I've spent a lot of time but still could not find a quality yet public/free data for causal inference (binary treatment; e.g., A=0 or 1, non-DCD vs DCD donations) with survival data. I know it's quite specific requirement, but need one for my master degree thesis.. Of course, one might recommend the one from 'survival' package in R.
But I really want to find good, real (if possible, not too old) data.
For example, the one from this journal looks perfect (but cannot access) Article Kidneys from Donors after Cardiac Death Provide Survival Benefit
.Can I get some advice or recommendation for such data?
Any comment is appreciated !
What is your research work on scientific research? How can I help?
Based on chapter 2 and 9 of the book called "Applied linear statistical Model" (John Netter), the regression analysis with observed data couldn't give information about cause-and-effect problems. However, if a regression analysis is performed with experimental data (experimental study), then the regression can give information about cause-and-effect relations. In such cases, the effect of latent variables on the response variable reaches its minimum due to randomization in the experimental study and the causality effect can be examined.
The survival analysis is the special kind of regression analysis and if the study is experimental, the causality inference could be analyzed.
There are sample datasets based on the Framingham Heart Disease Study for teaching purposes that have a number of covariates and the presence or absence of coronary heart disease (CHD) after ten years as the outcome.
A simple one is here:
https://www.kaggle.com/amanajmera1/framingham-heart-study-dataset
To obtain the full dataset, you would need to contact the researchers. They have an application process for doing so which is located here:
https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs000128.v3.p3
Mehdi Azizmohammad Looha Mehdi Azizmohammad Looha Thank you for your advice. I did not add on my post but actually adjustment of confounders(and making causal inference with observational data) is one of my key ideas. Thank you your detailed comments.
Philip Cochetti Thank you so much! Now I'm reading the description of the data and think it can be one of my datasets for application. :)
- Badges
- Science topic