Tuesday, August 19, 2008

UK Kids Make Up Nearly A Third of Stardoll's Userbase

Statistics are a funny thing - pretty useful for supporting claims or for justifying a particular subject/site of study, but also immensely malleable and oftentimes downright ambiguous in terms of how they can be mobilized, interpreted, reinterpreted, etc...depending on who is reading them, how they are read and categorized, and how they are framed within the surrounding discussion. As a qualitative researcher who focuses mainly on case studies, ethnographic research and design analysis, I have a pretty casual relationship with stats - I use them to support arguments, but don't generate any of my own. I'm also deeply suspicious of statistics, thanks in part to a wonderful stats course I took in my final year as an undergrad at the University of Ottawa, which focused as much on how stats can be misused and abused as it did on reading and analyzing them.

I was reminded of this when I read the recent batch of internet use stats published by Nielsen Netratings UK, which describes its findings as indicative of key differences in the internet preferences of UK kids vs. teens vs. adults. Their new study provides a top ten list for three age groups - kids under 12 years, teens aged 12 to 17, and then adults. The sites included in these top ten lists are ranked based on what percentage of the site's total audience belongs to each group (you can check out the complete listings over at Marketing Charts). This is a somewhat unique way of measuring "popularity" - for instance, some of the sites ranked low on the list actually attract a larger number of individual kids than the top ranked sites. From these stats, Nielsen Netratings concludes (as outlined in the press release as well as in media coverage of the report):
Entertainment sites have the greatest affinity with under 12s; for 12-17 year-olds it’s games sites and for 18-22 year-olds it’s student and video sites.

However, anyone familiar with these sites - particularly the kids' (under twelves) sites - will notice that the sites have been categorized in inconsistent or odd ways. For example, here's the list for the under twelves...notice how many of them are "games" sites, but aren't classified as such:

Websites with the greatest percentage of UK Unique Audience under 12 years.

1. Stardoll 32% (representing 108,000 child users) (categorized as a targeted "portal"...why not online game?)
2. Club Penguin 29% (196,000 users) (categorized as "kids, games, toys" - why not online game?)
3. Nick 25% (76,000 users) (also "kids, games, toys")
4. LEGO 25% (98,000) (also "kids, games, toys")
5. Cartoon Network 24% (107,000) ("kids, games, toys")
6. BBC CBBC 24% (318,000) ("kids, games, toys")
7. CBeebies 20% (293,000) ("kids, games, toys")
8. Disney Intl. 19% (146,000) (categorized as "multi-category entertainment", doesn't mention that it includes some of the most popular online games for kids around)
9. RuneScape 16% (87,000) (online game)
10. MiniClip 15% (239,000) (online game)

By my count every one of those sites includes a pretty significant gaming component, but for some reason, kids are described as having an affinity for "entertainment sites", whereas teens are the ones identified as favoring online games. For comparison's sake, here's the same list for 12 to 17 year-olds (also by Nielsen):
1. Frengo 26% (117,000 users) (mobile content download site))
2. RuneScape 25% (138,000) (Online Game)
3. Bebo 19% (919,000) (online community)
4. BBC Learning 5-19 19% (184,000) (educational content)
5. FreeOnlineGames 18% (112,000) (online game)
6. AddictingGames 17% (100,000) (online game)
7. Metro Lyrics 17% (108,000) (music)
8. MiniClip 15% (240,000) (online game)
9. LimeWire 12% (350,000) (downloading program)
10. Meebo 12% (87,000) (instant messaging)

I'm not sure, really - only three of these sites are online games, and they have pretty low percentages and rates of participation. What I'm mostly curious about here is how the categories used to make sense of the data were constructed, and why? My own interpretation of the data says that for kids, online gaming and cross-media entertainment (that incorporates different forms of content, gaming interactivity) is key, whereas teens have a much more diverse range of online usage habits but tend toward communication-based activities. Either way, the percentages provided are only somewhat useful...I'd much rather know the proportion of total child and teen users for each site (not just those from the UK), as well as the proportion of total UK users first, and then the percentage of UK kids and teens second, in order to get a better sense of national and international usage patterns, demographics, etc.

Of course, the point of a press release isn't to provide complete and comprehensive data...just to pique our interest enough to purchase the full report. But it disturbs me that so many media and news outlets take press releases such as this one on face value...I'm sure we'll be seeing lots of articles listing Stardoll as the most popular site among UK kids (though listed as an entertainment site and not as a game, which it clearly is), and "online games" as the preferred activity for UK teens, without any additional critical assessment or contextualization about what these findings might really indicate (or not) when examined in a tiny bit more depth. Meanwhile, the CBBC's amazing success at attracting nearly three times as many UK kids under twelve as Stardoll has (318,000 vs. 108,000) may very well go unnoticed.

No comments: