• Robert J Miller added an answer:
    How can I compare linear relationships?

    We have measured simple relationships between the size of different species and their weight (biomass).  They are linear regressions (sometimes semilog). The type of question I would like to answer is, for example, if I have 4 species of snail, each with a separate linear equation representing the relationship between size and biomass, are those relationships significantly different, or does one relationship suffice for all my snail species?  

    I'm not sure how to do this comparison. Each species often has different sample sizes. I thought about just doing the separate relationships, then pooling all samples and comparing the resulting slope to the species-wise relationships with t tests. However I'm worried that species with more samples will bias the results. I could randomly remove samples from those species to equalize them. Another alternative might be to do an ancova with species put in as a dummy variable and look for interactions with species as a test of parallelism. Does that seem reasonable?  It seems like a good idea to me because it will also be a test of whether the intercepts are the same.

    Robert J Miller · University of California, Santa Barbara

    Excellent points Timothy, thanks for your input.

  • Bruce E Oddson added an answer:
    Does anyone have suggestions for reporting a robust ANCOVA?
    I'm following the example in Andy Field's R book where he suggests that after failing the test for homogeneity of regression slopes, one might do a robust ANCOVA ala Wilcox 2005. I'm able to run the tests no problem, and interpreting them is also not an issue, but for output of the following nature (see below), does anyone know of a standard way to report this data?

    I think a way to start at least will be to report the standard ANCOVA up to the point where the interaction is significant and then say robust procedures were followed, how to report these though are a bit beyond me.

    ancova(covGrp1, dvGrp1, covGrp2, dvGrp2)
    [1] "NOTE: Confidence intervals are adjusted to control the probability"
    [1] "of at least one Type I error."
    [1] "But p-values are not"
    X n1 n2 DIF TEST se ci.low ci.hi p.value crit.val
    [1,] 10.30 20 12 -22.166667 2.7863062 7.955575 -47.42320 3.089867 0.0213100575 3.174696
    [2,] 11.30 28 17 -19.184343 2.7536447 6.966891 -39.98396 1.615273 0.0167914292 2.985495
    [3,] 12.45 32 23 -20.350000 3.9162704 5.196270 -35.02758 -5.672423 0.0008787346 2.824637
    [4,] 14.00 27 34 -8.314171 1.4638404 5.679698 -23.71193 7.083583 0.1524122220 2.711016
    [5,] 16.10 14 17 3.431818 0.3796813 9.038682 -22.28197 29.145604 0.7085490133 2.844860

    ancboot(covGrp1, dvGrp1, covGrp2, dvGrp2,tr = .2, nboot=2000)
    [1] "Note: confidence intervals are adjusted to control FWE"
    [1] "But p-values are not adjusted to control FWE"
    [1] "Taking bootstrap samples. Please wait."
    X n1 n2 DIF TEST ci.low ci.hi p.value
    [1,] 10.30 20 12 -22.166667 -2.7863062 -47.00379 2.670459 0.0355
    [2,] 11.30 28 17 -19.184343 -2.7536447 -40.93482 2.566135 0.0185
    [3,] 12.45 32 23 -20.350000 -3.9162704 -36.57264 -4.127360 0.0015
    [4,] 14.00 27 34 -8.314171 -1.4638404 -26.04606 9.417719 0.1525
    [5,] 16.10 14 17 3.431818 0.3796813 -24.78674 31.650380 0.6980
    Bruce E Oddson · Laurentian University

    Dear John,

       If you are going to be fair (depends how you look at it) to other robust techniques, then I would say you report it as simply as a regular ANCOVA. You state which package and assumptions you used. You give the p values and associated CIs for each statistic of interest. Although the additional information provided by the procedure is potentially helpful, nobody asks for it when (often incorrectly) "standard" procedures are used. 

  • Scott Taylor Barrett added an answer:
    How do you conduct a mixed-factors ANCOVA with a time-dependent covariate in R?

    This might be a long shot, but I thought I'd give ResearchGate a chance to prove itself.

    I'm having trouble trying to conduct an ANCOVA in R when one of my variables is a time-dependent covariate. For simplicity sake, let's say I have variables Y, A, B, and X, where Y is my dependent variable, A is a between subjects factor with two levels, B is a within-subjects factor with 6 levels, and X is a continuous variable I want to add as a covariate and is measured at all levels of A and B.

    Any help on how to conduct this in R using either lm(), lme(), aov(), ezANOVA(), or something similar would be very helpful.


    Scott Taylor Barrett · University of Nebraska at Lincoln

    Mixed-factors as in a between subjects factor crossed with a within-subjects factor (http://en.wikipedia.org/wiki/Mixed-design_analysis_of_variance).

    X is a covariate that is measured within-subjects (like the DV) at each condition of B (the within-subjects factor). X is time-dependent (i.e. time-varying) in that it is not stable across repeated measurements. 

  • Dasapta Erwin Irawan added an answer:
    How can we determining significance of independent variables in logistics regression?

    I am trying to do Logistic Regression in R.

    My data set contains more than 50 variables. Some of them are factor (qualitative variable) and others are independent variable(quantitive ). I would like to get the significance of the variables from their p-value.

    So far I came to know, I can do ACNOVA test to calculate p value of the factors. It (ACNOVA) combines features of both ANOVA and regression. It augments the ANOVA model with one or more additional quantitative variables, called covariates, which are related to the response variable.

    How can I calculate P-value of quantitive variables? Or if I am wrong about ACNOVA, what other possibilities are available?

    Any suggestion or help will be appreciated.

    Dasapta Erwin Irawan · Bandung Institute of Technology

    Thank you Oliver. 

  • Javier Miguelena added an answer:
    ANCOVA or repeated measures ANOVA?

    My experiments was mainly a split-plot design with water treatment as the main factor. Measurements were taken in 6 harvests from Dec 2012 to May 2013.  In Dec, all plots under irrigation; from Jan to March, half plots under irrigation and the other half withholding water; from Apr to May, all plots under irrigation. In each harvest, plants were defoliated  after other measurements were taken. I do not know whether I should do an ANCOVA with measurements in the first harvest as covariates, or do a repeated measures ANOVA?

    Javier Miguelena · The University of Arizona

    Interesting problem. REML means restricted maximum likelihood, it refers to a way of partitioning variation when you are including random effects. You absolutely need to include random effects in your model. The way I understand it, you are asking if there is an effect of changing the irrigation treatment. I would use a model that includes month (since measurements taken the same month are not fully independent due to cliamte) and plot (since plots might have different "personalities" that show up as you repeat the measurements) as random effects.

    The fixed effects should be treatment (continuous vs interrupted irrigation) and treatment time (before vs after irrigation change), as well as an interaction between the two. The interaction term tests the hypothesis that either of the treatments changes its response more than the other after the date when irrigation was interrupted. Your prediction, I think, is that the plots in the "interrupted irrigation" treatment will change more than the control.

  • Andrew Ekstrom added an answer:
    Any suggestion about using ANCOVA with repeated measures?

    My consulting adviser said that we can't use covariance method when there are more than 2 time points. But I'm not sure about it again!

    What's your idea about that?

    Andrew Ekstrom · University of Michigan

    Hey Roshanak,

    What do you consider the covariate and what do you consider factors/variables of interest?

    As a statistics student, I have taken stats classes from several different departments, psychology, education, stats, industrial engineering, Biostats, etc. What one person calls a covariate, another person calls a factor/variable. I was taught in my stats and IE classes that a covariate was a thing you want to test, that is not under your control. In my biostats and psychology classes, a covariate was a thing we wanted to test but was not of interest in our analysis. We included a covariate to remove some of the variability within the analysis.

    If I was analyzing your data in a stats class, I would use ANOVA with repeated measures. Depending upon what you call your variables, I would use a Fixed Effects, Mixed Effects of Random Effects model.   

  • Roshanak Soltani added an answer:
    When can I use ANCOVA?

    I know that we can consider pre-intervention amounts as covariate variable if we want to control initial differences. My question is: should these differences be statically significant or not? I mean, when there is a difference between pre-test scores, but it isn't statically significant, can I consider pre-test scores as covariate yet?

    Roshanak Soltani · University of Tabriz

     I have one more question.

    My consulting adviser said we can't use covariance method when there are more than 2 time points. But I'm not sure about that again!

    What's your idea about that?

  • Roshanak Soltani added an answer:
    What is the exact name of my test in spss?

    I have a question about the name of a specific test in spss.

    I have 4 separated groups (4 different interventions), and I measured the dependent variable over 4 time points. I considered my intervention groups as "between subjects factor" and the time points as "within subjects factor". In this case the suitable test would be "mixed ANOVA with repeated measures", right?

    Now if I have a covariate factor, what would be the NAME of the test? "mixed ANCOVA with repeated measures"? or what?

    Roshanak Soltani · University of Tabriz

    I have one more question.

    My consulting adviser said we can't use covariance method when there are more than 2 time points. But I'm not sure about that again!

    What's your idea about that?

  • Bruce Weaver added an answer:
    Can one use multiple logistic regression to estimate possible confounding effect?

    Multiple logistic regression to estimate possible confounder effect?
    We revealed that a protein A has significantly higher concentration in patients than in controls, but there might be a potential confounding variable B, which is also significantly different between controls and patients. I’d like to assess how important the effect of the variable B on concentration of protein A is. Is it OK when I compare simple logistic regression with the diagnosis (0=controls, 1=patients) being the dependent variable and the concentration of protein A being the independent variable with data from multiple logistic regression with variable B added among dependent variables?

    Bruce Weaver · Lakehead University Thunder Bay Campus

    Jurah, regarding your 3rd point, note that for ordinary least squares (OLS) models, it is the ~errors~ (not the outcome variable) that are assumed to be normally distributed.  For further discussion and commentary, click the link below.  HTH.

  • Emir Veledar added an answer:
    How can you compare two groups that are initially distinct in two moments of time?
    ANCOVA or difference between pre-and post? Some authors use ANCOVA with the variable measured in initial time as a covariate, and others employ the difference between pre-and post analyzed via t-test.
    Emir Veledar · Baptist Health South Florida
    Your question is formulated on a very vage way.
    I copy paste question from the top:
    "How can you compare two groups that are initially distinct in two moments of time? "
    So if we have two groups and they are initially distinct, it means that they are distinct at the baseline, so comparison on baseline is allready done.
    If they are "initially distinct in two moments of time" then it means that we compared them in 2 moments and we found them distinct.
    Can you re ask your question?
  • Simona Katholnig asked a question:
    What does it mean if a covariate turns the effect of an IV on DV from significant to insignificant in an ANCOVA?
    How should I interpret this? Does it mean it is a moderator? Or does it mean that there are just no effects of the IV (experimental manipulation)? Thanks!

Topic Followers (15) See all