Depending entirely on the scantron machine
Most of the methods discussed in this module I have used in my teaching/grading procedures. I find that with very large classroom enrollment, the essay exam method is not going to work for me.
Almost all my quiz/tests are given on a scantron sheet. I have found that I've procrastinated in grading these simply because the machine in the main office is broken (more often than not). Next thing I know I have a huge pile...and unfortunately sometimes I have to do it the long way...by hand with a key or punched out scoring sheet. I have decided to make every attempt to use technology and web-based testing (such as used in this course) for my future assessment practices.
Joshua,
You bring up a great point. I am sure that other instructors have experienced time delays that relate to scantron machine problems, long lines to use the scantron machine, and errors in coding the key that cause errors in grading. I have experienced all of these in my teaching career. To your point, we must do more than find technology to automate the grading process. We must find the appropriate technology – that which will truly save us time and effort.
This is a great point. We have a fairly small student population, and my largest class has 20 students. While essay grading is extremely time-consuming, grading multiple choice/True False exams are often easier to grade by hand than to wait for the machine. Also, errors in machine grading are a major headache. I'd much rather see the student's answer for myself so that I can recognize errors in the key or unclear wording of test questions as I go.
We do not have a scantron machine. Instead we use computer software for creating test banks. From the test banks we create a test. We create several versions for the same test or create a new test every time we give it. If we use written versions we have an answer sheet key to use for grading. The same test can be put on-line and pushed out to the students via the network. This version is scored by the program immediately. Some of our publishers supply us with textbooks, test banks, software to test online, lecture presentations, web sites for extra learning, and a training/assessment network that scores results immediately. Many of us still use paper tests which are mostly Mutiple Choice, True/False, Fill-in-the-Blank which are faster to grade by hand. Some tests are application or production tests--student is given a list of tasks to perform on the computer. The saved document or printouts are graded according to the criteria set in the instructions. I think I would be really frustrated if I had to use a scantron machine that maybe was working and maybe not every 5 weeks.
I wanted to explore a related but more pedagogical topic.
Scantron machines are convenient. However, they are susceptible to failure as we've seen in the above comments. They also allow us to fall into a common trap among introductory level classes.
Automated grading forces students into a tight construct that measures the instrument and not actual learning. This is true with true/false and multiple choice quizzes. These exams scratch the surface of learning with nominal- and ordinal-level testing instruments. This does not allow students to show the richness of their synthesis. It simply shows how well we've taught to the test.
I really do belive in essay tests, especially those that hinge on real-world concepts where students may borrow from their prior knowledge to create new knowledge. They take a long time to grade, so they must be spaced well throughout the term. However, they seem to work best for my students.
In a large lecture class, grading these exams is best left to the TAs. If there aren't any, or are not enough, then there can be a mix of focused essays, presentations, games, and group work. This will allow us to better gauge their learning.
The types of questions I use on exams are of a wide variety. I have used everything from scantrons to multiple choice, short answer, fill-in-the-blanks, and essays. I tend to use essays more for the College and Expository Writing courses that I teach, as the vast majority of the course hinges on essay writing.
I find that scantrons and multiple choice questions do limit our ability to perceive what students have learned. If they get the question right, it might have just been pure chance when they didn't know the answer and took a shot in the dark to pick the right answer. However, they may have truly known the answer.
That is where I bring in short answer questions and fill-in-the-blanks. This allows students to show a bit more of what they have learned by unlocking their memories and understandings and bringing them to the forefront on the test. In this way, I feel I am truly seeing that my students have learned and applied knowledge, rather than just possibly guessed at answers.
I have used all of the above methods that I mentioned, so I am not totally opposed to any of them. I just feel there are some that work better than others in terms of the way I feel my students should be assessed.
