Questions tagged [matching]
Matching refers to a process in experimental design in which observations are sampled in a systematic, non-random fashion to be analyzed more efficiently with special statistical methods.
599 questions
0
votes
1
answer
23
views
What are the differences between PS-match and adjusted Cox regression?
This is more like an extension of the following question: Propensity Score Matching with Cox Regression
I am wondering what are the differences between these:
matching patients with PS and running ...
0
votes
1
answer
25
views
Primer on Entropy Balancing
I've scoured the internet, but I can't for the life of me, find a comprehensive primer on entropy balancing.
I am currently in the process of cleaning data in order to create weights for the part of ...
0
votes
1
answer
34
views
covariate balance loveplot and distance
What is the distance parameter that is plotted on the love.plot ? I am not able to find any reference or explanation for this in the cobalt package. Any help understanding this value "distance&...
1
vote
1
answer
42
views
MatchThem and Mixed-Effects Logistic Regression in R
MatchThem and Logistic Regression in R
I am conducting a non-randomized study to evaluate the differences in complication rates between two treatments. The observed complication rates are ...
0
votes
0
answers
16
views
Inconsistent Matching Results with Matchit After Excluding One Patient
I used the R package Matchit for 1:1 nearest neighbor matching without replacement in my study. Here’s the code I used:
R code:
...
0
votes
0
answers
12
views
Matching control group and treatmeant group period in staggered difference-in-differences
I am investigating how different types of electoral systems systems Proportional Representation (PR) or Majoritarian System (MS). influence the level of clientelism in a country. I want to investigate ...
1
vote
0
answers
51
views
What should it be if the treatment group is larger than control groups in Propensity-Score-Matching (PSM) -DiD
In an effort to do Difference-in-Differences (DiD), to mimic the Randomized Controlled Trials, we normally use the Propensity Score Matching (PSM) to eliminate the difference between control and ...
3
votes
2
answers
72
views
Limitations of propensity score matching
While studying propensity score matching, I was struck by the following thought:
When we are running a logistic regression model to estimate $p(Z=1∣X)$ through some form of parametrization and we are ...
1
vote
0
answers
49
views
Statistical analyses after propensity score matching with matchit - Survival analysis- adjustment?
For an analysis of a data set I applied propensity score matching (matchit in R, 5:1, without replacement), which went very well- I dod not lose too much of the sample and got a balanced data set out ...
0
votes
1
answer
47
views
MatchIt Marginal Effects
I am trying to estimate the ATT of a treatment with MatchIt and am unsure if it is a marginal effect or a conditional effect. In the vignette for estimating the outcome, it says:
In addition, with ...
0
votes
0
answers
62
views
How to determine the treatment time for untreated in difference-in-differences design
I'm estimating the effect of treatment initiation on health outcomes using a Difference-in-Differences (DiD) design. The timeline is defined relative to the treatment date, not calendar time. For the ...
0
votes
0
answers
36
views
Coarsened exact matching (CEM) on longitudinal data
I want to compare the effect of 2 educational programs across time. In other words, I have tested 2 groups of students at 3 different time points (T1, T2 and T3). I have performed CEM on the data of ...
0
votes
0
answers
20
views
Propensity score matching - How to select observation for matching among repeated measures
I am unsure how to resolve the following problem and am hoping someone might have a suggestion.
We have repeated measures data for MS patients and we want to look at potential differences in an ...
0
votes
0
answers
23
views
(Propensity Score) Matching for multiple outcome variables
I have multiple outcomes (dependent variables) of interest. Before estimating the effect of receiving treatment on these multiple outcomes, I want to use propensity score matching as my control and ...
3
votes
1
answer
88
views
How to adress potential issues with interpretation/generalizability after discarding observations in exact matching?
Observations are discarded if they have no exact match, which potentially leads to a different distribution of the covariates in the matched sample compared to the original sample.
I assume that at ...
1
vote
1
answer
56
views
Implications of using a caliper when matching: Would it ruin ATE claim?
I've started to learn about matching methods using MatchIt package, and read "Choosing the causal estimand for propensity score analysis of observational studies" (Greifer & Stuart, 2021)...
