Toward Real-Time Infoveillance of Twitter Health Messages

• DOI: 10.2105/ajph.2018.304497
• Publication Date: 2018-06-21T20:02:52Z
• Authors: Jason B. Colditz; Kar-Hai Chu; Sherry L. Emery; Chandler R. Larkin; A. Everette James; Joel Welling; Brian A. Primack
• Keywords: 03 medical and health sciences; Data Collection; Humans; Public Health Surveillance; Health Promotion; 0305 other medical science; Social Media; United States; 3. Good health

There is growing interest in conducting public health research using data from social media. In particular, Twitter "infoveillance" has demonstrated utility across contexts. However, rigorous and reproducible methodologies for are not yet well articulated, particularly those related to content analysis, which a highly popular approach. 2014, we gathered an interdisciplinary team of science researchers, computer scientists, methodologists begin implementing open-source framework real-time infoveillance messages (RITHM). Through this process, documented common challenges novel solutions inform future work collection subsequent human coding. The RITHM allows researchers practitioners use well-planned processes retrieving, storing, filtering, subsampling, formatting topics interest. Further considerations coding include coder selection training, representation, codebook development refinement, monitoring accuracy productivity. We illustrate methodological through practical examples formative hookah tobacco smoking, reference essential methods literature understanding data.