Establishing proper guidelines requires statistical expertise
Identification of a societal problem
The increase in internet penetration and evolving technological possibilities have made online and/or self-administered survey data collection easily accessible. While these methods can reach an impressive sample size (i.e., a sufficient number of respondents), the reliability and generalizability of the data are significantly lower compared to probability sampling methods (such as face-to-face surveys). In Hungary, given the current internet coverage, online-collected samples lead to unreliable estimates in 80% of cases and introduce significant biases when generalizing the results to the entire Hungarian population. This is because the internet-accessible and internet-inaccessible segments of the population differ significantly in the majority of the investigated questions. In recent years, a substantial portion (40%) of social science research has relied on low-quality online-collected samples, and the majority (70%) of online news articles also report statistically weak quality data. This problem is not only scientific in nature but also relevant from a public perspective due to the increasing intensity of data generation, as misleading research results can occur without proper handling and interpretation of the data. In social sciences, adhering to appropriate data collection protocols and generalizing the findings to the entire population only under strict conditions are crucial for further development. While international literature extensively addresses the methodological foundations of innovative data collection techniques, similar protocols are currently unavailable in Hungary. However, establishing proper guidelines requires statistical expertise, data collection experience, and sufficiently detailed data on the internal demographic distribution of the Hungarian population. With our research, we aim to address this problem and fill this gap by taking the first steps towards establishing local protocols.
Expected Social/Economic/Environmental Impact of the Research
The research results will benefit both the academic and market sectors, indirectly leading to improvements for the general population in the long term. For the academic sector, the results will serve as tangible and tailored guidelines based on the demographic parameters of the Hungarian population, providing more precise considerations in research design and more reliable results. The results will also be valuable for the market sector. The data collection market currently performs poorly in terms of quality compared to international competitors, and data collection companies strongly feel the lack of specialized statistical guidelines and analyses adapted to the new environment.
Contact
Blanka Szeitl (Eötvös Loránd University, Faculty of Social Sciences)
E-mail