This reply takes a few steps back from Chris' reply . . .
Answers:
A1. Yes; just about any size sample can be used for a statistical analysis.
But to be accurate in your analysis and reporting, you should also disclose
the base population size along with the survey sample size. Question may
become, of the 3,500 requests received, how many are from unique
individuals? (If you receive a ton of requests from 30 people, then a
sample of 25 would be pretty good. But if you receive 3,500 requests from
over 3,000 people, then a sample of 25 would be relatively poor. Catch my
drift?) Results might vary month to month but I suspect you'll see an
smoothing effect over the course of a few months.
A2. You may want to perform an analysis of variance (ANOVA) or another form
of factorial analysis.
A3. When you want to include timescales, you'll may want to perform some
form of regression analysis.
Good news is, this is easier than it may sound! Feel free to contact me
directly if you need additional information or would like to discuss how to
get started.
-- Bing
Bradford Bingel ("Bing")
ITM3 California
http://www.itm3.com/
[EMAIL PROTECTED] (email)
925-260-6394 (mobile)
-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED]
Sent: Monday, August 14, 2006 1:45 PM
To: [email protected]
Subject: Statistical help needed
Hello, suppose you have 4 questions in a survey based upon a request closed
within the last week.
A ?
B ?
C ?
D ?
Each question can be answered from 1-6 (1 is very unsatisfied).
Each survey is connected to a Customergroup (20 different are possibly)
Each survey is belongs to an 'Productarea' (15 different are possibly)
We send 25 surveys each week, 80% are answered.
We close about 3500 requests each week.
Question 1: Is 25 each week enough for statistical calculations after a
couple of months when we will have some hundred answers.
Question 2: We would like to find our bad area, is it question C for a
special Customer, or is it related to only an Productarea/maybe in
combination with a Customergroup, or is it question A for all Areas?
Excel for reporting (pivotable?) or is there a better way?
Of course it's important to also have time included in the report, are we
better the last 3 months than the 3 before?
Of course we take care of the answers each week, and handle unsatisfied
answers, works, but we miss a way to clearly report the surveyresults with
focus on 'This is our poor area'.
Any comments will be appreciated. The survey itself works perfectly.
L ars Pettersson
Vattenfall Business Services Nordic AB
Action Request System Engineer (Tabasco)
S-461 88 Trollhättan
Sweden
Phone +46 520 888 35 Mobil +46 70 608 99 95
e-mail: [EMAIL PROTECTED]
____________________________________________________________________________
___
UNSUBSCRIBE or access ARSlist Archives at http://www.wwrug.org
_______________________________________________________________________________
UNSUBSCRIBE or access ARSlist Archives at http://www.wwrug.org