1
vote
1
answer
56
views
Mahalanobis distance calculation in MatchIt function
I've started to use MatchIt package recently and it's great! I'm also learning a lot by reading the documentation associated with the package. My question is about calculation of Mahalanobis distance. ...
1
vote
1
answer
52
views
Applying weights instead of matching
I want to conduct a Difference-in-Differences (DiD) design study. I have two groups (which you can see in the sector_eu_without variable) that I want to keep equal regarding certain variables: sex, ...
2
votes
2
answers
93
views
Estimating Individual-Level Causal Effects Under Selection on Observables
The value of a randomized controlled trial is that the random assignment of treatment ensures that no confounding is biasing the relationship between treatment and outcome and, therefore, a simple ...
1
vote
1
answer
40
views
Which kind of matching can i use and why doesnt my propensity score matching work?
I am currently trying to find out if a certain disease characteristic affects long term survival. I have observational data and i am trying to match my cohort, as there are huge baseline differences ...
0
votes
2
answers
140
views
What are the possible solutions for performing full matching on a dataset with a size of 61,000? [closed]
I am working with a dataset of approximately 61,000 observations (with 27,686 treatment and 32,405 controls) and need to perform full matching using the MatchIt ...
0
votes
1
answer
141
views
Use SMD or raw difference in proportions when comparing balance of binary covariates used in propensity matching?
I am a junior in the medical field, and have been tasked with performing a propensity matched analysis of some data, using R.
I am looking at comparing the balance of my covariates (both before and ...
0
votes
0
answers
18
views
Matching on actual earnings versus matching on the kind of unexpectedness in earnings
For simplicity, lets assume this is a question about linear modeling, although I am actually looking at some non-linear models and am willing to consider other models if they would be more ...
1
vote
1
answer
62
views
How to compare cases and controls after adjusting for age differences?
I have a dataset in which I have different measures (a, b, c) and two independent variables: D which is an independent binary variable representing cases (1) and controls (0), and age is another ...
2
votes
1
answer
47
views
General question about exact matching on variable/s
I haven't used matching a lot but have certainly read many papers (some quite old and obviously before propensity scores were used much) that have talked about 1:1 matching on age, sex, etc variables ...
0
votes
0
answers
14
views
How to deal with exposure status change and outcome contribution in different exposure status?
I want to estimate the risk of cardiovascular events in patients with diabetes (exposed group) compared to patients without diabetes (unexposed group). How do we deal (in the analysis) with patients ...
1
vote
1
answer
23
views
In Matching, are capped confounders a legitimate method of improving balance for matching with highly skewed continuous confounders?
In matching (or similar confounder-control methods such as weighting), are "capped metrics" a "legitimate" method of improving balance for highly-skewed continuous confounders?
...
1
vote
1
answer
66
views
After matching: How do I interpret the value of the type ‘distance’ (=Propensity score) in the balance measures table of the r-package cobalt bal.tab?
I have used the R-package ‘MatchIt’ to perform (1) a nearest neighbour propensity score matching (NNM) based on the Framingham Heart Study and (2) for comparison, an optimal PS matching (OM) for the ...
1
vote
0
answers
53
views
When does Matching result in an ATE vs. ATT on observational causal studies?
I have read that matching nearly always yields the $ATT$ effect, but that subclassification matching can yield the $ATE$. I am therefore wondering what is a heuristic for determining what kinds of ...
1
vote
1
answer
272
views
In propensity score matching, is the estimand to be estimated the ATE or the ATT?
In the propensity score matching literature (Central Role of the Propensity Score by Rubin), the treatment effect estimand is referred to as the "Average Treatment Effect" (ATE). However, in ...
0
votes
2
answers
63
views
Propensity score matching with near-perfect differentiation
I am working on an analytics task for my job to match control units to treated units to monitor the effectiveness of a new marketing initiative. We decided to use a propensity score matching method ...
2
votes
1
answer
42
views
How to compare initial means between weighted cases?
I want to check if the mean value of a certain personality trait differs between young adults in organization X and young adults outside of the organization, who are identical in terms of education ...
1
vote
1
answer
88
views
Is including weights in g-computation not the same as a plug-in doubly robust estimator?
In the R package vignette for WeightIt(), in the section "Modeling the Outcome", it explains that (assuming I'm reading correctly) that the purpose of applying g-computation after creating ...
