Variable difficulty tests
Teaching May 9th. 2008, 10:44pmMaking a test for students can be a hit or miss performance. I had to make several tests today, and I went about making them in a different way than I had before. It worked out pretty well.
In my lowest class, I set up a modified multiple choice test. The twist was that the multiple choices were definitions of different words. The listening part of the test was me saying the word that they had the definition for. The students copied it down for each number as a spelling test on top of choosing the definition. This has several advantages over a standard multiple choice test.
First of all, none of the students could go faster than I let them, so I could give all the students time to look over their answers and try to find the correct response. It also tests multiple skills at the same time. Listening, reading, spelling, and comprehension. Even if students couldn’t spell the word, they had enough time to pick the correct definition.
Also, because of me being the only person knowing the answers before I started the test, I could control the difficulty of the exam as it progressed, question by question. Since I pulled all the definitions from their books, I knew all the answers were valid. Some words are much harder to spell than others, and some have much harder definitions. By manipulating the difficulty from question to question, I made sure every student was being challenged without frustrating anyone or causing them to quit.
The downside was that I had to make the key as I was announcing the words for each question. When you make the test yourself this isn’t so difficult, but I also loaned my test to my coworker, who might have a harder time figuring out which hints correspond to the student’s vocabulary in the book.
In my other classes, I used a repeating table structure that modeled one whole question and answer before I left gaps that the students had to fill by modifying their patterns. This helps remind students what I am looking for them to do.
Just to see if the example made a big difference, I left one table in the test without a set of examples, and gave it more complete directions with English and Korean. Students had more trouble with the section with Korean directions and no examples than the sections with English directions and simple examples. This is because they don’t READ directions, but DO use the examples. In the future, I’ll include examples if suitable for the questions and levels.
I had a frantic day of testing, then grading in the next class. I handed tests back to the students as they headed out the door to give them homework. They had to get their tests signed. There was a very typical spread for all my classes, but there were less people completely bombing out, so I think my tests were better explained this time around.
One Response to “Variable difficulty tests”
Leave a Reply
You must be logged in to post a comment.










May 10th, 2008 at 1:59 am
Wow. Impressive. That’s quite an advanced assessment technique. You’re more capable than two-thirds of the folks out there with education degrees, I reckon.