Traditional sample surveys, emphasizing large sample sizes and robust finite population estimates through probability sampling designs, have been a staple in official statistics. However, modernization efforts are underway as statistical agencies explore the integration of big data and web panel information for near real-time estimations. This paper reviews the challenges associated with these endeavours, particularly addressing statistical biases linked to under-coverage in big data and errors in data variables. Kim and

Tam (2020) introduced data integration methods, treating big data as an incomplete sampling frame and utilizing calibration weighting. The paper systematically reviews various integration methods, including mass imputation, propensity score, and calibration weighting. Additionally this paper concludes with a simulation study evaluating the performance of Kim and Tam's (2020) proposed estimators, assessing Bias, Standard Error (SE), and Root Mean Square Error (RMSE). The findings contribute to the on-going discourse on modernizing survey methodologies and leveraging diverse data sources for more efficient and timely official statistics.

Highlights

• The findings contribute to the on-going discourse on modernizing survey methodologies and leveraging diverse data sources for more efficient and timely official statistics.