There has been a lot of talk about online tests and how much time we should give students to take a quiz or exam. When we think about it our students have their books, computers, phone, etc. to look up answers so why give them a time limit? They are going to look up the answers anyway. Who really has the time to generate 50 or 60 questions for an exam to make it difficult for the students to lookup? Finally, if a student wants to cheat, they will and there is nothing we can do to stop it. Thoughts?
If one has all kinds of information at one's finger tips instantaneously, what does cheating even mean? We must make different kinds of tests. Sure as hell will be harder to grade. :-)
Analogy: Once pocket calculators became cheap, what was the point of testing addition, subtraction, and the rest?
You bring up a great point. There some online courses that have over 50 students, what kind of test are you talking about?
Details will be discipline dependent.
To use the arithmetic example, "word problems" already do what we need to do. I think that captures the drift more or less adequately. These problems cannot be solved with calculators alone.
Quote from: dismalist on March 03, 2020, 07:31:35 PM
Details will be discipline dependent.
To use the arithmetic example, "word problems" already do what we need to do. I think that captures the drift more or less adequately. These problems cannot be solved with calculators alone.
From previous discussions here, these are a class of "application" questions which make them difficult to google since they aren't about simply regurgitating facts. Another type of question is one which asks students to integrate information from two sources, such as comparing the ways character A from source A dealt with a similar situation to character B from source B. If sources A and B are far enough apart culturally, google isn't going to find any connection.
You bring up some great points in a perfect world, but who has the time to sit and develop these types of questions? And then you have to spend time developing a rubric and grading these questions.
Quote from: marshwiggle on March 04, 2020, 04:05:45 AM
Quote from: dismalist on March 03, 2020, 07:31:35 PM
Details will be discipline dependent.
To use the arithmetic example, "word problems" already do what we need to do. I think that captures the drift more or less adequately. These problems cannot be solved with calculators alone.
From previous discussions here, these are a class of "application" questions which make them difficult to google since they aren't about simply regurgitating facts. Another type of question is one which asks students to integrate information from two sources, such as comparing the ways character A from source A dealt with a similar situation to character B from source B. If sources A and B are far enough apart culturally, google isn't going to find any connection.
Quote from: HigherEd7 on March 04, 2020, 06:30:50 AM
You bring up some great points in a perfect world, but who has the time to sit and develop these types of questions? And then you have to spend time developing a rubric and grading these questions.
Well, that's called class prep-- it's part of your job. We all spend time making good assessments.
In some types of courses online exams are proctored to reduce the possibility of cheating.
In my graduate online courses for MLS we never even had exams. It was all projects and papers.
I use online quizzes as a way for students to demonstrate having done the reading. I set the timer usually as 1 minute per question +1 minute (all MC/ fill in the blank etc. questions). This gives them not enough time to find the answer if they haven't done the reading. Early on in the semester, I do have students who get "timed out" of the quiz. Then they usually catch on.
I don't think timing would work well for longer free response type questions
Another tip: (Note this has some accessibility ramifications, but anyway...)
Use screenshots of text instead of straight text so they can't cut and paste into google. (Especially for quotations from sources.)
How do you use this tool?
Quote from: marshwiggle on March 04, 2020, 07:50:27 AM
Another tip: (Note this has some accessibility ramifications, but anyway...)
Use screenshots of text instead of straight text so they can't cut and paste into google. (Especially for quotations from sources.)
Quote from: HigherEd7 on March 04, 2020, 08:01:23 AM
How do you use this tool?
Quote from: marshwiggle on March 04, 2020, 07:50:27 AM
Another tip: (Note this has some accessibility ramifications, but anyway...)
Use screenshots of text instead of straight text so they can't cut and paste into google. (Especially for quotations from sources.)
Just take a screenshot of text (for instance using the "Snipping tool" in Windows), and then in your LMS upload the image. (We use D2L, and you can insert an image as part of a question in a quiz.) Then in the quiz question, say something like, "In the text sample shown,....."
This won't work for students who are visually impaired and need a screen reader, but otherwise may be OK.
I use them and in the end I feel they are a time saver for me and are a decent way to test students, but I would not want to make it worth too much of their final grade. The reasons I like this approach include:
1) It is actually a time saver in the long run because you can generate and build a library of test questions over years. While they do take time to create you can re-use the questions from year to year and slowly build the database, adding new questions and getting rid of problematic or outdated ones. Since you can have the questions randomly selected from the pool (or randomly selected from various pools to ensure different topics are adequately tested), there is no major issue of the test being the same as last year and students getting old copies of tests. Obviously this can be done manually with paper quizzes, but computers automate this and are better at it.
2) The other reason this is a major time saver is that they can be automatically graded and you can choose to allow students to see where they went wrong etc. This is a huge time saver and more than makes up for the extra time taken to create the tests. Students also like this because they get to see the results quickly instead of waiting for me or a TA to grade and return tests.
3) While there is still room to look up answers, the time limit makes this difficult and I have found that it is not a major issue. If it were a major problem, all students would get exceptionally high grades, but that has not been my experience with them. I think the key is to create good questions that require the student to synthesize facts and theories. If you are asking them the definition ox X, it will be easy to look up. If you ask them what the influence of X would be on process Y, it would be difficult to look up within the given time frame. So, I think this approach to testing is relatively good provided you design the questions well. I find it much more challenging to develop good MC questions than I would have anticipated when I was a student, but with a little effort they are effective for evaluating students and more fair than longer answer questions.
Good question. Do you ever add test questions from the test banks?
Quote from: Kron3007 on March 04, 2020, 08:57:52 AM
I use them and in the end I feel they are a time saver for me and are a decent way to test students, but I would not want to make it worth too much of their final grade. The reasons I like this approach include:
1) It is actually a time saver in the long run because you can generate and build a library of test questions over years. While they do take time to create you can re-use the questions from year to year and slowly build the database, adding new questions and getting rid of problematic or outdated ones. Since you can have the questions randomly selected from the pool (or randomly selected from various pools to ensure different topics are adequately tested), there is no major issue of the test being the same as last year and students getting old copies of tests. Obviously this can be done manually with paper quizzes, but computers automate this and are better at it.
