How Many Samples do I Need? A Formula for our Quality Meters!

October 21, 2020 at 9:25 pm | Updated October 21, 2020 at 9:25 pm | < 1 min read

Have you looked at our Produce Quality Meters including the Avocado, Mango, and Kiwi specific models but aren’t sure how many samples you would take?  Here is an excerpt from an article written by Jeremy Walsh and Kerry Walsh of Central Queensland University that covers this topic:

So how many samples should be taken, i.e., how many fruit should be measured in an orchard to get a reliable estimate of that orchard?  Well the answer is easy if all fruit had the same DMC – only one fruit would need to be sampled!  But as variation in fruit DMC increases, the number of samples needed increases.  There is a little  formulae to estimate the required number of samples, n, given the variation (standard deviation, SD) of fruit DMC, the acceptable error (e) and the desired probability    level.  For a 95% probability and an acceptable error of 1% DMC, the formulae is:

 

Now a typical orchard has a SD of DMC between fruit of 2.5 %, for which case the required number of samples is n = 25.  Odds are few Australian producers assess 25 fruit per orchard! But assessment of 25 fruit is easily possible with NIRS technology.

To continue reading this article entitled “How Good is That?” please click here.