There are circumstances where speed is genuinely part of competent clinical performance.
Consider an acutely deteriorating patient. Recognising the problem matters; but sodoes acting promptly. Similarly, clinical practice requires prioritisation. A doctor cannot gather unlimited information, consider every conceivable diagnosis and postpone a decision indefinitely. In these situations, the ability to process information and make an appropriate decision within a reasonable period may be part of the capability we want to assess.
Time pressure can therefore add validity when it reflects a meaningful aspect of the task. It can also prevent a different problem: giving candidates unlimited time to search through every possible line of reasoning may produce an assessment that bears little resemblance to the capability we are interested in. The principle is relatively simple: If speed is part of the capability, assessing it may make the assessment more valid. But there is an important qualification. The speed required in clinical practice is not necessarily the same as the speed imposed by an exam.
When time pressure distorts the assessment
Time pressure becomes problematic when it introduces factors that we did not intend to measure. For example, imagine a candidate who understands the clinical problem but reads more slowly than another candidate. Or a candidate who knows the relevant facts but takes longer to consider competing diagnoses. Or someone who reaches the correct answer only after carefully checking the information provided. A highly speeded exam may penalise these candidates. Their lower score may then reflect not simply what they know or how well they reason, but also how quickly they can read, process information and respond in an artificial examination environment.
This is the problem of unintended measurement. We think we are measuring clinical knowledge and reasoning. But the exam may also be measuring processing speed, reading speed, test-taking strategy and the ability to work under exam pressure.Sometimes those things really matter. Sometimes they don’t. The key thing is to recognise which situation applies.
Clinical reasoning makes this particularly difficult
Clinical reasoning provides a useful example because it doesn’t have a single relationship with time. Some clinical decisions are appropriately rapid: a recognisable emergency may require an immediate response. Other decisions benefit from deliberate consideration. A patient with several possible diagnoses may require the clinician to weigh evidence, consider alternatives and decide what information is still needed. Good clinical reasoning is therefore not simply a matter of reasoning as quickly as possible. This creates an important distinction: is the time pressure a feature of the clinical task, or a feature of the exam?
If we want to assess the ability to recognise and respond rapidly to an emergency, a time constraint may be entirely appropriate. If we want to assess the ability to reason through a complex diagnostic problem, an arbitrary time limit may tell us more about the candidate’s speed than their reasoning.
Question the clock
Perhaps the simplest issue to address in this area is that exam timing is often inherited rather than designed. An exam has always been three hours long, so it remains three hours long. A paper has always contained 120 questions, so it continues to contain 120 questions. But tradition is not necessarily an assessment rationale. When reviewing an exam, we should be prepared to ask: What evidence is there that this amount of time, or this number of questions, is appropriate? That doesn’t mean we should make exams shorter, or longer, by default. It simply suggests treating the time limit as a design decision that deserves justification.
Digital assessment gives us more options
This is where digital assessment becomes particularly interesting. Traditional exams often use a relatively blunt timing model: You have X minutes to answer Y questions.Digital assessment offers the possibility of making time a more deliberate part of the design. We could, for example, distinguish between:
- overall examination time
- time allocated to particular tasks
- deliberately timed clinical decisions
- staged release of information
- sections where candidates are given sufficient time to demonstrate reasoning without speed being the primary concern
These approaches will not all be appropriate. But the point is that digital technology gives us more choices. And those choices should be driven by the assessment objective, rather than by what the software happens to make possible.
Three questions for the assessment designer
Perhaps there are three simple questions to ask whenever a time limit is introduced.
First: Is speed part of the capability we are trying to assess?
If it is, then time pressure may be justified.
Second: How much does speed matter?
Does the difference between 20 seconds and 40 seconds matter clinically? Or is the important distinction between two minutes and ten minutes?
Third: What happens if we give candidates more time?
If their performance improves substantially, what does that tell us about the original assessment?
These questions do not produce a single, universal answer. They do, however, invite us to justify the clock. Don’t remove the clock — justify it
The answer is not that medical examinations should simply give candidates more time. Nor is it that doctors need to work quickly, so exams should be deliberately speeded. The issue is more fundamental:
The time allowed should be defensible in relation to what we are trying to measure.
Digital assessment gives us an opportunity to rethink this, rather than simply reproduce traditional timing models on a screen. The question is not: Should an exam be timed?, but: Can we explain the purpose of the timing?
If the answer is yes, time pressure may be an important and valuable part of the assessment. If the answer is no, perhaps it is time to reconsider the clock.