Analysis of the Heterogeneity of Apartments and Its Impact on Rental Prices in Prague Using Web Scraping and Hedonic Regression Model

Adam Slavíček, Stanislava Hronová, Richard Hindls
Statistika, 105(4): 433–447
https://doi.org/10.54694/stat.2025.45

Abstract
The study analyzes the heterogeneity of the housing stock and its impact on rent levels in Prague. Using web scraping and hedonic log-log regression on data from the SReality portal (20 184 apartments offered), it quantifies the impact of layout, size, and amenities (elevator, terrace, garage, facilities, new construction) on the price per square meter and examines the spatial and temporal variability of these factors. It compares classic indices (Laspeyres, Paasche, Fisher) with their hedonic variants to assess whether changes in the structure of supply distort aggregate statistics. The results show significant spatial heterogeneity of parameters, but only slight differences between traditional and hedonic indices in short-term monitoring.

Keywords
Rent prices, hedonic regression, heterogeneity of housing stock, web scraping