I'm trying to write a paper based on my PhD data. I think the data describes several interesting effects, but the significance levels are all coming out between 0.06 - 0.09. Is this likely to be due to the design being 'under-powered' i.e. not enough subjects? is it legitimate to still describe these effects and what they may mean but to add as a caveat that the significance levels are only 'trend'?
All advice welcome.