Since the nineties there has seen a striking increase in the number of compulsory admissions in the Netherlands. Van der Post and colleagues aimed to highlight changes in emergency psychiatry in Amsterdam which have contributed to the rise in the number of acute compulsory admissions. They compared a cohort of consultations conducted by the city crisis service in 1983 (n = 460) with a similar cohort of consultations conducted in 2004-2005 (n = 436).

Penterman and Nijman tested the predictive validity of the Checklist of Risks to the Crisis team (crc). This instrument was specifically designed to assess the risks of aggression from severely disturbed psychiatric patients before they are seen by the crisis team. Any perceived aggression following outreach contact was recorded by means of the Staff Observation Aggression Scale-Revised (soas-r).

The authors somehow should have corrected for multiple correlated testing because for each item of the checklist, e.g. the person or organisation reporting the crisis situation, all categories were tested separately. However, correcting for multiple tests is not straightforward when items are correlated. In this study six checklist topics were tested with 38 Chi-square tests (see Table 1 below), of which 20 were statistically significant at the 5% level and 12 at the 1% level. Without explicit null hypotheses and low power, test results are unclear and risk of erroneous conclusions is high (Murphy en Myors, 2004).

Hornsveld and Kraaimaat explored the results of outpatient group treatment of violent adolescents (aggression control therapy) in the forensic outpatient clinic ‘het Dok’ in Rotterdam. They compaired patients with a group of secondary vocational students and treatment completers with patients who dropped out. Patients traits and problem behaviour were assessed at three moments: during uptake, at the start and at the end of treatment.

Not surprisingly, patients compared to students scored higher on anger- and on hostility scales and lower on social anxiety. Patients who completed the therapy scored lower on psychopathy and demonstrated significant decrease in physical aggression compaired to the patients who dropped out. The nul-hypotheses of "no difference" are highly unlikely, therefore Type I errors are not an issue but Type II errors are very likely and chances are increased because corrections for multiple testing were applied. 

Null- en Nil-Hypothesis

feed-image Follow new articles