Development and Validation of the Douyin Algorithm Filtering Bubble Susceptibility Scale

Algorithmic recommendation has become the primary means of information delivery in today's social media landscape. Prolonged exposure to algorithms can lead users to develop cognitive biases, algorithmic dependency, and content homogeneity, potentially contributing to the "filter bubble" effect. When I.E. Pariser proposed the filter bubble effect, it was merely a hypothesis and not empirically tested. Whether filter bubbles exist and to what extent have become a hot topic in communication research. This study aimed to develop a robust and effective scale to measure the "filter bubble" effect of algorithmic recommendation on the Douyin platform. Through extensive literature research and questionnaire surveys, this study refined the "Information Cocoon Effect: A Three-Level Dependent Variable Measurement Scale" proposed by Yu Xin's team at Tsinghua University. Based on social cognitive theory, the scale proposes three dimensions: content homogeneity, negative interference, and cognitive bias, totaling nine items. Measurements revealed that the three factors of the scale exhibited good reliability and validity, and exhibited a strong exploratory factor clustering effect. Factor loadings were all greater than 0.7, the cumulative variance explained after rotation was 73.440%, the average variance extracted (AVE) for each factor exceeded 0.5, and the composite reliability (CR) exceeded 0.7. The structural equation was consistent with the initial hypotheses. The results show that the scale has good reliability and validity, and can effectively measure the existence and extent of the filter bubble effect, providing a valuable tool for empirical research on filter bubbles.