Abstract

CaSty is a Python-based software tool for calibrating stylized survey items with grouped or point-valued response categories. It addresses two related problems: the direct calibration of stylized items that may generate structural distortion, and the calibration of stylized questionnaire data using diary-based reference values. Version 2.4 revises the calculation procedures and data-checking routines and introduces the computation of the Structural Bias Component (SBC). The program is controlled through a plain-text input file and produces a structured Excel workbook and a log file. CaSty implements four calibration models, reports cumulative or retro-cumulative calibrated estimates, and provides graphical outputs, checked data, and diagnostics such as the Error Index and SBC. CaSty is intended for survey researchers who need a transparent and reproducible workflow for improving the quantitative interpretation of stylized survey measures. In this version, the associated EVS replication data (3.EVS_SBC.zip) has been updated to include the input files and corresponding outputs used in the latest empirical applications.