2

I have to prepare an online survey. Most of the questions will be along the line of:

Which of the following sentences exists in 'X' book?

A

B

C

D

I would like to limit cheating as much as possible. Since the survey is conducted online, I can't have any type of physical restrictions for the users who answer the question.

My main concern is that many people will use a search engine to see which of the four answers exist in the book.

In order to prevent that, I would like to set a time restriction for answering the questions. But the problem is that users may have different reading speeds they may miss some or all of the information in a question.

How can I balance between providing enough time to read a question and restricting the time to prevent cheating?

  • You might have better luck asking at Academia. This seems like less of a user experience question and more of a how to minimize cheating question. – maxathousand Jul 12 '17 at 17:33
  • 1
    Can you clarify whether this is a survey in which case @jazZRo's answer of providing a "Don't know" option seems the right thing to do, or a quiz/test where you actively don't want cheating? (Even then, a "don't know" might be appropriate). – TripeHound Jul 13 '17 at 8:21
1

If this is really a survey and there is no right or wrong answer, the reason that someone would look up the answer is that they don't know the answer but still have to give one.

A possible way to overcome this is to give an extra option "don't know" or something.

EDIT:
This doesn't answer the question how to estimate a time. But what happens when the time limit has passed? Also I think it will make the UX worse and likely less people will finish the survey. If this is about getting lots of results this is important. But if it is some sort of competition I doubt a time limit is fair: Think about people with dyslexia, slow computers/browsers etc. In a non-controlled environment adding such time limit maybe isn't the way to go.

  • That depends if the time limit is revealed to the user or just stored in the data for analytics - If the user is unaware then it should have no impact but the analyst can still have a guess that the user has at least left the page for a while - However, this could mean that they went to make a cup of tea while they thought about the answer. – Andrew Martin Jul 13 '17 at 7:33
  • @AndrewMartin, Good point. Either way a time limit won't be easy for implemenation nor analytics. – jazZRo Jul 13 '17 at 7:41
0

You could try some warning or browser prompt that informs the test taker when they open a new browser or tab, that it will shut the test off or invalidate it in some way. I don't know how that would work technically but worth a shot?

  • I'm sure you can use javascript to detect a shift of focus from one tab to another and then, when focus returns, mark the question as cheated. – Andrew Martin Jul 13 '17 at 7:27
0

You can give penalties and/or bonuses related to time/delay. This way the final score will be affected by delay.

For example:
Set the average delay for the first question to 90 seconds.
If delay is 100'' subtract a point from the final score.
If delay is 50'' add a point to the final score.

Another example:
Set the average delay for the first question to 90 seconds.
If delay is bellow 90'' its ok, final score is not affected.
if delay is between 90'' and 180'' (2 x average) subtract a point from the final score.
if delay is between 180'' and 240'' (3 x average) subtract 2 points from the final score.

0

Delays are a bad way to test page presence. You will run into all sorts of accessibility cases where users with motor impairments are penalised for not being able to move the mouse fast or accurately enough, users with cognitive impairments are penalised for taking too long to think about the question, even users with visual impairments are penalised for the speed their screen reader functions at...

Time is an incredibly bad way of testing if people are leaving the page to look up answers elsewhere.

As @edwardn suggests, you could monitor the focus of the page and mark any times that the focus has shifted away from your page but even this is fallible: if this is a simple survey, how do you know that your users are not simply getting bored and nipping over to check their Facebook accounts?

If this is a test or exam then you need to make it clear to your users that they should not leave the page or switch to another task until the test is complete. This, in turn, also means that you should give them a clear indication of how long you expect the test to take before they start answering questions and what the penalties for leaving the page are.

Once they have accepted that they are under test conditions then you can start checking the page focus.

But don't use time.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.