Errors in stats... World Internet Use

I think the results released a couple of weeks ago by Nielsen on global home internet access are probably more of an example of errors in data collection or a general error rate of about 2% than a demonstration in a drop in home internet access use among some of the most digitally sophisticated countries in the world. While I don't doubt there is some drop off in home internet access due to removals, expenses and so on, I can't quite believe that 3% of Australians and 3.5% of all Swiss netizens decided to switch off home internet access in February this year. It looks more like the sample Nielsen used was shoddy to me.