2) The other reason this is a major time saver is that they can be automatically graded and you can choose to allow students to see where they went wrong etc. This is a huge time saver and more than makes up for the extra time taken to create the tests. Students also like this because they get to see the results quickly instead of waiting for me or a TA to grade and return tests.
3) While there is still room to look up answers, the time limit makes this difficult and I have found that it is not a major issue. If it were a major problem, all students would get exceptionally high grades, but that has not been my experience with them. I think the key is to create good questions that require the student to synthesize facts and theories. If you are asking them the definition ox X, it will be easy to look up. If you ask them what the influence of X would be on process Y, it would be difficult to look up within the given time frame. So, I think this approach to testing is relatively good provided you design the questions well. I find it much more challenging to develop good MC questions than I would have anticipated when I was a student, but with a little effort they are effective for evaluating students and more fair than longer answer questions.
Quote from: HigherEd7 on March 04, 2020, 09:39:18 AM
Good question. Do you ever add test questions from the test banks?
Never. I would never use test bank questions on an exam. Maybe a few for a quiz or other formative assessment where they need to know the basics to come to class prepared.
For a few reasons:
One, they are almost always simple recall type questions that can easily be answered with Google or the textbook. I'm not going to put a factual recall question on an summative assessment like an exam.
Two, the answers are almost certainly available online (even for application or higher level questions) and savvy students will find them (Course Hero and other "study help" websites have exactly this sort of material available for free).
Three, the questions can be low quality/not on essential concepts/poorly written. I would have to go through each question to make sure it's not terrible.
Just write your own questions. You can use test bank questions as a resource for images/types of questions/etc., but take the time to write your own.
Yes, include a time limit for the reasons other folks have stated.
Thank you great points
Quote from: the_geneticist on March 04, 2020, 10:11:44 AM
Quote from: HigherEd7 on March 04, 2020, 09:39:18 AM
Good question. Do you ever add test questions from the test banks?
Never. I would never use test bank questions on an exam. Maybe a few for a quiz or other formative assessment where they need to know the basics to come to class prepared.
For a few reasons:
One, they are almost always simple recall type questions that can easily be answered with Google or the textbook. I'm not going to put a factual recall question on an summative assessment like an exam.
Two, the answers are almost certainly available online (even for application or higher level questions) and savvy students will find them (Course Hero and other "study help" websites have exactly this sort of material available for free).
Three, the questions can be low quality/not on essential concepts/poorly written. I would have to go through each question to make sure it's not terrible.
Just write your own questions. You can use test bank questions as a resource for images/types of questions/etc., but take the time to write your own.
Yes, include a time limit for the reasons other folks have stated.
Quote from: the_geneticist on March 04, 2020, 10:11:44 AM
Quote from: HigherEd7 on March 04, 2020, 09:39:18 AM
Good question. Do you ever add test questions from the test banks?
Never. I would never use test bank questions on an exam. Maybe a few for a quiz or other formative assessment where they need to know the basics to come to class prepared.
For a few reasons:
One, they are almost always simple recall type questions that can easily be answered with Google or the textbook. I'm not going to put a factual recall question on an summative assessment like an exam.
Two, the answers are almost certainly available online (even for application or higher level questions) and savvy students will find them (Course Hero and other "study help" websites have exactly this sort of material available for free).
Three, the questions can be low quality/not on essential concepts/poorly written. I would have to go through each question to make sure it's not terrible.
Just write your own questions. You can use test bank questions as a resource for images/types of questions/etc., but take the time to write your own.
Yes, include a time limit for the reasons other folks have stated.
This is excellent advice for someone who has lots of time and no other priorities.
Writing test questions is hard work and very time consuming. I have often looked back at the old ones I created and thought that they were weak. Some test bank questions are good, some are not. If you are teaching a course each semester for years, you can gradually build up your own test bank with your own questions. It can be helpful to adapt test bank questions from elsewhere.
I use test bank questions all the time in online courses. It is generally very time consuming for student to cheat, and if they have to be switching between different screens to google stuff, they mess up.
If a student does copy and paste their answer from elsewhere, and you catch them (often rather easy), you can give them a zero on the test. That is satisfying.
Given how much demand there is for test banks, it is clear that many faculty use them, even if they won't admit it publicly.
As I often say, when teaching it rarely makes sense to aim to be the ideal as part-time faculty. Just aim to be good enough.
Quote from: downer on March 04, 2020, 10:52:22 AM
Quote from: the_geneticist on March 04, 2020, 10:11:44 AM
Quote from: HigherEd7 on March 04, 2020, 09:39:18 AM
Good question. Do you ever add test questions from the test banks?
Never. I would never use test bank questions on an exam. Maybe a few for a quiz or other formative assessment where they need to know the basics to come to class prepared.
For a few reasons:
One, they are almost always simple recall type questions that can easily be answered with Google or the textbook. I'm not going to put a factual recall question on an summative assessment like an exam.
Two, the answers are almost certainly available online (even for application or higher level questions) and savvy students will find them (Course Hero and other "study help" websites have exactly this sort of material available for free).
Three, the questions can be low quality/not on essential concepts/poorly written. I would have to go through each question to make sure it's not terrible.
Just write your own questions. You can use test bank questions as a resource for images/types of questions/etc., but take the time to write your own.
Yes, include a time limit for the reasons other folks have stated.
This is excellent advice for someone who has lots of time and no other priorities.
Writing test questions is hard work and very time consuming. I have often looked back at the old ones I created and thought that they were weak. Some test bank questions are good, some are not. If you are teaching a course each semester for years, you can gradually build up your own test bank with your own questions. It can be helpful to adapt test bank questions from elsewhere.
I use test bank questions all the time in online courses. It is generally very time consuming for student to cheat, and if they have to be switching between different screens to google stuff, they mess up.
