What is the main issue with the following inference from a sample? A large web-based news outlet wants to know what changes to make to improve reader-satisfaction among their subscribers. It uses analytics data to identify subscribers who regularly close an article without reading the full article. The news outlet polls these subscribers. They asked them "what is the most common reason you have for closing an article before reading it in full?". Options included reasons like "The article was too boring", "I disagreed with the opinions presented in the article" and "The article was too long". The news outlet received 6,000 responses and of these, 4,180 respondents said that the most common reason they closed an article before reading it in full was that the article was too long. The news outlet concludes that roughly 70% of its subscribers find their articles too long and, in response, it implements new editorial guidelines around the length of articles. Question 1 Answer a. The sample only includes subscribers who regularly close an article without reading the full article. It doesn't include people who typically finish articles. This latter subset of subscribers might think that the articles are the right length. This is an issue with measurement bias. b. People are unlikely to respond truthfully when asked why they closed an article before reading it in full. This is an issue with selection bias. c. The sample only includes subscribers who regularly close an article