Building multiple choice questions that don't waste everyone's time
Most chemistry MCQs are poorly constructed, and the people writing them usually don't know it. I spent several years grading exams and building question banks, and the pattern is always the same. Good distractors require actual thought about what students get wrong. Bad ones just fill space. Here is how I approach it, starting with the mechanics before the theory.
Creating Reliable Multiple Choice Questions In Chemistry
The first step is picking a single, unambiguous concept. Not "explain the reaction." Something like "What is the major organic product when 2-methylpropene reacts with HBr in the presence of peroxides?" That has one correct answer and the distractors can map directly to common misconceptions. The distractors matter more than the correct answer. A well-written question needs wrong options that represent real errors students make. For the example above, the right answer is 1-bromo-2-methylpropane due to anti-Markovnikov addition. A weak distractor would be something random like "butane." A useful one is "2-bromo-2-methylpropane," which is exactly what a student gets if they forget about the peroxide effect and apply standard Markovnikov rules. Another solid distractor: "2-bromo-1-methylpropane," which catches people who mix up the carbon numbering. I once spent three days debugging a question about acid-base equilibria where every student picked the same wrong answer. The problem wasn't the chemistry. The stem accidentally implied a specific temperature condition that changed the pKa values we were using. I had written "at standard conditions" but the reference table I built the question from used 25 degrees Celsius while my mental model assumed 20. Both answers were technically defensible depending on interpretation. The fix was rewriting the stem to explicitly state the temperature and updating the reference values. That question sat in a draft folder for six months after that incident.
There is a practical workflow I stick to. Write the question first. Then write three distractors based on specific student error patterns you have actually observed. Only then go back and verify the correct answer independently. If you write the answer first and work backward to the distractors, you will unintentionally make the right answer too obvious or the wrong ones implausible. Chemistry MCQs run into trouble in three predictable areas. Thermodynamics questions suffer from rounding differences. A calculation that gives 4.32 kJ/mol versus 4.35 kJ/mol depending on when you round changes whether your answer choices overlap. Keep at least two significant figures between every option. Equilibrium problems hit the same wall with K values expressed in different formats. Spectroscopy questions tend to produce ambiguous answers when the structure has symmetry that makes two positions equivalent. Students who notice this will mark both correct options and demand curve adjustment. I avoid these by using constraint-based question design. For calculation questions, I pick the numbers first so the answer lands cleanly on one of my choices with no rounding ambiguity. For structural questions, I deliberately choose asymmetrical molecules. It is slower to write questions this way, roughly doubling my initial drafting time, but it cuts my revision cycle down to almost nothing because the questions actually work the first time.
Get the Full Details
The biggest mistake people make is treating multiple choice as a content coverage tool rather than a diagnostic one. You can write fifty MCQs that cover every chapter and learn nothing about whether students understand anything. Five carefully constructed questions that target the specific misconceptions in your recent lectures will tell you far more. The metric that matters is not coverage. It is whether the distractors separate students who understand the concept from students who are guessing or applying a superficial rule. I use a simple validation check before releasing any question bank. Run through every option and ask which specific error leads a student there. If you cannot name the misconception behind at least two distractors, those questions need work. A question with three random wrong answers and one right answer is just a trivia prompt dressed up as assessment.
