Negative Marking v5
1 like1,141 views
Negative marking on multiple choice tests assigns 3 marks for correct answers and deducts 1 mark for incorrect answers. This discourages random guessing and makes test scores more realistic and reliable. Without negative marking, test takers who randomly guess could score 25% on average just by chance. Negative marking cancels out the benefit of lucky guesses, ensuring scores better reflect a test taker's actual knowledge.
Negative Marking v5
- 1. Negative Marking Guidance for Test Takers: Select the option that seems to you the most likely to be correct for each question. Only one option is correct per question. Three marks are awarded for every correct selection, minus one for every incorrect selection. Random guesses are equally likely to result in a lower or higher test score, so it is safer to skip questions unless you feel that you can at least make an “educated” guess.
- 2. Rationale: Any randomly selected option is three times more likely to be incorrect than it is to be correct. Therefore, by awarding +3 for every correct selection and -1 for every incorrect selection we ensure that Test Takers have nothing to gain, on average, by making purely random selections.
- 3. Negative marking within multiple- choice tests has two benefits: 1) Since pure guesswork is no longer advantageous, the test scores are more realistic, 2) less guesswork means the test scores are more reliable.
- 4. Want to know more? If so, read on...
- 5. A good test should yield a score that is proportional to each test taker's knowledge, comprehension, or aptitude; whatever it is the test has been designed to assess.
- 6. However, if test takers guess some of the answers this will lead to unrealistically high test scores if there is no negative marking.
- 7. If each question has 4 answer options, the expected score for a test taker who has zero knowledge, comprehension, or aptitude is 25%. This is an average score, since each test taker's actual score will be variable.
- 8. There are two distinct problems here. The first problem is that pure guesswork leads to unrealistically high test scores. This is not a major problem, because the scores can be scaled down to compensate for the inflation. The second - and much more serious - problem is that pure guesswork leads to variable test scores. This means that traditional multiple- choice tests without negative marking are inherently unreliable!
- 9. To illustrate the problem, consider a test consisting of 25 multiple-choice questions, each with 4 25 Qs answer options. ✔ Suppose that a particular 5 Qs test taker knows the right answer to the first 5 questions, and that they make purely random ? guesses for the remaining 20 Qs 20 questions.
- 10. If each question is worth 3 marks, the test taker will get 15 marks for the first 5 questions. 25 Qs (3 marks per Q) (You'll soon see why we ✔ need to have 3 marks per 5 Qs 5 x 3 = 15 marks question rather than just one mark.) ? 20 Qs
- 11. If they then make random guesses for the remaining 20 questions, on average they will get 25 Qs (3 marks per Q) 5 of them right, which ✔ will give them another 15 5 Qs 5 x 3 = 15 marks marks. ? 20 Qs 5 x 3 = 15 marks
- 12. This will give them 30 marks altogether, on average, out of the 75 marks that are available. 25 Qs (3 marks per Q) So their expected score ✔ is 30 out of 75, which is 5 Qs 5 x 3 = 15 marks 2/5, or 40%. (Their actual score may Score = 30 out of 75 (expected) = 40% of course be greater than or less than 40%, ? 20 Qs 5 x 3 = 15 marks with equal probability.)
- 13. In other words, a test taker who knows the answers to 20% of the questions has an expected score of 40%, since on average they will guess the right answers to a quarter of the remaining 80% of the questions.
- 14. Now we get to the negative marking! If we deduct one mark for every question answered incorrectly these negative marks will (on average) cancel out the positive marks gained through lucky random guesses. The idea is that for each lucky guess that adds 3 marks to the score, there will be (on average) 3 unlucky guesses that each deducts one mark from the score. Consequently it is not advantageous (on average) for test takers to engage in pure guesswork.
- 15. So in our example the expected score is now 20%, which is in line with the profile of a "good" 25 Qs (3 marks or -1 mark per Q) test (which we saw in the ✔ 5th slide). 5 Qs 5 x 3 = 15 marks Research has shown that negative marking does in Score = 15 out of 75 practice discourage test (expected) = 20% takers from engaging in ? pure guesswork. 20 Qs 5 x 3 = 15 marks This means that the actual scores will be less 15 x -1 = -15 marks variable, and therefore the use of negative marking increases test reliability!
- 16. To conclude, here’s the third slide again... Negative marking within multiple- choice tests has two benefits: 1) pure guesswork is no longer advantageous, therefore the test scores are more realistic, 2) less guesswork means the test scores are more reliable.
- 17. Thank you for viewing this presentation! www.ProperTests.com