+ Reply to Thread
Results 1 to 2 of 2

Thread: What test to compare bacterial growth in three areas with repeating variables?

  1. #1
    Points: 10, Level: 1
    Level completed: 19%, Points required for next Level: 40

    Posts
    1
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Smile What test to compare bacterial growth in three areas with repeating variables?




    I have a dataset of number of bacterial colonies grown from 11 water samples taken from 3 different watersheds. I want to know if the land area, population, number of waste water treatment plant outfalls, and number of septic tanks in the watershed have any effect on the bacterial growth. The samples are skewed, with 8 from one watershed A, 1 from B, and 2 from C.

    Essentially I am trying to compare 3 treatments by categorical numbers, since the land area is one of three numbers, right? I am confused and don't know which test to use, please help! I've attached a file of the data table if this post is confusing. I am grateful for any advice.
    Attached Files
    Last edited by valentine35; 11-15-2013 at 10:15 PM. Reason: attachment of data table

  2. #2
    Points: 501, Level: 10
    Level completed: 2%, Points required for next Level: 49

    Posts
    12
    Thanks
    1
    Thanked 3 Times in 3 Posts

    Re: What test to compare bacterial growth in three areas with repeating variables?


    Summarizing your question: You have 4 variables and you want to know whether they are related to bacterial growth.

    Looking at your data, for all variables have 3 distinct values, corresponding 100% with with land location. So in fact you only have 3 groups, regardless of which variable you choose to group on. So you basically just want to compare the growth between A, B and C. Which variable is responsible for any difference in growth on each site can not be determined, you just know that there is a difference between the sites.

    But the pain in the arse is your sample size per location. For A you are fine with 8 (HIGHLY VARIABLE!!) measurements, but for B and C (n=1 & 2) you can simply not test any association between site (=all variables in this case) and growth.

    I would make a box plot of the measurements on site A and scatter plots of the measurements on site B and C. Site C seems to be relatively low, but still within the range of A, while B is much higer (single measurements). This visualization will emphasize this without assigning any statistical value (significance).

    Best, Martin

  3. The Following User Says Thank You to MARijlaarsdam For This Useful Post:

    GretaGarbo (11-20-2013)

+ Reply to Thread

           




Tags for this 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