The Learning Blend
One of the questions I get most often when training teachers is, “How do I prevent cheating on a Blackboard test?” I’ve come up with a pretty thorough presentation for teachers with such concerns, and I know that I’m pretty confident about the practices I used in my own class when I taught.
So, when I was recently charged with designing blendedschools’ upcoming professional development course, Teaching in a Blended Learning Environment, I knew that I would need to address this. As I wrote this lesson, I figured it would be useful to link some sort of article bringing together all of the tricks for preventing cheating in Blackboard.
After a fruitless Google search, I found that there was no in-depth guide for Blackboard Learn 9.1 (blendedschools’ current version). What’s more, the guides that did exist for previous versions were very brief. It occurred to me that K-12 teachers (especially those who will be taking my course) need more than technical tips (though that’s helpful as well); they need practical advice on test design, pedagogy, and management to address their specific situation.
So, I’ve set out to write this blog post bringing together all of the Blackboard anti-cheating tech with my best advice for practices garnered from my experience and what I’ve learned talking to others. This article assumes that we are talking about relatively high-stakes summative assessments. I will get it out of the way and say that I’m not a huge fan of such assessments playing a big role in K-12 education. That being said, many people use them (I have a few times myself), and so I’ll share what I know. I’ll be sure to include a rant about this issue near the end of this article, but for now, lets get to the tools and techniques for preventing cheating in Blackboard (many of which will apply to other LMS’s).
Technical solutions are one piece of the puzzle, and they’re very effective. A few of these are simple “click this button” sort of ideas, whereas others are pretty advanced tools that are typically reserved to “power users”. I will cover all of these, starting with the simpler tricks and working up to the more involved ones.
Many of the most basic tips are touched on in these tutorial video (Creating Your Own Blackboard Test and Deploying a Test). This blog will cover the basics again but also address classroom management and test design concerns for the teacher using online tests for a blended classroom.
Randomizing the answers on some of your questions will prevent two kinds of cheating. First, glancing over someone’s shoulder to get his or her answers is difficult when they’re in different orders. Second, when students attempt to exchange answers (i.e., a hastily scrawled note reading, 1. A, 2. C, 3. D, 4. C, etc.), such answers will be useless. Even if students are allowed to work together, randomizing answers can be nice because the students will need to discuss the content when working on their questions instead of just exchanging letter and number combinations (obviously, there are better forms of collaboration than multiple choice tests, but this may have its place).
Each time you create a multiple answer or multiple choice question in a test, there is a checkbox under part two of the form that reads, “Show Answers in Random Order”. Simply check this box to have Blackboard randomize the order the answers are shown each time a student loads the question.
Be careful not to do this with “all/none of the above” type of questions (though you could still use “none of the others” or “all of the others”, but, logically, not both).
Randomizing Question Order
Cheating by glance or answer exchange becomes even harder with randomized questions, with the added advantage that it works for all question types (not just multiple choice and answer). Each student who loads the test will get a randomly generated order of questions. Keep in mind that this won’t necessarily be a unique order with a small number of questions (in a classroom of twenty-two students taking a five question quiz, there will probably be a repeat).
This level of technical cheating prevention takes place on the test options level. You can toggle this option when you click the “Create Assessment” button to first deploy a test or anytime after that by clicking “Edit the Test Options” from the test’s drop-down menu. In part 6 of the form, check the box that reads “Randomize Questions”.
This cheating prevention tool will not work for a test where order is important. For example, if you have the students read a passage in the text of question one, then answer four more questions about that passage, it will be out of order. You can, however, have the students open a reading passage in another window or use a reading passage on paper. However, if there are multiple passages on the one test, I would not randomize under any circumstances, as it would be pretty harrowing to have to jump from passage to passage.
If you are testing your students remotely, setting a time limit can reduce the students’ ability to consult other sources. That being said, it doesn’t prevent it (especially for a technologically savvy student using multiple computers). If a test is given remotely, I prefer using the pedagogical cheating prevention tools (given near the end of this article) to design a test with higher-order thinking questions that assume the student has his or her materials open. Our students are likely to live a lifetime with nearly constant connection to the Internet, so we should design tests that challenge their learning under these circumstances.
That being said, I promised to save my soapbox for further down the page. Here’s how to set a time limit if you want to. On part two of the test options form, check the box labeled “Set Timer”. You can then use the drop-down boxes to set any time limit you like. Remember the old rule of thumb, your students will probably take about three times longer to do something than it takes you. This is, however, teacher folk wisdom and certainly not perfect. You’ll have to use your judgement on appropriate time.
