Don't want to mark answers with small typos as incorrect? The threshold function in our editor might be what you're looking for!
With this function, it is possible to set a range of correct answers, rather than manually marking a single or several answer(s) as being correct. The option for setting a threshold can be found below the answer box, by clicking on 'edit additional answers and specific feedback', and is available for the numeric and one-word answer types.
Read more on the four different answer types in this article.
For numeric answers, it’s possible to set a threshold for what is considered a correct answer. When students give an answer that is (very) close to correct and shouldn’t be considered wrong, you can use this function to set a range of answers that will be deemed correct in our platform. The value you enter for ‘threshold’ is the percentage that students are allowed to be off without their answer being assessed as incorrect. Below, you can see an example of a question that is assessed as correct, even though the answer is not exactly the same.
By default, this range is set to 5% of the correct answer.
One-word (open) answer
It can be frustrating for students when their seemingly correct answer is marked as incorrect because of a typo. In order to avoid the answer check from being overly stringent when it comes to spelling, the threshold function can be used to mark slightly misspelled words as being correct, without you having to manually add them to the answer model. For this answer type, the threshold function works slightly differently. Here, the higher you set the value for 'threshold', the closer the answer has to be to the answer model to be deemed correct. The default setting for one-word answers is 75.
What to avoid
When using the threshold function, it is important to also take into consideration what should be regarded as a wrong answer, especially when the wrong answer is close to the correct answer. consider the following example: With the question 'What type of research is conducted by testing hypotheses using numeric datasets?', a clear wrong answer would be 'Qualitative'. If you set the threshold to a lenient setting (75 in this example), this answer would still be marked as correct.
Changing the threshold setting to a higher value (90 in this example) will give you the intended result.
If you have any questions about this function, please let us know via the chat icon on the bottom-right.