+ Reply to Thread
Results 1 to 3 of 3

Thread: which model to use? Help for newbie

  1. #1
    Points: 16, Level: 1
    Level completed: 31%, Points required for next Level: 34

    Posts
    2
    Thanks
    0
    Thanked 0 Times in 0 Posts

    which model to use? Help for newbie




    Hello! I am trying to determine which model would be best to use to solve the following problem. I am looking at the occurrence of patient to patient violence in a psychiatric hospital. I have the time of day that these events occurred for the past 3 months (about 30 events total) and I need to do a data analysis to determine if there is trend for what time these incidences occurred or a period of time during which these fights are occurring more frequently, so that potentially additional staff members can be added to that time period. However, I have no idea which model would be best because I do not understand much of what I am reading. My assumption is that patient violence = dependent variable and time of day = independent variable. (?) Would a regression model be best? If so, would that be considered linear regression? Or would this be a t-test? Or something else? Please have mercy on my statistically challenged soul. I am trying to help a friend in need. Also, how would I look for statistical significance in this data?

  2. #2
    Omega Contributor
    Points: 38,289, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,992
    Thanks
    397
    Thanked 1,185 Times in 1,146 Posts

    Re: which model to use? Help for newbie

    Well Kate, I will tell you this doesn't sound like an easy model to build. There is much information out there on violence models and I am not familiar enough with it. Issues that make it hard, you have repeated observations for patients and certain patients may interact with each other differently.


    Yeah, you will likely need to run regression. You can always try to do a generic version in regression. I think your outcome as binary, fight: y/n. I would try to find an article comparable to the one you are try to analyze, but I would imagine there are so many contributing factors (meds, diagnosis, interaction, time of day, day, etc.). As mentioned, you may be able to get away with a very generic model were you just look at time or make a run chart, if you just what crude results.
    Stop cowardice, ban guns!

  3. #3
    Points: 16, Level: 1
    Level completed: 31%, Points required for next Level: 34

    Posts
    2
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: which model to use? Help for newbie


    Thank you so much! You are exactly right about just wanting to analyze the "crude" time of the incident to determine, for example, if the hospital could benefit from adding extra staff for a shift from
    7-11 (for example) due to a trend of having most of the occurrences happening during this time period.

+ 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