Hacker News new | comments | show | ask | jobs | submit login

Small percentages like these aren't necessarily worth reading much into, and they don't necessarily have anything to do with the question being asked.

It could very well be that lower-income, older people (who would disproportionately live in the southwest) are disproportionately likely to make mistakes when filling out surveys.

Use of a letter-grade system also introduces error, since the value of a particular letter grade is subjective and has varied over time. It could very well be that older people, who received their own grades before the grade inflation of the last generation, have stricter standards when it comes to assigning letter grades to things.

Didn't dive into it in the post, but we settled the survey-entry-mistake hypothesis to our satisfaction by looking only at the responses that were gathered via phone (about 1/4 of the dataset) and finding that the same trends still held true.

Also, I really wanted the grade inflation hypothesis to be true, but I just took another look at the data and age turns out not to be consistently indicative of lower or higher grades for city services.

Thanks for the comment. Both really valid questions/comments.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact