I think the survey was probably done with control questions, that's why we get jaywalking as the worst habit. For example, the survey question could simply be like:
What do you consider to be the worst habit of Singaporeans?
a) Jaywalking on a busy road
b) Slurping soup in expensive restaurant
c) Talking loudly in library
d) Others (Please specify:_______)
If the questions was simply just a blank for the person to fill up, there will be probably be
too many different replies to properly tabulate. Alternatively, by giving the respondents limited choices which don't seem too bad, they can simply make the survey look less negative.