If a student does copy and paste their answer from elsewhere, and you catch them (often rather easy), you can give them a zero on the test. That is satisfying.
Given how much demand there is for test banks, it is clear that many faculty use them, even if they won't admit it publicly.
As I often say, when teaching it rarely makes sense to aim to be the ideal as part-time faculty. Just aim to be good enough.
I have never used test bank questions, but would consider it if I were teaching a course that used a textbook that offered them. I would definitely use them selectively and perhaps re-work them a little as I saw fit, but it seems reasonable to use them.
That being said, it does take time to create good questions but if you will be teaching a course repeatedly it is an investment.
Quote from: downer on March 04, 2020, 10:52:22 AM
Quote from: the_geneticist on March 04, 2020, 10:11:44 AM
Quote from: HigherEd7 on March 04, 2020, 09:39:18 AM
Good question. Do you ever add test questions from the test banks?
Never. I would never use test bank questions on an exam. Maybe a few for a quiz or other formative assessment where they need to know the basics to come to class prepared.
For a few reasons:
One, they are almost always simple recall type questions that can easily be answered with Google or the textbook. I'm not going to put a factual recall question on an summative assessment like an exam.
Two, the answers are almost certainly available online (even for application or higher level questions) and savvy students will find them (Course Hero and other "study help" websites have exactly this sort of material available for free).
Three, the questions can be low quality/not on essential concepts/poorly written. I would have to go through each question to make sure it's not terrible.
Just write your own questions. You can use test bank questions as a resource for images/types of questions/etc., but take the time to write your own.
Yes, include a time limit for the reasons other folks have stated.
This is excellent advice for someone who has lots of time and no other priorities.
Writing test questions is hard work and very time consuming. I have often looked back at the old ones I created and thought that they were weak. Some test bank questions are good, some are not. If you are teaching a course each semester for years, you can gradually build up your own test bank with your own questions. It can be helpful to adapt test bank questions from elsewhere.
I use test bank questions all the time in online courses. It is generally very time consuming for student to cheat, and if they have to be switching between different screens to google stuff, they mess up.
If a student does copy and paste their answer from elsewhere, and you catch them (often rather easy), you can give them a zero on the test. That is satisfying.
Given how much demand there is for test banks, it is clear that many faculty use them, even if they won't admit it publicly.
As I often say, when teaching it rarely makes sense to aim to be the ideal as part-time faculty. Just aim to be good enough.
You bring up some great points. Writing test questions are very time consuming depending on how you write them. I have been using a 1 minute per question, and I am wondering if that is to much time. I have heard people using 45 seconds a question if my math is correct would be under 5 minutes for 10 question quiz.
45 second per question, for ten questions, is 7 1/2 minutes. 5 minutes for 10 questions would be 30 seconds per question.
In my experience, 1-2 minutes is the right amount of time per question, if the questions are fairly analytical. You can also check Canvas and see the average amount of time used per test, and adjust accordingly. If they all finish it very early, reduce the time. If they mostly take all the time, allow more time.
Quote from: Hegemony on March 04, 2020, 07:35:04 PM
45 second per question, for ten questions, is 7 1/2 minutes. 5 minutes for 10 questions would be 30 seconds per question.
In my experience, 1-2 minutes is the right amount of time per question, if the questions are fairly analytical. You can also check Canvas and see the average amount of time used per test, and adjust accordingly. If they all finish it very early, reduce the time. If they mostly take all the time, allow more time.
Thank you for the correction, and I agree. I might need to take a math refresher LOL, how did you figure that out?
I used a few test bank tests from the text this term as open-book quizzes. The quizzes weren't worth much in the overall score- the bulk of the grade in class comes from thoughtful reflection and original creation of material- and I don't think it's necessary for students to memorize these particular facts. However, this text is a good resource, and I want to make sure they know what material is in there. This class is a new prep for me, and I'll have to decide if I want to use them again next time I teach this course. Most of what I want them to do in this class is to create their own materials, and they can use things like this text as a resource for that.
Quote from: HigherEd7 on March 03, 2020, 06:21:48 PM
Who really has the time to generate 50 or 60 questions for an exam to make it difficult for the students to lookup?
I have time to generate 10-20 or so each time I teach the class. After a few runs, that adds up to a formidable bank of questions.
Well it depends on what you want from your online tests-- but yeah timed tests are a waste of time. A googling contest at best and a mechanism of rewarding dishonesty at worst.
Great by points by all. So what is the solution online quizzes and exams are being used in online and face to face courses? Are we really educating students if they have all the resources at home?
Quote from: HigherEd7 on March 08, 2020, 06:44:08 PM
Great by points by all. So what is the solution online quizzes and exams are being used in online and face to face courses? Are we really educating students if they have all the resources at home?
This is the "homework" conundrum that has always existed in education. When assessment is done away from a regulated classroom (or proctored) environment, there are certain curriculum aspects improved (e.g. extended time, time flexibility, additional resources), and certain curriculum aspects that are made worse (e.g. plagiarism, cheating, poor time management, unequal resource access).
As professional educators, we pick and choose our individualized curriculum models to best fit how we want students to learn, and how we want to assess that learning. Responsible and experienced educators will identify the advantages and weaknesses of "homework", optimizing advantages and reducing weaknesses as much as possible.
Online assessments are just another form of homework. Use online exams with the same amount of caution as you would with traditional take-home tests. There are some perks with online that you can't get so much with take-home (e.g. automated software processes like timers, proctoring, plagiarism-tracking), and there are also more potential risks with online that are worse than traditional take-home (e.g. internet outages, technology access, resource inequality).
For fully online curricula, there are a couple of models that most educators should just outright avoid.
- MOOC's
- Automated Courses in a Can
And then there's just some
random advice that most of us already know but sometimes just needs reminding.
*Public test Banks are often awful. But sometimes they are necessary and useful.
*People are far more distracted and more easily distracted than they've ever been, and it's especially bad outside the classroom
*A lot less college students have personal computers than most professors believe (no, a smartphone does not count)
*A "digital native" is not the same thing as being "technology competent". Most of today's students know less of operating professional office software than students of 15-20 years ago.
