3
$\begingroup$

I have run 15 experiments to compare the effect of different hormone combinations on the maturation on Xenopus oocytes (immature eggs). I am hoping to find the best performing variable.

I have 4 hormone conditions:

(1) Progesterone

(2) IGF-1

(3) Both

(4) No hormone

I have 7 treatment groups:

(1) 1 µM Progesterone

(2) 5 µM Progesterone

(3) 100 nM IGF-1

(4) 200 nM IGF-1

(5) 100 nM IGF-1 + 1 µM Progesterone (Both)

(6) 200 nM IGF-1 + 1 µM Progesterone (Both)

(7) No hormone

There are 15 tests for each treatment group and the results are % maturation (the amount of eggs which matured by counting a white dot).

For treatment groups 3, 4 and 7 I observed 0% maturation in all oocytes. Therefore running classical ANOVA is difficult and the data assumes unequal variance. I am unable to run unequal variance tests in SPSS when using all the groups as they cannot compute 0.

The data is not normally distributed either. Would it be sensible just to run non-parametric Willcoxon tests on appropriate treatment groups (e.g. progesterone against progesterone) to compare one another. Each test was conducted at the same time for each treatment group?

I was hoping to compare the best performing hormone and treatment combination i.e. comparing Progesterone against IGF-1, combination and no hormone and to compare the different concentrations. Any advice would be appreciated as I am a beginner when it comes to statistics and SPSS. Thank you !!

Improve this question
$\endgroup$
5
  • 2
    $\begingroup$ What kind of data is your response variable? Is it a count of matured out of a total number of cells? Is it an unscaled count? Is it a continuous percentage? Something else? $\endgroup$
    – gung - Reinstate Monica
    Commented Aug 7, 2015 at 15:11
  • 1
    $\begingroup$ % sounds as a continuous response, even a ration. However, if your ratios are like 3 out of 5, you might as well look at it as a categorical or two ordinal outcome. So, question, what is the distribution of your outcome? (In SAS, that is the result of a proc freq with table denominator * numerator;) $\endgroup$
    – Dirk Horsten
    Commented Aug 7, 2015 at 17:58
  • $\begingroup$ % is just the amount of eggs which matured over eggs which didn't around 150 for each treatment $\endgroup$
    – Lauren Neill
    Commented Aug 9, 2015 at 17:47
  • $\begingroup$ When you say that you have 15 tests for each treatment group, does that mean that you have 15 groups of eggs, or 15 eggs? So do you have 15 "yes/no" data, or do you have 15 "% fertilized data"? Can you give us a snapshot of the what your data look like? $\endgroup$
    – user88719
    Commented Sep 24, 2015 at 20:04
  • 2
    $\begingroup$ I’m voting to close this question because the OP has not responded to questions that are needed for an answer. $\endgroup$
    – Peter Flom
    Commented Dec 7, 2023 at 22:51

0

Reset to default