|
The numbers still feel pretty outlandish to me. The biggest factor in your count, and I think it is the one with the highest ceiling, is smart devices. Trouble is, even by sources like https://www.consumeraffairs.com/homeowners/average-number-of..., around half of all households still have zero, and the average household has only 2.6 people. In this thread (from its root), we have various users defending the reasonableness of the numbers, some providing numbers in their own houses: 10, 11, 14, 17, 19, 23, 28, 34, over 50, 60+. Averaging, I’ll say, about 27, and that’s with two pretty big outliers—if you excluded them (maybe reasonable, maybe not), you’d be down to 19.5. And these sorts of users are already likely to be above-average, it’s the nature of HN, compounded by them being the ones commenting (confirmation bias). Yet already (with the fiddling of removing what I’m calling outliers) they’re under the claimed average. And for each one of them, there’s another household with zero smart home devices; and the 20% of the population with no broadband are, I imagine, effectively using zero wifi devices, though discounting in this way is a little too simplistic. However you look at it, the average will drop quite a bit. In fact, if you return to the original 27 and simplify the portion of the population without smart home devices to a 30% zero rate (mildly arbitrary, but I think reasonable enough as a starting point) and let the other 70% be average… your 27 has dropped to about 19. In order to reach the 21 across the population, you’d need to establish these HN users, defenders of high wifi device counts, to be below average users of wifi devices, which is implausible. If the number was 10, I’d consider it plausible, though honestly I’d still expect the number to be lower. But I think my reasoning backs up my initial feeling that 21 is pretty outlandish for your national average. I’d like to see Deloitte Insights’ methodology; I reckon it’s a furphy. I bet it’s come from some grossly misleading survey data, or from sales figures of devices that are wifi-capable even though half of them never get used that way, or from terrible sampling bias (surveys are notorious for that), or something like that. Wouldn’t be the first wildly wrong or grossly misleading result one of those sorts of companies have published. |