A generalization of linear regression allowing for nonlinear relationships via a "link function" and for the variance of the response to depend on the predicted value. (Not to be confused with "general linear model" which extends the ordinary linear model to general covariance structure and ...

learn more… | top users | synonyms (1)

0
votes
0answers
12 views

Linear model in R - 2 fixed and one random factor

I have performed an experiment on plants and have some trouble analysing the data. For the experiment, we have three replicates (in time), two species and three treatments. In each experiment, we ...
0
votes
0answers
6 views

Glm gaussian Vs Glm Binomial Vs s log-linked GLM Gaussian

I am trying to do a study on death of malaria in certain in order to estimate the best way to predict how dangerous is this disease. I don't have a strong background in statistics, I am auto-learner ...
0
votes
0answers
11 views

What specification when dependent variable is a fraction?

Basically what the title says; I'm carrying out an exploratory regression on the determinants of board independence. Is using a GML regression the way to go? Thanks in advance
0
votes
0answers
19 views

I'm in over my head, need help on specification

I am carrying out an exploratory regression looking at the determinants of corporate governance. The final model is: Board independence= Country Dummies+ Democracy level+ Sector dummy+ log GDP+ ...
2
votes
2answers
125 views

Difference between regression methods

I have a set of data where the response is a proportion. For each event in the experiment, a system will correctly tag X of Y ...
8
votes
6answers
159 views

Reference Request: Generalized Linear Models

I am looking for an introductory to intermediate level book on Generalized Linear Models. Ideally, in addition to the theory behind the models, I would want it to include applications and examples in ...
3
votes
0answers
23 views

How can IRT-Models be understood in GLM/ SEM Framework? (Predict Learning with added Paradata-Covariates)

I'll be working with data from an intelligent tutor system similar to one studied in the KDD-Cup 2010 on student performance prediction and plan to use IRT models to infer item and ability parameters. ...
1
vote
2answers
48 views

Understanding dummy (manual or automated) variable creation in GLM

If a factor variable (e.g. gender with levels M and F) is used in the glm formula, dummy variable(s) are created, and can be found in the glm model summary along with their associated coefficients ...
4
votes
1answer
24 views

Using log-linear models for presence/absence data in wildlife

I'm working on a project wherein I compare the presence/absence of a number of bird and herptile species between wetlands that have received three different treatments. The populations were surveyed ...
0
votes
1answer
52 views
1
vote
2answers
57 views

Homoscedastic and heteroscedastic data and regression models

How to understand the homoscedasticity and heteroscedasticity in context of regression models? Is there a way to check these properties in R?
0
votes
0answers
21 views

How do you extract confidence intervals and OR out of the step() function in R?

I've been wondering something for a while. If you run a simple regression model in R and then perform a step-wise selection (it doesn't have to be the way I typed the code below), how do you extract ...
0
votes
1answer
28 views

How to interpret coefficients in a Poisson regression with interaction terms?

This question is a prolongation of this question: How to interpret coefficients in a Poisson regression? If we follow the (almost) exact same routine, but we add correlation between the variablese ...
1
vote
0answers
23 views

Compare averaged GLM with boosted regression trees using cross validation : d2 and RMSE calculation

I want to compare BRT and averaged glm models on test sets by calculating the explained deviance and RMSE. How can I calculate d2 and RMSE from predictions? I use the following functions: gbm1 ...
0
votes
0answers
34 views

Fitting non-normal data in lme4 with a family distribution

I'm currently working on fitting a model where we predict the level of some biomarker as a function of time (see image at bottom). I have two difficulties: Each person contributes 2-3 datapoints ...
2
votes
1answer
99 views

GLM for count data

I ran an experiment with an eye tracker and my data frame has this look: ...
0
votes
0answers
18 views

How to select a model in quasi-poisson GLM with interactions using drop1 command?

