Assessments form the basis for eLearning. Without them, it can’t be decided whether the learners are ready for work or not. But the assessments should be inclusive which means that nobody is biased against them due to personal problems, or lack of knowledge of local languages. There could also be other disabilities like dyslexia which could prevent a person from attempting a paper-based test and hence it should be taken care of.
Therefore, it’s the job of the L&D team of the company to design inclusive assessments.
Now what should be the characteristics of such assessments:
- Prepare tests in all formats and common languages:
The test should allow participants to provide oral answers to questions. This makes sure that participants with reading or speaking disabilities are also able to pass the exam with flying colors. Also, the test questions should be translated into different languages so that all the participants can answer them. If a participant has a writing disability like a fractured arm, he should be provided an aide to complete the test for him or he can be provided speech-to-text facility for the test.
In fact, visually impaired participants should also be given the same facility so that they have no problems doing the test. The test could be converted into the braille format for them and they must have a speech-to-text typewriter so that they can complete the test.
The test takers should not have problems attempting the questions due to the language. The language of the questions should be simple and comprehensible for all. Apart from that, the questions should not be specific to a particular region. For example, if the questions are being attempted by someone from a specific country, then the questions should keep in mind his cultural norms.
There should not be phrases used in such questions which such test takers are not familiar with.
- Relevant questions:
The aim of any good test is to the knowledge of someone who is going to work in a certain position in the company, whether he is apt for the role or not. The test creator must see to it that someone is not made to answer irrelevant questions in the test which are not related to his caliber to perform the job duties. A digital test is the norm these days since it can be easily evaluated by software and the test results are generated quickly. But someone, who is not required to type so much as a part of his/her role must not be given a test with a lot of typing requirements because it’s useless to check this skill.
Even if a candidate scores well on this test, it does not measure his ability to do good at his job which does not require typing. Therefore, a test must not suffer from construct irrelevant variance which means including impertinent questions in the test which don’t indicate anything in the context of certain skills.
There are also engineering test for employees to check whether they can fix machines in the future. But this test must not include questions in the local language containing intricate words which can’t be understood by someone who is not a fluent speaker. He is required to fix machines but not required needed to speak a language coherently.
Therefore, a test checking the engineering capabilities should be translated into foreign languages too for foreign engineers to attempt it. This is what the test creator has to keep in mind.
How to formulate test questions:
What is the construct of a test?
Whatever quality a test seeks to measure in a test-taker is its construct. Constructs are very important to be established before the test so that it does not stray away from its objective.
Hence, it should be the target of the L&D department to see that test content should be as per its construct. If a test content varies from the construct, which means it starts evaluating something other than the construct and this phenomenon is known as the construct irrelevant variance. For example, when a test requires engineering candidates to show an exemplary typing speed, it is varying from its objective of measuring engineering skills.
Then, as per these test results, the scores of the candidates will reflect their typing skills and not their engineering caliber.
Therefore, before designing any test it’s crucial that some steps are followed so that the test results reflect its purpose:
- Definition of the Construct:
The first step is that the construct of the test should be established clearly.
- Coming up with questions:
Designing the questions based on the construct of the test.
Checking the test’s validity is important so that if there are any problems they can be identified before the test is given to the real candidates. It’s better that before a test is given to a sample, it’s chosen carefully. Such a sample should consist of diverse candidates so that the final test is as inclusive as possible.
For example, there can be the inclusion of someone who does not speak the local language in this sample so that the use of any complex term in the local language is detected
Revise the test based on the feedback of the sample of the candidates chosen in the previous step.
Test anxiety is also a problem in which a test taker can have trouble answering questions because of the pressure to complete the test within a certain time limit. It’s better that the test is designed in mind, keeping the construct. If the construct is to check the competence of a test taker, there should be a liberal speed limit.