survey-methodology

Repository for a small package for joint calibration of totals, quantiles and other metrics

R/tidyverse pipeline for attributing single, holistic satisfaction ratings across multiple public services — solving the 100-service survey dilemma while staying compliant with Indonesia's Permenpan RB 14/2017. Presented as a Lightning Talk at useR! 2026 (Virtual Presentation Room).

This package tells you how burdensome your survey is. It works primarily with Qualtrics (.qsf export or API access), but can take your own question catalogue. Burden is scored using the GfS+ points scheme.

社会统计学知识体系｜MSc 社会研究方法与统计学 原创整理：研究设计→数据→分析方法(回归/因果推断/SEM/复杂抽样)→估计→解读，配 R & Stata 双轨实操与真实案例。Bilingual social-statistics knowledge base.

Dry-run-first Python toolkit for standardized scientific telephone surveys with a bilingual local workbench.

Survey methodology project evaluating ASHA response behavior, postcard reminder effects, and postal mail vs. web survey mode differences using R, bootstrap inference, and logistic regression.

This repository contains the data, code, and supplementary material for the paper "Vox Populi, Vox AI? Using Language Models to Estimate German Public Opinion" by Leah von der Heyde, Anna-Carolina Haensch, and Alexander Wenz, available at https://doi.org/10.1177/08944393251337014

End-to-end pipeline for collecting, labeling, and analyzing metaphor framing and stance in Reddit and news discourse using LLMs.

LLM-vs-digital-trace sentiment analysis project comparing electric vehicle public opinion across Reddit, news, and model-generated sentiment; selected for AAPOR 2025.

Personal website showcasing my work in survey methodology, machine learning, and computational social science.

Statistically defensible stratified sampling for an employee engagement survey: sample-size calc (95% CI + finite-population correction), proportional allocation, and representativeness validation in Python.

Market survey project on EV charging-service satisfaction in Suzhou.

Random digit dial sampling design for remote household surveys in Nepal, with stratified allocation by reporting domain

Anchoring vignettes applied to language models for the first time. Placing reference exemplars before a self-assessment transfers their rating onto it; reversing the turn order removes the effect. 5 models, 900 observations, execution-verified ground truth, all thresholds preregistered.

Preprint: Synthetic respondents in survey research — an empirical test of LLM 'digital twins' against a student survey (N = 1,795). Russian original + English version.

This repository contains results and reproducible code from the Mixed-data Stratified Sampling work conducted by Borros, Er and Salau. Applied results can be explored on our Shiny App hosted here.

autumn: Fast, Modern, and Tidy-Friendly Iterative Raking in R.

Analysis of the accuracy of U.S. presidential election polls (2008–2024) based on prediction error data. Compares ideal vs. practical polling methodology, presents three contrasting visualizations of the same dataset to illustrate framing effects, and identifies selection/measurement biases with recommendations to improve accuracy.

Replication code for Gieng (2026), "Measurement Is Not Passive: Identification of Treatment Effects When Outcomes Are Elicited Rather Than Observed."

Framework for testing risks and biases in silicon sampling as it relates to disability and hard-to-reach populations in survey research.