Live Test 1 Complete

Testing of The Echo Chamber Project, New Zealand 2017 General Election is complete.  The findings are as follows.

Workload: For the test, less than 2 hours per day to retweet and prepare daily bulletins.  The volume of tweets was light, given that parliament was transitioning into Christmas break.  The workload is likely to increase substantially during election season.

Technical: No problems were encountered in using Twitter, TweetDeck, or WordPress.  The volume of retweets did not lead to the accounts being flagged as suspicious, but this may change during election season.

Echo chambers: The Twitter accounts for each echo chamber appear to be the partisan media bubbles that they were intended to be, at least for those parties that were able to generate a sufficient quantity of tweets.

Retweeting: It seems to make sense only to retweet things that I think will affect voting decisions, but this means the echo chambers will be biased by what I consider to be important in making voting decisions.  The option remains to simply retweet everything, but this seems likely to dilute the focus of the echo chambers, not improve their quality.

Bulletins: Even more so than the retweeting, the content of the bulletins are heavily curated by me, and thus the most heavily influenced by my bias.  They are easier for readers to digest than the torrent of tweets in the echo chambers, but the bulletins unmistakably contain my perspective.

The test was a success.  It is anticipated that The Echo Chamber Project will be dormant from now until the announcement of the New Zealand 2017 General Election date, when the project will become active.


Beginning Live Test 1

The Twitter accounts for the New Zealand 2017 General Election echo chambers are set up.  Live testing will begin now, and go on for up to 3 days.

The aims of this test include:

  • Assessing which tweets should be retweeted into each echo chamber, if not all.
  • Considering if anyone aside from the seeds should be followed by each echo chamber, if this were during the election campaign.  It is unlikely that anyone new will be followed for this test.
  • Assessing the workload of maintaining the echo chambers.
  • Seeing if the activity required for maintaining the echo chambers cause the accounts to be flagged as bots by Twitter.
  • Creating an example of what the echo chambers will look like when active, which will be publicly visible for anyone who visits the corresponding Twitter pages.

All NZ2017 echo chambers will be active for the duration of this test, including:

Feedback and suggestions are welcome via the contact page.