THE METHODOLOGY FOR SELECTING AND INTEGRATING DATA SOURCES AND USING OFFICIAL STATISTICAL ENTERPRISE DATA, QUESTIONNAIRES, AND PROXY INDICATORS IN FORMING THE EMPIRICAL BASIS OF THE STUDY

Abstract

This paper develops a methodology for constructing an empirical research database through the systematic
selection and integration of multiple data sources, including official statistics, firm-level data (accounting, managerial,
transactional, and registry records), survey data, and proxy indicators for hard-to-measure constructs. A practical set of
criteria for data source selection is proposed, covering representativeness, accuracy, coverage, temporal and spatial
comparability, data preparation costs, legal and ethical constraints, and reproducibility. The data integration workflow is
described step by step, encompassing identifier standardization and record linkage, harmonization of measurement units
and classifications, panel data construction, metadata management, and version control. In addition, the study outlines
key data quality checks, such as missing data diagnostics, outlier detection, internal consistency analysis, selection bias,
and linkage bias, alongside documentation standards required for transparent and robust econometric research. The
Results section presents illustrative tables and conceptual figures, including a data source matrix, an integration map, a
proxy indicator catalogue, and a data quality scorecard. The paper concludes with practical recommendations for building
reliable, auditable, and replicable empirical databases in applied economic research.