Questions tagged [causality]

The relationship between cause and effect.

Filter by
Sorted by
Tagged with
2
votes
1answer
30 views

Front-door adjustment formula: difficulty in reconcile the two formula

In The Book Of Why, the author gives two formulas related to the front-door adjustment formula. On page 227 (formula 7.1), the front-door adjustment is given by $$P(Y|do(X)) = \sum_zP(Z=z|X)\sum_xP(Y|...
0
votes
1answer
27 views

How is this term "effect of treatment" justified?

In Mathematical Statistics with Applications, 5th Ed., by Wackerly, Mendenhall, and Scheaffer, Section 13.5, the authors present the Statistical Model For a One-Way Layout (p. 587): For $i=1,2,\dots,k$...
3
votes
3answers
80 views

Textbook recommendations covering machine learning techniques for causal inference?

Over the past 15 years there has been progress in adapting machine learning methods for causal inference. For example: targeted learning, double machine learning, causal trees. Is there a textbook ...
0
votes
0answers
17 views

Is it possible (reasonable) to weight the regularization for some variable in Ridge/elastic net based on their importance/causal effect

Say I have 100 predictor variables. And I have estimations from a causal inference method that indicates the causal effect size of each variable to the response variable. Then I want to build a linear ...
2
votes
1answer
41 views

Why is a' used in the proof of the front-door adjustment formula by Hernán and Robins in their book Causal inference What If [closed]

Please can I confirm what the meaning of a' is in the proof of the front-door adjustment formula. I assume it represents all other treatment levels not equal to the level of interest (a) but I am ...
3
votes
1answer
46 views

Why does propensity score matching fail to estimate the true causal effect when OLS works?

Consider the following model (DAG), where D is the treatment (exposure) and Y1 is the outcome. To estimate the causal effect of <...
2
votes
1answer
66 views

Is the DAG do calculus rule an axiom?

Using the 'do calculus' + DAGS framework for causal inference, is this If $(Y\perp X)_{G_{\underline{x}}}$ then $P(Y|X=x)=P(Y|\text{do}(X)=x)$ an axiom? Or can it be proven from first principles (...
0
votes
1answer
29 views

Adjustment set for the DAG

I have the following DAG From Dagitty tool, I am getting minimum adjustment set as Growth when the exposure is Treatment and outcome is dANB and UC is the unobserved confounder variable. If the ...
3
votes
2answers
126 views

Using a DAG to understand omitted variable bias in OLS vs Binary Dependent Variable Regression

Suppose I have three variables. $A$ and $U$ are continuous variables but $U$ is unobserved. $Y$ is the binary outcome. $A$ and $U$ are independent. Let the true model be from the typical probit or ...
5
votes
0answers
75 views

Derivation of a doubly robust estimator with clever covariate and inverse probability weighting

With notation: outcome $Y$, (binary) treatment $A$, and covariates $L$. In Hernan and Robins (2020) causal inference textbook: To obtain a doubly robust estimate of the average causal effect, first ...
0
votes
1answer
9 views

Difference-in-difference with discrete treatment intensity

In my context I have individuals who receive treatment with an intensity of 1 to 5 once a policy has been implemented. I've seen many posts on continuous treatment and how to estimate a DD model in ...
0
votes
1answer
22 views

Variable type name in causal inference

Causal inference language distinguishes different variable types: confounders, mediators, colliders, moderators. Some time ago I encountered quite rare variable name which I can not remember. The idea ...
1
vote
2answers
39 views

How can you express the average treatment effect on the treated (ATT) in Pearl's do notation?

How can you express the average treatment effect on the treated (ATT) in Pearl's do notation? Would it be $E(Y|X=1,do(X)=1)-E(Y|X=1,do(X)=0)$?
0
votes
1answer
27 views

Doubly robust learning with binary treatment and outcome

I'm trying to use doubly robust learning to estimate heterogenous treatment effects. My treatments T and outcomes y are both binary. I'm following the example listed under "How do I select the ...
2
votes
0answers
37 views

What can I do to get better overlap in propensity score distributions?

I would like to verify the positivity assumption to identify causal effects from observational data. My exposure prevalence is about 6%. When I included several potential confounders in my exposure ...
0
votes
1answer
48 views

Why isn't the Random Variable over which expectation is taken generally written down or noted as the argument of the expectation?

