"Statistical significance helps establish whether a result is reliable," writes Rachel Banawa, "while practical significance helps determine whether it is worth acting on." Banawa is writing in the context of user interface design, but there are of course implications of this distinction in wider learning and education research. An intervention, for example, might result in a 0.2 percent increase in test scores, but the effort to implement it might be too costly, or ethically questionable, or involving a score not really worth improving in the first place. That's why 'what works' research (related) needs to be taken with at least some scepticism. It's not good to be going faster if you're going faster in the wrong direction.
Today: Total: [Direct link] [Share]
