cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
The Discovery Summit 2025 Call for Content is open! Submit an abstract today to present at our premier analytics conference.
See how to use JMP Live to centralize and share reports within groups. Webinar with Q&A April 4, 2pm ET.
Choose Language Hide Translation Bar
View Original Published Thread

Which statistical test is recommended?

Erez
Level I

Hello everyone,

 

I performed an experiment that tests whether a honey bee tends to turn in the same direction or change direction (right or left). The bee flies through a maze with turns, each time the bee chose to change direction I added 1 to the counter. If the direction was not changed, the counter was at 0. Each honey bee was tested 8 times (8 trials), a total of 60 bees from 2 treatment groups and 5 hives per group (6 bees per hive). I would like to know if there is a group that changes direction more often, or if there are changes between trials.

 

I reference to bee and the hive as a random effect (bee nests in hive)

Which statistical test is recommended to use?

 

Thanks, Erez

 

 

 

 

1 REPLY 1
Phil_Kay
Staff


Re: Which statistical test is recommended?

Hi,

Because your response is a count, it should properly be modelled with a Poisson distribution. You also have random effects, so that would be a Generalized Linear Mixed Model, which is now possible in JMP Pro 17.

I suspect that Bee should be nested within Hive. There is a very similar example with nested random effects in the help documentation for JMP (Pro) 17.

Phil_Kay_0-1670251141964.png

 

I hope this helps.

Phil