Although a 5% difference in the proportion of heroin-positive urine samples in favour of methadone was found, the 95% CI were within the 10% non-inferiority margin.
When analysing complex multidimensional data, researchers can easily pick up small differences between samples in favour of their hypothesis (i.e. confirmation bias), while ignoring the major trends in the data.
As we showed in a former study of citizens of Hamburg, even recruitment strategies that did not disclose any relation to health topics led to samples extremely in favour of cancer screening.
After 10 months of recruitment, saliva sampling was suspended in favour of blood sampling.
Our online questionnaire requested information on the visitor's education level; even if our final sample was biased in favour of higher education levels, this information was taken into account during the statistical analyses.
This would suggest that the study sample was biased in favour of less busy physicians.
This sample was skewed in favour of high socio-economic status, and there was no comparison group.
It was assumed that for any new service to be termed as useful, at least three quarters (75%%) of the sample should be in favour.
Contrary to the result of the GRR test in sample 1, which suggested a recessive effect of the ITGAV rs3738919-C allele, the result of the larger combined sample is more in favour of a dominant effect of this marker.
