Why it matters when you are writing
Reviewers police this closely because overclaiming is the fastest way to lose credibility. A cross-sectional design cannot establish direction, a sample of undergraduates in one country does not license a claim about people in general, and a significant result in one study is not a settled fact. Hedges are how you say what you found without saying more than you found, and they are read as competence rather than as weakness.
The opposite failure is real too. Prose in which nothing is asserted — “it could perhaps be tentatively suggested that there may be some indication” — reads as having no findings at all. Each hedge should do a job. Stack two and you have usually hidden your own result. The strongest academic sentences state the finding plainly and qualify once, in the place where the qualification is genuinely needed.
Different hedges do different work, and it is worth choosing deliberately. Modal verbs qualify certainty (may, might, could). Reporting verbs qualify the strength of an inference (suggests, indicates, demonstrates). Scope qualifiers limit the population or condition (among first-year students, under laboratory conditions). Frequency qualifiers limit generality (typically, in most samples). Naming the right dimension is more useful than adding another “possibly.”