This repository contains code for a research project examining the division of parenting labor among Taiwanese physician families.
It demonstrates the data cleaning and construction processes used to build analysis-ready datasets from population-level administrative data.
-
00_1_collect_all_doc_id.R
Extracts all valid physician identifiers from outpatient visit datasets across medical systems. -
00_2_build_doc_child_relation.R
Identifies children whose parents are both physicians and constructs the parent–child linkage table.
Also produces annual summaries of doctor-parent births. -
01_1_build_doc_child_opdte.R
Loads outpatient visit records for doctor-parent children and compiles yearly visit lists. -
01_2_build_doc_child_ipdte.R
Harmonizes inpatient records for doctor-parent children and expands hospitalization episodes into daily-level panels. -
03_1_build_doc_diary.R
Constructs a daily-level panel for physician parents by integrating outpatient, inpatient, birth, marriage, and child illness records.
Adds demographic attributes and event indicators related to caregiving and work activity. -
04_construct_analysisData.R
Builds the main analysis dataset by expanding events into ±30-day windows, defining treatment and control groups, and preparing a clean, analysis-ready table.
Each script is modular and should be executed in order to reproduce the full workflow—from raw data extraction to the creation of analysis-ready datasets.