I hope you’re right. However, for me a ‘pathway’ is predicated on there being a significant difference in variables.I've read the OPs post a few times, and I don't see anything mentioning that she expects these things to be higher in military populations; to me it looks like she's seeing if the pathway towards risk taking behaviour can be predicted by a number of variables, including military service.
Also its an undergrad project, so needs to be nowhere near as sophisticated as your mention. It's likely they won't need to use anything more complicated than multivariate regression using the measures she is proposing, which knowing how most psych undergraduates are, is probably is enough to fill the OP with dread as it is - lets not try freak her out (or add to the pastoral workload of her tutor!).
OP - good luck with you project. I've not filled it out yet, but I will make the time in the next few days.
Edit: and in addition to here, I trust you have considered the various facebook service groups. To be honest you may get better data in those as usually members are (informally) vetted to ensure they have actually served etc.
Also as I say, I've not filled it in yet, so no idea if you ask this, but you may wish to distinguish between TA/regular soldiers - no idea if its too late to add this question on, or if you have to resubmit ethics.
That’s why it’s important that the OP discusses the scope with their tutor.
Lastly, the problem with self-selection bias still remains...