+ Reply to Thread
Results 1 to 2 of 2

Thread: How to find "active" time periods in my database

  1. #1
    Points: 4, Level: 1
    Level completed: 7%, Points required for next Level: 46

    Posts
    1
    Thanks
    0
    Thanked 0 Times in 0 Posts

    How to find "active" time periods in my database




    For my thesis I'm using a database of 9 years (unit of analysis = 1 day). The variable I want to examine is "number of actions per day" (a ratio variable with min = 0 and max =6). I'm working in spss statistics.
    I want to check for time periods with an unusually high number of actions (periods of 2-3 weeks).
    I really don't know how to begin doing this. I know the basics of working with spss but this is something I've never learned. I'm pretty sure there is an easy answer to this problem, but I'm completely stuck.

    Thanks in advance!

  2. #2
    TS Contributor
    Points: 14,811, Level: 78
    Level completed: 91%, Points required for next Level: 39
    Miner's Avatar
    Location
    Greater Milwaukee area
    Posts
    1,171
    Thanks
    34
    Thanked 405 Times in 363 Posts

    Re: How to find "active" time periods in my database


    A comment/question first. "number of actions per day" is count data, not ratio. Ratio data can be expressed in decimal form while count data is integer. Why is there a max of 6? Is this a capacity issue?

    They following suggestion is not an ideal solution because of your limit of 6, but should work for your stated purpose of identifying an unusually high number of actions. I am not an SPSS user, but a quick check shows that SPSS can create control charts. The correct type of control chart for count data is a c-chart (assumes constant opportunity), or a u-chart (non-constant opportunity). If the number of counts were large enough to approximate the Normal distribution, an individuals chart would be another option.

    Plot the data in time sequence and perform a test for runs above the mean. There are a variety of rules with different thresholds, but typically a run of 9 points above the mean warrants investigation. This is close to your two weeks.

    This is a very common tool in industrial statistics, but I rarely see it used outside of that specialty, so it may not be accepted in your thesis.
    Attached Images  

+ Reply to Thread

           




Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats