You might be in need to justify certain methods. Here are a few good quotes on justifying certain procedures or methodological appraoches.
From Lian HW etal (2012, JAP, page 110 :
Although Internet recruitment methods differ from more traditional recruitment methods, such as sampling college students or individuals from specific organizations, their use has been endorsed by the American Psychological Associa- tion’s Board of Scientific Affairs’ Advisory Group (Kraut et al., 2004), and they have been shown to produce data of equal quality to those from more traditional recruitment methods, with the added benefit of the sample being more diverse and hence more likely to generalize (Gosling, Vazire, Srivastava, & John, 2004).
From Moon 2001 JAP “Looking forward and looking back: Integrating completion and sunk-cost effects within an escalation-of-commitment progress decision” :
Limitations - The major limitation of this study was the use of student participants and a relatively simply worded scenario to model what is obviously a complex natural dilemma. I have noted some potential concerns with the sunk-cost manipulations of previous studies that attempted to explore the separate effects of both sunk costs and completion. These same studies indicated no sunk-cost or interactive effects when separated from completion; therefore, additional findings using a similarly structured experiment were a necessary first step toward pointing the literature toward a more complementary paradigm. This study successfully clarifies and advances the findings of earlier works using basically the same manipulation by adding specificity to the sunk-cost manipulation (years invested and a definitive statement of the level of sunk-costs incurred) and, maybe more importantly, controlling for perceptions of success (which enabled a more clear look into the true impact of the two variables of interest). Therefore, any limitation associated with the use of both students and simple scenarios was necessary to offer a reconceptualization of how these two phenomena, sunk-costs and need to complete, act both independently and in concert with a decision maker's proclivity to continue investing into a dubious progress-related project. That is, the purpose of this study was to find results, different from previous literature, within a controlled environment rather than address issues of generalizability.