Blog 6 - Construct-response and Fixed-response

Describe the differences between constructed-response and fixed-response written assessments. Describe the benefits of using both in eLearning. Finally, describe the necessity for a balance between teacher-graded and computer-graded assessment items.

As stated in blog #5 (http://maj-eln.blogspot.com/2014/02/blog-5-constructed-response-and-fixed.html), there are differences between constructed-response and fixed-response assessments.  Constructed-response items, often known as “fill-in-the-blank” allows the student must enter or write out their answer. Whereas in fixed test items the student selects their answer from the response options. The most common types of fix-response items are multiple-choice or true-false test items. However, variations of fixed-response include matching, ranking, multiple true-false, and embedded-choice items. (Refer to blog #5, http://maj-eln.blogspot.com/2014/02/blog-5-constructed-response-and-fixed.html, for further differences between these two response types.)

Both of these test construction options have benefits in the eLearning environment. With both constructed-response and fix-response the student enters their answer and the computer can automatically evaluate whether the student has answered the test item correctly or not and assign the appropriate score to that test item. The drawback, however, in a constructed-response is often computer programs use an algorithm that is based on letter recognition. If the student misspells the word or inserts an extra space, the computer would mark this as an incorrect answer. A way around this is for the teacher to have the ability to overwrite the computer’s score and assign the appropriate points following the test item’s scoring plan. In the online assessment program I’m familiar (Galileo, www.ati-online.com) with this is an easy process as I can see the students who have the incorrect answer and quickly evaluate the student’s response and enter the appropriate points.

In addition to a “fill-in-the-blank” type of constructed-response, a teacher may create a short or long essay to assess the student’s mastery of the stated performance objective. A short essay should be able to be answered in less than 10 minutes (Oosterhof, Conrad, & Ely, 2008). “How does the Moon’s appearance change during a four-week lunar cycle?” is an example of a short essay.  An example of a long essay test item is “What important contributions has space exploration contributed our everyday life? Provide at least two of these contributions. Explain why and how each contribution has impacted everyday life.”

There can be many fixed-response test items on a single test. A teacher can assess many instructional objectives since declarative and procedural knowledge is assessed. The computer scores these items quickly and accurately. Unfortunately, a student can easily guess and/or cheat on this type of test.

Even though a constructed response measures instructional objective more directly, a problem is that there should only be so many essay questions/prompts per test. Another problem with essay style test items in an eLearning environment is that students must type their answer and they might not be good typist. The students need to know how they are being assessed when answering such prompts, so a scoring plan should be provided to them upfront. The scoring plan must be well defined, not only for the student but also for the teacher (or whoever is reviewing and scoring the student’s response). High-stakes assessments, such as state standardized testing, has long written responses which computer algorithms can help score, but human intervention is still required. For a formative assessment, the teacher often must manually review this and score the student’s response. This can be a time-consuming process if there is a large class size. Additionally, inter-rated reliability is a factor. If there are four teachers giving the same test in which there are test items that must be manually scored by the teacher, will all four teachers be faithfully following the defined scoring plan? Should the teachers share the scoring (e.g., teacher A score teacher B’s test)? Will the scoring rotation delay the feedback to the student? Providing the student with timely feedback is important, not only in an eLearning environment but also in the traditional “brick and mortar” environment.

Speaking of feedback, the student can get immediate feedback on each test item or at the end of the assessment.  However, the test creator must ensure that the feedback is appropriate. In a constructed-response, the student is asked “What is 2 plus 3?” and they answer “5” the computer can display “great!” or “you’ve answered correctly.” If the student answered incorrectly, a message “Sorry, the answer is 5” along with an explanation of why. Providing correct/incorrect feedback can communicate misleading information to the students who select the correct answer, but for the wrong reason. Or at the end of the test, a final score is provided to the student.

The teacher must keep time constraints in mind. A constructed-response test is going to take longer for the student to complete than a fixed-response test.  A teacher must decide when it is better to have a test be administered online with all the test scoring conducted electronically and when a teacher-intervention is needed. Creating a test online takes time, depending on the teacher’s experience in test construction and their learning management system.

Some assessments are not suitable for an eLearning, such as in the case of needing to measure a performance based skill (e.g., dancing skills). A teacher needs to balance between constructed and fixed-response test types and those that might be more performance based. Performance based items will need to be teacher-graded with a clearly defined scoring plan.

When a test is created, the teacher should consider placing constructed response, especially essay types of test items, at the end of the test. This way the student can answer the easier items first and more challenging questions later. A written test should have a variety of fixed-response items followed by more challenging constructed-response items.

The test creator needs to have a balance between computer-graded and teacher-graded assessment items. This balance not only helps the educator but also the student. By providing a mixture of computer-graded and teacher-graded tests, the computer can quickly and accurately grade fixed-response and “fill-in-the-blank” constructed-response items, leaving the teacher time to review the only the incorrect constructed-response items and time to review and score the short and long essay constructed-response items. Students have the experience of easier to more challenging test items through the different types of test items as they progress through their test. It is important that both types of assessments provide the students with clear, meaningful, and timely feedback.

Work Cited

Oosterhof, A., Conrad, R.-M., & Ely, D. P. (2008). Assessing Learners Online. Upper Saddle River: Merrill/Prentice Hall.