A good friend emailed me a few days ago and asked for some suggestions on writing a test. Students were mostly getting good grades
on her tests but she wasn’t sure that they were learning as much and as deeply
as she wanted.
I think testing is a
teaching topic that we do not talk about frequently enough. Too often, we are probably afraid of
exposing our weaknesses. Few people are
trained to write good test questions. She
was creating an accounting test but I believe the basic discussions around
testing are pretty much universal across the various disciplines. Here were the thoughts that I sent to her.
Rule 1 – Testing is just
about the most important thing a teacher does each semester because it sets the
tone for the students. It tells them what you really want and
expect from them. If you test memorization, they will learn to
memorize. If you test mechanics, they will learn nothing but
mechanics. Think about what you want them to learn and then test that
way. Here’s where you need some type of mission statement – “I am
teaching my students to . . . “
Rule 2 – Because it is so very important, never turn your testing over to a busy grad student in some unknown college in some unknown place. That is who writes most test banks. For what you know, those people might well know less about excellent teaching and testing than my cat. If you were a basketball coach at one of the local universities, would you turn the practice sessions over to the janitor because you were uncertain about running them? That is nonsense. But you turn your testing over to someone who doesn’t know your school or your class or your students. Learning good testing takes some practice but you can always do it better than the grad student at an unknown college in an unknown place.
Rule 3 – A test has one major purpose – to differentiate the A students from the B students and so on. Differentiation is the purpose. And, of course, to have the students believe that your differentiation was fair. 1/3 of the questions should be workable by the A, B, and C students but not the D and F students. 1/3 of the questions should be workable by the A and B students but not the C, D, and F students. 1/3 of the questions should only be workable by the A students – that is how they prove they are A students. That is how you make them feel good about themselves.
Rule 4 – If you use a test bank, always realize that most test banks are for sale on the Internet. The students often buy them for practice purposes. No test bank is absolutely safe as far as I know.
Rule 5 – Always be willing to curve. I tell my students that I grade the tests and then I assess what is excellent work, good work, average work, and so on. I next curve those tests (and only those tests) that deserve to be an A so that they get an A. That is where my professorial judgment comes in to play. I decide what is excellent, what is good, and so on. If I judge a 78 to be excellent, I curve that to an A. If I judge a 95 to be good (a different test obviously), I curve that to a B.
Rule 6 – I am a big believer in the wonder of puzzles. Where possible, I try to write test questions that are basically puzzles. I also believe where possible that questions should resemble real life. These are 20 year old adults – they are old enough to vote and old enough to go to war. Don’t make test questions look like test questions from their high school days. Make them look like real life with some kind of twisted puzzle logic. Questions that incorporate “what if” are usually good as are questions that ask “how would this have changed” or “how would you decide between these two options?”
**Rule 2 – Because it is so very important, never turn your testing over to a busy grad student in some unknown college in some unknown place. That is who writes most test banks. For what you know, those people might well know less about excellent teaching and testing than my cat. If you were a basketball coach at one of the local universities, would you turn the practice sessions over to the janitor because you were uncertain about running them? That is nonsense. But you turn your testing over to someone who doesn’t know your school or your class or your students. Learning good testing takes some practice but you can always do it better than the grad student at an unknown college in an unknown place.
Rule 3 – A test has one major purpose – to differentiate the A students from the B students and so on. Differentiation is the purpose. And, of course, to have the students believe that your differentiation was fair. 1/3 of the questions should be workable by the A, B, and C students but not the D and F students. 1/3 of the questions should be workable by the A and B students but not the C, D, and F students. 1/3 of the questions should only be workable by the A students – that is how they prove they are A students. That is how you make them feel good about themselves.
Rule 4 – If you use a test bank, always realize that most test banks are for sale on the Internet. The students often buy them for practice purposes. No test bank is absolutely safe as far as I know.
Rule 5 – Always be willing to curve. I tell my students that I grade the tests and then I assess what is excellent work, good work, average work, and so on. I next curve those tests (and only those tests) that deserve to be an A so that they get an A. That is where my professorial judgment comes in to play. I decide what is excellent, what is good, and so on. If I judge a 78 to be excellent, I curve that to an A. If I judge a 95 to be good (a different test obviously), I curve that to a B.
Rule 6 – I am a big believer in the wonder of puzzles. Where possible, I try to write test questions that are basically puzzles. I also believe where possible that questions should resemble real life. These are 20 year old adults – they are old enough to vote and old enough to go to war. Don’t make test questions look like test questions from their high school days. Make them look like real life with some kind of twisted puzzle logic. Questions that incorporate “what if” are usually good as are questions that ask “how would this have changed” or “how would you decide between these two options?”
I recently gave a test
in Financial Accounting and another test in Intermediate Accounting II here at
the Robins School of Business at the University of Richmond. If you would like to get a copy of either of
those tests (just to see what I do), drop me a note at Jhoyle@richmond.edu.
**
Because I am writing above about puzzles, here is a quick story that I liked. A good friend of mine recently gave me the book Mindset – The New Psychology of Success by Carol S. Dweck. I opened the book to the first page and was fascinated to read the following which seemed to have come directly from my own experience:
“When I was a young
researcher, just starting out, something happened that changed my life. I was obsessed with understanding how people cope
with failures, and I decided to study it by watching how students grapple with
hard problems. So I brought children
one at a time to a room in their school, made them comfortable, and then gave
them a series of puzzles to solve. The
first ones were fairly easy, but the next ones were hard. As the students grunted, perspired, and
toiled, I watched their strategies and probed what they were thinking and
feeling. I expected differences among
children in how they coped with the difficulty, but I saw something I never
expected.
“Confronted with the
hard puzzles, one ten-year-old boy pulled up his chair, rubbed his hands together,
smacked his lips, and cried out, ‘I love a challenge.’ Another, sweating away on these puzzles,
looked up with a pleased expression and said with authority, ‘You know, I was hoping this would be informative.”
Yeah, puzzles can make
class and teaching a whole new ball game.
