Questions tagged [confounding]

In statistical models, confounding is said to occur when the apparent dependence of the response on a predictor is partially or wholly due to the dependence of both on a third variable not included in the model, or dependence on a linear combination of other variables included in the model. Confounding with a variable included in a model is often called multicollinearity. A synonym is *aliasing*, used in design of experiments.

402 questions
Filter by
Sorted by
Tagged with
47 views

How to deal with possibly important predictors omitted during the building of an OLS multivariate linear regression model?

I am building a descriptive model using OLS multivariate linear regression. I have a couple dozen candidate predictors, but only around 200 cases. Since I wanted at least 10 cases / variable for the ...
90 views

Shrinkage of covariates in the Cox model

In a regression model (e.g Cox model) when there are too few events to support modeling all desired covariates / confounders, a possible solution is to apply shrinkage / penalise all but the exposure(...
• 819
42 views

Three continous variables + 2 factors vs. five continous variables to control for confounders?

I am trying to make sense of the design for my Master's thesis. I am looking at how three different types of play relate to anxiety in children. So I have three continuous independent variables ...
• 31
53 views

Understanding "Multiplication/ Group Operation" in Fractional Experiments

There's a group operation, at least of sorts, in Fractional Factorial design that I'm trying to understand. For definiteness, let's say we have 3 factors; A,B,C , at two levels each . Please critique ...
• 579
52 views

Chi-Squared for demonstrating confounding in Logistic regression (or not...)

I am using logistic regression for inference and classification, using data from 190 X-rays/subjects. We want to see if certain X-ray measurements could predict development of a disease (Case vs ...
1 vote
80 views

When does Dose Response Function estimation work better than simple regression?

I have been recently asked what is the difference between a Dose Response Function (DRF) estimation (as the one proposed in this link and this paper) and a statistical regression method. I therefore ...
• 460
19 views

IPW recalculating with deterministic treatment

Let's assume I want to calculate the ATE for a certain deterministic treatment, such as surgery (i.e., one either had it or not), and I'm interested in a per-protocol analysis. Note that those who had ...
26 views

How can I test whether a moderation effect is only present due to confounding variables?

I plan to investigate the effect of a personality trait on reaction measures (emotional reactions and intention for political participation) in a vignette study with two conditions. I assume that the ...
• 1
43 views

How to account for confounders in a simple correlation analysis?

Beginner question sorry - I'm a coder and need stats advice. I have a dataset broken down by local area, with columns for the proportion of owners who are French, the proportion of owners who grow ...
• 683
32 views

A covariate as an inherent part of predictor

I want to compare brain volumes of two disease categories: young vs. old onset. I know that age, in general, is a covariate for brain volume. That is, the older the age, the smaller the brain. However,...
236 views

Adding and interpreting covariates in logistic regression

I have a dataset and I want to do a logistic regression between the continuous variable "A" and the categorical variable "B". However, I also wanted to include "age" and &...
18 views

Confounding variables in a t-test [duplicate]

Suppose you want to compare a certain score (dependent variables) in two groups A and B (independent variable), to see if one group has significantly better scores. You can run a t-test. Now what if ...
• 101
1 vote
29 views

Can I simulate an unobserved confounder to test sensitivity?

I have a linear regression model Y ~ b0 + b1*X, with X and Y as continously measured variables (say age -> income). Assume I know from theory that there is an unobservable variable Z that is ...
• 11
239 views

Why does controlling for a collider open a path, while controlling for a confounder closes a path, if there are relations to third variable for both?

Collider bias occurs when there is no association between X and Y but when a third variable which is caused by both X and Y is controlled for, this "opens a path" between X and Y and leads ...
• 939
56 views

Logistic regression - When to include or exclude a confounding variable from the model?

I am working on determining if there is an association between a medication (yes/no) and a health outcome (yes/no). However, the lines between what is a confounding variable to include in the model, ...
1 vote
28 views

Is it a problem to correlate X with changes in Y caused by X and Z?

Imagine you have three variables: X, Y, and Z. X and Z are both separate causes of Y. You want to know if changes in Y caused by Z correlate with X. So, you manipulate Z to cause changes in Y. Let's ...
61 views

Bias results for mismeasurement of continuous confounders

Consider data generated from a model $Y = \alpha A + \beta U$, where $U$ is a confounder, i.e. $\langle A,U\rangle \neq 0$. We don't measure U, but rather a noisy version of it, $U' = U+\epsilon$, ...
• 495
59 views

How do random effects affect fixed effects (or just other coefficients) within a model

I have found that random effects terms can affect other coefficients within a model from here. I see how in this example the coefficients change with the addition of a random effect; I'm still not ...
• 651
637 views

Should You Specify a Curvilinear/Non-Linear Effect If You Suspect It is Spurious?

Consider the following (simplified) example of a project I am working on: I assume that $X$ has a linear effect on $Y$. However, after plotting the relationship on a scatter plot, it looks like the ...
31 views

I have some epidemiological data, relating to the prevalence of obesity at UK local authority level. For the purpose of exploration I want to derive the median obesity prevalence. However, I need to ...
30 views

Confounding by indication in observational cohort study: how to address?

Background We are conducting a register based cohort study into effect of pain killer medication in cancer patients. Index date is date of diagnosis of early cancer in a defined study population. We ...
• 819
47 views

May unobserved variable confound or create open backdoor paths, why didn't controlling for the collider O make bad?

Is the U, the unobserved creating an open backdoor path or confounding? Why condition on the collider Occupation good here?
16 views

If you condition on O, you stop the flow of information from which arrow into or out of O? [duplicate]

Why does conditioning on O open up this second channel below?
• 1
18 views

How to test for effect modification and mediation by time-varying factors in a longitudinal context?

I have longitudinal data with measures of an exposure, outcome, variables that may be mediators or effect modifiers, and confounders at 3 time points. I would like to account for time-invariant and ...
• 471
1 vote
28 views

Comparing reduced, intermediate and complex model in logistic regression?

How should I compare my logistic regression models? My study is about the association between diabetes (dia), and a genetic variation (GV) I'm adjusting for confounders such as gestational age (GA) ...
• 333
1 vote
122 views

Different results with splitting data vs. adjusting

I have a question regarding the results that I have achieved from my analysis. I'm new to statistics and the understanding of epidemiology. Please, help me interpret this better. I know what a ...
• 333
1 vote
139 views

Are we allowed to use classifiers to estimate the outcome in the first stage of Double Machine Learning when the outcome is binary?

It is clear to me how to proceed when the outcome is continuous, since the EconML and all other references I checked work with this type of examples (continuous outcome case). We simply apply a ...
• 11
18 views

• 495
111 views

Compare two groups with differnt treatments and different populations

I will simplify my research question. I have 40 patients (group 1) with assessed a variable v1 (e.g. their blood preassure) at timepoint 0 ( v1(t0) ). After treatment (treatment 1 at timepoint 1) I ...
• 11
95 views

How control for a pre-treatment outcome $Y_0$ if is a strong confounder while avoiding regression to the mean bias for treatment effect on $Y_1$?

I'm facing a dilemma in a pre/post cohort matching analysis for a healthcare intervention: Matching on the pre-treatment outcome $Y_0$ (a continuous variable) will likely lead to regression to the ...
• 6,134
45 views

Antimalarials as confounding variables in a prevalence study

I am performing a multi-level logistic regression analysis of cross-sectional survey data, and I am curious as to whether "taking antimalarials" is a confounding variable in the following ...
• 171
1 vote
15 views

Is a high prediction performance of allocated treatment evidence for lack of unobserved confounding?

Example Say Investigator 1 is looking to investigate the effect of fish consumption (a binary variable $A$ indicating whether they eat or do not eat fish) on cardiovascular health (a continuous ...
• 1,471