Causal inference tries to quantify the effect of a change in $X$ on $Y$ whilst holding constant or eliminating all other relevant factors which might influence this relationship.

learn more… | top users | synonyms

2
votes
2answers
24 views

Understanding the direction & strength of a correlation, & causal status, from a chi-squared test

I am analyzing Stack Overflow Posts. So I have a database with 1000000 questions, their current score (upvote or downvote) and a flag, that there is a source code part in the question (or not). So I ...
2
votes
1answer
35 views

Robust Coefficients For Differences in Differences

I have a panel data set which I am looking to analyze for relationships/causality using the OLS differences-in-differences method. The panel data includes multiple observations over time for various ...
0
votes
0answers
39 views

Causal Trees to Estimate Heterogenous Treatment Effects: Transformed Outcomes [Machine Learning in Python]

I am interested in using off-the-shelf tools like scikit-learn for Python to implement the Athey-Imbens recommendation for estimating treatment effect ...
3
votes
1answer
111 views

Questions about the paper “Inferring causal impact using Bayesian structural time series models”

I am simulating this model in the paper Brodersen et al. (2015) and found I have difficulties on figuring out how to choose the starting point of Bayesian inference. To be explicit, to my ...
5
votes
2answers
92 views

Representing interaction effects in directed acyclic graphs

Directed acyclic graphs (DAGs; e.g. Greenland, et al, 1999) are a part of a formalism of causal inference from the contrafactual interpretation of causality camp. In these graphs the presence of an ...
1
vote
2answers
770 views

What test is this for endogenous variables?

Can somebody tell me whether the following R code (for econometrics endogenous variables) is for a Hausman test, a Nakamura test, or some other test? ...
2
votes
0answers
266 views

group fixed-effects, not individual-fixed effects using plm in R

I am analyzing some data to evaluate the impact (causal effect) of a program that is delivered at group level (a village). The outcome of interest is measured at the individual level (individuals ...
1
vote
0answers
38 views

fixed effects question

I am analyzing survey respondent data. Respondents are nested within regions and are surveyed 4 times. Over the 4 time periods, different regions undergo a policy change (go from 0 to 1) randomly (or ...
0
votes
0answers
10 views

why are these uplift estimates different?

I recently stumbled upon an ODSC presentation: Machine Learning Based Personalization Using Uplift Analytics: Examples and Applications Uplift wherein two methods for "uplift" estimation are provided: ...
0
votes
2answers
56 views

CausalImpact - Should I use more than one control?

In the intro document (https://google.github.io/CausalImpact/CausalImpact.html) it suggests that using one predictor is not ideal. Am I current in understand that they mean one control? If so, should ...
3
votes
1answer
364 views

How do I interpret a “difference-in-differences” model with continuous treatment?

How do I interpret the ATE coefficient (i.e., the post-treatment indicator interacted with the continuous variable)? Does it make sense? Should I break it down into subgroups and just run a fixed ...
0
votes
0answers
220 views

Is there evidence of mediation? Need help with interpretation of mediation analysis results

I have performed a mediation analysis. I have an independent variable T, a mediator M, and outcome Y. (All 3 variables are binary, and I use logit.) (While I used Stata ...
1
vote
1answer
36 views

Challenging Propensity Score/Causal Inference Problem

I am reaching out to the Cross-Validated statistical community seeking suggestions on a challenging problem on which I'm working. I've been asked to look into a problem related to electronic ...
0
votes
0answers
35 views

Google trends data for interest

I was discussing about the popularity of some terms and used google trends to conclude in the decrease of their popularity. Here is an exemple of the queries for some of the biggest french ...
1
vote
1answer
39 views

Causality studies on observational data: DID with 2SLS to compliment Rubin causal model

A study with observational data has treatment and control group but the assignment is not randomised: some chose to be in the treament, some otherwise. But the choice had been made before the ...
0
votes
0answers
13 views

How to incorporate characteristics of the object of outcome variables as predictors in model

I'm working with survey data for a social science research topic. The particulars aren't super important, so I'll use a simplified version to make it easier to understand what I'm trying to do. I say ...
0
votes
1answer
55 views

CausalImpact - Is pre-processing of covariates required?

