Tests are not on the Level.
Tests, particularly those of the Multiple Choice variety, are a rigged games that claim to measure student understanding but are really a brutal shell game in which test writers show the correct answer and then hide it within three other ‘distactor’ choices.
Students take all sorts of ‘tests’ from the lowly quiz, to longer unit tests, to High Stakes End of Course tests; then there is the mother of them all – the ACT and SAT. To me the most frustrating part of testing is that students assume it’s all on the level. They assume that if they get a 100% they are smart and if they get a 50% they are not. Even worse is when parents, teachers, and administrators (of all types) believe the same thing. Rather than a tool to assess and improve student understanding, most testing is used to simply to rank students. The problem with this is that eventually “the test“ becomes the reason for the teaching and not a tool to help improve instruction.
My Sales Pitch to Students
I try to break the testing anxiety with students by telling them it is all a game designed to give them anxiety. I start with this – “Did you know that the official title of a Test Writer is Psychometrician? That’s right! They are literally psychos.” I follow this up by asking how many answer choices are on a typical MC test question. Since they all know that there are 4 answer choices and only 1 correct answer I hit them with this – “So that means that they want you to get the question wrong 3 times more than get it right!?” And then “If I have done my job you should be able to get the correct answer pretty easily which means they have to try to TRICK YOU INTO PICKING THE WRONG ANSWER!!!”. At this point I see lightbulbs going off in their heads. The students begin to realize that they might know the right answer and still get the question wrong. Those “psychos” are trying to play on their pre-existing anxiety and self doubt to trick them.
Look, I understand there is a science behind test creation but lost in the science is the lived experience of the students. I’m sure that test creators are doing the best that they can but I do not think it is a coincidence that as the era of high stakes tests has progressed I seem to have far more students each year claim to have test anxiety. When a field test question for our US History End of Course exam is thrown out if 80% of the students answer it correctly you cannot convince me that the question was eliminated for any reason other than they need to rank the students – not assess what they know.
False Objectivity of the Test
I have a required Standard 1 Common Assessment coming up next week. The test will consist of 55 Multiple Choice questions. Nearly all will have 1 correct answer out of 4 choices but a handful will have multiple correct answers out of 6-7 choices. When students complete the test they will be given a percentage score and some will score well and some will not. What does that number represent? Is it their understanding of the roughly 200 years of American History from European discovery to the War of 1812? What if there was information on the test that I didn’t teach well? What if they knew the answers to questions that were not included on the test? What if the test creators were really good at creating distractor answer choices? What if they didn’t sleep/eat/go to the bathroom before the test? Do we really think that an 83% on a single common assessment means that the student knows 83% of the content of that particular unit?
I think that the specificity of the multiple choice test allows teachers to feel a false sense of objectivity.
With this specific number in hand we can turn the score back on the student and tell them you didn’t study hard enough. In reality that number cannot possibly be that specific though. Sometimes students guess correctly. Sometimes students misread a question. Sometimes the vocabulary term I used in class is different than the one on the test.
For the past year I have been attempting to change this dynamic with my testing. My first goal was to turn the Multiple Choice tests into a formative test rather than a summative. This means led to several important realizations.
- Tests need to be broken up into smaller units.
Since my 55 question tests are mandated I can’t just skip them but I can make the data more useful for me and the students. I do this by breaking the test down by indicator or substandard. Standard 1 had 6 substandards and each question is linked to 1 or more indicators. This allows me to give 6 grades rather than 1. (I use a “Powered Up XP” system you can read about here.) This allows students to see where they actually struggled and where they excelled.
2. Tiers work better than raw percentages.
I can determine the number of questions missed by a student for each indicator. Based on the raw percentage score I assign them a 1-4 tiered (level) score using this chart.
I use XP grading and on this chart I referred to XP as “LP” for reasons we don’t need to discuss here. The point is that based on the percentage score I awarded between 150 and 300 XP. Because I know that a student might miss a question they should have gotten correct because the “psychos” successfully tricked them I give a bit of grace; students can receive full credit for scoring above a 90% raw score for the indicator. Students also get at least 1/2 credit (150XP) regardless of the score between 0-60%. Tiers give a bit of a cushion for the anxious but also don’t reward the good guessers as much.
3. Reduce the stress
I see many students struggle because they become focused on answering EVERY question correctly which then leads them to mental exhaustion later in the test. The tiered system gives students permission to miss a question or two and still score well. This allows them to not become overly focused on a single question which I hope allows them to score better on the other questions. This also reduces some anxiety because they know that I know they will make some mistakes and its ok. Believe it or not I don’t think that a student intentionally wants to do poorly on a test.
Use the Dang Data – Allow Credit Recovery and Retakes
What is the point of taking a test if we simply factor it into the overall grade point average and then move on. Allow the students to learn from their mistakes! Allow them to do something with the information that they have learned! Allow them to analyze the data on their own test and to grow from it. My wife used the term fail forward the other day and I think that is a correct term.
I allow students to recover their full credit as long as they demonstrate to me they have learned from their mistake and now understand the information. Again, this reduces the test day anxiety. It also lets students know that I see them as individuals and value their humanity. They are not just a test score. I have modified a Blake Harvard (@effortfuleduktr) form that forces students to work with the questions they did not answer correctly. You can download the form and make a copy if you would like. Turn these traditionally summative end points into a stop along the way. Make them a growth document rather than a dead document.