Then looking at that through an ethical lens, an anonymous survey would be the standard procedure. You would have to think about what questions you were going to ask based on what exact issue you were looking to research (attitude to suicide or reasons for committing suicide and not exploring alternative options for example), then you would need to think about the phrasing of the questions and whether the most appropriate response type was a Likert scale (5 point, 10 point etc) or free-text answers etc. You also need to make clear to people at the outset what the nature of your research is, include a disclaimer that personal questions relating to 'abuse' or whatnot will be asked which some people may find triggering (so people should be comfortable talking about this or decline to participate), and you should have the ability for people to opt-out at any time, or withdraw from the study after they've given their answers (sometimes people have a change of heart and want to delete their answers). Those are matters of ethics and indeed, if you are studying through a university, the school will have their own guidelines on such matters that you will need to follow and include a declaration that you have read and observed those standards.
What it sounds as though you were looking to do is more of a clinical interview through PMs, and again I can't see that being against the rules but the appropriateness would depend on your level of training and experience, and again, I would personally have some misgivings about that given how vulnerable people here are, the environment in general and the fact that you yourself are anonymous - that in itself raises ethical issues.
But as I have said above, I don't believe you have any bad intentions. I just worry that this is something you haven't really thought through - there's no shame in that, I was a student once, been there, done that, got a bollocking for it. But please do double-check any of your university's guidance on ethics and research, the absolute last thing you want is to involve someone on here in some research (even informally) and have something go wrong, perhaps they have a suicide attempt (whether successful or otherwise), which whilst unrelated to you still prompts an investigation by authorities or family members. You don't want your name associated with that, a complaint made to your university and being shouldered with any blame. I know that it seems to be an unlikely scenario, but I have seen it happen in other contexts.
It's one thing to do such research on Reddit, there is little harm in that, but the nature of this place @MrBigSad, it demands a far greater degree of caution and sensitivity. There are creeps here and people who want to do other members harm, there is perilously little to allow members to distinguish those who wish to identify the vulnerable and cause them harm, from the genuinely well-intentioned. I'm not going to continue to lecture you beyond this, I just hope that you'll pause, take a breath and give this project a little more thought.