*A lot more college students cheat and cheat all the time than most professors believe
*Teaching is hard, but the crafting, grading, and securing of assessments is often a lot harder
*Trying out new pedagogy techniques is inefficient in the short term but sometimes it leads to better outcomes
*Disney ruined Star Wars movies
Quote from: larryc on March 07, 2020, 11:00:29 PM
Well it depends on what you want from your online tests-- but yeah timed tests are a waste of time. A googling contest at best and a mechanism of rewarding dishonesty at worst.
I think your first sentence and your second are contradictory.
You have asserted that there is no utility (or validity) to online testing. I disagree.
It can be done well, or poorly. So can f2f tests. What you are testing, and the design of your questions matters.
Quote from: Aster on March 11, 2020, 10:09:14 AM
*A lot less college students have personal computers than most professors believe (no, a smartphone does not count)
*A "digital native" is not the same thing as being "technology competent". Most of today's students know less of operating professional office software than students of 15-20 years ago.
Public libraries see a lot of distance-education students who come to use library computer and internet resources because they don't have them at home. Some of them need quite a bit of help from library staff to do their work. It is likely that public libraries will see a LOT more of this need now that most of the nation's higher ed institutions are in panic mode and abruptly trying to push all of their instruction online.
Fishprof U "requires" students to have laptops.
We're about to learn how serious we are about that requirement.
In my CC, there are students who don't know how to copy even the answers are in front of them! Based on experience, regardless of whether you ask 2-5 questions or 50 questions, the marks are similar.
Each online quiz requires a password. Even I told my students not to send the password to those who are not in the classroom, they continue to do that. Last time I told them that if they were not in the class, even they got 100%, I would give them a zero. A guy who got a perfect score came to me asking me to give him a second chance. I asked the same two simple questions and he could not answer. If he did not cheat, how did he get 100% when he did the test outside the classroom.
Quote from: hamburger on March 12, 2020, 10:06:50 AM
Each online quiz requires a password. Even I told my students not to send the password to those who are not in the classroom, they continue to do that. Last time I told them that if they were not in the class, even they got 100%, I would give them a zero. A guy who got a perfect score came to me asking me to give him a second chance. I asked the same two simple questions and he could not answer. If he did not cheat, how did he get 100% when he did the test outside the classroom.
Question how is giving students a password going to prevent cheating? Also, this could become a headache if you have a large class and you have to issue a password to each student.
Quote from: HigherEd7 on March 12, 2020, 03:47:02 PM
Quote from: hamburger on March 12, 2020, 10:06:50 AM
Each online quiz requires a password. Even I told my students not to send the password to those who are not in the classroom, they continue to do that. Last time I told them that if they were not in the class, even they got 100%, I would give them a zero. A guy who got a perfect score came to me asking me to give him a second chance. I asked the same two simple questions and he could not answer. If he did not cheat, how did he get 100% when he did the test outside the classroom.
Question how is giving students a password going to prevent cheating? Also, this could become a headache if you have a large class and you have to issue a password to each student.
It is just a function that allows the professors to limit access to the questions before the test. With online learning and testing, my students can cheat in whatever way they like.
Quote from: hamburger on March 14, 2020, 08:00:41 AM
Quote from: HigherEd7 on March 12, 2020, 03:47:02 PM
Quote from: hamburger on March 12, 2020, 10:06:50 AM
Each online quiz requires a password. Even I told my students not to send the password to those who are not in the classroom, they continue to do that. Last time I told them that if they were not in the class, even they got 100%, I would give them a zero. A guy who got a perfect score came to me asking me to give him a second chance. I asked the same two simple questions and he could not answer. If he did not cheat, how did he get 100% when he did the test outside the classroom.
Question how is giving students a password going to prevent cheating? Also, this could become a headache if you have a large class and you have to issue a password to each student.
It is just a function that allows the professors to limit access to the questions before the test. With online learning and testing, my students can cheat in whatever way they like.
We used to do this and we had two people who took the test while not in the room. We were able to have the test software publisher pull IP logs for us so we could verify that and assign zeros.
After that, we found out that we could lock the test down by IP so that only computers in that room (and one other, but we didn't tell the students that ... the chances they'd find the other classroom accidentally and be able to login during the test were virtually nil) could access the test. See if you can do that.
Quote from: kiana on March 14, 2020, 08:13:13 AM
Quote from: hamburger on March 14, 2020, 08:00:41 AM
Quote from: HigherEd7 on March 12, 2020, 03:47:02 PM
Quote from: hamburger on March 12, 2020, 10:06:50 AM
Each online quiz requires a password. Even I told my students not to send the password to those who are not in the classroom, they continue to do that. Last time I told them that if they were not in the class, even they got 100%, I would give them a zero. A guy who got a perfect score came to me asking me to give him a second chance. I asked the same two simple questions and he could not answer. If he did not cheat, how did he get 100% when he did the test outside the classroom.
Question how is giving students a password going to prevent cheating? Also, this could become a headache if you have a large class and you have to issue a password to each student.
It is just a function that allows the professors to limit access to the questions before the test. With online learning and testing, my students can cheat in whatever way they like.
We used to do this and we had two people who took the test while not in the room. We were able to have the test software publisher pull IP logs for us so we could verify that and assign zeros.
After that, we found out that we could lock the test down by IP so that only computers in that room (and one other, but we didn't tell the students that ... the chances they'd find the other classroom accidentally and be able to login during the test were virtually nil) could access the test. See if you can do that.
How does this help if they are at home in an online course?
Well, given how many of us are now teaching/developing/planning online courses, the answer is "we are going to be doing online tests. Now to figure out how to make this work best"
My online grad course in a healthcare field has been using online tests for about 10 years. We developed the questions on our own from the material to resemble national boards. It takes 2-3 semesters to "test" a new question. Right now I have a test back of about 1000 questions. The exam randomly picks from the topics in the test bank and even if students are sitting together taking the test, they are unlikely to receive all the same questions. It's about 40 questions and they have 2 hours - some questions are fairly hard so that's why the long time frame.
