Re: Validity of averaging 'point in time performance indicators

From: Rod (RodRodRodRod_at_Hotmail.com)
Date: 11/23/04 

Date: Tue, 23 Nov 2004 10:42:45 GMT

How about fitting an exponential (A*exp(b*t)) to all the monthly data and
then reading off the best fit figure for "b" from which an annual increase
can be calculated..

"Jayne" <j_clarke@postoffice.utas.edu-dot-au.no-spam.invalid> wrote in
message news:41a27697_1@Usenet.com...
> I am trying to ascertain whether it is appropriate or statistically
> valid to average the figure of an indicator which measures total
> number of people on waiting list at a particular point in time. I
> want
> to look if the waiting list has increased or decreased over the
> years.
>
> For example, Total Number on Waiting List- figure is reported on a
> monthly basis- and is a figure ‘as at 30 of the month'. This figure
> is merely a snapshot at a particular point in time and therefore does
> not take into account the number people who are added or removed
> from this waiting list during the month.
>
> If I wanted to derive a representative figure for the purposes of
> examining percentage change 2003-04 and 2002-03 , which is more
> appropriate and why:
> a) Is it valid/ appropriate for me to sum the monthly figure and then
> find the average of the 2003-04 period (this would be the average
> number of people on a waiting list as at 30 day of the month) and
> then
> compare that with the same figure calculated for 2002-03; or
> b) Or would it be more appropriate to look at the percentage
> difference between say 30 June 2004 and 30 June 2003?
>
> Averaging a point in time number is not sitting well with me at the
> moment- but I am not sure why?
>
> Any help which could be provided would be greatly appreciated.
>
> Regards
>
> Jayne
>
> *-----------------------*
> Posted at:
> www.GroupSrv.com
> *-----------------------*