I want to evaluate the effect of three factors (one categorical, and the other two continuous) on the response variable, which is a count data. I have performed 7 candidate GLM models with ...
1
vote
0answers
21 views

Back Transforming Rates in Poisson GLM with Box and Cox Transformation

Suppose I have fitted a Poisson GLM to model rates as follows: > fit.1=glm(response~X1+X2+ offset(log(population)),family=poisson,data=...) I can get the ...
0
votes
0answers
11 views

Standard error of mean vs. standard error of coefficient in Gaussian glm

I'm having trouble understanding why calculating the standard error of the mean: $$ \widehat{SE} = \frac{\widehat{SD}}{\sqrt{n}} $$ gives me a different SE estimate than the SE estimate of a ...
0
votes
1answer
35 views

glm inflated error…why?

I'm pretty new to stats, so this may be dumb. I've been running a bunch of models on randomly generated data to try and develop my understanding of type 1 error. I've noticed that using ...
0
votes
0answers
7 views

Best method to fit a GEV distribution with generalised linear modelling of parameters?

I need to fit a generalised extreme value distribution to my data but I want the ability to perform generalised linear modelling of the parameters, particularly the location. Can anyone recommend the ...
0
votes
0answers
15 views

Including squared predictors in model matrix [migrated]

I have the following code x <- c(1, 2, 3) y <- c(2, 3, 4) z <- c(3, 4, 5) df <- data.frame(x, y, z) model.matrix(x ~ .^4, df) This gives me a model ...
1
vote
1answer
36 views

Poisson GLM vs Quasi Poisson GLM

I have a Poisson GLM in R that is over dispersed, so I fit a quasipoisson GLM, however the residual deviance nor the degree of freedom change. Can that happen? What does it mean in that case? Thank ...
2
votes
2answers
62 views

Same dataset analysed with four different linear models

I've analysed the same dataset (diamonds from ggplot2) in R with four linear models. Each model has a different error structure. ...
3
votes
2answers
180 views

Checking residuals for normality in generalised linear models

This paper uses generalised linear models (both binomial and negative binomial error distributions) to analyse data. But then in the statistical analysis section of the methods, there is this ...
0
votes
1answer
39 views

Effectively using coefficients from poisson regression

This is maybe annoyingly easy for some, but I am completely new to regression. As an example, I shall use the data set in R, called mtcars. I am interested in the ...
1
vote
0answers
33 views

Comparison between normal glm and glm.nb regression with quadratic term?

Let's say I have a function to simulate data for negative binomial regression: ...
2
votes
0answers
19 views

What difference (if any) exists between the Response Distribution and Error Distribution in GLMs?

Ok, forgive my ignorance, but I keep getting confused about something at the core of GLMs. Some textbooks describe the two main parts of a GLM as the link function and the distribution of the error ...
2
votes
1answer
46 views

Why NB and Poisson performs superior than ZIP, ZINB and Hurdle in presence of lots of zeros?

I am working on a data which contain nearly 80% of zeros and positive counts as large as 7. The dataset is very large, nearly 16,000 cases. It is a health related data. I have fitted ZIP, ZINB and ...
1
vote
0answers
31 views

What algorithms are available for logistic regression?

I am trying to implement a logistic regression function in c++, and not sure what algorithm to use. So far I have heard of these: Newton-Raphson IRLS Gradient descent Are there other algorithms ...
1
vote
1answer
44 views

Analysing overdispersed data with generalised linear models

Let's say I have an explanatory variable and a response variable that represents counts. I want to see if the explanatory variable can predicts counts. I'm aware the response variable is ...
0
votes
0answers
31 views

What is wrong with this implementation of logistic regression (using iterative reweighted least squares)?

