Table 1

The characteristics of the three types of measurement in health care6

Aim/IntentNew generalizable knowledgeComparison, choice, reassurance, spur changeImprovement of processes and outcomes
Observability of testingBlinded or controlledNo test, observe current performanceTests are observable
BiasDesign to eliminate biasMeasure and adjust to reduce biasAccept consistent bias
Sample sizeAs many as possible for sufficient powerObtain 100% of available data‘Just enough’ data with sequential samples
Flexibility of hypothesisFixed hypothesisNo hypothesisHypotheses are flexible and change as learning takes place
Testing strategyOne test or several large testsNo testsSequential tests
Determining if a change is an improvementHypothesis testing, statistics (eg, t-test, χ2, ANOVA)No change focusStatistical process control charts