Hi there, I am running a penalty analysis on some sensory data and am a bit confused about "Threshold for population size: Enter the % of the total population that should represent a category to be taken into account for multiple comparisons."
I have data for 4 different food products, does this make my percentage 25%??
Thanks, Serena
Up
0
Down
By Efthalia |
Sep 26, 2017 05:05PM CEST |
XLSTAT Agent
Dear Serena,
Penalty analysis is based on multiple comparisons between the different levels on the JAR scale in order to detect if they are related to the liking scores.
By default, the threshold is fixed at 20%. This means that for all levels representing less than 20% of the total sample size, comparisons tests won’t be performed because they might not be reliable enough.
The higher the value you enter the more you increase the statistical power of the test.
Hi there,
I am running a penalty analysis on some sensory data and am a bit confused about
"Threshold for population size: Enter the % of the total population that should represent a
category to be taken into account for multiple comparisons."
I have data for 4 different food products, does this make my percentage 25%??
Thanks,
Serena
Up
Down
Dear Serena,
Penalty analysis is based on multiple comparisons between the different levels on the JAR scale in order to detect if they are related to the liking scores.
By default, the threshold is fixed at 20%. This means that for all levels representing less than 20% of the total sample size, comparisons tests won’t be performed because they might not be reliable enough.
The higher the value you enter the more you increase the statistical power of the test.
Regards,
Post Your Public Answer