Best Statistics Test for Comparing Categorical Data Sets

Discussion in 'Plants: Science and Cultivation' started by gopat, Nov 27, 2010.

  1. gopat

    gopat Member

    Likes Received:
    Grand Rapids, Michigan USA
    Local Time:
    6:35 AM
    I conducted an indoor rooting test involving 9 treatments (including 1 control). These tests were repeated by using 2 clones (siblings) and replicated across 2 flats with the plant trays (3" x 3") in random locations. Each flat contained 18 test trays with 2 sibling clones each having 9 treatments. Each tray contained 3 cuttings.

    I scored the rooting results for each tray into 3 categories:
    1. The total number of all roots that are longer than 1 mm (abv. "Roots>1mm").
    2. The percent of cuttings with root lengths longer than 20 mm (abv. "%Roots>20mm).
    3. The percent of cuttings that have any roots > 1 mm (abv. "%Rooted).

    Both flats contain the same tests so in theory, both should have similar results for each clone/treatment. I need to apply a statistical test that will compare the Flat 1 and Flat 2 values for the SAME clones and provide a P value that the 2 results might be the same. I am not even sure what P value I should be testing against (.95 ?).

    Below is the data for the 1 control clone:

    TreatmentName -  Flat1  -  Flat2
    %Roots>20mm     0.66        0.66
    %Rooted          1          1
    Roots>1mm       31         19
    Note that the only difference is that the Flat2 samples had 12 less roots than the Flat1 samples. What statistical test would be appropriate to compare these 2 sets of results?
    Thanks in advance!

Share This Page