isomeric log ratio transformation for compositional data

[cmidega]

Hi,

How does one construct a procure statement for isomeric log ratio transformation for compositional data? Example below

Time spent by a parasitic wasp in each arm of an olfactometer

Treatment (olfactometer arm)	Proportion time (sec)
Treated with extract1	20.84
Treated with extract1	6.64
Treated with extract1	29.61
Treated with extract1	23.44
Treated with extract1	32.47
Treated with extract1	23.82
Treated with extract1	32.97
Treated with extract2	15.83
Treated with extract2	12.32
Treated with extract2	18.38
Treated with extract2	30.44
Treated with extract2	30.08
Treated with extract2	30.38
Treated with extract2	24.45
Blank control	15.83
Blank control	63.33
Blank control	81.04
Blank control	52.01
Blank control	46.12
Blank control	37.45
Blank control	45.81

[Rick_SAS]

What scientific/statistical question are you trying to answer? Your question usually applies to data that represent proportions, but I don't see any proportions here. This looks like standard ANOVA data with three treatments. What am I missing?

[cmidega]

Dear Rick,

Thank you. The data represent proportions of time spent by a parasitic wasp in three different olfactometer arms and by all means the data is compositional. I intend to run ANOVA but think this should be run on log ratio transformed data. Kindly help

Regards