NO MORE UNIT TESTS...What do you think?

My students are bright and motivated. Most work hard and prepare for class and tests. They perform extremely well on district-wide tests and my own classroom tests. However, I see real weaknesses on cumulative assessments requiring high levels of application. My students simply do not retain the content knowledge. I want to restructure my course to exclude "unit tests" and include only cumulative assessments. I'll share my early ideas here, and I would love to hear your experiences.

In the past, I have given a traditional unit test at the end of each topic of study. The tests are comprised of 12-16 multiple choice questions and 5-7 free response questions. Students are not allowed to use calculators on the multiple choice, but they are free to use them on the free response. I have used online resources, exam test prep books, textbooks, and workbooks to get ideas for the questions. I also use released AP exam questions on the topic if they are available. In the end, I have between 20 and 25 questions to assess their understanding of that particular unit of content, for example, atomic structure.

My revised idea is regularly scheduled comprehensive tests assessing the student's ability to apply all of the content learned so far in the semester. The test period is 90 minutes. I envision my new tests having 20 multiple choice and 5 free response questions. The trick will be to write or find good questions that apply multiple concepts without requiring knowledge of ALL of the semester's content. I hope to find many in the released AP exam questions, but there simply aren't enough released questions for the re-written exam.

I would love to hear your thoughts on any of the following questions.

  • Are your tests cumulative?
  • How long are they?
  • How often do you give a test?
  • Where do you find the best questions?
  • Do you have other tips for helping students retain knowledge?

 

Join the conversation.

Comments 9

Deanna Cullen's picture
Deanna Cullen | Mon, 07/18/2016 - 11:46

Hi Allison

Several years ago, as part of a district directive, we developed 6 week assessments for my courses. We have 90 minutes for the semester and final exams, which are cumulative. We only have 60 minutes for the others. 80% of the test is from the previous 6 weeks and the other 20% is cumulative. This seems to help students find more success on the semester and final exams. I think that by constantly revisiting and reviewing past material, retention is improved. 

 

 

Allison Tarvin's picture
Allison Tarvin | Wed, 08/10/2016 - 07:01

Deanna, how do you select the topics for the 80% review? I have selected commonly missed questions or misconceptions in the past to check for new understanding. Is there one idea that you put on every test? How much of your test is free response?

Sorry for the tons of questions!

Deanna Cullen's picture
Deanna Cullen | Thu, 08/11/2016 - 09:19

Good questions. I keep a bank of previously used questions and try to hit a variety of learning targets on the 20% review portion of the test. If there are topics that stand out as a weakness on a test, then I will try to focus on reteaching/reinforcing those topics so those questions will likely end up on one of the upcoming tests. I do not tend to reuse any specific questions on every test. As far as free response, I try to make those questions about 25% of the overall grade. 

 

 

Karen Collins | Wed, 08/03/2016 - 10:25

I try to do this within some questions - ie. in gas law questions, do some stoichiometry, or in periodic table questions, some mole conversions.  I would be afraid some students wouldn't see this as "fair", at least initially, and it would take a lot of changing of perception in the classroom.  But I do think it would reinforce the idea that student need to learn the material and that it doesn't just go away.  

To that end, I have also been looking into Standards Based Grading, where students have to meet benchmarks by the quarter end and have multiple times to demonstrate mastery, which would also accomplish this goal, although they would have to be motivated by grades to do this.  While most Honors kids are, many general chemistry kids are perfectly satisfied with a C or a high D, and I fear they would not take advantage of this opportunity.

Allison Tarvin's picture
Allison Tarvin | Wed, 08/10/2016 - 07:04

Thanks for your feedback. I have been using Standards Based Grading for the last few years after a district-wide initiative. Our assessments haven't shifted well. I'm hoping that the more cumulative style test will be more in-line with my gradebook now. I'm working very hard to "sell it" to my students and their parents.  In my Pre-AP class, the kids are buying in as they know they'll see the content again soon.  In my standard honors class...I've been too chicken to change this semester! :-(

Sean Fisk's picture
Sean Fisk | Mon, 08/08/2016 - 16:04

In a class where they are expected to take and pass a cumulative exam, like AP, this is a great idea! My only concern is that there is enough info in a unit that needs covering, can you viably squeeze review in on top without sacrificing the unit?   

My tests are slightly cumulative.  Anything they've done is fair game, but the focus is generally on the unit.  I have 83 minutes on a rotating block, but I generally plan on 70 minute tests in the AP format. ( 10-15 MCQ, no calculators, followed by several short and long FRQ)  I got in the habit of using released AP test questions last year, and I like it.   Before that I was using review book questions, and they just aren't the same.  The AP questions have a feel to them and are quite wordy and often cover more than one topic, so they are difficult to replicate.

I tell my kids to buy a review book early and to study study study, though they are very grade driven and it's easier to get them to study than when I was in public school in the states.  I give tests every 3-4 weeks.

Chem 1 is a different story though.  Tests tend to be unit only.   And you are right, abysmal cumulative test scores.  Something to consider.  Thanks.

Allison Tarvin's picture
Allison Tarvin | Wed, 08/10/2016 - 07:09

Sean,

Thanks for your feedback. I appreciate your ideas!  When you choose released AP questions, do you stick strictly to the redesigned questions? If so, where do you find enough? If not, how far back is reasonable, in your opinion?

 

Lori Maffe | Tue, 08/09/2016 - 13:30

Hi Allison,

We're trying something new in gen chem this semester. Each unit exam will have 5 MC questions related to previously taught material and 15 MC questions on new material; free response will only cover newly taught material. The key change however, is that questions on the final exam will be pre-coded according to the unit in which the concept was taught. Students who perform better on final exam questions from a particular unit than they did in the original unit test will be awarded "recovery points" to show that they've mastered the material, although a bit later than desired. 

Research has shown that this ability to "recover" on the final exam serves to motivate students to continue to learn and reduces failure rates. 

Allison Tarvin's picture
Allison Tarvin | Wed, 08/10/2016 - 07:13

Lori,

I love it! It is especially great to hear from a college professor. I'm going to try to figure out a way to incorporate this into my AP class. It seems like a much better motivator than applying a curve to a test.  I really hate curves. With a chance to recover, students will likely strive to learn the concept or correct a misconception rather than simply walk away grateful for a few extra points handed out rather than earned.