| In principle this is anything that biases what research or other findings get published. However, it generally refers to the propensity for “positive” results to be published more readily than “negative” ones. |
Details #
The problem with this bias in publication is the knock on effects it has for meta-analyses and systematic reviews. This is particularly important as these are given the highest weight in defining “evidence based treatments” or “empirically supported treatments” in national guidelines. There are now diagnostic tools to detect whether the distribution of findings in what has been published suggests that there has been publication bias, however those tools will only have substantial strength to detect bias if there are a substantial number of publications on the same question. Where there aren’t many these tools can have very little power to detect bias.
Chapters #
Sadly, dropped to meet the word count as the book developed!
