Blog Entry 5: Describe the differences between constructed-response and fixed-response assessments. When would you use each type of assessment in eLearning? Why?
Constructed Response assessments are essay and completion type formats. The advantage of constructed response is that the student writes out their response rather than selecting an option provided by the test question as in fixed response assessments such as multiple choice or true false formats. Another advantage of completion constructed response questions are they are relatively easy to construct. Essay questions can reflect the higher end of Bloom's taxonomy such as create, analyze, compare, prove, defend and contrast. Essay questions reflect the ability of a student to write clear organized thoughtful answers. A student can explain their logic in solving a problem. I find it fascinating to see how different student's thought processes are even though they all can solve a problem.
The disadvantage of a constructed response completion questions are they are generally limited to the lower end of Bloom's taxonomy in the recall area of information. Sometimes there can be more than one right answer especially if the subject of the question is generic rather than very specific. Spelling and capitalization can make a huge difference. Once when I was helping one of my math students who was extremely frustrated because the student knew their answer was correct but the computer would not accept the answer. In a math problem there are many ways to write a correct answer. Such as say for example: Factor the number 12:______.
A student could be perfectly correct in answering 12x1, 6x2, 3x4, 12*1, 6*2, 3*4, (12(1), (6)(2),
(3)(4). 3*2*2*1, (3)(2)(2)(1), 3x2x2x1 or 1,12,3,4,6,2 depending on what mathematical symbols you wish to use. Any of these is a perfectly correct answer mathematically. Although once students get into high school they have to stop using x for multiplication, because the symbol x can get mixed up with the variable x. In addition to all these symbols for multiplication and factoring, I have seen another symbol used for multiplication - a solid large black dot.
The limitations of an essay questions is that it takes longer to grade. Also usually a few essay questions will take the whole test period. Grading an essay question can be subjective unless a rubric is used, and sometimes another teacher's opinion is necessary especially on the borderline answers, and the grey areas. Most of the time it's easy to tell excellent from poor answers, but to tell a high good, medium good, or low good is hard. Even to tell apart a superb excellent from an excellent is difficult.
Just to contrast the difference between constructed response assessments and fixed response questions I as a student am very good at taking multiple choice fixed answer response questions, however I cannot take a true/false test to save my life. I read too much into a true/false test.
When I took the teachers licensing exam I saved 1 hour of the test just for the essay question and it was a doozer! In order to answer the question, I had to know formulas, know geometry, know angles, logic, trigonometry, algebra, and almost every math subject just in order to solve the problem.
With fixed response assessments such as multiple choice and true/false a writer can test many items in a single exam. They can be easily graded on a computer. I used to do an item analysis on every test question I wrote. If practically every student got a question wrong I would not count that question in the grade. I figured either I didn't teach the concept, the students misunderstood what was asked, the question was ambiguous, or the numbers weren't good numbers. I believe these tests should use relatively easy numbers 2, 5, 10, etc not 9, 7 11, weird numbers. If I want to specifically test if students can multiply or divide by higher numbers that should be on a separate test question. Personally I don't give true/false tests. I think they are too susceptible to all or nothing judgments. I can see the exception to every rule.
No comments:
Post a Comment