Search results

  1. GretaGarbo

    Logistic Regression Interpretation help

    I can't see any problem with Spearman's rho.
  2. GretaGarbo

    Logistic Regression Interpretation help

    What about Spearman's correlation?
  3. GretaGarbo

    Constrained design of experiments

    Skip an other level of the three levels and you will have just two levels. That is your two factor experiment is redused to one factor. To do factorial experiments is to do compromises. You can't get everything.
  4. GretaGarbo

    Permutation test with skewed data

    I am sorry but I don't understand exactly what Obh mean. My point was that when two data sets have different "spread" (but the same mean) then the WMW does not have the promised quality of an error rate of 0.05. The error rate will be larger. And I believe that most people interpret a...
  5. GretaGarbo

    Don't understand the meaning of ''training a dataset''

    The "training" is just to estimate some parameters.
  6. GretaGarbo

    Two-way ANOVA or one-way ANOVA for time-course experiment with different drugs?

    To me it looks like one way anova with repeated measures. Search for it. Here is one example.
  7. GretaGarbo

    Analyzing data with continuous variables + discrete outcomes

    How about a two way anova with time as dependent variable and competition as one factor and person as an other factor? (otherwise: rank as ordered logistic and competition as one factor and person as an other factor. But anova is much easier and more statistically efficient.)
  8. GretaGarbo

    PCA pre-treatment: Centering or Standardizing?

    It is called standardizing. ("autoscaling" is a phrase from a strange software. The subject should define the software, not the other way around.)
  9. GretaGarbo

    Possible dumb question

    Some example data would make it easier for us to understand and to suggest something.
  10. GretaGarbo

    Possible dumb question

    Well, hmm, ehh,.... of course anybody can have whatever opinions they like. As someone said "you can have any opinions but you can not have any facts". But of course you can ask about any facts, especially here at talkstats. Someone said that they wanted this to be a very frindly place. But it...
  11. GretaGarbo

    Missing Data Imputation - How to choose the best method?

    Here is some introductory info about imputation and here is a paper. @jgeconolove What is EMB? A variant of EM, Expectation Maximization algorithm?
  12. GretaGarbo

    LPA assumuptions

    And what is that? And what are " the assumption of justly identified or over identification"?
  13. GretaGarbo

    Interpreting principal components

    Is there a question or comment that MalikMc wish to do here?
  14. GretaGarbo

    LPA assumuptions

    LPA? "Thanks for any help provided."
  15. GretaGarbo

    Best statistical method for CFU/ml?

    What is "CFU/ml"?
  16. GretaGarbo

    Caclulating Sample Size

    Look at this wikipedia page about the power of test at the example section. With sigma/sqrt(n) you can replace that with (sigma/sqrt(n))*sqrt((N-n)/(N-1)) where N is the populations size. But if N is large (say larger than 1000) and n is small (say 50) then you are essentially dividing 1...
  17. GretaGarbo

    Is Multivariate Analysis possible?

    OK! Prove it! Or show a published paper that support that claim!
  18. GretaGarbo

    Is Multivariate Analysis possible?

    I agree with most of what J58 has said. But not with this statement. I believe that it come from one publication where in that application it happened that the sample size was 10 times the number of explanatory variables. In reality they have no relation. Compare the Placket Burman design where...
  19. GretaGarbo

    Number of measurements

    Hmm, usually the sample size is equal to the number or measurements. What do you (altf2) mean by this? Also, one way a sample can be "representative" is when every individual has the same chance of being included, like in simple random sampling.
  20. GretaGarbo

    Reaction times analysis with Mixed Models

    I would guess that reaction times would be a skewed distribution. So maybe a gamma distribution fits. Or maybe a lognormal distribution. Maybe they don't have that on the standard software. Then try to take the logarithm (so that data transforms from the lognormal to the normal distribution).