A city wants to estimate the average number of library visits per month. They take a random sample of library card holders. Which group might be missed, causing undercoverage?