logotip carneta reference center
  Self-assessment and Summative Assessment in E-education     english
logotip referalnog centra
 
 

Reference centers / Self-assessment and Summative Assessment in E-education / Knowledge assessment theory / Writen Examination / Types of Tests

Types of Tests

 


  Additional possibilities of e-test evaluation

 Using permutable questions - Student is offered more than one correct answer to one or several questions, and is required to choose the right combination. That significantly increases the number of possible answers, while the probability of guessing the right answer is decreased.

 Choosing the order of "accuracy" - Student has to list the answers based on their accuracy and wins points according to the position of the right answer on the list. For example if a students places the right answer on the first place in a question with 5 distracters, he wins five points, for second place he wins 4 points etc.  Clearly, questions with several correct answers should be adapted to such examination by making one answer the most correct. However, students don't like such questions and see them as unfair!

 Certainty evaluation questions - Student chooses the correct answer, but also has to state their level of certainty that this answer is correct (most often by using 3, 2 or 1 point). If they answer correctly they win the same number of points they stated for the level of certainty, but when their answer is incorrect, they will lose that amount (usually -3, -2, -1, and according to some authors even double, e.g. -6 etc.

 Normalisation and negative points - Conducted to reverse the effect of guessing. A general formula for MCQ with n number of answers is:

Z = (P - 100/n)(n/(n-1)

If we use negative points, the correct answer gets "n-1", and incorrect is "punished" with 1 point.
The effect is almost the same as with normalisation. 

 Liberal tests - The students can be allowed to choose more answers, with each answer carrying a certain reward or penalty. This also rewards partial knowledge students might have. The simplest way is for students to get information about whether their answer is correct immediately after answering. They keep answering the question until it is answered correctly. The process of answering is awarded points similarly as in the case of choosing the order of accuracy.  Feedback can also be more complex, but make sure that too complicated rules don't draw students' attention from the content of the questions! Most acceptable solution could be to allow students to answer a certain question 2 (or 3) times.

 
 

Copyright © CARNet
design: logotip kimajaka