Re: proportions: (a,b,c)/3 vs d?

Posted by Salbod on Nov 06, 2014; 1:20pm
URL: http://spssx-discussion.165.s1.nabble.com/proportions-a-b-c-3-vs-d-tp5727795p5727820.html

Thanks for getting back to me and for taking the time to consider the problem. You’ve given me another lens to view my problem through. Let go back to the drawing board. --Steve

 

 

From: Rich Ulrich [mailto:[hidden email]]
Sent: Wednesday, November 05, 2014 9:49 PM
To: Salbod, Mr. Stephen; SPSS list
Subject: RE: proportions: (a,b,c)/3 vs d?

 

As I read this, for the statement of the problem as, "I want to test
whether the IV was working"...  there is one correct answer for the
particular message.  Like on an IQ test, the important information is
not a matter of which of the *wrong* answers was given.  If you are
doing scale development, you may care about which categories are
confused most, when it comes to figuring out improvements, etc.

This is hardly a situation for a *test*.  You should not be happy with
something that is a mere 0.05 or even 0.01.  If your actor does not do
a good job showing the intent -- something a lot better than chance --
then the item will not work for the future purposes.

Is 76% a good fraction?  Well, you can put a CI on it to see how wide
it is.  If you have a number of these as potential items, your best
display might be to list them in increasing number of Wrongs. 
The Odds Ratio for 75% is 3.0, if you want another way to describe
the effectiveness.

--
Rich Ulrich

> Date: Wed, 5 Nov 2014 20:57:09 +0000
> From: [hidden email]
> Subject: Re: proportions: (a,b,c)/3 vs d?
> To: [hidden email]
>
> Hi Art, 2/111 = .018.
>
> I might not be asking the right question. One hundred and eleven students heard an actor give a video message and selected the emotional tone of the message. The message was a single sentence. The actor had been given the intent message. I want to test whether the IV was working. I thought a contrast within each message should do the trick? Is there another way to skin the cat?
>
> --Steve

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD