Written examinations feel as natural to students as the walls of a classroom, but they are an invention with a date and a purpose. The earliest large system of written tests appeared in imperial China, where candidates sat for days in small cells to compete for posts in the civil service. The point was not education. It was selection: a way of choosing officials by demonstrated skill rather than by birth, in a state too large for personal recommendation to work.
European universities borrowed the idea slowly. For centuries their examinations were oral; a student defended propositions aloud before masters who could ask anything. The shift to written papers in the eighteenth and nineteenth centuries was driven less by a theory of learning than by arithmetic. Numbers of students were rising, and a written script can be marked by several examiners, stored, compared and appealed. Fairness, in other words, arrived through paperwork.
Every examination system faces the same two problems, and both appear in school life today. The first is validity: does the test measure the ability it claims to measure, or merely the ability to take that test? A paper that rewards neat memorisation will produce candidates who memorise neatly. The second is backwash: once a test matters, teaching bends towards it. If speaking is not examined, speaking is not taught, however often a curriculum insists it should be.
Reformers have tried several escapes. Coursework spreads assessment across months but is harder to verify. Portfolios show development but are slow to mark and easy to dispute. Computer-adaptive tests adjust difficulty to each candidate and shorten the exam, yet they require large question banks and make the scoring process opaque to the people being scored. Each solution trades one virtue for another; none removes the underlying tension between measuring quickly and measuring well.
The history is useful for anyone designing a scoring system, including a school sports league. A score is a promise that something real has been measured. The promise is kept only if the rules are published, if different judges would reach the same number, and if the number cannot be raised by behaviour nobody wanted to encourage. These are old problems. They were old when the first candidate walked into a cell in Nanjing with a brush and an inkstone, and they are no newer because the examiner is now a camera.
Vocabulary in this text
- appear /əˈpɪɹ/ — to become possible to see, or to seem
- candidate /ˈkændədeɪt/ — a person being considered for a job or prize
- cell /ˈsɛl/ — the smallest unit of a living thing
- demonstrate /ˈdɛmənstɹˌeɪt/ — to show clearly that something is true or how it works
- skill /ˈskɪl/ — the ability to do something well
- borrow /ˈbɑɹˌoʊ/ — to take something for a short time and give it back
- century /ˈsɛntʃɚi/ — a period of one hundred years
- shift /ˈʃɪft/ — a period of work in a factory or hospital
- theory /ˈθɪɹi/ — an explanation supported by evidence
- compare /kəmˈpɛɹ/ — to look at how two things are alike or different