They can use all the sources but the questions are pretty much all application of complex concepts. If they have not read the material and studied well beforehand, they are not going to pass. Some learn the hard way the first time and don't heed my advice about that. Live and learn I guess.
From a teaching standpoint, the most difficult part is going through the questions every year before the exams to ensure that the match the content. Sometimes healthcare info changes (rapidly!) and our content changes.
I suspect that except one student, the rest of the class cheated. Now that everything is online, it is a paradise for them.
Quote from: hamburger on March 17, 2020, 02:36:25 PM
I suspect that except one student, the rest of the class cheated. Now that everything is online, it is a paradise for them.
I assume they are dealing with anxiety and uncertainty like all of the rest of us. This seems like a good time to drop the bad attitude towards your students. Seriously either make exams that will be harder to cheat on online or stop obsessing over it.
If one person has taken this course at this institution multiple times, then clearly failure is an option. I agree with Caracal that you, hamburger, have the option to stop taking student behavior personally. I disagree with just letting the behavior go unchecked.
Start documenting . What are students doing to cheat that can be observed? What evidence is created that can be submitted to academic standards board beyond what you need to record the F for academic dishonesty?
One technique is to make everything an open-ended answer that requires several lines to answer.
I used to give open-book, open-note, open-neighbor tests that were a combination of essays and problems in physics, engineering, and math. The submitted scratch/note paper in addition to any electronic test will demonstrate uniqueness (or not). If necessary, students can take a picture of that paper with their phone to upload/email or physically mail it to you. That paper can also help show who is really doing the work (tend to have many more steps, a little planning on the side or erasures/crossouts) and who just wrote down some random intermediate steps to look like they were doing the work. Asking questions that require scratch paper even for experts tends to cut down dramatically on claims of "I didn't need any scratch paper".
A computer code of more than a handful of lines tends to also look unique; consider the variations between v1, var1, varX, X1, variable1, density, Den, d, and density_input_before_unit_conversion. The basic operations are likely to be the same for a simple program, but individuals tend to have a personal style in naming variables along with commenting style. Always require comments as though this were a production code that someone else must maintain. That's a way to cut down on easily observable cheating as well as a service to the broader community.
Asking open-ended questions and allowing people to consult references can be defended pedagogically as how people use their knowledge and skills gained from courses (and all the prerequisite/corequisite courses) outside the formal classroom. These tests are an opportunity for students to really strut their stuff for those who are learning. Being able to reproduce pre-chewed material is generally far less useful for later life than being able to consult multiple sources, consolidate/integrate/extend/interpolate/extrapolate that information, and then answer an open-ended question with a range of right enough answers.
Or, you can just rip off your students and the public by letting everyone who signs up and submits any work just pass. Apparently, that's a thing that some faculty members encourage because it's about sustaining the system, not helping people become educated.
Quote from: polly_mer on March 18, 2020, 05:31:18 AM
If one person has taken this course at this institution multiple times, then clearly failure is an option. I agree with Caracal that you, hamburger, have the option to stop taking student behavior personally. I disagree with just letting the behavior go unchecked.
That wasn't actually what I was proposing. I specifically design my exams so that cheating is likely to be of relatively minimal benefit. I do this because I don't want students to cheat, but also because I don't want to have to create a culture of distrust in my class. If a student asks to go to the bathroom during the exam, I'm going to say yes and assume they need to do one of the things humans do in there. If, they really are going to go duck in the bathroom and try to google my essay question on their phone or ask their friend in the other section, that would be cheating and if I learned of it, I'd report it, but it also would be pretty ineffective. I'm not asking for random facts or a one sentence answer, but an argument that puts together course themes and backs them up with evidence. Three minutes scrolling on the phone and then coming back to class is going to lead to something pretty incoherent.
Last year, a former student told me that they could buy solutions, hire somebody to do it, or get the answers for free from GitHub. I guess if 20-30 students share the cost of hiring one person to do the work, the cost would be low. For sure, somebody outside the class helped them as in the last lab, all of them started to panic when I told them that they read the question wrong. They started sending messages from their phone, left the room immediately or doing nothing but starring at the screen. Only one student attempted to do the work by himself. In the middle of the lab, I went to the toilet and when I returned, an unknown person was in the room helping a student. This also happened one more time in the past. In another lab, some students had their programs working perfectly but they did not even know how to enter five numbers from the keyboard. Instead of entering: 1 2 3 4 5. Two or three students typed: 12345.
Due to a lack of participation, it is not uncommon to get emails about departments inviting colleagues from other departments to attend meetings to deal with cheating. I attended such meeting as a "free service" but it did not help to get a full time job.
Quote from: hamburger on March 18, 2020, 06:28:19 AM
Last year, a former student told me that they could buy solutions, hire somebody to do it, or get the answers for free from GitHub. I guess if 20-30 students share the cost of hiring one person to do the work, the cost would be low. For sure, somebody outside the class helped them as in the last lab, all of them started to panic when I told them that they read the question wrong. They started sending messages from their phone, left the room immediately or doing nothing but starring at the screen. Only one student attempted to do the work by himself. In the middle of the lab, I went to the toilet and when I returned, an unknown person was in the room helping a student. This also happened one more time in the past. In another lab, some students had their programs working perfectly but they did not even know how to enter five numbers from the keyboard. Instead of entering: 1 2 3 4 5. Two or three students typed: 12345.
Due to a lack of participation, it is not uncommon to get emails about departments inviting colleagues from other departments to attend meetings to deal with cheating. I attended such meeting as a "free service" but it did not help to get a full time job.
Look, you need to have clear expectations and stick with them. Are they allowed to use outside help or not? Can they use their phones or not?
And it sounds like you need to write better assignments. If I could complete your class using Google, why bother to be in the room.
