I am looking to analyse a three-way choice experiment but am struggling with the non-independence of the data. The data is generated by animals choosing to spend time in three environments within a set trial period. The total trial period is divided between the three environments. Any suggestions would be great!
1. INTRODUCTION TO
BIOSTATISTICS
SECOND EDITION
Robert R. Sakal and F. James Rohl
2. INTRODUCTORY
BIOSTATISTICS
CHAP T. LE
Distinguished Professor of Biostatistics
and Director of Biostatistics
Comprehensive Cancer Center
University of Minnesota
3. INTRODUCTION TO
PROBABILITY AND STATISTICS
FOR ENGINEERS AND SCIENTISTS
■ Third Edition ■
Sheldon M. Ross
Department of Industrial Engineering and Operations Research
University of California, Berkeley
What's the goal? Total time in each environment? Number of transitions?
I chi-square test of proportions maybe; but some more details would be helpful. You could, for example, show that the animals spent significantly more time in environment B than in environment A or environment C.
Histograms, chi squared, multiple regression with categories in order to compare the three groups in terms of means etc
The goal is to find the preferred environment, ie how much time is spent in each environment. Following this I would also like to test whether certain attributes of the individuals influence which environment they spend time in. I would usually do this using a glmm but don't want to violate assumptions of independence. Thanks for your help!
Gemma,
If I understand correctly, it boils down a degrees of freedom question. The total of the three time-spent variables must be equal to the total time. Thus you have 3 quantities and 1 constraint (that they sum to the total), meaning you really have two unique bits of information. I think if you switch to percent time, so that the same total applies to all rats, a multivariate model (such as MANOVA) with two outcomes would be appropriate, contrasting to a reference environment.
Pat
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics