Akamai has released its latest State of the Internet report, this one for the second quarter of 2011. As readers of this blog will know, I've periodically commented on and linked to these reports, as they are one valuable source of empirical data about the speed of broadband internet connections in countries around the world. However, as with all statistics, they can be either used or misused.
One of my alerts sent me to an article on ReadWriteWeb entitled "S. Korea Loses Top Spot According to Akamai's State of the Internet Report." This was news to me, so I decided to read the new Akamai report. In fact, it shows that, on average, South Korea still has by far the fastest average broadband internet speed in the world. Indeed, the headline, as it stands, is an example of misleading journalism. As shown in the graphic to the left, taken directly from the Akamai report, the average download speed in South Korea during the second quarter of this year was 13.8 mbps, far higher than that of the Netherlands, at 8.5 mbps. Note that Korea did experience a year-on year decrease in average download speed of over 17 percent.
What the ReadWriteWeb article seized on for its headline was a small section of the Akamai report devoted to what is called "global high broadband connectivity," devoted to an analysis that looks only at connections at speeds higher than 5 mbps. On this one particular measure, the Netherlands recorded a 40% year on year increase, so pulled out ahead of Hong Kong and South Korea, which ranked second and third, respectively. Clearly the headline used is misleading, given that Korea still appears at or near the top of virtually all the tables presented in the Akamai report.
If you've read this far, go ahead and link to the ReadWriteWeb article, but only to see a blatant example of misuse of statistics!