Search results

  1. M

    Two ways to include temporal autocorrelation

    Hi, assume that we have time series with autocorrelated values described by the regression model Y_i ~ X_i. As far as I understand, in AR-regression models, this correlation is considered by assuming that residuals are autocorrelated. Instead, I could introduce additional predictors...
  2. M

    Variance of a sum of predictions

    Hi, I predict the spatial distribution of a species on a regular grid using a regression model. Species numbers vary with environmental covariates. Now I want to predict the overall size of the population within a given area, which means technically that I sum up all predicted values. But...
  3. M

    Dealing with temporal autocorrelation

    Hi, assume we have a measured value for each year from 1990-210, and we want to estimate a linear trend. Now there is a problem of autocorrelation, i.e., the measured values do not scatter randomly around the fittet line but rather resemble a "smooth function" wriggeling around the linear fit...
  4. M

    Regression with additive and multiplicative predictors

    Hi, we are analyzing pollutants [µg/kg] in mussels with regression models. Let's say we have only the two predictors (1) "year" and (2) "month". However, the effect of "year" on the outcome is assumed to be additive, i.e., in avergage a constant amount of poullutants is added each year...
  5. M

    Regression with data from different sources: Nested fixed effects possible?

    Assume I have collected my data with 3 different methods, each method share the same outcome Y as well as the predictor X. I am interested in the dependency of Y on X via regression analysis. However, each method has additionally a unique set of fixed or random effects (A,B,...), which are...
  6. M

    Combining data from different sources

    Hi, I have data from two very different sources (bird counts from ships and bird counts from the shore) which aim to estimate the same population/outcome. Both data have different covariates/predictors and thus it is difficult to combine both data in the same regression model. Does...
  7. M

    Normality / QQ-Plot in Generalized Linear Models (GLM)

    Hi, sorry if this is a frequent question, but I did not find an appropriate answer: If I have a GLM with a non-normal distributed outcome (such as Poisson- or Binomial distributed), does it make sense to check normality of the residuals? Intuitively I would say no, since I explicitly...
  8. M

    Interpretation interaction term in glmer()

    Hi, I am using the glmer() function from the package lme4 for a mixed logistic regression model. Here, the formula is Y ~ X + Z + X:Z, where Y is the binomial outcome, X is a categorical predictor with 3 levels (X1, X2, X3, where X1 is the baseline), and Z is a continuous predictor. In the...
  9. M

    Regression with stratified data: What to do?

    Hi, I want to model a bird population in a country via regression analysis (Poisson regression due to count data). In order to optimize the sampling effort versus precision, I use a stratified random design: The country is splitted into 10 different habitats, and from each habitat I draw...
  10. M

    Partial effect plot: Which options do I have?

    Hi, I have a multiple regression model, and I would like to do an effect plot with confidence bands regarding one of the predictors. One solution I see is to create a fake-dataset, where all other predictors are constant. However, this produces pretty large cofindece bands, since they are...
  11. M

    Prediction Plot with CI's: What to do with covariates?

    Hi, I have a multiple regression model (especially a GLMM), with - outcome Y, - a fixed effect predictor wich is in the main focus of my interest (time variable: "YEAR") - an additional fixed effect covariate, which is only in the model to prevent bias ("LOCATION" - the sampling...
  12. M

    nested fixed effects possible?

    Hi, I investigate a problem via a regression model. Firtsly, I have two different ways to measure my outcome, which I incorporate as a two-level factor variable X. However, each of these methods is again influenced by a factor variable with two levels, but these factor variables differ for...
  13. M

    Trend analysis / regression model with missing data

    Hi, I want to apply a Poisson GLM to count data, analyzing trends. However, there are missing counts (i.e. missing values in the outcome variable) in the dataframe. Is there any recommended method how to deal with this? I know there exists "multiple imputation methods" to fill the gaps in the...
  14. M

    More comlex regression models: Terminology and Software

    Hi, I would like to couple two regression models and solve them simultaneously. That means: I have one regression equation: \hat Y_i = \hat \alpha_0 + \hat \alpha_1 X_i with Y_i \sim N(\hat Y_i, \sigma_1^2) and \hat Y_i is again the predictor in a second regression model: \hat...
  15. M

    Unimodal Link-function possible?

    Hi, using a GLM model (e.g. a Poisson regression), there are various link functions available in statistical packages. However, I don't find an unimodal depedency ("Normal/Gaussian link"), and I always have to approximate an unimodal relationship with polynomial terms. Does somebody know why...
  16. M

    Multivariate Statistic: Linear or unimodal relationship?

    Hi, if I want to analyze a multivariate dataset, I have to choose between methods assuming a linear relationship between the outcome and underlying gradients (e.g. in PCA or RDA) or an unimodal relationship (e.g. in CA or CCA). Does somebody know if there are appropriate test statistics...
  17. M

    ANCOVA versus regression: Independence of the covariate and treatment effect.

    Hi, if you read about the assumptions of an ANCOVA, you find the following assumption: "independence of the covariate and treatment effect". That is, the treatment effect should be randomly distributed between groups representing the treatment, and can be checkt via a T-Test. I dont...
  18. M

    Violating Homogeneity

    Hi, if in regression models homogeneity is violated, the variance can in principle depend on both: Predictor and outcome variable(s). In case of a dependency on categorical or continuous predictor variables, there are many ways to model different dependencies, e.g. using GLS models. In...
  19. M

    Appropriate link-function in a Poisson Model

    Hi, does somebody know why the log-link function is always assumed to be the "standard" link function in Poisson models? E.g. regarding count data (such as numbers of individuals), under which circumstances does it make sense to assume that the response variable increases exponentially with a...
  20. M

    Automatic addition of contrasts

    Hi, if I define contrasts for a factor variable with n levels, e.g. contrasts(x) <- c(-2,1,1) and I subsequently display these contrasts via contrasts(x) I obtain - beside the self-defined contrasts- (n-1) additional contrasts (second row): [,1] [,2] -2...