Hi,

I'm using LimeSurvey as part of a course I am teaching. I assign individual tolkens to each student and allow them to save and restart their survey later in the General Settings. However, I find that a number of students submit two surveys. One which is largely incomplete and the other which is complete. This is wreaking havoc on my grading scripts since there are two entries per student in some cases. Is there a way to avoid this problem such as forcing students to pick up where they left off rather than being able to start over?

This is weired. There should be different entries with the same token. Do you allow the usage of a token more than once?

One other thing: As a quick solution (until we find out what goes wrong) you could always delete the responses that are largely incomplete beofre you run your grading scripts.

I am having a very similar problem. I am using Version 1.91+ Build 120302
I have a non-anonymous survey, cookies set =yes, participant may save and resume later=yes, allow editing after completion=yes, enable token based persistance =yes.
I am still receiving duplicate responses in answer table.
I have tried to recreate it myself but cannot seem to figure it out. One will be incomplete and one complete.

Did you check:
Enable token-based response persistence?

My issue is similar.

Version 2.00+ Build 130129

anonymized responses = no
cookies set = no
participant may save and resume later = yes
allow editing after completion = no
enable token based persistance = yes

The first time I noticed the problem was when a user reported an error and I noticed that the response had been completed even with the error, however they were able to go back in and complete another response.
I thought this was a browser problem so I set
enable token based persistance = yes (it was originally = no)
This did not stop the problem.
I then used a token that had already been used and found I was able to create another response using the completed token.
I checked the Token list and found that none of the completed survey's tokens had been set down to 0 uses. Thinking this was odd I manually set the used token to 0 and found I was unable to access the survey as i expected.

I therefore expect the problem is simply because the token use is not decrementing. The odd thing is that the token list was created using the CPD panel and when I had uploaded my lists from csv the token uses decremented no problem. I also noted that when I manually added several tokens into the same CPD list (that is not working)these tokens decremented as expected.

Have I set something wrong or is there some other issue?

Just checking now to see if bug report 07278 is the issue. found usesrow on line 958 of frontend_helper.php and replaced it with the new variable name.

Did a couple of checks and this appears to solve the problem.

I also can fill out the survey multiple times with the same token.

Version 2.00+ (130206)
Anonymous response: Yes
Cookies: Yes
Save and resume later: Yes
Editing after completion: No
Enable token based persistance: Yes / No (tried both)

Bug report 07278 doesn't seem to apply to me as there's no new variable name for "usesrow".

Please advise.

Thanks,
Karen

interesting, I upgraded to 130206 thinking the fix had already been applied (changed usesrow to new var) but it hadn't. I still had to go in and change this.
Is your token uses number decreasing after each use by the user?
Does it indicate that the survey has been completed?
Did you user get an error message when they submitted?

So far mine appears to be working as expected but I have not had a lot of responses to my survey since I made the fix so I am still checking into this.

Kim

Hi Kim,

Yes, I only allow single use of the token and it will decrease after the survey has been completed so that part works as expected. But I've since found out that it will allow you to submit many partial surveys until one is completed. I'm not sure how that's happening as I've just tested that if I save the survey (to resume later), I will get back to the same partially completed one whether I click on the invite link or the saved link (from resume email). I think it's when you don't save the survey but the partial response is actually saved although you can't access it again later, that's all.

I'll have to test this further. But if what I've described is correct, LS shouldn't keep the partially done survey if user has not clicked resume later since there's no way they'll be able to get back to that survey so that record is 'orphaned'. I'll update after testing the different scenarios.

Cheers,
Karen

I was under the impression that
Enable token based persistence: Yes
would deal with stopping multiple instances of the same token entry being created. It appears to work with my survey. I have had several students just close out of the survey without actually ending it by saving or submitting and I do not get any new instance being created.
The only time I was getting multiple entries for the same token was when the token table did not register a decrease in the token use to 0. Since the fix i mentioned above it has been working as expected.
I was surprised that your 130206 update included the fix and mine did not. I have updated to 130213 and it does have the changed file.

Good luck,
Kim