I'm using the CausalImpact to evaluate the effect of a programme. My covariates are seasonal and I wonder whether I need to deseasonalise/ detrend the regressors before using the R package? Hal ...
0
votes
0answers
21 views

Using spike-slab to fit log-link GLM Gamma

I am attempting to model the causal impact (using CausalImpact package in R) of a know discrete event on the change in medical expenditures. I have 12 pre and 6 post period observations and upwards of ...
0
votes
1answer
68 views

Regression discontinuity design with multiple cutoffs

When we have an RD with multiple cutoffs, and we pool all observations and estimate the treatment effect across cutoffs, what does the pooled estimate identify? I have found one paper using a RDD ...
0
votes
0answers
17 views

Variables with low coefficients/high p-values in multiple linear regression highly correlated to significant predictors

This is possibly a very naive question: I am watching Hastie and Tibshiranie's class ("Introduction to Statistical Learning"), and at the end of this lecture, Tibshirani gives example of a multiple ...
3
votes
1answer
255 views

Is it okay to run CausalImpact in R on successive portions of a time series?

We’re doing some advertising tests with test and control groups very similar to the example in the Google Research Causal Impact publication except we’re doing state tests and not DMA. I just have a ...
1
vote
1answer
124 views

CausalImpact on single time series

Today I have tried to play a little with CausalImpact R-package https://google.github.io/CausalImpact/CausalImpact.html (Brodersen et al. 2015) to explore the impact of some decissions in a sales data ...
1
vote
1answer
38 views

How can I analyze / compare point-of-sales data between stores with different product offerings?

Let's say I have two stores, A and B, and let's say I have two products in the same department: product 1 and product 2. Let store B have product 2 which store A doesn't have, both stores have product ...
0
votes
0answers
14 views

Modeling complex mechanism with regression (or similar method)

There are six variables A, A', B, C, D, and D' and they are related as shown in the figure below. I want to know the effect of A on C through B. I firstly thought to use 2SLS, but since A and C are ...
2
votes
1answer
33 views

How to test whether it is valid causal inference

definition of average causal effect $$ ACE = E(C_1) - E(C_0) $$ $$ ACE = E(Y|X=1) - E(Y|X=0) $$ given the condition that $$ X \bot (Y(0), Y(1)) $$ So if I have a regression ...
5
votes
1answer
374 views

Confidence interval for average treatment effect from propensity score weighting?

I am trying to estimate the average treatment effect from observational data using propensity score weighting (specifically IPTW). I think I am calculating the ATE correctly, but I don't know how to ...
4
votes
1answer
135 views

What is the best way to visualize difference-in-differences (multi-period) regression?

What's the best way to visualize difference-in-differences for both binary and continuous treatment? Do I regress the outcome variable on the set of controls but exclude the treatment variable and ...
3
votes
3answers
1k views

Formula for one-sided Hodrick-Prescott filter

I am not very familiar with filters. The Hodrick-Prescott filter as one can find it e.g. in wikipedia is two-sided. I also found an R implementation for this in the R package mFilter. There the filter ...
4
votes
0answers
36 views

Who conjectured that every correlation is caused by causal mechanisms?

I remember reading about this conjecture in Causality (Pearl, 2000). It states that every dependency between random variables can be explained (or originates from) a purely causal model. Of course ...
13
votes
5answers
16k views

What do “endogeneity” and “exogeneity” mean substantively?

I understand that the basic definition of endogeneity is that $$ X'\epsilon=0 $$ is not satisfied, but what does this mean in a real world sense? I read the Wikipedia article, with the supply and ...
0
votes
0answers
17 views

How to show influence of a treatment group?

Instrumental variables regression is often used in economics to infer causal relationships. For instance, if we are interested in the effect of remittances (X) from abroad on home consumption (Y), we ...
3
votes
3answers
323 views

Transfer function in forecasting models - interpretation

I am occupied with ARIMA modelling augmented with exogenous variables for promotional modelling purposes and i have hard time explaining it to business users. In some cases software packages end up ...
2
votes
2answers
194 views

Causal inference from Cross sectional study

As far I know that, causal inference can make for longitudinal study. Is their anyway way to make causal inference from cross sectional study design? If yes, how can I do this? Please share if any ...
2
votes
0answers
23 views

How can I compute cross-correlation and auto-correlation in R using pooled data?

