ABSTRACT
A Type I error is a false positive, which occurs when the null hypothesis is rejected but a true effect is not actually present. This would correspond to the situation where the evidence seems to indicate an intervention is effective, but in reality it has no benefit. Unfortunately, though, false positives and non-replicable results are far more common in the literature than they would be if our scientific method was working properly. A related reason is a selective focus on positive findings within a study. Consider a study where the researcher measures children's skills on five different measures, comprehension, expression, mathematics, reading, and motor skills, but only one of them, comprehension, shows a statistically significant improvement after intervention that involves general “learning stimulation”.
