+ Reply to Thread
Results 1 to 3 of 3

Thread: Reporting t-test (table) + boxplot (mean, range and outliers)

  1. #1
    Points: 2,096, Level: 27
    Level completed: 64%, Points required for next Level: 54

    Posts
    12
    Thanks
    5
    Thanked 0 Times in 0 Posts

    Reporting t-test (table) + boxplot (mean, range and outliers)




    Hi
    I am looking for a reasonably appropriate way to report t-test along with a summary statistics. Initially I only wanted to go with box plots, and now I have to add t-test results as well. Is it a good idea to report the t-test in a table with a box plot for summary statistics as below image points out?

    Since simple SD does not work for me (range and outliers are important), I chose box plots instead of simply reporting mean and SD. I just want to make sure if it is not very odd to report the statistics (and results) in this way.
    Your comments will be very helpful.

  2. #2
    Omega Contributor
    Points: 38,432, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,006
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: Reporting t-test (table) + boxplot (mean, range and outliers)

    This is interpretable, though you may put the values in the box for the reviewers, in that "Non-significant" could mean anything from 0.055 to 0.99. This could already be in the not included text, but I always enjoy knowing sample sizes as well. Lastly, if you are making 4 comparable comparisons here, you may need to correct for false discovery, though we don't know the context of your project!
    Stop cowardice, ban guns!

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

    true_friend (03-13-2017)

  4. #3
    Points: 1,741, Level: 24
    Level completed: 41%, Points required for next Level: 59

    Posts
    230
    Thanks
    37
    Thanked 68 Times in 59 Posts

    Re: Reporting t-test (table) + boxplot (mean, range and outliers)


    I just wanted to add on a bit. I would recommend including p-values to about 4 decimal places. This is generally the advice I've gotten from statisticians. It avoids the binary, black-and-white mentality about p-values and allows the reader to judge for themselves the strength of contradiction between the data and the null hypothesis. See the ASA's Statement on P-values if you'd like more background.

  5. The Following User Says Thank You to ondansetron For This Useful Post:

    true_friend (03-14-2017)

+ 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