FORMULA SHELF
12 of 370
The arithmetic behind the work: what to measure, what to type in, and what the answer comes out as. Every formula says what the symbols mean and gives a worked example, so you can check it against a job you have already done.
Statistics
12 formulas
The range the true average is likely to lie in. Use 2.58 for 99%.
Mean 50 and SE 2 gives 46.1 to 53.9.
Comparing variability between groups with different averages, such as two flocks.
SD 2 kg on a 40 kg mean is 5% CV.
Whether two things move together. Correlation is not proof of cause.
r = 0.8 is a strong positive relationship.
Fitting a straight line through data to predict one value from another.
Feed in versus weight gain can give gain = 0.2 + 0.8 x feed.
The everyday average. Sensitive to outliers, so check the median as well.
Yields 4, 5, 6 and 9 t/ha average 6 t/ha.
A fairer average when one very high or low reading would distort the mean.
4, 5, 6, 9 has a median of 5.5.
The most common size, score or category in a batch.
Sizes 2, 3, 3, 4 have a mode of 3.
Smoothing noisy daily readings such as milk, rainfall or prices.
A 7-day average smooths the weekly milk curve.
How many samples you need before an average is trustworthy.
SD 10 and margin 2 needs about 96 samples at 95%.
How precisely a sample mean estimates the true mean. More samples, smaller error.
SD 10 with 25 samples gives SE 2.
Convenience only. Double-check the result against the worked example or a second method before you use it.
How spread out the readings are. The square root is the standard deviation.
Values 2, 4, 4, 10 have a sample variance of 12.
How many standard deviations a reading sits from the average. Useful for grading and outliers.
A 70 kg lamb in a 60 kg flock with SD 5 has z = 2.
This page is an AI-organised index. The label means the grouping and cross-links were compiled with AI; the information itself comes from the public-domain books, reports and manuals listed below. Open any source to read the local copy or follow the link to the original publication.