I’ve been dealing with this problem for quite a time. I’m running an experiment with plants growing them with Compost A and Compost B. I’ve been measuring the pH each day at the same time and keeping most variables constant. I have one month worth of data( i.e 30 measurements for 6 Compost A subjects and 30 measurements for Compost B subjects). To each subject I have substracted a reference measurement which is the pH before a plant is introduced, to allow me to calculate the “change” as times passes.

I want to show that the change in ph across the days vary with the compost used.

I tried to use a graph showing the coefficient of variation but had no luck since the changes in pH are positive and negative in the first days. I tried to use an absolute value CV but it doesn’t feel right.

I wanted a more robust method to demonstrate that the difference it’s more obviously as the time goes by. I was thinking of using ANOVA but I don’t know how well suited it is for this type of data. I thought of maybe doing longitudinal analysis but I’m not sure if 30 days is enough data.

Would love to hear some suggestions.

| cite | improve this question | | | | |
  • $\begingroup$ from your writeup you have 30 vales for each of 12 subjects ... 6 for each type of compost A & B .. identify the series as A1,A2,A3,A4,A5,A6,B1,B2,B3,B4,B5,B6 and post a 30x12 csv file and I and others might try and help you. $\endgroup$ – IrishStat Feb 1 '19 at 21:51
  • 1
    $\begingroup$ Can't you model your data via a linear mixed effects model? Use change as the outcome variable and allow the model to include compost type, time and their interaction as fixed effects? The model can also include a random intercept for subject as well as a random slope for time. You may need to model the effect of time as nonlinear rather than linear, depending on what your data look like. $\endgroup$ – Isabella Ghement Feb 3 '19 at 3:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.