turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Stat Procs
- /
- ANOVA with repeated measures and correspondence an...

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

06-20-2016 02:48 PM - last edited on 06-20-2016 03:19 PM by Reeza

Dear Everyone,

I need you help.

This is a study where we have applided 4 treatmente, and we register data every 10 min during 10 hours. Each data is about animals activities (I mean behaviour activitis, there where 4 types of activities 1, 2, 3 or 4). Dependente variable is a categorical variable.

For Example.

Animal | Treatment | Time | Activity |

1 | 0 | 1 | 1 |

1 | 0 | 2 | 1 |

1 | 0 | 3 | 1 |

1 | 0 | 4 | 1 |

1 | 0 | 5 | 1 |

1 | 0 | 6 | 1 |

1 | 0 | 7 | 4 |

1 | 0 | 8 | 4 |

1 | 0 | 9 | 1 |

1 | 0 | 10 | 1 |

1 | 0 | 11 | 1 |

1 | 0 | 12 | 4 |

1 | 0 | 13 | 4 |

1 | 0 | 14 | 3 |

2 | 0 | 1 | 1 |

2 | 0 | 2 | 1 |

2 | 0 | 3 | 1 |

2 | 0 | 4 | 1 |

2 | 0 | 5 | 1 |

2 | 0 | 6 | 1 |

2 | 0 | 7 | 4 |

2 | 0 | 8 | 4 |

2 | 0 | 9 | 1 |

2 | 0 | 10 | 4 |

2 | 0 | 11 | 4 |

2 | 0 | 12 | 4 |

2 | 0 | 13 | 1 |

3 | 0 | 1 | 1 |

3 | 0 | 2 | 1 |

3 | 0 | 3 | 1 |

3 | 0 | 4 | 1 |

3 | 0 | 5 | 1 |

3 | 0 | 6 | 1 |

3 | 0 | 7 | 4 |

3 | 0 | 8 | 4 |

3 | 0 | 9 | 4 |

3 | 0 | 10 | 3 |

3 | 0 | 11 | 1 |

3 | 0 | 12 | 4 |

3 | 0 | 13 | 1 |

3 | 0 | 14 | 4 |

3 | 0 | 15 | 4 |

3 | 0 | 16 | 4 |

3 | 0 | 17 | 1 |

3 | 0 | 6 | 4 |

3 | 0 | 19 | 4 |

3 | 0 | 20 | 1 |

3 | 0 | 21 | 4 |

3 | 0 | 22 | 4 |

3 | 0 | 23 | 4 |

3 | 0 | 24 | 4 |

3 | 0 | 25 | 4 |

3 | 0 | 26 | 4 |

3 | 0 | 27 | 4 |

4 | 0 | 1 | 1 |

4 | 0 | 2 | 1 |

4 | 0 | 3 | 1 |

4 | 0 | 4 | 1 |

4 | 0 | 5 | 1 |

4 | 0 | 6 | 1 |

4 | 0 | 7 | 3 |

4 | 0 | 8 | 1 |

4 | 0 | 9 | 1 |

4 | 0 | 10 | 4 |

4 | 0 | 11 | 3 |

4 | 0 | 12 | 4 |

5 | 0 | 1 | 1 |

5 | 0 | 2 | 1 |

5 | 0 | 3 | 1 |

5 | 0 | 4 | 1 |

5 | 0 | 5 | 1 |

5 | 0 | 6 | 1 |

5 | 0 | 7 | 1 |

5 | 0 | 8 | 1 |

5 | 0 | 9 | 1 |

5 | 0 | 10 | 1 |

5 | 0 | 11 | 4 |

5 | 0 | 12 | 1 |

5 | 0 | 13 | 1 |

I need to do an ANOVA with repeated measured. And also a Correspondence analisys with the treatments and the animals activities.

Can you help me please?

Accepted Solutions

Solution

07-06-2016
11:06 PM

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Ksharp

06-28-2016 09:20 AM

Repeated categorical measures are difficult to model, but could be done. If there is no reason to expect any odd distributions, a general logit multinomial model could be fit. If the OP has access to SAS/STAT14.1, I would strongly recommend looking at Example 43.6 GEE for Nominal Multinomial Data in the PROC GEE documentation. This example should get you started. A true correspondence analysis does not consider repeated measures, and so I would hesitate to go that route.

You might start with the following untested code:

```
proc gee data=yourdata;
class animal treatment time activity;
model activity=treatment/dist=multinomial link=glogit;
repeated subject=animal / within=time;
lsmeans treatment/diff means oddsratio cl;
run;
```

Steve Denham

All Replies

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Anita2016

06-20-2016 03:22 PM

1. ANOVA w repeated measures

http://www.ats.ucla.edu/stat/sas/library/repeated_ut.htm

2. Corresponding Analysis - PROC Corresp

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Reeza

06-20-2016 03:29 PM

Dear Reeza, the data is about categorical variables like differents activities, EAT, RUMIA, DRINK, OTHER. The info that I have is de frec of this activities long the time (10 hours).

I do not have a script that use a categorical dependent variable.

Did I explain me?

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Anita2016

06-20-2016 03:39 PM

Well then you can't do an ANOVA since that's for continuous variables. What's your question/hypothesis your trying to test?

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Reeza

06-20-2016 03:49 PM

It is an study with animals, we want to know it the treatment (Control, Treatment of 12, Treatment of 24, Treatment of 36), make changes in animal behaviour. We want to know if animales activities (eat, drink wather, rumia or others) change after the aplication od treatments.

The activities are evaluate counting the observations of the activities.

During a time period , it was measured every 10 minutes what activity takes the animal at that point. Therefore the study is a longitudinal repeated measures over time and the response variable is not continuous, is qualitative because is a activitie count.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Anita2016

06-20-2016 03:55 PM

Shouldn't treatment vary in your data then?

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Anita2016

06-21-2016 01:51 AM

Check PROC GLM .There is an example about it.

Firstly Change your data structure like :

Treatment Activity Time1 Time2 ... Time14

1 1 6 (count) 12 21

1 2 6 12 21

1 3 6 12 21

1 4 6 12 21

2 1 6 12 21

2 2 6 12 21

2 3 6 12 21

2 4 6 12 21

```
proc glm;
class Treatment Activity;
model Time1-Time14 =
Treatment Activity Treatment*Activity / nouni;
repeated Time 14 / summary printe;
run;
```

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Anita2016

06-21-2016 01:58 AM

I think it agian , maybe you need MIXED model, since your have longititude data.But I have no enough knowledge about it.

@SteveDenham maybe give you some other code .

Solution

07-06-2016
11:06 PM

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Ksharp

06-28-2016 09:20 AM

Repeated categorical measures are difficult to model, but could be done. If there is no reason to expect any odd distributions, a general logit multinomial model could be fit. If the OP has access to SAS/STAT14.1, I would strongly recommend looking at Example 43.6 GEE for Nominal Multinomial Data in the PROC GEE documentation. This example should get you started. A true correspondence analysis does not consider repeated measures, and so I would hesitate to go that route.

You might start with the following untested code:

```
proc gee data=yourdata;
class animal treatment time activity;
model activity=treatment/dist=multinomial link=glogit;
repeated subject=animal / within=time;
lsmeans treatment/diff means oddsratio cl;
run;
```

Steve Denham