+ Reply to Thread
Results 1 to 2 of 2

Thread: ANOVA on paired data

  1. #1
    Points: 8,343, Level: 61
    Level completed: 65%, Points required for next Level: 107

    Thanked 7 Times in 7 Posts

    ANOVA on paired data

    Hello all,

    I am trying to design an experiment, and in order to dig a sample size I need to know how the data will be analyzed.

    This is an experiment on rats. Each rat will get some treatment, and for each rat 2 measurements will be taken, 1 before the treatment and 1 after (paired design).

    There will be 4 different treatments, 3 new ones differing in some way, and 1 control.

    My aim is to check if at least one of the new treatments is better than the control.

    The lower the difference between after and before, the better !

    Another problem: The measure being taken is angles (continuous), however, the measurement error can be up to 10 degrees.

    I was wondering, will it be correct, if for each rat I calculate D=after-before (or the other way round), and I run a simply one way ANOVA on the D's ? Is there a better way?

    If the measurement error is 10 degrees, will my analysis be OK if I only look at differences bigger than 10 degrees?

    Thank you !

  2. #2
    Points: 3,730, Level: 38
    Level completed: 54%, Points required for next Level: 70

    Thanked 30 Times in 29 Posts

    Re: ANOVA on paired data

    A better method would be to perform ANCOVA, adjusting for the baseline (first) measurement. Also, the 10 degree error should be taken into consideration when you are calculating your sample size so that you use the appropriate number of rats.

+ 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