"Note also that the original data set (if there
were one) should
probably be analyzed as a time series with a 1-week
cyclic component.

-Robert Dawson"

Yes, I ignored this possibility in trying to get the
right null.  But if the data is not handled properly
doesn't matter what the null is, the results are
incorrect.

Michael 

****************************************************
Michael Granaas                 [EMAIL PROTECTED]
Assoc. Prof.                    Phone: 605 677 5295
Dept. of Psychology             FAX:  605 677 3195
University of South Dakota
414 E. Clark St.
Vermillion, SD 57069
*****************************************************

----- Original Message -----
From: "Robert J. MacG. Dawson" <[EMAIL PROTECTED]>
Date: Tuesday, February 24, 2004 12:54 pm
Subject: Re: [edstat] question on hypothesis testing

> It was asked:
> 
> > How would you test the following hypothesis:
> > A fast-food restaurant claims that 75% of their
revenue is from the
> > "drive-thru". Suppose you have 50 days of
receipts from the
> > restaurant. Each days' receipt shows the total
revenue and the
> > "drive-thru" revenue for that day.
> 
> and various people responded (eg):
> 
>         compute the proportion from the drive
through for each 
> day, and
> > then test if that were 75%.  Then, instead of
two numbers for 
> each day,
> > you have only one, 
> 
>       Yes...
> 
> > and that's the one you're interested in.
> 
>       No, it isn't!   The mean of the daily proportions
is not in 
> general the
> same thing as the overall mean. This is closely
related to Simpson's
> paradox.
> 
>       Drive   Table   Proprortion
>       through         DT
> 
> Tue   30      10      .75
> Wed   30      10      .75
> Thu   30      10      .75
> Fri   30      90      .25
> Sat   30      90      .25
> Sun   30      90      .25
> ___________________________
>                    #   0.5
>                    =======  
> Total 180     300 | 0.375
> 
>       Note also that the original data set (if there
were one) should
> probably be analyzed as a time series with a
1-week cyclic component.
> 
>       -Robert Dawson
> .
> .
>
=================================================================
> Instructions for joining and leaving this list,
remarks about the
> problem of INAPPROPRIATE MESSAGES, and archives
are available at:
> .                  http://jse.stat.ncsu.edu/     
              .
>
=================================================================
> 

.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to