When research is claimed to be “approaching significance”, it means that the level of significant is very close to the point where we can claim a significant effect. For example, if a researcher conducted a study and found a significance of p=0.056- the researcher would claim that this result was ‘approaching significance’, however we all know that for a result to be significant it must be p=0.05 exactly (or less).
Some would argue that this result was so close to the required level of significance that it shows that there is an effect but due to other circumstances like the potential of outliers or maybe the fact that the sample is too small, the significance level isn’t quite enough, however the effect is evidently there, therefore claiming significance is perfect acceptable. However, this is where researchers should look at the power behind their research- the effect is obviously there but the power of the data isn’t enough and on most occasion all this would take is more participants and the level of significance would come right down and the effect would be indisputable.
However one has to ask, if we claim that this is “pretty much” significant then where do we stop- do we stop at p=0.06? or is a level of 0.062 to be claimed as approaching significance then? How far will we go? Significance is a pretty important brick in the foundations of psychological research and I believe that if we were to discount this and view this as something we could move around to suit our research then it has the possibility of bringing the whole field down with it. Imagine if significance became something that we discounted, we would end up taking all research results as effective- even those with a level of p=2 etc?
I think that this is something that has to be upheld- if it comes to it that your research reports a level of p=0.056- then i think this should show that you need to add more people to your sample and if that doesn’t change it then maybe the effect isn’t there after all. I think that the claim of significance should only be placed on p levels of p=0.05 or less- specifically.
The term approaching significance should only be seen as a cue to do more research or increase sample size or improve methods in research.