Evaluating the representativeness of mobile big data: A comparative analysis between China's mobile big data and census data at the county level

Mobile big data has emerged as an essential tool for various scientific research fields. However, the credibility of mobile big data and the extent to which it can represent the real-world population remain unclear. This study evaluated the representativeness of mobile big data by comparing it to the most recent census data at the county level in China. Using power-law and multiple linear regression models, we aim to determine the accuracy and reliability of mobile big data in reflecting the population dynamics and characteristics of different geographical areas. Our results indicate that disparities among individuals with different socioeconomic statuses, demographic characteristics, or geographic locations may contribute to biased estimations of the actual population density. Higher illiteracy rates and median ages may be associated with underestimating population density. In contrast, higher GDP per capita, elevated urbanization levels, and larger percentages of the 15-64 year age group may be associated with overestimating population density. Our research highlights the importance of cross-validating population estimates and offering practical statistical methods for addressing potential biases and estimating population dynamics in future applications of mobile big data.