You can use the auto-submit option to decide what happens when the timer expires. If it’s off, students can continue, but their submission will be marked with the amount of time it took them to complete the test. If it’s on, then the students will automatically submit whatever’s finished when the time runs out.
Decide what your policy will be regarding timer questions in advance to prevent conflict later. If a student tells you that he or she accidentally loaded the test and let the timer run down or that he or she did not know there was a timer, what will you do?
Most of these cheating prevention methods addressed so far has been geared towards preventing “in the moment” cheating. However, with an online course or blended course with online assessments, it’s very important to have a mind towards preventing premeditated cheating such as previewing or printing the test. The options you select for Password, Test Availability, and Feedback are essential to make sure you do not have copies of your test floating around among students.
I always set passwords for my tests in K-12 classes. It’s important to set passwords that are not easily guessed. You might think it’s cute to use your mascot or one of the topics addressed in that unit, but this is a great way to make your password nearly useless. You also probably don’t want a password that’s difficult to type. I would tend to use fairly rare but easy to spell dictionary words like “invincibility”. The likelihood of a student guessing this is near zero, but it’s fairly easy to communicate at test time.
In a blended classroom, once the password is distributed, I would treat it the same way that I would treat the distribution of paper tests. The password is written on the board, and, from there on out, students cannot talk.
If you are giving the same test at multiple times, be sure to change the password before the first group finishes the test to avoid the password being distributed to subsequent groups.
Setting the password is simple. In part 2 of the test options form, the last checkbox is labeled “Password”. Check that box and type your password in the text field below.
A second option for controlling previewing and printing is controlling test availability. You can use this instead of or in addition to password protection. I preferred password protection myself, because I taught my course sequentially with my Blackboard course serving as a sort of schedule. This way, students knew when a test was coming up but were unable to access it.
The option to manually make your link available or unavailable is found in the same place as the option to have it appear and disappear at pre-scheduled times. In part 2 of the test options form, “Make the Link Available” is your first option. Further down, you can check “Display After” or “Display Until” for pre-scheduled releases.
As with the scheduled availability everywhere in Blackboard, I encourage you to avoid setting up too much of this too early in your course. You don’t want to be caught in a situation where cancellations force you to change the display properties on dozens of items.
The way you display feedback is one of the more difficult decisions in preventing cheating. For low-stakes formative assessment (a much better role for Blackboard tests, in my opinion), you can probably trust your students to work on the honor system. In this case, you get the huge benefit of letting students know right away how they did while their performance is fresh in their minds.
But, if you are working in a high-stakes environment on a summative assessment, you have several concerns. First, students in a blended environment might be able to look over one another’s shoulders if they finish at different times. They can also exchange answers between classes. It’s even worse in a remote environment as students can print out or screen capture all of the correct answers and exchange them at their leisure.
So, if you are concerned about exchanging answers, I would advise you to leave all of the boxes in part five of the test options form unchecked. This means that students will get zero feedback when they submit the test. Then, once all of your students across all sections have completed the test, you can give them the feedback.
If you plan to use the same test in subsequent years, the way you handle this feedback is important. If you click the boxes and let students look at the feedback whenever they like, it’s just like letting your students take the test home with the correct answers on it. In future years, students will have tests from older siblings and friends.
To avoid this, proctor test feedback the same way you do test taking. Only make the test available while students are in the room with you and turn the feedback off again when they leave. You can even use adaptive release to assure that only the students in your current section can see the items.
You can present questions one at a time and prohibit backtracking as well. This will make it nearly impossible to look at another student’s screen for answers. These options can be found under part six of the test options form.
That being said, I am hesitant to put a student in the situation where they can’t learn something during the test and go back. It seems like a recipe for very high-test anxiety.
Mel Brooks famously suffered from high anxiety...
Source: Towpilot, wikipedia
I do not recommend using this option. This option makes it so a student can only load the test once. In theory, this prevents a student from loading and previewing the test, then going out and studying some more.
In practice, what this does is create headaches. We all know that browsers crash and people accidentally hit backspace and leave tests (oh, how I hate the backspace as a hot key issue). With Force Completion enabled, however, every error becomes a request to the teacher for assistance in resetting the test and the possibility of starting from scratch.
If you insist on using it, however, it’s on part two of the test options form.
And now we come to the aforementioned “super user” tools. I hope that won’t scare you off, but these are the sorts of tools I tend to see teachers using after they’ve become familiar with the basic interfaces of Blackboard. In essence, I see this being a logical step to take in the second or third year of using a course.
