TStat - TStat Training

Day 2 will be dedicated to our conference workshop Automating Your Research in Stata. A unique opportunity for doctoral students & early-career researchers looking to strengthen their Stata programming skills. Workshop leader: Giovanni Cerulli (IRCrES-CNR, Rome)

The conference will feature presentations and discussions showcasing new #Stata commands, applications, and research. Together with Tips and Tricks to elevate your Stata experience. A space to exchange ideas and strengthen collaboration across the Stata community.

We are delighted to announce the first Greek Stata User Meeting 🇬🇷 📍Athens | 🗓27–28 January 2027 at the AUEB. The conferences represents a new forum for Greek Stata users across academia, research & government to share research and ideas. Info at bit.ly/4w72fos

to implement hierarchical clustering using TF-IDF vectorization & cosine distance. Allowing users to identify patterns on data, uncover latent themes & organize information for macroeconomic forecasting, sentiment analysis, or real-time policy monitoring. bit.ly/42bAyxD

Combining native Stata capabilities with Python-enhanced workflows, Carlo Drago offers an accessible and policy-relevant toolkit for unsupervised text classification in multiple domains. This applied text mining & clustering framework enables users to classify & interpret text

Loreta Isaraj illustrating automated data extraction from unstructured textual sources using a scalable fully integrated workflow employing Large Language Models (LLMs), specifically @ChatGPT 4.0 via @API, in conjunction with @Python & @Stata. Details @ bit.ly/481VM4C

Covariate selection is performed through a Lasso-based local estimation, ensuring valid inference under approximate sparsity. The command generates output variables that can be used for further post-estimation analysis within the same session. More @ bit.ly/3KdQY2n

Marianna Nitti currently illustrating her NEW @stata command rdlasso, which implements Regression Discontinuity Designs with high-dimensional covariates in @Stata. The procedure extends the methodology developed by Kreiss and Rothe (2023) to both sharp and fuzzy designs.

The new @Stata command outdetect by Giulia Mancini et al., offers researchers faced with extreme values in #surveydata a simple univariate detection procedure, with useful diagnostic tools for #sensitivityanalysis, for outlier detection. Detailed info @ bit.ly/48EXJEe

bootstrap sampling, and optimized #Mata routines to compute the co-association matrix, ensuring computational efficiency.. Download Carlo’s presentation @ bit.ly/4gJO8OE

co-association matrix, which allows @stata users to address the possible inherent instability of partitioning-based #clustering by aggregating results from multiple #bootstrap samples, improving robustness and reproducibility. Carlo uses a combination of existing Stata routines,

Are you trying to group data into sets of similar observations without relying on predefined labels? Then Carlo Drago’s work on consensus clustering in @Stata 19 is likely to be of interest. Carlo’s approach combines bootstrapped k-means with hierarchical clustering based on a

@StataCorp’s Gaby Ortiz currently putting the spotlight on @Stata 19’s #metanalysis capabilities. An excellent prep for tomorrow’s Introduction to #Bayesian #MetaAnalysis course! Presentation available online at bit.ly/4nOfzcf

Maria Elena Bontempo illustrating how her NEW flexible panel data command pantarei extends the existing commands xtsum & loneway to consider, amongst other things, the role of CCE inside the within variability. Presentation at bit.ly/4gGHgBt

Working on #EventfHistory (#SurvivalAnalysis) in @Stata? Then David Bussi’s splitting command, stsplit, will facilitate the breaking of your time axis into episodes in order to include time-varying covariates in your analysis. Command details at bit.ly/4pDmZRu

xtcorri calculates unit specific correlations or covariances. More details @ bit.ly/4nwFFRi- Commands available on request!

Jan Ditzen's panel data xttools suite of commands opening our Tips & Tricks session this morning: xtpcaget obtains principal components from multiple variables; xtplot2 identifies panel structure for variables

& xtcorri calculates unit specific correlations or covariances. More details @ bit.ly/4nwFFRi - Commands available on request from Jan!

Testing for or estimating #structuralbreaks in #timesseries or #paneldata? Then check out Jan Ditzen et al. NEW @stata command xtbreak to detect for breaks, determine their number and location, and provide break date confidence intervals! Presentation at bit.ly/4pDkXRg!

Working on #CausalInference & #TreatmentEffects in @stata and need to estimate Conditional Average Treatment Effects (#CATEs)? Then Giovanni Cerulli's @cerullig discussion of Stata 19's NEW CATE command offers an indispensible step by step guide!! More at bit.ly/48Ewxp8