During the household survey, we profiled lactating cows;
lactation length,
month and year of the last calving,
average milk production per day, milk production at calving, peak, yesterday and late lactation.
Hello Joseph Onam Auma
please check resources
www.sciencedirect.com/science/article/pii/S0022030278838107
www.fao.org/docrep/008/y4705e/y4705e16.htm
There are a number of scientific papers on methods for doing this. I have ATTACHED a file from an older paper that may be useful. If you would like more recent publications, please let me know.
Joseph -
I noticed in this question thread that among your questions you asked for the "average milk production per day." Would that be for the entire farm/household so that you could obtain a total volume by household farm? If not, if you had that for each cow at the household, you would have that value. If you now wanted to know total milk production for the current population, you might use these data, total previous milk production per household farm, as predictor (independent variable) data, if you have this for every household in the population of interest, and a relatively small current sample of the largest cases, i.e. farm households, plus perhaps a small systematic sample of the rest of the current population, for the response data. Thus you might relatively quickly obtain an estimate (actually a "prediction," not a forecast, since Y is a random variable) of the total milk production for your current population.
I promoted this model-based approach for finding total energy production, for various categories, at the US Energy Information Administration, and realized this could be applied elsewhere, whenever there is a census followed by a sample or samples, say an annual census survey and 12 monthly samples. The systematic sampling part of the sample here would make it possible to check that the regression coefficient estimated here is good enough for the entire population, or to perhaps stratify otherwise (as in the Karmel and Jain reference found in the first paper to follow). I know that systematic sampling is chapter 8 of both the 1st and 3rd editions of Cochran's Sampling Techniques, Wiley, but the estimation here would instead be by a model-based (prediction) method.
So, if the possibility of relatively fast and relatively inexpensive predicted/estimated total milk production would be of interest to you, here is a guide:
https://www.researchgate.net/publication/303496276_When_and_How_to_Use_Cutoff_Sampling_with_Prediction
Here is a way to estimate sample size needs with a model I think may be appropriate for you:
https://www.researchgate.net/publication/261947825_Projected_Variance_for_the_Model-based_Classical_Ratio_Estimator_Estimating_Sample_Size_Requirements
This presentation may have some practical advice:
https://www.researchgate.net/publication/263927238_Cutoff_Sampling_and_Estimation_for_Establishment_Surveys, using prediction,
and this was an invited presentation for mathematical statisticians at the US Energy Information Administration:
https://www.researchgate.net/publication/319914742_Quasi-Cutoff_Sampling_and_the_Classical_Ratio_Estimator_-_Application_to_Establishment_Surveys_for_Official_Statistics_at_the_US_Energy_Information_Administration_-_Historical_Development, 2017, using prediction.
I have a project on this topic, as follows:
https://www.researchgate.net/project/Cutoff-and-quasi-cutoff-sampling-with-prediction-for-Official-Statistics, with updates in reverse chronological order.
Cheers - Jim
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publication