In the 2nd half of 2016, when reviewing research papers on marketing applications of social media at a US university, I raised the need for validity checks on social media accounts.

The researchers were using test populations, semi-randomly selected within certain interest groups among Facebook accounts (by brokers).

I noted that persons could have multiple accounts, could easily give false information about themselves (for different reasons, incl. with criminal intent), and that the social media have very limited tools to verify the fidelity of the accounts (in other words, they are overselling the value of their accounts for e.g. commercial use). For example, I could easily create a 2nd account calling myself Napoleon or presenting myself as a member of an (not well known) Hells Angels branch.

Note:     the recent revelation of Senator Mitt Romney’s secret Twitter account "Pierre Delecto" is just a, relatively innocent, demonstration of this.

The discussion following resulted in two things:

  • The agreement that the research papers would
    • include a disclaimer stating the limited possibilities for validity checks on social media accounts
    • use of populations, where suitable, selected from an environment where social control of the members would likely add a level of validity checks (e.g. selecting accounts of persons belonging to a group with likely enough social contacts)
    • We started an effort to model the discrepancy between the visibility that a social network has on a person’s information and the actual information (this work is ongoing, now including sociological expertise)

About 2 months thereafter, the US elections took place, and the first rumours of the possible role of social media in the election result surfaced. Next, we found ourselves flooded by reactions from people that realized that in the process we had ‘uncovered the dangers’ before the election.

To set the record straight, we did not uncover something others had not seen. Instead, we simply proposed some measures for application in a limited domain - research papers on the marketing use of social media - and initiated work to model the modalities of visibility and validity checks.

Some of us had been worried about misinterpretation and misuse of social media, but we also did not realize the impact that this could have on an election in a country like the US.

Since, the role and possible misuse, abuse of and collaboration by social media have been increasingly important items for news and study; this includes the use of social media to selectively broadcast to large audiences, and in the process bypass legislation applied to broadcasters and press.

Contribution uses work by Lanting, Lokshina* and Thomas* (* SUNY Oneonta, NY)