iSixSigma

How to do a Gage RR on survey data

Six Sigma – iSixSigma Forums General Forums Methodology How to do a Gage RR on survey data

Viewing 5 posts - 1 through 5 (of 5 total)
  • Author
    Posts
  • #53479

    Nick
    Participant

    I’m working a project to improve the customer satisfaction with a service provided. I understand that there are a great many challenges associated with using this particular metric but the bottom line is that it is the only thing that we have that gets close to measuring the process.

    A 3rd party agency periodically surveys the customers by asking them to rate their level of agreement with a statement from 1 (strongly disagree) to 5 (strongly agree). Since the rating is subjective and there is no ‘right’ answer I am at a loss about how evaluate this measurement system.

    One idea I am considering is building multiple scenarios (i.e. a fictional situations where the service was provided at various levels of quality) and having two to three personnel answer the survey question based upon how they would feel if they were the customer in the fictional scenario. I’m not very fond of this particular idea but don’t have anything better. Any ideas?

    0
    #190288

    Strayer
    Participant

    Kappa is a better MSA tool for such a survey since you are dealing with attribute data obtained from subjective ratings. Don’t try to force fit Gage R&R.

    0
    #190289

    bbusa
    Participant

    Kappa method is for nominal data ( 2 categories) . I thought the correct method ordinal data is the ICC – but it is not going to be easy .

    Just my opinion

    0
    #190294

    Nick
    Participant

    Is there a quick way to get the ICC in Minitab? I’ve been searching the help files and haven’t found it yet… Found a few references on the internet that directed me to do some manual calculations based upon other Minitab results.

    0
    #190295

    Just another MBB
    Participant

    If your baseline is from the 3rd party resource, that may be the best place to start. The approach you’ve outlined below with your internal personnel is a good approach, but won’t take the place the actual feedback you’ve already gotten through the 3rd party resource (assuming adequate sample size and response rate). I would either use the existing survey results or create antother to serve as the baseline. I would, post process improvement, ask the same questions (either 3rd party or newly created survey) to begin to measure the impact.

    Bottom line – I’m not sure if you’re over complicating the project by firguring out a way to conduct a Gage R&R as opposed to understanding the perception, figuring out root cause, and then driving the needed change.

    0
Viewing 5 posts - 1 through 5 (of 5 total)

You must be logged in to reply to this topic.