Choosing p=0.08 as marginally significant just because it has a medium effect size is not acceptable.
← all excerpts
A sense of belonging: The role of higher education in retaining quality STEM teachers.
4
—
—
The sentences
There is really no such thing as marginal significance.
The words "approached significance" would be acceptable, unless your a priori alpha is 0.08, which would need to be justified as it is non-standard.
With medium (η 2 ≥.06) to large (η 2 ≥.14) effect sizes, data reported were practically significant, offering evidence that connections to STEM professional networks impacts STEM teachers’ perceptions of their professional efficacy.