10 Tips to Improve the Validity of Your Elearning Assessments
This format lists options as say A.B,C, and D, without limiting the permutations and combinations of the options. From the learners point of view, such framing of an assessment creates a dilemma and does not factor in the possibility that partly correct knowledge can yet lead to correct performance in real life situations.
Recommendation: Either limit the permutations and combinations of options by providing a hint on the number of correct options or frame the same question as multiple choice assessment rather a multiple answer assessment. For e.g. restrict the options to AB, BD, AC, or D only.
This format lists options as say A.B,C, and D and also limits the correct choice to only one option. This template defines the problem space for the learner in clear terms and also casts an obligation on the assessment designer to provide feedback which is unique to the choice made by the learner.
Recommendation: Not only should this template be treated as an assessment device but also as learning interaction, which offers an opportunity for providing choice-specific feedback to the learner.
This template allows learners to match a list of items say A,B,C, and D with a list of options say E,F, G, and H; and also limits the correct match to only one combination.
Recommendation: Do not use this to combine two or more multiple choice questions into one matching exercise. All matching items should follow a parallelism and each column should be category. For e.g. Country Vs. Language. It’s a best practice to use category names as headings above both columns for the learner interaction to yield maximum value.
True and False:
This template allows learners to select either the true or false option in the context of the provided statement.
Recommendation: Such items are highly prone to guessing with 50 per cent probability that the guess may be correct. Combining two or more statements in a multiple answer items is much better option. Note however that deficiencies of multiple answer assessment items still need to addressed.
This template allows learners to order a list of items with the limitation that only one order is correct.
Recommendation: Avoid using match-with-order-position template for such assessment; prefer the simpler drag and rearrange template in elearning.
Fill In The Blanks:
This is truly a supply-type assessment with the learner typing in the response.
Recommendation: The most difficult item to frame; it should be principally used to test knowledge of definitions, with the answer being preferably being one word. Most assessment templates allow up to 10 variations of the correct answer.
This template allows the user to supply (not type) a word or phrase from the available options, with the limitation that only one is correct.
Recommendation: This template should be preferred over the Fill In The Blanks template especially when there is greater uncertainty over the correct answer or greater probability that the answer may be misspelt.
This template allows user to select a particular area of an image with the limitation that only one area has to be selected.
Recommendation: If clear spatial demarcation of different options is not possible, convert such an assessment into a multiple choice item with reference of a labelled diagram.
This supply-type template allows the user to enter a numeric value. It could be a specific number, a range of numbers, or any combination thereof.
Recommendation: This assessment item inevitably makes the Iearner to think that a calculation is required; if it is not required, prefer the multiple choice template instead.
Randomisation of Assessment Items:
This template solution allows the assessment developer to randomize the assessment items from within an assessment bank.
Recommendation: Most elearning development tools provide the developer the capability to randomize a question bank and eliminate the role of memory effect in the repetition of assessments. This capability is critical to test the learner’s performance and improve the validity of test results.