Episode 786 of the For Immediate Release podcast includes a segment on the role of Twitter for opinion measurement. The discussion generally pivots on the volume of traffic and whether an appropriate analysis is possible given the “echo chamber” effect.
This is a challenge we’ve considered in research projects at Full Duplex. Our Idle No More at Six Months report is our the first public (and free) report that includes a deep analysis of Twitter for opinion and issues. The argument is that by organizing participants into groups based on participation rates, we could isolate the echo chamber, the engaged public and the general public. Thus, we could more easily isolate biases and gain a better understanding of trends in public opinions, sub-issues, demographics and geographics.
While we could have gone completely crazy with our analysis, we chose to demonstrate the concept by organizing Twitter participants into three primary groups:
- Low – Those who issued 1-15 tweets during the six month period. This group generally represents those with a passing interest in Idle No More online chatter and accounts for a larger number of participants and, potentially, a better representation of public opinion.
- Medium – Those who issued 16-500 tweets during the six month period. This is the group of people and organizations with a greater interest in the movement and its activities (enough to issue a greater number and mix of tweets, retweets and replies) and features a number of journalists, media organizations, academics, analysts, political figures and engaged members of the public.
- High – Those who issued 501 or more tweets during the six month period. This is the echo chamber where committed points of view, both in favour and critical of the movement, are expressed with fervor.
You can read more in the report. It’s available as a free download. And, it includes some amazing graphics that summarize our analysis.
The upshot is, the true benefit in using Twitter for intelligence gathering and opinion measurement can only be realized by slicing and dicing the data to extract the signal from the noise. It’s a big job. It can be done. And the findings are worth the effort when it comes to campaign and counter-campaign efforts.