Considerations for Online Quizzing and Testing | University of Illinois Springfield

Why Use Quizzing and Testing?

A primary reason is to assess students’ progress. Note that this reason is different from the need to assign a grade at the end of the term.
Quizzing and testing also encourage progress. Students who know that the final grade will depend in part on the quiz may become motivated to learn.
The results of quizzing and testing provide feedback to the instructor about their success in helping students to identify the key issues.

It is useful to identify what students know. Prerequisites don’t always provide an indication that students are ready for your course.
Introduction of topics through early testing may help students focus on important topics that will be introduced in the course.
An exam before instruction begins establishes a base line performance against which performance at the end can be compared.

Quizzing and Testing after instruction may measure student learning. Creating quizzes and tests that measure learning and not some other outcome is, of course, not easy.
In some cases the results can be used to determine whether or not a student qualifies to take a more advanced course.
Demonstration that a body of material has been learned may show that the student is ready to apply that knowledge.

Students receive immediate feedback if the software that is used automatically grades the quiz.
Students can take online quizzes and tests from anywhere they have an Internet connection.
All students need not take the quiz at the same time, allowing flexible schedules.
For quizzing programs that automatically grade quizzes, the scores are automatically recorded.
For paragraph or essay questions, the responses are legible since they are typed and not hand written.

The first line of defense is to use a quizzing program that requires user IDs and passwords.
Control the pressure to cheat by making each assessment worth only a few points.
Offer the quiz in a proctored lab where IDs can be checked.
Use alternative assessment methods that don’t rely on multiple choice, TF and fill in the blank questions.

TF questions are relatively easy to construct.
Guessing produces a score, on average, of 50%. By most assessment standards, this is not a passing score.
It is difficult to write TF questions that test “gray” areas.
The software is easy to write since there are few choices to make. If the question is properly written, the answer a student has selected is either correct or incorrect.

Good distractors are hard to write. Non-trivial MC questions are extremely hard to write, and are prone to misinterpretation.
MC questions reduce guessing.
MC questions can provide diagnostic information about the state of student’s knowledge.
Some people believe MC questions are less effective at testing higher order thinking.
Software is easy to write.
Quality software can allow for more than one correct answer per question, and can assign point values to each choice in the question.

Essays can provide in-depth assessment of knowledge and understanding.
Guessing is reduced.
Good for assessing higher level thinking.
Software to grade is very difficult to write.
Some programs generate a score based on counting key words.
Some newer programs are using artificial intelligence to attempt to judge writing.
At a minimum, software should collect student essays online and present them to the instructor for grading.

Software should allow instructor to specify more than one correct answer. There are very few short answers for which there is only one possibly correct answer.
In some instances you want the answer a student types to be judged without regard to case. The answer judging engine should ignore the case of the students’ answers. For example, “CLINTON” and “Clinton” should be judged as correct.
In some instances you want the answer a student types to be judged on the case of the letters. The answer judging engine should compare the case of the students’ answers to the case of the correct answer. For example, “NaCl” is correct while “NACL” is not.
The software should trim away extra spaces and nonprinting characters such as Tabs that students have accidentally inserted in their answers. For example, “Clinton” and “Clinton ” should both be marked correct.

Most quizzing and testing packages do not reside on the instructor’s computer. They require a College Web Server.
Quizzing and testing software package.
Faculty and student connections to the Internet.