Quote from: the_geneticist on March 18, 2020, 09:33:56 AM
Quote from: hamburger on March 18, 2020, 06:28:19 AM
Last year, a former student told me that they could buy solutions, hire somebody to do it, or get the answers for free from GitHub. I guess if 20-30 students share the cost of hiring one person to do the work, the cost would be low. For sure, somebody outside the class helped them as in the last lab, all of them started to panic when I told them that they read the question wrong. They started sending messages from their phone, left the room immediately or doing nothing but starring at the screen. Only one student attempted to do the work by himself. In the middle of the lab, I went to the toilet and when I returned, an unknown person was in the room helping a student. This also happened one more time in the past. In another lab, some students had their programs working perfectly but they did not even know how to enter five numbers from the keyboard. Instead of entering: 1 2 3 4 5. Two or three students typed: 12345.
Due to a lack of participation, it is not uncommon to get emails about departments inviting colleagues from other departments to attend meetings to deal with cheating. I attended such meeting as a "free service" but it did not help to get a full time job.
Look, you need to have clear expectations and stick with them. Are they allowed to use outside help or not? Can they use their phones or not?
And it sounds like you need to write better assignments. If I could complete your class using Google, why bother to be in the room.
I clearly stated both verbally and in writing that no phone and internet access were allowed in the test. They did not listen. As for somebody showing up in the lab "to help" them to do the work, never had this happened in other places I taught.
Some students do not know how to use google to get information. Colleagues told me that some students do not have the mental capacity to do that.
Quote from: hamburger on March 18, 2020, 09:45:15 AM
Quote from: the_geneticist on March 18, 2020, 09:33:56 AM
Quote from: hamburger on March 18, 2020, 06:28:19 AM
Last year, a former student told me that they could buy solutions, hire somebody to do it, or get the answers for free from GitHub. I guess if 20-30 students share the cost of hiring one person to do the work, the cost would be low. For sure, somebody outside the class helped them as in the last lab, all of them started to panic when I told them that they read the question wrong. They started sending messages from their phone, left the room immediately or doing nothing but starring at the screen. Only one student attempted to do the work by himself. In the middle of the lab, I went to the toilet and when I returned, an unknown person was in the room helping a student. This also happened one more time in the past. In another lab, some students had their programs working perfectly but they did not even know how to enter five numbers from the keyboard. Instead of entering: 1 2 3 4 5. Two or three students typed: 12345.
Due to a lack of participation, it is not uncommon to get emails about departments inviting colleagues from other departments to attend meetings to deal with cheating. I attended such meeting as a "free service" but it did not help to get a full time job.
Look, you need to have clear expectations and stick with them. Are they allowed to use outside help or not? Can they use their phones or not?
And it sounds like you need to write better assignments. If I could complete your class using Google, why bother to be in the room.
I clearly stated both verbally and in writing that no phone and internet access were allowed in the test. They did not listen. As for somebody showing up in the lab "to help" them to do the work, never had this happened in other places I taught.
Well, then that's an easy F to put in the grade book.
After interacting with my students in the first onlne class, I had a big headache the rest of the day and I feel disguisted.
My students had two online quizzes. 2 MC questions/quiz. What else can they ask for? If they listened and studied, they could have answered all questions in less than a minute. I gave them 15 minutes. Unlike previous quizzes, I shortened the time and also randomized the questions. Moreover, only one question was shown on the screen at any time. Many students failed. While they took the quizzes, I observed their behavior. Similar to last time when they did it in class, only a few individuals took the quiz first. The rest were just waiting for the answers.
After the quizzes, two students (A & B) told me that they could not take the quizzes due to "IT problem". Student A told me that he knew the answers. He asked me if he could submit his answers by email. Interestingly, he said he could not access the quizzes due to IT issues but he knew the answers! Also, the record showed that both of them got 0 in the first quiz and did not answer any question in the 2nd quiz.
After the quizzes, I gave them a review for the upcoming test. Only a few students joined the online class. A few of them joined for a minute or so as a way to have a record that they "attended" the class. This included the jer* who told me that he paid for my salary. The first thing they asked was whether or not I would tell them the questions and answers of what would be on the test. I said no. Then, I asked them if they had read the learning materials I posted for them to study last week. A few said no and the other remained silent. I asked them if they had completed the assignment to be due in two days. Again, a few said no and the other remained silent. I used chat to ask those remained silent but no answer. So they just logged in and went somewhere else to pretend that they were there.
In the class, I gave a full review of things important to the course. They assigned one student (Student A) as a representative to interact with me. I asked him over 10 questions. Even with the answers right in front of him, he could not answer my questions. I guided him to find the answers himself. Obviously this guy did not study nor has a brain. For example, after guiding him to answer a question, he got the answer right. Three seconds later, I asked him the same question and he got it wrong. This happened many times.
This group of students is the worst academically. They are not stupid as humans but they are stupid academically. They don't have the will nor ability to learn. As many colleagues said, students in my school are there not to learn but for a certificate. There are some students who use the internet to badmouth their professors if the professors do not do everything they want. They gave poor evaluations and wrote bad things about their professors in RMP but we can't do the same. There are students who put all the blames on the professors when they did not get the marks they deserved.
I could make easy MC questions for the remaining of the course to make it easy for them to cheat and get high marks. I could make questions to fail the entire class. What would you do?
Hamburger. See what efforts are being made to monitor your online class. Probably none. So you can relax.
Probably the students will be allowed to do a Pass/Fail option. Tell them to do that.
Give the students a couple of easy tests and pass them all.
Then go do something you enjoy.
One might suspect you enjoy complaining. But try something different from that.
Quote from: downer on March 24, 2020, 06:45:00 AM
Hamburger. See what efforts are being made to monitor your online class. Probably none. So you can relax.
Probably the students will be allowed to do a Pass/Fail option. Tell them to do that.
Give the students a couple of easy tests and pass them all.
Then go do something you enjoy.
One might suspect you enjoy complaining. But try something different from that.
Thanks for the suggestion. Some bad students not only want to pass but also want high marks. They did not study and failed the tests. Two students decided not to take the second test after they got zero in the first test. Then, they claimed that they could not take both tests due to IT issue. Now one of them is making all sorts of accusations and sending me lots of emails. Better not to check work email account.
