Re: Comparing proportions?

Robert Lundqvist wrote:
Could anyone please help me find a suitable approach to the following
problem:

*I have got k persons (typically 3), all of whom are observed in m (about
5) sessions of constant length.

*In each session, the part of the session they spend working with one out
of 4 methods is observed and recorded.

The data looks like:

Teacher Session M1 M2 M3 M4
1 1 20 30 35 15
. . . . .
1 5 25 35 20 20
2 1 30 30 30 10
. . . . . .
. . . . . .
3 5 10 40 40 10

The M:s all sum up to 100%.

What I would like to do is to see if there are any significant differences
between teachers, between methods and possibly also an interaction between
the two.

The problem was posed by one of my colleagues who has studied "teacher
behaviour". At first it sounded like some ANOVA setup, but then there are
these dependent M:s. Some multinomial approach?

Any suggestions?

Robert
********************
Robert Lundqvist
Dept of mathematics
Lulea University of Technology
Sweden

http://www.ltu.se/web/pub/jsp/polopoly.jsp?d=4171&a=10697

Yes, you can do an anova. See the sci.stat.consult thread
"Within-subjects proportional data", Sep 20-28, 2005:
http://groups.google.com/group/sci.stat.consult/browse_frm/thread/d6b4b2c1d6180943/

.

Relevant Pages

  • Re: How best to use php5 objects between pages?
    ... $_SESSION array between pages that were navigated on the site. ... representing students that referenced teacher objects. ... found much redundant data was stored in each file about the teachers ... way -- reading large session files each time the user navigates to a ...
    (comp.lang.php)
  • Comparing proportions?
    ... *In each session, the part of the session they spend working with one out ... What I would like to do is to see if there are any significant differences ... between teachers, between methods and possibly also an interaction between ... The problem was posed by one of my colleagues who has studied "teacher ...
    (sci.stat.consult)
  • Re: how to prevent security violation when users share URL with cookieless session id?
    ... I recently even saw a web page with a cookieless session indexed on google:) ... colleagues click on the link, the colleagues end up piggybacking on ... validation we perform against our registration database? ...
    (microsoft.public.dotnet.framework.aspnet.security)
  • Re: how to prevent users from sharing their cookieless session id?
    ... this is the main disadvantage of using the url for session id. ... session variables with a new visitid, and a unique visitorid, and then ... door, pastes an inner page's URL into an email message and sends it to his ... If the session hasn't timed out, the colleagues who receive the ...
    (microsoft.public.dotnet.framework.aspnet)
  • how to prevent users from sharing their cookieless session id?
    ... We authenticate the username and password against our database, and if they match, we let the user in the door, so to speak, by assigning session variables with a new visitid, and a unique visitorid, and then redirecting the user to our internal pages. ... How can we stop the practice where a user, who has made it through the door, pastes an inner page's URL into an email message and sends it to his or her colleagues? ...
    (microsoft.public.dotnet.framework.aspnet)