3
votes
1
answer
81
views
What are downsides to "genetic matching," particularly outside of causal inference settings?
Multivariate matching methods typically involve two steps. First the user computes $D$, a matrix of the multivariate distances between units. Second, the user applies a matching function (e.g., 1:1 ...
1
vote
0
answers
108
views
Assessing whether the probability of being assigned treatment is equal (or reasonably close) between two individuals/groups [closed]
I'm currently studying the textbook Design of Observational Studies, second edition, by Rosenbaum. Chapter 3 Two Simple Models for Observational Studies says the following:
3.1 The Population Before ...
3
votes
1
answer
165
views
What are the main pros and cons of the most commonly used weighting methods?
There are many methods to generate balancing weights in observational studies (see, for example, the many methods implemented in the amazing WeighIt package).
I have seen some great discussions about ...
0
votes
0
answers
49
views
When to use paired samples tests? Specific case in clinical research
I am working on the analysis of an observational clinical study. It is basically a control/intervention study, where the intervention group received a novel protocol. In order to assess its ...
0
votes
1
answer
82
views
Propensity score matching R swap control and treatment
I am analyzing observational data and want to perform propensity score matching. I would like to compare males and females matched for age,smoking status, bmi, etc. When I coded the male gender as 0 ...
1
vote
0
answers
22
views
How to construct a common control group in two cohorts with different disease outcomes?
Question details:
I have cohort data with a total of 100,000 participants. For some reason, I need to select 2000 participants as controls to construct a control group. I have 3,000 cases of disease A ...
1
vote
0
answers
21
views
How to Evaluate Interaction Effects in Propensity Score-Matched Samples:
Suppose I want to study the association between an exposure X and outcome Y, and I have used the propensity score to match each exposed subject with those unexposed but with similar characteristics ...
1
vote
0
answers
23
views
Matching numerical dataset to closest set in library
I have numerical, 1D datasets from measurements that I each want to match to the closest dataset in a library containing similar datasets. Coding is done in Matlab. Unfortunately I am not too ...
1
vote
1
answer
52
views
Average treatment effect (ATE) estimation via matching method while outcomes of control population are constant
I want to estimate the average effect of a treatment that was given with a selection bias. To do this, I'd like to use a matching method. Basically, this method involves finding, for each treated ...
0
votes
0
answers
62
views
Propensity score matching - within subsample analysis
I am trying to understand the impact of a graduate degree in mathematics on the wages of students compared to other degrees. I have been running a probability score matching model using a logarithm ...
0
votes
1
answer
42
views
Do I have to use the exact same variables in each step, if I have a two-step propensity score match followed by a regression?
I am using the propensity score match to grow my sample based on a smaller dataset of existing units that received the treatment. The match will find more likely units from its large population that ...
2
votes
1
answer
111
views
How to obtain the (weighted or unweighted) L1 imbalance measure for "raw" data
I have two questions regarding the JASA paper, Multivariate Matching Methods That Are Monotonic Imbalance Bounding, by Iacus et al. (2011), the authors produce Figure 2 (left panel) where they compare ...
1
vote
0
answers
48
views
Is it advised to do treatment-covariate interactions to estimate the average treatment effect when using exact matching?
I am using method = “exact” from the MatchIt package. In the vignette of MatchIt, it is ...
5
votes
1
answer
38
views
Should I use independent- or propensity score matching?
I work on data form an observational study (n=34) and want to compare the outcome of the individuals to an historic cohort (n=600). Subjects in the observational study received additional treatment ...
1
vote
0
answers
36
views
estimating effect with marginaleffect package
I want to estimate ATE. first of all I used MatchIt package for full matching for propensity score and then I used logistic regression with all of variable in propensity score model after that I used ...
1
vote
0
answers
64
views
Estimating treatment effect with G computation
I used the following code to obtain ATE. Before using the 'marginal effect' package, I examined the coefficients of the model and noticed that the 'symptom' variable has a high standard deviation. Is ...
0
votes
0
answers
29
views
When to calculate Outcome value in Propensity Score Matching with staggered adoption?
Sorry if this is a basic question (I'm no statistician): In a propensity score matching study, WHEN do you calculate the outcome values (for each treated and control unit) needed to compute the ...