In my school, there is a tendency that students, professors and staff dislike each other. Some students even complained on the school's SNS that the school only cares about money.
You obviously work in a cesspool yet you are constantly remarking that you are surrounded by shit.
Quote from: Hegemony on March 04, 2020, 07:35:04 PM
45 second per question, for ten questions, is 7 1/2 minutes. 5 minutes for 10 questions would be 30 seconds per question.
In my experience, 1-2 minutes is the right amount of time per question, if the questions are fairly analytical. You can also check Canvas and see the average amount of time used per test, and adjust accordingly. If they all finish it very early, reduce the time. If they mostly take all the time, allow more time.
How are you figuring 45 seconds per question, 7 1/2 minutes? Unless I am doing something wrong 45 x 10 = right around 4 minutes and 50 seconds.
Quote from: HigherEd7 on August 18, 2020, 05:05:29 AM
Quote from: Hegemony on March 04, 2020, 07:35:04 PM
45 second per question, for ten questions, is 7 1/2 minutes. 5 minutes for 10 questions would be 30 seconds per question.
In my experience, 1-2 minutes is the right amount of time per question, if the questions are fairly analytical. You can also check Canvas and see the average amount of time used per test, and adjust accordingly. If they all finish it very early, reduce the time. If they mostly take all the time, allow more time.
How are you figuring 45 seconds per question, 7 1/2 minutes? Unless I am doing something wrong 45 x 10 = right around 4 minutes and 50 seconds.
Yes, you are doing something wrong. Seconds and minutes are in base 60. 45x10=450 seconds. Divide result by 60 seconds/minute. That results in 7.5 minutes.
Quote from: arcturus on August 18, 2020, 05:28:33 AM
Quote from: HigherEd7 on August 18, 2020, 05:05:29 AM
Quote from: Hegemony on March 04, 2020, 07:35:04 PM
45 second per question, for ten questions, is 7 1/2 minutes. 5 minutes for 10 questions would be 30 seconds per question.
In my experience, 1-2 minutes is the right amount of time per question, if the questions are fairly analytical. You can also check Canvas and see the average amount of time used per test, and adjust accordingly. If they all finish it very early, reduce the time. If they mostly take all the time, allow more time.
How are you figuring 45 seconds per question, 7 1/2 minutes? Unless I am doing something wrong 45 x 10 = right around 4 minutes and 50 seconds.
Yes, you are doing something wrong. Seconds and minutes are in base 60. 45x10=450 seconds. Divide result by 60 seconds/minute. That results in 7.5 minutes.
Thank you for the response and information!! Be safe
Timed tests and screenshots are not accessible to some students. I hope you are working with accessibility services or the equivalent to ensure tests are accessible from home. Similarly, not allowing students to go back and review their answers impacts students with disabilities more than students who are not disabled. It's sad to see so much ableism in higher ed.
Also, IT issues disproportionately affect students who live in rural and remote areas, who live in lower SES neighbourhoods, or who are Indigenous. There are students who have taken exams in the parking lot of a Tim Horton's or other location with free wifi because they don't have reliable internet access at home. Yes, some students will cheat, but maybe I'm naive but I prefer to give my students the benefit of the doubt. I've been that disabled Indigenous student so I know the challenges that exist with online assessments.
Quote from: MarathonRunner on August 22, 2020, 07:16:29 AM
Timed tests and screenshots are not accessible to some students. I hope you are working with accessibility services or the equivalent to ensure tests are accessible from home. Similarly, not allowing students to go back and review their answers impacts students with disabilities more than students who are not disabled. It's sad to see so much ableism in higher ed.
Also, IT issues disproportionately affect students who live in rural and remote areas, who live in lower SES neighbourhoods, or who are Indigenous. There are students who have taken exams in the parking lot of a Tim Horton's or other location with free wifi because they don't have reliable internet access at home. Yes, some students will cheat, but maybe I'm naive but I prefer to give my students the benefit of the doubt. I've been that disabled Indigenous student so I know the challenges that exist with online assessments.
Access to reliable internet is the biggest inequity for online learning. I know that the majority of my students are completing all of their coursework on their phones. Why? It's the only internet-capable device that they don't have to share with parents, siblings, or room mates.
I hope I've set up my exams so that they are a test of student understanding and not a test of their internet access speed.
My campus is offering free wifi hotspots & chrome books, but the students have to be close enough to campus to come and get them.
There are also Xfinity wifi hotspots nationally and they have been made free for the pandemic. The map is here: https://hotspots.wifi.xfinity.com/
They're not absolutely everyplace, but they've been helpful to a number of my students.
Quote from: MarathonRunner on August 22, 2020, 07:16:29 AM
Timed tests and screenshots are not accessible to some students. I hope you are working with accessibility services or the equivalent to ensure tests are accessible from home. Similarly, not allowing students to go back and review their answers impacts students with disabilities more than students who are not disabled. It's sad to see so much ableism in higher ed.
Also, IT issues disproportionately affect students who live in rural and remote areas, who live in lower SES neighbourhoods, or who are Indigenous. There are students who have taken exams in the parking lot of a Tim Horton's or other location with free wifi because they don't have reliable internet access at home. Yes, some students will cheat, but maybe I'm naive but I prefer to give my students the benefit of the doubt. I've been that disabled Indigenous student so I know the challenges that exist with online assessments.
You are definitely naive. Most students will not cheat, but it in any larger group, there will be a subset that will. Providing tests where cheating is possible/easy will disproportionately benefit this subset and reward the behaviour. Timed tests will not eliminate this as students are smart, but it will reduce it.
As for accessibility, I believe it is better to address this directly. We generally increase the time limit for students with disabilities (or accomodate in another more appropriate fashion), with technology this is quite simple. Regarding internet/technology access, that can also be addressed but is obviously a challenge. However, this goes much deeper than timed tests...
I think it is also important to realize that Covid has put everyone in tough positions. Virtual learning and timed tests have issues, but they are the best solution available.
