政府调查的响应率显著下降
在一份通讯中看到这张图表，令我颇感惊讶：美国的调查响应率低于70%，而2013年时超过90%。我认为这是“技术进步”以及如今沟通便利的一个有趣且负面的副作用。我们被信息淹没，只想关闭一切。更不用说无休止的垃圾电话（我不接任何不在我通讯录中的号码，因此调查员无法联系到我）。
这是一个严重的问题，因为如果调查响应率低，就会增加数据的误差范围，同时也意味着政府与公民之间的沟通减少了。
有一篇论文探讨了（部分）原因。
根据这篇文章，英国也面临同样的问题。
人工智能只会让这个问题更加严重，因为它使得在没有人参与的情况下发起垃圾电话变得更加容易。
查看原文
Saw this graph[0] in a newsletter[1] and was quite surprised: Survey response rates in the US are below 70% vs over 90% in 2013. I think it's an interesting and negative side-effect of "technological progress" and the ease of communication these days. We're inundated and just want to shut everything off. Not to mention the endless spam calls (I don't answer any number not in my contacts, so a surveyor wouldn't be able to reach me.)<p>It's a significant problem because if survey responses are low that increases the error margin of the data, and it also means that the gov is just not hearing from its citizens as much.<p>There's a paper[2] on (some of) the reasons for this.<p>According to this[3], the UK has the same problem.<p>AI is only going to make this worse with the ease at which spam calls can be made without a human in the loop.<p>[0] https://www.bls.gov/cps/methods/response_rates.htm<p>[1] https://kyla.substack.com/p/chicago-fed-president-austan-goolsbee<p>[2] https://academic.oup.com/jssam/article-abstract/12/5/1138/7658685?redirectedFrom=fulltext<p>[3] https://www.census.gov/programs-surveys/cps/about/modernization.html