The use of social media to measure public opinion faces major unresolved challenges. While potentially promising for trend-spotting, modeling or qualitative analysis, current research indicates that social media is not appropriate for quantitative analysis of broader public attitudes, particularly when the research aim is to estimate population values. …
Difficulties in reliably using social media to discern broader public attitudes include the biasing nature of self-selection in participation and subject-specific use of social media; challenges in determining the meaning of often cryptic social media comments; sampling problems relating to keyword selection and variance in the volume of postings; the reported presence of fabricated, orchestrated and/or automated postings; and the lack of adequate and reliable demographic or other attitudinal data with which to produce meaningful analysis. [cont. – PDF]
Langer Research Associates