Quote from: Kron3007 on August 27, 2020, 04:45:26 AM
Quote from: MarathonRunner on August 22, 2020, 07:16:29 AM
Timed tests and screenshots are not accessible to some students. I hope you are working with accessibility services or the equivalent to ensure tests are accessible from home. Similarly, not allowing students to go back and review their answers impacts students with disabilities more than students who are not disabled. It's sad to see so much ableism in higher ed.
Also, IT issues disproportionately affect students who live in rural and remote areas, who live in lower SES neighbourhoods, or who are Indigenous. There are students who have taken exams in the parking lot of a Tim Horton's or other location with free wifi because they don't have reliable internet access at home. Yes, some students will cheat, but maybe I'm naive but I prefer to give my students the benefit of the doubt. I've been that disabled Indigenous student so I know the challenges that exist with online assessments.
You are definitely naive. Most students will not cheat, but it in any larger group, there will be a subset that will. Providing tests where cheating is possible/easy will disproportionately benefit this subset and reward the behaviour. Timed tests will not eliminate this as students are smart, but it will reduce it.
Well it depends on the kind of test and the sort of material. If you're talking about multiple choice, or fill in the blank exams, then, yes, probably having timed exams is the only way to try to limit cheating. However, you can go in the other direction and have longer, open note exams. That may or may not work for certain fields and it may not be feasible depending on the workload.
I have many students like this student you`ve just described. This is a well known phenomenon and it is called developmentally underdeveloped that cause the learners to become underachievers. This is the result of environmental retardation and the lack of intellectual stimula during early years which result in these folks to become an underachievers. You may like to consider changing your approach and lowering your expectations from them, because developmentally they are at 8th or 9th grade level at best.
Quote from: hamburger on March 24, 2020, 06:24:07 AM
After interacting with my students in the first onlne class, I had a big headache the rest of the day and I feel disguisted.
My students had two online quizzes. 2 MC questions/quiz. What else can they ask for? If they listened and studied, they could have answered all questions in less than a minute. I gave them 15 minutes. Unlike previous quizzes, I shortened the time and also randomized the questions. Moreover, only one question was shown on the screen at any time. Many students failed. While they took the quizzes, I observed their behavior. Similar to last time when they did it in class, only a few individuals took the quiz first. The rest were just waiting for the answers.
After the quizzes, two students (A & B) told me that they could not take the quizzes due to "IT problem". Student A told me that he knew the answers. He asked me if he could submit his answers by email. Interestingly, he said he could not access the quizzes due to IT issues but he knew the answers! Also, the record showed that both of them got 0 in the first quiz and did not answer any question in the 2nd quiz.
After the quizzes, I gave them a review for the upcoming test. Only a few students joined the online class. A few of them joined for a minute or so as a way to have a record that they "attended" the class. This included the jer* who told me that he paid for my salary. The first thing they asked was whether or not I would tell them the questions and answers of what would be on the test. I said no. Then, I asked them if they had read the learning materials I posted for them to study last week. A few said no and the other remained silent. I asked them if they had completed the assignment to be due in two days. Again, a few said no and the other remained silent. I used chat to ask those remained silent but no answer. So they just logged in and went somewhere else to pretend that they were there.
In the class, I gave a full review of things important to the course. They assigned one student (Student A) as a representative to interact with me. I asked him over 10 questions. Even with the answers right in front of him, he could not answer my questions. I guided him to find the answers himself. Obviously this guy did not study nor has a brain. For example, after guiding him to answer a question, he got the answer right. Three seconds later, I asked him the same question and he got it wrong. This happened many times.
This group of students is the worst academically. They are not stupid as humans but they are stupid academically. They don't have the will nor ability to learn. As many colleagues said, students in my school are there not to learn but for a certificate. There are some students who use the internet to badmouth their professors if the professors do not do everything they want. They gave poor evaluations and wrote bad things about their professors in RMP but we can't do the same. There are students who put all the blames on the professors when they did not get the marks they deserved.
I could make easy MC questions for the remaining of the course to make it easy for them to cheat and get high marks. I could make questions to fail the entire class. What would you do?
Quote from: Caracal on August 27, 2020, 06:14:45 AM
Quote from: Kron3007 on August 27, 2020, 04:45:26 AM
Quote from: MarathonRunner on August 22, 2020, 07:16:29 AM
Timed tests and screenshots are not accessible to some students. I hope you are working with accessibility services or the equivalent to ensure tests are accessible from home. Similarly, not allowing students to go back and review their answers impacts students with disabilities more than students who are not disabled. It's sad to see so much ableism in higher ed.
Also, IT issues disproportionately affect students who live in rural and remote areas, who live in lower SES neighbourhoods, or who are Indigenous. There are students who have taken exams in the parking lot of a Tim Horton's or other location with free wifi because they don't have reliable internet access at home. Yes, some students will cheat, but maybe I'm naive but I prefer to give my students the benefit of the doubt. I've been that disabled Indigenous student so I know the challenges that exist with online assessments.
You are definitely naive. Most students will not cheat, but it in any larger group, there will be a subset that will. Providing tests where cheating is possible/easy will disproportionately benefit this subset and reward the behaviour. Timed tests will not eliminate this as students are smart, but it will reduce it.
Well it depends on the kind of test and the sort of material. If you're talking about multiple choice, or fill in the blank exams, then, yes, probably having timed exams is the only way to try to limit cheating. However, you can go in the other direction and have longer, open note exams. That may or may not work for certain fields and it may not be feasible depending on the workload.
Yes, you can go that route. My main concern with that is consistency in grading. I saw a talk on testing methods and their valifity, and long answer questions did not do well for reliable and reproducible grading. This is especially problematic in larger classes where multiple people are grading. I have done this in the past and don't doubt that the grading was not 100% consistent. For example, I'm sure my standard drifted over time despite trying not to do so. For this reason, and the workload, I prefer timed MC and shirt answer. In some cases, I include images and ask questions about the image, making it difficult to google an answer.