The quality of an essay exam depends heavily on the quality of the question. A vague prompt such as "Discuss Tvrdica" invites rambling answers that are difficult to grade, while a focused question produces responses that can be compared fairly. For a play like Tvrdica, a good question might ask how Molière uses Harpagon's relationship with his children to criticize greed. The prompt should signal exactly what the student must do and what kind of evidence is expected.

Use clear task verbs such as analyze, argue, compare, or evaluate, and avoid vague verbs like discuss. Each verb implies a different kind of thinking, and students respond best when they know which one is required. A prompt that asks students to evaluate how effectively the final scene resolves the play's conflicts calls for judgment and evidence. Precision in the question leads to precision in the answers.

Consider the time students will have and set the scope accordingly. A question that requires coverage of the whole play may be too much for a forty-minute essay, while one focused on a single act is manageable. Offering a choice between two prompts can help students play to their strengths, though it complicates scoring. If you offer choices, make sure both prompts demand similar skills and effort.

Building the Scoring Guide

A scoring guide should be written before the exam is given, ideally alongside the question. List the elements a strong answer should include, such as references to specific scenes, an arguable thesis, and analysis of Molière's comic techniques. Also describe common weaker responses so graders can recognize them. A guide that anticipates the range of answers makes scoring faster and more reliable.

• State the specific skills the question is designed to assess
• List the key points a strong response is likely to include
• Describe what mid-level and low-level responses look like
• Specify how many points each criterion carries
• Note acceptable alternative interpretations to avoid penalizing originality

Allowing for Multiple Valid Answers

Literature questions rarely have a single correct answer, and a rigid guide can punish students who read the play differently than the teacher. Build flexibility into the guide by describing qualities of reasoning rather than specific conclusions. For example, instead of requiring that students call Harpagon a villain, require that they support their characterization with accurate evidence. This approach rewards thoughtful reading regardless of the specific view.

Test the guide on a few sample responses, if possible, before using it. Write or collect a strong, a medium, and a weak answer and see whether the guide produces the scores you expect. If it does not, revise the language. This small investment prevents scoring problems later.

Scoring at Scale

When you have a hundred or more exams, consistency becomes the main concern. Score all responses to the same question together, shuffle the papers periodically, and revisit your first few scored responses to confirm that your standards have not shifted. If multiple graders are involved, hold a norming session using a few sample papers. These routines keep scoring fair.

AI grading tools can support this process by applying the scoring guide uniformly and producing criterion-level notes for each essay. Teachers review and adjust the scores, particularly for unusual or high-quality responses. The speed gained lets exam results return to students sooner, when feedback is most useful. Combined with a well-designed question and guide, this approach makes essay exams practical even in large courses.