Question

Topic: Research/Metrics

2x2 Matrix With Satisfaction And Importance

Posted by Anonymous on 5/21/2008 at 6:03 AM ET 125 Points
Dear Experts,
I am trying to plot the mean score of a seven variables (close-ended questions) on a Y-axis against the correlation of coefficient between overall satisfaction with the seven variables on the X- axis. The mean score came from five-point scale. I have the following questions:
1. What will be the Y and X scale axis, is it 1-5 for Y-axis and -1 to 1 for the X-axis?
2. I want to add a horizontal and a vertical line through the middle of the plot points, but I don’t know where to put them, can I use 2.5 for the Y-axis and zero for the X-axis?
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 dmmarathe on 5/21/2008 at 6:14 AM Accepted
    Y-axis scale can be from 1 to 5 & X-axis scale can be from -1 to 1.You can draw middle line at the centre.
  • Posted by Jay Hamilton-Roth on 5/21/2008 at 9:00 AM Accepted
    Instead of 1-5 on the Y-axis, use the same wording you used for your questionnaire (ex: 1=highly satisfied...5=highly unsatisfied"), since it'll make the graph interpretation easier.
  • Posted by koen.h.pauwels on 5/21/2008 at 9:58 AM Member
    I agree with Jay; better keep the questionnaire wording for ease of graph interpretation. As to your other question though:

    1) for the X-axis, zero is the logical middle point for a line, as it seperates satisfaction promotors from detractors

    2) however, for the Y-axis, 2.5 does not make much sense. Instead, I would use the median score (which may be e.g. 3 or 1), which thus seperates the higher from the lower scores. Alternatively, you can use a company benchmark if there is one. For instance, if 4/5 is the norm/target for thes scores, I would use 4 as the line, coloring everything above as 'green', and everything below as a 'red light'
  • Posted by koen.h.pauwels on 5/25/2008 at 4:33 PM Accepted
    Hi Omar,

    As the X-axis represent the correlation coefficient of each variable with overall satisfaction, a negative value means the variable is negatively correlated with (and thus detracts from) satisfaction. Of course, if each of your variables is positively related to satisfaction, it is better to use e.g. the median correlation as the dividing line
BACK TO TOP

Post a Comment

Related Discussions

Community Info

Top 25 Experts
(Research/Metrics)