On Severe Tests of Trend Hypotheses in Psychology