iSentia New Zealand Election Index

Both Whale Oil and Homepaddock have blogged about the iSentia Media New Zealand Election Index. I suspect others have as well, but that is the only place I have seen it so far. It looks at both mainstream media and social media coverage. I will ignore the mainstream media part. There are people better qualified to talk about that part of it.

First lets look at the graphic:

Let’s firstly assume that the figures are correct, what do they tell us? For the top three the figures are roughly in proportion to the number of followers each account has on Twitter. So the proportion of the mentions seems about right. But this doesn’t tell us if the mentions are active or passive, positive or negative. By active I mean things like retweeting something from the account, or someone actively mentioning the account as part of say asking a question. By passive I mean the account getting tagged in a bunch of tweets that are part of a conversation that the account is only a side part of. There is also no mention of positive/negative sentiment. I don’t expect something like that to be done on a survey like this, but it should serve as a note of caution when assessing this information.

Howver, I am not sure how they are measuring mentions, but some of the numbers seem rather low. I just had a look at Metiria Turei’s mentions in Tweet deck and counting back from around 1030 today to when my paper.il tweet, that mentions her, came out at 0845 yesterday, there were at least 110 mentions of her. That did include retweets, which they may not be counting. But even if we say that 1/3 were retweets and don’t count them, that is still 74 mentions in around 24 hours. Times that by 7 and you have 518 Which is markedly different to the number that iSentia have given her. As well, Metiria has a similar number of followers to Russel, so why are their numbers so different? So if these aspects are out, what else is out? This assessment also totally leaves out any mention of Facebook, Instagram, Youtube, which draws the figures into doubt even more.

I plan on trying to contact iSentia on Monday to see if they will provide more detail how they measure their social media interactions. These sorts of metrics, which take a lot of time and effort to compile, offer the potential to tell us a lot about what is happening on social media, but without knowing exactly what is being measured, there is no use to them.