Student Assessment Considerations

Assessment, Grades, and Homework

by  Harvey Craft

Student assessment methods need consistency across the nation. The typical method of assessment is a matter of grading student work and averaging the grades, but assessment involves much more than numbers. 

Assessment Quick and Dirty

Teachers are invited to read the points below, read the articles and links in this channel and offer opinions. Perhaps your contributions will help others improve grading and assessment. The points below are intended for kick-starting your assessment process-- how and why you assess.

Students Assessment Issues

What is the basis for your method of grading? (I.e., did you invent it, adapt it from another source, base it on reading and/or research?)

• Do you adjust grades for student behavior like being late to class?
• Do you have a test/quiz make up policy?
• Do you give extra credit to allow students to improve a period grade?
• How often do you give homework?
• Do you include homework as part of a period grade? If so, how much does it count?
• Do you grade all student work?
• Do you know the difference between formative and summative assessments?
• How many grades per student do you record in a typical grading period?
• Do you ever give true/false items?
• Do you give zeros?
• Do you give a variety of items on major tests like multiple choice, essay, short answer, or other tasks?

Perhaps I omitted something that you would like to include or encourage others to discuss. Leave us a comment? 

Problems with True and False Tests

Teachers should know how to write and grade tests. Regrettably, they don't always use good assessment methods. 

True/False Tests and Guessing

Please answer the following true/false item:

___T ___F   Neils Bohr accepted the principal of complementarity as a necessary part of quantum theory.

If you actually participated in my one-question true/false "test" you had about a 50-50 chance of getting it right (unless you are quite knowledgeable regarding quantum theory). The question was easy for me to write-- I modified a phrase from a website. 

True/false tests have perhaps two advantages over other types of tests: (1) they are quickly written, and (2) they are quickly scored. 

These two advantages are both for the convenience of the teacher and do not offer students much. The trap for both students and teachers lies in the 50-50 chance of a correct or wrong answer. This encourages guessing-- not a test-taking skill to be encouraged. Also, because true/false items generally address simple concepts a large number of them is necessary to produce valid results. 

The chance factor is also important in scoring. Teachers typically grade them as they do other test items-- i.e., 20 items will count 5 points each (100% / 20 = 5). Thus a student who misses 10 of 20 will receive a score of 50. But if the items are fairly written a chimpanzee  trained to use a pencil could be expected to get half (10) of the items correct. Did the chimp know the answers? Likely not. Should the chimp receive 50 points based on chance? Well, if the teacher wants to grade according to knowledge the answer would be "No."

The correct way to grade true/false test and score for the "guess factor" is to multiply misses by two (the number of choices) and subtracting the result from the percent right. Each item counts 5 in a 20-question test. 

"Traditional" scoring simply deducts 5 points for each one missed:

100% - (5 X 10) = 50%

Scoring to compensate for guessing is different (and correct):

10 right X 5 = 50%;

10 wrong X 5 = 50%;
% right - % wrong + 50% - 50% = 0

Another view of the same "correction score" method is:

100% - two times the percent wrong;

100% - 2(50%) = 0

Students who score around 50% with traditional scoring techniques are demonstrating a score predicted by guessing! Of course, the same issue arises with any multiple choice test. After all, a true false test is a multiple choice test with 2 choices. The effect of guessing is reduced as more choices are added. 

Do True/False Tests have Value in Real Assessment?

The short answer is "Yes." But if a they must be used, consider them for formative (ungraded) activities designed for teaching simple concepts and adding variety to instructional methods. True/false worksheets can be an good exercise for homework.

True/false tests can be improved by having students justify their answers, but why not simply ask a short-answer question to begin with? 

By the way, the sample item at the beginning is "true."