|
I do think high and low ball outliers are worth noting, but - like you - I think it's deceiving to average them into your overall average parameter values, when they don't reflect normal experience with the product.
My last post was a statistics digression:
When you describe a set of data, if I only get to have one number - the average is the best one.
If I only get 2 numbers - I want the average and standard deviation
If I get 4 numbers to characterize a set of data, then I'd want average, standard deviation, maximum and minimum.
And your test kit showdown helped me make sense of stupid magnesium readings.
I'm mostly following this thread for the oddball trace element/metals contents of various salt mixes. You're doing awesome stuff around here, really valuable info.
Sent from my SAMSUNG-SGH-I337 using Tapatalk
|