This is page 266 of the Causal Inference book by Rubin/Imbens. Note that the argument of the expectation is written down below $E_w$ but not in the first set of equations on the page. IMHO, it makes ...
1
vote
3answers
64 views

Causal inference in python - where to start?

Point 1: I'm not sure if this question could be asked here, as it is may not seem to be about the "science" itself at the first glance! At the second glance though, I think in practice ...
1
vote
0answers
19 views

Doubly robust learning with same features influencing treatment and outcome

I'm looking at some of the examples in the econML package for double machine learning. Specifically, the example found here (code below). In the example W is the features which might influence both ...
17
votes
1answer
2k views

Why do we do matching for causal inference vs regressing on confounders?

I'm new to the area of causal inference. From what I understand, one of the main concerns that causal inference tries to address is the effect of confounders! For the sake of reference, let's denote ...
1
vote
1answer
42 views

Is there a way to know X is really causing Y?

The real question here is if there is a way to know X is causing Y rather than simply being associated with it. In my example Y seems to be telling more of X(if not causing it) than X says about Y. ...
0
votes
0answers
8 views

Groups Impact on Other groups in Twitter Data

I'm looking at twitter data and have: Time format (YYYY-MMM-DD HH:MM:SS) Retweets Indegree outdegree between centrality weighted outdegree weighted indegree I have 5 groups [A, B, C, D, E]. What ...
3
votes
2answers
75 views

Determining direct cause with some "realization" of `V \ {X, Y}`

I'm bouncing around "Causality" by Judea Pearl. On page 222 it offers this definition of a direct cause: "$X$ is a direct cause of $Y$" if there exist two values $x$ and $x'$ of $...
0
votes
0answers
31 views

Interpretation of the E-value

A recent method for sensitivity analysis is the E-value (VanderWeele and Ping, 2017). Yet, I'm still struggling with the interpretation of such a value. Let's say one has an E-value of 1.75. Could ...
1
vote
1answer
29 views

Dividing data set based on the Dependent Variable, for interpretation of the coefficients

I have a data set that has as DV the preference of spatial reproduced audio files (OLE) and as IVs the preference of only their content and the sensation of envelopment. All the variables are ...
1
vote
1answer
39 views

Pearl's Front-door and Back-door

I've encountered lots of causal inference terms and jargons (under the Neyman-Rubin potential outcome framework), and I had a question regarding how Pearl's DAG restrictions relate to ignorability and ...
0
votes
0answers
27 views

Causal Inference: Ignorability and Collider

I've encountered lots of causal inference terms and jargons (under the Neyman-Rubin potential outcome framework), and I had questions regarding ignorability. Is it the case that ignorability is always ...
0
votes
1answer
40 views

Causal Inference: Moderation and Mediation

I've encountered lots of causal inference terms and jargons (under the Neyman-Rubin potential outcome framework), and I had a question regarding mediator and moderator. Is it the case that moderation /...
1
vote
0answers
38 views

Causal Inference: Selection Bias and Endogeneity [closed]

I've encountered lots of causal inference terms and jargons (under the Neyman-Rubin potential outcome framework), and I had questions regarding their relationships: I know that exogeneity E(e|X) = 0 ...
1
vote
0answers
26 views

Determination of treatment effect on the nodes of generalized random forests

I am reading Athey paper on generalized random forests that you can find for instance here https://arxiv.org/abs/1610.01271. Now I do not understand formula (6) where it is defined an approximation of ...
0
votes
0answers
23 views

Why test.modmed (mediation package - moderated mediation) does not find covariate?

I am trying to do a moderated mediation analysis in R with the mediation package, test.modmed function. In summary, my dataset consists in: Y = cell_size (continuous); X = body_size (continuous); ...
2
votes
1answer
25 views

When is the knowledge of the causal mechanism useful for pure prediction?