I am trying to implement logistic regression using the following algorithm: fit a simple linear model $y \sim Xb_0$ calculate $W = \frac{e^{Xb_0}}{(1+e^{Xb_0})^2}$ calculate $z = Xb_0 + y \cdot ...
1
vote
1answer
42 views

Link Functions where MLE=OLS

This is a follow-up of a question that I posted previously. I'm trying to get parameter estimates from two different SAS functions (Proc REG and ...
0
votes
0answers
50 views

Treating numeric as categorical variable in regression

I need a little bit of help and confirmation that I have the right idea. I have some fake data of 8 tribes; within each tribe members work hard to gain food for their own tribe. No one can speak to ...
1
vote
2answers
27 views

T test with one group

I think it is a simple question, but I got many different results and now I am not sure which test to use. I have one group and 2 different scores (higher scores means higher either positive or ...
3
votes
0answers
24 views

Using residualized predictors outside the linear model context

Can anyone point me towards a good explanation of when a residualized variable in a regression will give you the same answer as using a non-residualized variable with controls? For instance, say I ...
1
vote
1answer
63 views

Statistically prove the effectiveness of a treatment using GLM repeated measurements

I have two lots of samples: one is the control lot and the other undergoes some treatment. I did three measurements for the samples: one at the initial time (T1) ...
3
votes
0answers
95 views

How to choose the cutoff probability for a rare event Logistic Regression

I have 100,000 observations (9 dummy indicator variables) with 1000 positives. Logistic Regression should work fine in this case but the cutoff probability puzzles me. In common literature, we ...
0
votes
0answers
12 views

What makes the canonical link function special in GLMs? [duplicate]

Why is the canonical link function used so frequently with GLMs? What makes it "natural"? Is there any reason to think that, $Q(\theta _i)$ (where $Q$ is the canonical link function, and $\theta _i$ ...
2
votes
1answer
47 views

Generalised linear model fitted values

I have ran this model in R: glm(alert ~ water.height + ssp*ssp.zone + log(count) + ssp*days, family=quasibinomial, data=ScanSampling_sub_alert) ...
2
votes
1answer
34 views

Should anova(model, test=“Chisq”) not produce a test statistic as well as a P value?

Doing a poisson regression like this: model<-glm(y~x*z,family=poisson) with one predictor being a factor, I would use ...
0
votes
1answer
114 views

Logistic regression using penalized likelihood (lasso?) in Matlab/R

I am trying to use logistic regression in a scenario where there are very few positives. I'm aware that maximum likelihood suffers from small sample bias. So ...
2
votes
1answer
52 views

R code for zero inflated Poisson

I have used glm() to model some data I have. The code looks like the following: ...
1
vote
1answer
44 views

Using GLM with transformed data

I have a dataset with outcomes from a Vasicek distribution (see this pdf) and some covariates. Re-expressing the Vasicek's pdf into the exponential family form requires me to transform my data, i.e. ...
3
votes
0answers
39 views

A lot of iterations before converging - GLMM vs GEE question

I'm wondering if there is any implication if a model takes 100+ iterations to converge. Can I still trust the results? I'm running cumulative logit with random intercept in proc glimmix in SAS. ...
3
votes
0answers
53 views

R: Problem with NaNs in gamma regression using glm

I have found a problem with NaNs when I try to fit the model from example 5.1 in the book Generalized Linear Models. With Applications to Engineering and Life ...
3
votes
1answer
83 views

Why are there no one-inflated count data models?

I am working on zero-inflated count data models using the pscl package. I am just wondering why there is no development of models for one-inflated count data ...
0
votes
0answers
19 views

Model building based on preliminary data analysis

I have a extremely simple data set which consists of one independent variable (five sampling sites as factors) and the outcome variable which is disease status (positive/negative) in crabs from each ...
0
votes
0answers
30 views

How do I test the difference between two dependent correlation coefficients whilst controlling for other variables?

How do I test the difference between r(X,Y) and r(X,Z) [Williams Test?] whilst controlling for demographic variables that were measured in addition to X and Y and Z [Partial Correlation]? Thanks, ...
0
votes
0answers
18 views

z-test in paper false? can Williams Hotelling test be used instead?

I am looking at the statistics of this paper: http://emotional.intelligence.uma.es/documentos/6-Extremera2011Emotional.pdf I am wondering how you can use the z-test to compare the two correlation ...