In statistics, if something is significant then it means it is not likely to have happened by chance, if the data produce an insignificant result we must accept the null hypothesis and admit that we have found absolutely nothing. But have we found nothing? Just because SPSS tells us the value of some test is under 0.whatever that apparently makes our research worthless, but thinking about what significance means in the real world, our research can still be relevant and important to other psychologists, adding to the research in that field.
But if your research is based on a solid theory, then it should probably produce statistical significance and so no significance means the researcher has two things to consider-
Firstly, you can assume that you must have made an error in data collection or analysis, overlooked outliers, errors with measures and validity, unfair population sample etc. Any number of things could affect the significance of your results. So the best thing to do would be to repeat the experiment again and hope for significance. Secondly, if after repeated experiments you still find no significance then maybe you haven’t made any errors at all, and now you should question the theory itself. Is the theory really that solid? A solid theory is a theory that is the best for now at explaining why and how something happens, but that doesn’t mean it is true. The most well known and obvious example here would be the theory of evolution, supported by a lot of evidence and makes the most sense out of any other theories for why we are here, yet it’s still only a theory and any one piece of strong opposing evidence could destroy the whole theory. A piece of research finding no significance could also disprove a solid theory, so maybe this solid theory needs to be revised or scrapped altogether as a result of your insignificant research, in which case get it published immediately!
psud1a
April 18, 2012
You make an interesting point that even if research is insignificant that it should be published, one which I totally understand and on the most part agree with, however there are a few problems with this. One in particular is the credibility of the researcher. If you had indeed based your research on a “solid” theory, we’ll use the theory of evolution as an example, and you found conflicting evidence against it, surely you would want to be sure about it? It may well be the case that you adhered to all the correct standards and measures etc. but without additional research and a strong enough claim, then surely your findings would be subject to scrutiny? I’m not saying that there is a problem with conflicting theories or evidence, but it is important to be 100% confident in your findings before publishing them. With regards to publishing, slow and steady wins the race 🙂