Choose the correct Statistica test

[Ana2017]

Dear:

I followed your suggestion and I found significance differences between treatments in each activity. Using both ANOVA and MANOVA.

Eat: 0 . 0 0 0 1

Drink: 0. 0 3 9 7

Sleep: <0 . 0 0 0 1

Now, how can I find the differences between treatments? Because for example, for eat maybe there is a difference between treatment 4 and 2, but not between treatment 4 and 3.

[Ana2017]

Dear:

I followed your suggestion and I found significance differences between treatments in each activity. Using both ANOVA and MANOVA.

P-values:

Eat: 0 . 0 0 0 1

Drink: 0. 0 3 9 7

Sleep: <0 . 0 0 0 1

Now, how can I find the differences between treatments? Because for example, for the activity EAT maybe there is a difference between treatment 4 and 2, but not between treatment 4 and 3.

[StatDave]

As sld mentioned, the response is multinomial - it is an unordered, categorical response. Further, you have repeated measures per animal over time. An appropriate model can be fit using PROC GEE using the DIST=MULT and LINK=GLOGIT options in the MODEL statement and including a REPEATED statement. It might look like this if you want Treatment as the only predictor. Assuming you want comparisons among the treatments, the LSMEANS statement provides that. 

proc gee;
class treatment animal;
model behavior = treatment / dist=mult link=glogit;
repeated subject=animal;
lsmeans treatment / diff;
run;

[Ana2017]

Hi:

Do you meant this way? Changing in the place of behaviour, each activity one by one (eat, drink and sleep)?

proc gee;
class treatment animal;
model eat= treatment / dist=mult link=glogit;
repeated subject=animal;
lsmeans treatment / diff;
run;

proc gee;
class treatment animal;
model drink = treatment / dist=mult link=glogit;
repeated subject=animal;
lsmeans treatment / diff;
run;

proc gee;
class treatment animal;
model sleep = treatment / dist=mult link=glogit;
repeated subject=animal;
lsmeans treatment / diff;
run;

  I got this error.... please look...

ERROR: Only the Independent working correlation type is available for multinomial models.

NOTE: The SAS System stopped processing this step because of errors.

NOTE: PROCEDIMIENTO GEE used (Total process time):

Regards. 

[StatDave]

No, I mean you should use the Behaviour variable which has values Eat or Sleep or Drink and to run PROC GEE just once with Behaviour as the response variable as I showed.  The default should be the independence correlation structure, so I don't know why you got that error unless you specified the TYPE= (or CORR=) option in REPEATED with a value other than IND. To be sure, specify TYPE=IND in REPEATED:

repeated subject=animal / type=ind;

[Ana2017]

StatDave:

I am afraid that I cannot use the variable Behaviour directly, because my data is as follows. A frequency of activities. 

DATA Behaviour;
INPUT Treatment Animal Eat Sleep Drink;
DATALINES;
1 1 52 4 5
1 2 19 2 2
1 3 20 4 2
1 4 21 23 2
1 5 23 78 1
2 6 23 11 1
2 7 22 78 2
2 8 24 1 1
2 9 23 1 1
2 80 22 1 0
3 11 90 2 0
3 12 25 1 1
3 13 19 2 20
3 14 20 2 10
3 15 21 1 15
4 16 24 87 20
4 17 23 0 10
4 18 24 0 18
4 19 23 14 19
4 20 24 12 263 
RUN;

So if I include just the word behaviour the system says:

ERROR: Variable BEHAVIOUR not found.

 😞

[StatDave]

I was referring to this form of the data that you mentioned earlier and which I assume you have:

Moment                 Animal             Treatment          Behaviour

0                            1                      1                        Eat

10'                         1                      1                        Sleep

20'                         1                      1                        Sleep

.                              .                       .                        .

.                              .                       .                        .

.                              .                       .                        .

(0 to 180´)            (1 to 20)       (1 to 4)                   (Eat, Sleep or Drink)

[Ana2017]

ahhhh, OK! thanks for that. I just changed the table, reporting the global frequencies. 

[Ana2017]

So the Proc gee is not suitable to be used with the new arrangement of the data that I have now?

proc gee;
class treatment animal;
model sleep = treatment / dist=mult link=glogit;
repeated subject=animal;
lsmeans treatment / diff;
run;

[StatDave]

You could use it for a binary response like that (though you would then specify DIST=BINOMIAL and LINK=LOGIT), but then you would presumably have to fit three separate models. That could complicate testing for Treatment differences. It's usually better to fit a single model when that is possible as it is here.