eLearning: the keys to a successful assessment

There are many benefits to a well-designed assessment. Learn how to measure competency and design a successful assessment using cognitive science

Measuring a skill is never easy, and anyone who has had to design tests can attest to that. However, There are many benefits of a well-designed evaluation! While it is accepted that they make it possible to estimate the level of learners, less well known is that assessment is also a valuable aid that improves learning. Proposing a relevant evaluation is therefore a real challenge for design effective training !

Thankfully, teachers and trainers can rely on docimology, or exam science. This discipline focuses on identifying problems in assessment methods, and advising the evaluator in constructing and correcting a test. She tries to avoid any biases that could compromise the accuracy of the assessment such as, for example, the way in which the instruction is formulated, the time or order in which the correction is made or the adjustment of the scale according to the level of the group [1]. Docimology So gives us valuable keys to improving learning, some examples of which are as follows.

Above all, clearly define/dissect what you want to test

To avoid these biases, you must first of all precisely define the skills targeted. All the evaluator needs to do is ask himself a few questions: “What do I expect from the learner?” , “How can the test lead the learner to the expected performance?” , “What are the skills to mobilize?” He will then be able to better target the exercises that mobilize the skills in question.

For example, the ways to assess language level are multiple because several components come into play: oral comprehension, written comprehension, phonetics, written expression, etc. Segmenting the different components to be tested thus makes it possible to create exercises adapted to the nature of the expected performance [2], and to best correspond to effective skills.

Each type of test has its own strategy!

In the 1980s, Elana Shohamy, professor of educational sciences at Tel Aviv University, conducted an experiment testing the impact of different methods of evaluating the same competence on the results obtained by The learners. They had to answer either a multiple choice questionnaire (MCQ) or several open questions. On the same subject and requiring the same skill, the marks obtained on these two types of tests were however different. Was the learner more ignorant in one case than the other? Well no, it's actually the type of test that has an influence on his ability to mobilize his knowledge! [3]

In neuropsychology, there are two ways of restoring information that we have in our memory: cued recall and free recall [4]. In the cued recall, the restitution is favored by a word or other clue, which is absent in the case of the free recall. The more a test promotes free recall, the more certain we are that the feedback is due to good memory and not to a simple deduction thanks to the clue. Moreover, recalling information means improving its memory (see also our article on the testing effect). Since free recall involves a greater effort in remembering, it also allows better anchoring in memory (see also our article on desirable difficulties).

Open-ended questions ask the learner to produce the correct answer himself, while the MCQ makes him recognize it among a set of answers. In absolute terms, Therefore, open-ended questions should be preferred over the MCQ., since they require more memory than the latter. However, let's not throw the MCQs out too quickly: the precision and the high degree ofautomation Of their Feedback in fact, they are a relevant learning tool, as long as they respect a number of constraints that we will see now. This is especially what makes it possible to do The Didask eLearning solution with its interactive scenarios.

‍

A clear instruction to explain expectations

It is through the instruction that The learner Will try to identify what the evaluator expects of him. Faced with a poorly defined instruction, the learner could make a mistake even though he basically knew the right answer. He can also use a strategy other than the one that the trainer wanted him to use: when it is necessary to link concepts together, the learner can proceed by elimination to find the right answer. Under these conditions, it is then difficult for the trainer to know whether this result is due to a real mastery of knowledge or if it is due to a simple logical deduction [5]!

When the trainer carries out his evaluation, he must therefore ensure that the only factor necessary to answer correctly is the mastery of this concept. A good strategy is to start from the plausible errors around the concept being tested, and to design the evaluation based on them. For example, for an effective MCQ, these errors will make it possible to propose a choice of answers that all have the same clues: to discern the correct answer, the learner will then have to assess the plausibility of each proposal by mobilizing additional knowledge.

Mobilizing skills from every angle

Finally, to ensure the correct assessment of the learner's real level of mastery, it is desirable to diversify the tasks to be carried out for the same skill. Indeed, mobilizing a skill through different modalities allows the reviewer to ensure that the observed result is due to the mastery of the latter, and not to a bias in his test. In short, it is a way of cross-refering your sources!

On the other hand, this diversification also helps learning. It makes it possible to use various cognitive resources such as memory, metacognition, understanding... in order to reinforce and anchor learning in itself [6].

Care must therefore be taken to build the test in such a way that it sufficiently guides the learner without saying too much. Thinking about the construction of a test is like putting yourself in the shoes of the person being tested to predict the potential strategies that they will use.. It is also about identifying the target skills that you want him to use and building the test in such a way as to discourage him from using alternative strategies. In addition, varying the pleasures by diversifying the types of exercises will allow the corrector to ensure the solidity of the knowledge, and will also help the learner to consolidate his knowledge and improve himself.

You now have the keys to a successful evaluation, An essential first step towards sustainable learning. Can also be used as a self-assessment or group exam.

‍

‍‍

REFERENCES:

[1] Leclercq, D., Nicaise, J., & Demeuse, M. (2004). Critical docimology: difficulties in writing down copies and assigning grades to students.
[2] Kraiger, K., Ford, J.K., & Salas, E. (1993). Application of cognitive, skill-based, and affective theories of learning outcomes to new methods of training evaluation. Journal of Applied Psychology, 78 (2), 311.
[3] Shohamy, E. (1984). Does the testing method make a difference? The case of reading comprehension. Language testing, 1 (2), 147-170.
[4] Van der Linden, M., Coyette, F., Coyette, F., Poitrenaud, F., Poitrenaud, J., J., Kalafat, M., Calicis, F., F., Wyns, C., & Adam, C., & Adam, S., & Adam, S. (2004). The 16-item free recall/reminder test (RL/RI-16) (pp. 25-47). Solal.
[5] Raby, F., & Baillé, J. (1997). The ergonomic approach to learning strategies in institutional foreign language learning. The APLIUT notebooks, 16 (3), 10-21.
[6] Senkfor, A.J., Van Petten, C., & Kutas, M. (2008). Enactment versus conceptual encoding: Equivalent item memory but different source memory. Cortex, 44 (6), 649-664.