The Use of Scanner Data for Price Indices Calculation: the Case of Poland

In the context of the growing use of scanner data in inflation measurement, a key methodological challenge remains the choice of the appropriate price index formula and data filtering method. This article aims to investigate the impact of these choices on the results of price dynamics measurement. The paper analyses transactional data collected from point-of-sale terminals in Polish supermarkets for selected product groups: coffee and milk. Various price index formulas – including unweighted bilateral (Dutot, Carli, Jevons), weighted bilateral (Törnqvist, Fisher), and multilateral (Geary-Khamis, CCDI) – were compared, and the influence of different scanner data filtering methods was tested. The results show that both the choice of the price index formula and the applied scanner data filtering method have a measurable and significant impact on the assessment of price dynamics. These differences can lead to diverse interpretations of inflation, which underscores the need for conscious and justified methodological choices when using scanner data in price statistics.