Here's some useful research from Royal Holloway, University of London's Department of Psychology:
Research published in the journal Psychological Science [DOI: 10.1177/0956797616661523] [DX] has shown that judgements of attractiveness vary depending on who is nearby, and how good-looking they are in comparison. A person will rank higher on a scale of attractiveness when compared alongside less attractive people, than they would when judged alone. Popular opinion points to a person's perceived level of attractiveness as somehow fixed. However, research from Royal Holloway, University of London shows that context is key to assessing attractiveness.
[...] Participants in the study were asked to rate pictures of different faces for attractiveness, one by one. They were then asked to assess the same faces, placed alongside ones perceived to be undesirable. When adding these 'distractor faces', the attractiveness of the same faces increased from the first round of ranking.
Participants were then shown two attractive faces, alongside a 'distractor' face and asked to judge between them. The presence of the less attractive face was found to make the viewers more critical between the attractive face, as Dr Furl explained: "The presence of a less attractive face does not just increase the attractiveness of a single person, but in a crowd could actually make us even more choosey! We found that the presence of a 'distractor' face makes differences between attractive people more obvious and that observers start to pull apart these differences, making them even more particular in their judgement."
The DUFF (2015). You may also be interested in this study suggested by Medical Daily:
Attractive Female Romantic Partners Provide a Proxy for Unobservable Male Qualities: The When and Why Behind Human Female Mate Choice Copying (open, DOI: 10.1177/1474704916652144) (DX)
(Score: 0) by Anonymous Coward on Monday October 03 2016, @02:37PM
Nevertheless there is a strong general consensus on attractiveness.
So what?
There is no particular reason why someone taking part in a survey like this should want to give false answers.
The mere fact that you cannot conceive of a reason that someone would lie about something such as this doesn't mean that one doesn't exist. There may even be multiple reasons. In any case, assuming they're not lying is not very scientific.
There is also the possibility that the people taking the survey cannot accurately translate how attractive they believe someone is into a number or similar.
No more than any other survey.
Unless you can objectively verify the accuracy of the information, your survey is probably unscientific.