In many settings, we are only interested in building a good predictor: e.g. $E(y_t | x_{t-1})$, where $y_t$ and $x_{t-1}$ are vectors of arbitrary dimension. However, sometimes we are also given, or ...
0
votes
0answers
17 views

Partial Dependence Plots for Predictors that Cause Other Predictors in the Model

A while ago it was asked Is it a creditable approach to use Random Forrest Variable importance for causal inference? The recommendation was to use partial dependence plots (PDP). The referred paper (...
1
vote
0answers
25 views

Endogeneity in Discrete Choice Models

Suppose we are trying to estimate a model to predict failure of a company as a function of some firm-level and macroeconomic variables using a logistic regression $Y_{it} = X_{it} \beta + \epsilon_{...
1
vote
1answer
51 views

Matching is not recovering the true effect in simulated data

I am trying to recover the true (simulated) effect of a treatment Z on an outcome Y, which is set to ATE = 5 (the csv file for the data is located here: https://www.dropbox.com/s/92obn9hsu3tqy92/...
4
votes
1answer
49 views

mediation vs interaction - R application

I am struggling to understand how/if the interaction is connected to mediation. I understand that the interaction in a regression indicates that a variable Z influences the effect of a variable X on ...
1
vote
0answers
21 views

Confusion on Ablation test (Ablation Experiment or Ablation study)

I followed the steps of the ablation test to calculate the feature importance one by one. In Table 1, row 1 presents the model prediction performance of using full features. Regarding rows 2-4, these ...
0
votes
1answer
18 views

Min-Max Scaler for fixed effects regression?

I m currently doing some social science research using panel data to determine the impact of budget cuts on financial vulnerability. I have decided to use a fixed effects model to determine this ...
0
votes
1answer
22 views

How to interpret output from "mediation" package - can I claim full mediation?

I have an outcome binary variable Y, a continuous mediator M, a binary treatment T and some covariates C. I have a linear regression model for the mediator: $ M = \alpha + \beta T + \delta C $ and a ...
2
votes
1answer
23 views

Is the "constant additive unit causal effect" assumption really needed to interpret a regression coefficient as the ATE?

I am reading Unpacking the black box of causality. At page 768 there is written that, in order to uncover the ATE: In observational studies, slightly more complex calculations may be needed, although ...
1
vote
1answer
37 views

Estimate Causal Effect of Countinuous Treatment on Binary Outcome

I am trying to estimate the causal effects of a continuous treatment variable $T$ on a binary outcome $Y$. I have a set of nuisance variables $X$ that I know effect both $T$ and $Y$. I have read about ...
1
vote
1answer
30 views

Why is the inverse probability weighting still defined when positivity does not hold?

Specifically, in Technical point 3.1 of the book “Causal inference What If” by Hernán and Robins, the authors note that when positivity does not hold the standardised mean is not well defined however ...
4
votes
1answer
67 views

Causal inference - difference between blocking a backdoor path and adding a variable to regression

I have just started this introductory course to causal inference. The DAG approach is completely new to me even though I come from an econometric background (though that dates back to 15 years ago). ...
4
votes
1answer
80 views

causal graph - counting the number of backdoor paths in a DAG

I am following "A Crash Course in Causality: Inferring Causal Effects from Observational Data" on Coursera. I am struggling at correctly identifying backdoor paths in causal graphs (or DAG ...
0
votes
0answers
18 views

Reverse causality in panel data with count outcome

I have panel data on several thousand individuals and I am trying to estimate a model to look at the effect of a variable $x_i$ on an outcome $y_i$. There is definitely a reverse causality ...
5
votes
1answer
44 views

Cyclicality in causal relationships

Causal graphs are an increasingly popular tool for causal inference. The underlying understanding of causality is deterministic. In the popular directed acyclic form of causal graphs, we assume that ...
1
vote
1answer
106 views

Time fixed effects and entity fixed effects?

I am currently trying to find if there is a causal relationship which exists between cuts to government spending and county court judgements. I have panel data covering multiple regions across 10 ...
1
vote
0answers
17 views

causal graphs in non-linear physics [closed]

Can we learn emergent non-linear physics using causal graphs? Causal graphs are well used in linear systems, I'm wondering if it's too risky to use the approach for non-linear dynamics, such as ...
0
votes
1answer
28 views

Confused with logistic regression concept (vs. linear regression) based on causal thinking [closed]

I always thinking about regression model is based on Y occurs given X. It means Y is always occur after X shown. linear regression Like this... example1. price of egg = b0 + b1*(chicken's age) + b2*(...
1
vote
0answers
40 views

CausalImpact package - results summary inconsistent with the result plots

I'm trying to get some experience with the `CausalImpact` [package][1] in python. I'm using this seemingly simple example: ...
1
vote
1answer
91 views

Classical difference-in-differences: Coding the time (post) variable when treatment starts at different times

I have panel data with 40 treated cases and 40 control cases. I thought about the application of the 'classical' difference-in-differences (DiD) equation with the following linear regression model: $y ...

1
2 3 4 5
24