Question

Topic: Research/Metrics

Monitoring Changes In Mean Score Over Time

Posted by Anonymous on 10/5/2008 at 4:17 AM ET 125 Points
Dear all,
If the mean score (resulted from five-point scale question) between two periods as follows:

Dimensions Jan 08 Feb 08
One 3.9 3.5
Two 4.2 4.1
Three 3.2 3.5
Four 3.6 3.9
Five 2.2 1.7

Kindly, how can we deal with the above results?
Can we, for example, consider our performance in dimension one as a bad performance (area for improvement), and dimension four as satisfied performance.
Many thanks in advance
Omar
To continue reading this question and the solution, sign up ... it's free!
SUBSCRIBE
Or, SIGN IN

RESPONSES

  • Posted by Jay Hamilton-Roth on 10/5/2008 at 10:05 AM Accepted
    You need more than 2 data points to make an judgment, since you're not seeing the big picture. What's the standard deviation of the scores? Also, ensure you're comparing statistically significant sample sizes.
  • Posted by Frank Hurtte on 10/5/2008 at 11:41 AM Accepted
    Without a list of the questions and an understanding of the questions it would be difficult to put a finger on the problem. However, something has changed. It could be one of the following:
    1) Service Level - Product Level
    2) Competition has turned up the heat
    or
    3) Environment of the responders has changed - ie economy, consumer confidence, etc.
  • Posted by dmmarathe on 10/5/2008 at 11:18 PM Accepted
    Dear, For 2 factors use ANOVA Two Way if data points are at least 3 for each. Else use ANOVA One Way for any one factor at a time.
  • Posted on 10/6/2008 at 8:21 PM Accepted
    As mentioned above, it's not possible to give you an answer unless you provide sample size and standard deviation data.

    Rather than looking at these as being defined as "bad" or "good", it is more important to regard these measurements as benchmarks for the future. I think you will want to observe additional data points before drawing any conclusions.

    Finally, don't forget to factor in external factors that may be influencing these scores.
BACK TO TOP

Post a Comment

Related Discussions

Community Info

Top 25 Experts
(Research/Metrics)