+ Reply to Thread
Results 1 to 4 of 4

Thread: Comparing Measured and Predicted body Mass - Can I use the T-Test

  1. #1
    Points: 86, Level: 1
    Level completed: 72%, Points required for next Level: 14
    Gnu's Avatar
    Location
    Namibia
    Posts
    8
    Thanks
    5
    Thanked 0 Times in 0 Posts

    Comparing Measured and Predicted body Mass - Can I use the T-Test




    I have measured the body mass (Mb) of several wildebeest in kg.
    I have also measured other parameters including Girth (G), Length (L) and shoulder Height (SH) in meters.

    I have some equations that were derived by another researcher on another wildebeest subspecies. These equations can estimate body mass from the other measurements.

    In order to test whether these equations are useful for my wildebeest subspecies I would like to compare the body masses predicted by the equations to my measurements of mass. My null hypothesis would be that there is no difference between the measured and predicted masses.

    My supervisor suggests that I use the T-Test. I am worried that this isn't the appropriate test because:

    The body masses of wildebeest in a population consisting of all wildebeest of all ages are not normally distributed. Therefore using the T-test which compares measurements to the mean of the sample seems like a poor option.

    Would the Wilcoxon-Mann-Whitney test, a distribution-free, nonparametric test be more appropriate?

    Furthermore my data are collected in one of three seasons and from both sexes. In order to test whether sex or season has an effect on my data I have derived equations* to predict body mass for the data from one sex or season and then used that derived equation to predict body masses for the other data.

    * I derived my equations by using simple least squares to plot a linear relationship on log-transformed data.

    Can I use the same methods to compare the calculated and predicted masses in theses groups?

    Thanks for taking the time to read and think about my problem. Any suggestions or advice would be very welcome.

    Have a great day

    Gnu

  2. #2
    Super Moderator
    Points: 31,766, Level: 100
    Level completed: 0%, Points required for next Level: 0
    bugman's Avatar
    Posts
    2,255
    Thanks
    290
    Thanked 324 Times in 265 Posts

    Re: Comparing Measured and Predicted body Mass - Can I use the T-Test

    A collegue did an allometric study on ant populations a few years ago and I remember we applied breakpoint regression to his data. I think this would be relivant to your study. BP regression lets you split a single linear regression into 2 or more points - the points being definded by your morphotype. However, becasue you have errors on the x and y axis (presumably) you may need / want to look into major axis regression to account for this.
    The earth is round: P<0.05

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

    Gnu (03-17-2015)

  4. #3
    Points: 86, Level: 1
    Level completed: 72%, Points required for next Level: 14
    Gnu's Avatar
    Location
    Namibia
    Posts
    8
    Thanks
    5
    Thanked 0 Times in 0 Posts

    Re: Comparing Measured and Predicted body Mass - Can I use the T-Test

    Thanks Bugman.

    I'd like to describe the relationship over the whole range of wildebeest masses. An entire lifetime. The data do not show changes at different growth stages.

    I'll have a look at Major axis regression too.

    Regards
    Gnu

  5. #4
    Super Moderator
    Points: 31,766, Level: 100
    Level completed: 0%, Points required for next Level: 0
    bugman's Avatar
    Posts
    2,255
    Thanks
    290
    Thanked 324 Times in 265 Posts

    Re: Comparing Measured and Predicted body Mass - Can I use the T-Test


    No problem, good luck!
    The earth is round: P<0.05

+ 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