, as the animals were kept together. How do I get the standard deviation of the FCR?
You cannot infer this from your data. One might think of assuming some distribution of food intake (per animal), but this is likely to be closely correlated with liveweight gain - and that is what you actually want to know. This cruical information is missing, it is not in the data, and it cannot be constructed from the data, If you really want to know the FCR, get proper data.
If you have good reasons to believe that all animals eat the same amount (up to some negligible differences), you can calculate the FCR from the averages. The variance of the ratio (food intake/liveweigt gain) may be propagated from the variance of the liveweight gains (Wiki "Fehlerfortpflanzung"). However, this makes sense only if the assumption of equal food intake makes sense. It cannot be derived from the data.
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology
- Research Papers