I'm trying to perform a lagged linear regression on time series data sourced from ~10,000 hospital patients, for the purpose of estimating causal relationships between administration of a drug and a ...
0
votes
0answers
22 views

Dealing with constraints in probabilistic graphical models

Suppose I have a graphical model with i.i.d. $\Lambda_i \sim exp(\lambda),\ \ i = 1,...,n$, and $\bf{\Lambda} = \sum_{i=1}^{n}\Lambda_i$. Imagine that that these $n$ lambda and this capital ...
0
votes
0answers
25 views

How to identify the underlying mechnism in regression (to preserve the causality)?

everyone,I recently encounter a problem when doing some empirical work. Say, with a cross section dataset,I'm studying the effect from A to Y, including a branch of control variables X,of which A is a ...
1
vote
1answer
32 views

How to do regression if one is not sure of dependent variable?

I have data on age, gender, height and weight of subjects, and also levels of 2 chemicals in blood of these subjects (chem1 and chem2). It is not clear whether chem1 affects chem2 or chem2 affects ...
1
vote
0answers
13 views

Isolating influence of sampling from actual change

Say I want to evaluate teams' batting coaches in a hypothetical baseball league. It's an unusual league in that there is no control over (and large fluctuation within) the number of at-bats each ...
9
votes
2answers
216 views

Why is using cross-sectional data to infer / predict longitudinal changes a Bad Thing?

I'm looking for a paper which I hope exists, but don't know if it does. It could be a set of case studies, and / or an argument from probability theory, about why using cross-sectional data to infer / ...
12
votes
1answer
206 views

What are the main differences between Granger's and Pearl's causality frameworks?

Recently, I ran across several papers and online resources that mention Granger causality. Brief browsing through the corresponding Wikipedia article left me with the impression that this term refers ...
0
votes
1answer
73 views

Should I use the t-test or the Wilcoxon rank sum test given these qqplots and Shapiro-Wilk stats [duplicate]

I have 4 groups and I want to test if the pairwise difference in means are significantly different. There are 6 pairwise differences. The QQnorm plots of the 4 groups look like this: The ...
3
votes
2answers
46 views

Names for some canonical directed causal graphs/illustrations of some canonical causal relationships?

Certain names are used for structures or node relationships that appear in acyclic, directed graphs (DAGs). Often these DAGs are interpreted causally. Here's a partial list for relationships that ...
1
vote
1answer
29 views

Establishing causality under conditions of certainty

I'm currently reading "Causality: Models, Reasoning, and Inference" by Judea Pearl. Early on, he states that the development assumes that there are no certain entailments, no 1 or 0 probabilities -- ...
2
votes
3answers
105 views

Can two linear regression variables be perfectly correlated but not share a single causal chain ancestor?

A causal chain lists event (or fact) $y$ with all its causal antecedents. We make a model of the following form: $$ y = \beta_0 + \beta_1x_1 + \beta_2x_2 + \epsilon $$ $\hat\beta_1$ has a p-value ...
1
vote
0answers
25 views

What are the best empirical studies comparing causal inference with experimental, quasi-experimental, and non-experimental techniques?

The Issue: People attempt to draw causal inferences using many different statistical techniques (e.g. regression, propensity score matching, regression discontinuity, instrumental variables, etc.). ...
0
votes
0answers
41 views

Necessary conditions for causality

E. Tufte writes: 'Probably the shortest true statement that can be made about causality and correlation is "Empirically observed covariation is a necessary but not sufficient condition for ...
2
votes
1answer
199 views

Does rejection of null hypothesis in multiple regression entail causation?

We make a model of the following form: $$ y = \beta_0 + \beta_1x_1 + \beta_2x_2 + \epsilon $$ and with $n=1,000$, $\hat\beta_1$ has a p-value <0.001. If our data and data collection meets ...
1
vote
0answers
15 views

Need a statistic for comparing “strength” of Markov blankets in a Bayesian network

Working with Bayesian networks. I take a given network structure and fit its parameters on data. I am looking for a statistic based on those parameter estimates that allows me to compare Markov ...
0
votes
0answers
73 views

Time domain regression - determining lagged predictors

Determining lifestyle factors affecting a medical symptom I have a dataset with n=200 records corresponding to contiguous days and consisting of 1 continuous output variable (a medical symptom) and ...
0
votes
0answers
24 views

Measuring effects with longitudinal data

Problem: I have sales data through time e.g. how much each user spent on each shopping trip. I am interested in certain events (think users switching to Amazon Prime for instance). I know the date ...