+ Reply to Thread
Results 1 to 2 of 2

Thread: Comparing proportions

  1. #1
    Points: 4, Level: 1
    Level completed: 7%, Points required for next Level: 46

    Posts
    1
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Comparing proportions




    I am planning ahead before starting an experiment and want to be sure that I can analyze the results statistically. I have two related questions, below:

    Question #1:

    I have two genotypes of fruit flies to compare. I would like to compare egg-to-adult viability for these two genotypes.

    For each genotype, I set up some number of independent replicate vials, and in each vial I place a specified number of eggs and then count the number of individuals that either survive or die during development from egg to adult. The vials are the sampled units and, for each vial, I end up with a proportion of eggs that survived from egg to adulthood.

    What statistical test would I use to compare the egg-to-adult viability between these two genotypes?



    Related Question #2:

    What if I have two independently generated copies of each genotype listed in Q#1, above? In particular, these genotypes are generated with a complicated transgenic technique, and I have gone through the extra work to generate two copies of genotype #1 and two copies of genotype #2, so I can make sure that what I measure is repeatable across independent replicate copies of each genotype.

    Now I have FOUR genetic lines to compare, with two of them being genotype #1 whereas two of them are genotype #2.

    Again, I would like to compare egg-to-adult viability between these two genotypes. For each genetic line (genotype #1 A and B, plus genotype #2 A and B) I set up some number of independent replicate vials, and in each vial I place a specified number of eggs and then count the number of individuals that either survive or die during development from egg to adult. The vials are the sampled units and, for each vial, I end up with a proportion of eggs that survived from egg to adulthood.

    Now what statistical test would I use to compare the egg-to-adult viability between the two genotypes?

    Thanks in advance for any advice.

  2. #2
    Points: 3,882, Level: 39
    Level completed: 55%, Points required for next Level: 68

    Posts
    47
    Thanks
    2
    Thanked 13 Times in 13 Posts

    Re: Comparing proportions


    Quote Originally Posted by mpal View Post
    I am planning ahead before starting an experiment and want to be sure that I can analyze the results statistically. I have two related questions, below:

    Question #1:

    I have two genotypes of fruit flies to compare. I would like to compare egg-to-adult viability for these two genotypes.

    For each genotype, I set up some number of independent replicate vials, and in each vial I place a specified number of eggs and then count the number of individuals that either survive or die during development from egg to adult. The vials are the sampled units and, for each vial, I end up with a proportion of eggs that survived from egg to adulthood.

    What statistical test would I use to compare the egg-to-adult viability between these two genotypes?
    I thought I would add my 2-cents regarding question #1. I believe something similar can be done for question #2 but wanted to start with the simpler scenario.

    I would probably approach this using a logistic regression model. The dichotomous endpoint would be egg death yes/no and genotype would be a fixed effect. This would provide an estimate of the genotype odds ratio (OR) for death...that is, the odds of death occurring in genotype #1 relative to genotype #2 (or vice-versa depending on which one is set to be the reference level). I would include vial as a blocking effect and treat it as a random variable. So my suggestion would be to treat this as dichotomous data analyzed with a logistic regression model using generalized linear mixed models (GLMM) methods.

    For question #2, I am not sure if you have 4 groups or just 2 groups. I am also uncertain if the analysis should be done with all 4 or just on the new replicate. If the aim of question 2 is to see if results are consistent with those in #1, then I think you would just perform the same analysis but only using the replicate data. Like I said, I am uncertain about this and would like to hear your thoughts as well as others.

+ 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