If you’re very concerned about cheating, you might want to manually create multiple forms and have students take different forms of the test. This video tutorial explains how to do this.
This approach certainly works, and is also useful for differentiation. However, if you really want to go high-tech on the test forms, consider using random blocks...
Random blocks are sets of questions that are randomly pulled from a pool of questions in Blackboard. With this tool, students won’t just get questions in different orders, but they’ll get completely different questions.
This is a several step process that may be appropriate for a high-stakes test (like your course final) that you give each year for several years. This is especially apt if you have several teachers delivering the same course in Blackboard.
The first step is to build pools of questions. To do this, follow this procedure:
From here, you have several options. You can either make the questions here by clicking "Create Question" (like you would in building a test), or you can use the Find Questions button at the top of the canvas to bring in questions from pre-existing tests.
Ideally, you want to build several pools of questions as appropriate. For example, for a final in an English class, I might have four pools from which I draw for my final exam:
Each of these pools has several dozen questions in them (in total, much more than the sixty questions I plan to put on my final).
Once my pools are complete with all of my questions, I then go through the normal procedure of setting up a test, but instead of clicking Create Question, I click the Reuse Questions button. From that dropdown, I select Create Random Block. For the example above, I’m going to start by selecting “Reading” from the “Pools” listed at the left, and then “All Pool Questions” from the list of “Question Types” below that.
The random block will be added to my test. There, I can set the number of questions to appear in that block, and the number of points to assign to each question. For this example, I might have four blocks of fifteen questions each. That way, each student gets fifteen questions one each subject, but they all get different questions, making any attempt to share answers pointless.
Using this process and keeping a deep pool of varied questions is really the most thorough solution to preventing students from cheating off of old tests in your course.
Solutions Through Design and Management
Technology can only go so far. Effective classroom management (in a blended classroom) can go even further.
If at all possible, I think any important test should be proctored. In a blended classroom, this is easy. For a virtual program, your school’s policies may not make it possible. If working together, using notes, or using Google makes your test invalid, then it is an invalid test without proctoring.
Good proctoring means moving around the room with as many computers in sight as possible and watching for the following:
If you skipped down to this point, consider checking out the advice above on test availability and passwords. Managing these is very important for proctoring on-site online tests.
Well written higher-order questions are much more difficult to cheat on than a typical lower-order question. Check out these links for some solid advice on higher-order questioning in objectively-scored exams (if you’re reading this article as part of my blended learning course, you’ve already seen these links in Unit 2, Lesson 2):
I try not to make a habit of making moral judgements about students. Most of us do some dumb things when we’re kids. While cheating seems awful, most people have a psychological need to see themselves as basically moral. So, how do kids justify cheating to themselves?
It all goes to pressure. Kids feel the need to get good grades because of parents, sports, college, whatever. Given this pressure, cheating seems like a good way to alleviate that, especially on tests which they know to be inauthentic and easily cheated.
So, we have two approaches. The first is reducing pressure. Why are tests worth so much in your class, given those keywords above (inauthentic and easily cheated)? I’m not saying don’t use them, but I’m saying that cheating will happen if students’ scores are primarily determined by tests.
The other thing we have an obligation to do is to teach good citizenship (both analog and digital).
As promised, I’ve saved most of my proselytizing for the end. If you’re still with me, scan back up the page. Look at all of the steps being invested in preserving the sanctity of your high-stakes test. Doesn’t it seem wrong somehow to spend so much energy in what’s basically an antagonistic relationship with students?
I’m not saying all testing is bad, but think about your own relationship with testing.
Standardized Tests -- How well do you think the questions on your teacher certification test reflected what you need to know in your day to day job? How good did it make you feel as a human being? How professional did it make you feel?
Personality Tests -- If you’ve done online dating through any of the less sleazy sites (OKCupid or EHarmony), you know that people enjoy testing when it tells them useful information about themselves and the things they want to do.
Self-Check Quizzes -- When we’re learning something important, we might do some self-quizing to make sure we know what we’re talking about. When I first learned to train in Blackboard, I gave myself the Blackboard Basics Quizzes to make sure there weren’t any gaps in my knowledge. I didn’t mind this, because I knew I would have to give an authentic performance later.
So, what are your tests more like? Are they like standardized tests: difficult hurdles carrying little resemblance to the real world but holding a huge influence over a student’s grade? Or are they like the good tests: assessments that can be taken repeatedly to aid a student’s metacognition and certify him or her as ready for further learning?
Students aren’t going to cheat on the latter form of test, and you’ll spend a lot less energy fighting against them when you should be fighting for them.