Please help us help you and fill where relevant:
Your LimeSurvey version: url=https://community.limesurvey.org/]LimeSurvey Community Edition[/url] [url=https://edu.freezhao.com/survey/index.php?r=questionAdministration/view&surveyid=565433&gid=64&qid=939#modalSystemInformation]Version 6.0.6+230508 [/url
Own server or LimeSurvey hosting: Self-hosting
Survey theme/template: Fruity
==================
I have a group ask a before/after question, what …… before and what …… after?
But same participant choose the same answer in both, which means no different between them, but It has to be different, base on the privous conditional question is Yes/No different.
Can I set the condition for that? Many Thanks!

Validation equation in the second question
self.NAOK!=Q1.NAOK

Or do not display the chosen answer of Q1 in Q2.

Please explain more detailed, best send a lss export of the relevant questions.

Joffm 

Let's say we have 3 products, ad we will do a longitudinal research. Time1. and Time2 ……，
Time1 participants will rank the product.
When the Time2 survey start, the first question will ask them if they change order of ranking. If they choose yes, then the before/after question will show up.
They need to answer what they ranked before, what they would rank now. So the answer should be different.
But when I analysis the Time 2 data I found many participants answer the before/after question as same. That would be a pity if I delete all the before/after-same cases.

That's why I ask for help, and thanks so much Joffm, Can you help me more about that?

Hi,

They need to answer what they ranked before

Do you really think, they remember? Okay, if there are really only three products they may remember.

Or do you have a token-based survey and store the result in an additional attribute with plugin "update TokenByResponse"?
So you can export the token table and import it into your second survey.
Show it at the beginning, ask if there changed something, and - if "Yes" - display the ranking question again.
But seeing your URL ("edu") I doubt that you have rights to install plugins.

Now, without this and a simple open survey

1. you can to validate in the second question that not all ranks contain the same value.
That's easy:
It's fine
if
the value in 1st place of ranking 1 is unequal the value of 1st place in ranking 2 OR
the value in 2nd place of ranking 1 is unequal the value of 2nd place in ranking 2 OR
the value in 3rd place of ranking 1 is unequal the value of 3rd place in ranking 2 OR
...

And if three products are realistic you only have to compare 1st and 2nd place.

2. you can compare and in case there is an identical ranking, set the "Yes/No" question to "No" and also clear the ranking questions by an equation.
​​​​​​
Joffm


 

 I try to find the plugin called "updateTokenByRespons" seems it not default install in the LS, and I didn't find the official LS link to download it. Can you tell me the official link of updateTokenByRespons?

By my understanding, the Token should be send to participant one by one as a ID to distinguish data, In my context, I sent to survey link to a chat group, and for the best experience to encourage participant to complete the whole survey, I have to make it anonymous. And the participants do not use email so much. So Can  "updateTokenByRespons" help me in this senario?

I will try the 2 solution from your help, thanks first! 

Hi,
if you only send the generic link to a chat group you conduct an open survey.
This is always anonymous as you do not have the least idea who answers.
And you have no tokens. So you can't use the plugin.
And I did not include the link because I said "But seeing your URL ("edu") I doubt that you have rights to install plugins."
If it is an installation of a university  or similar, many times the admins don't allow the installation.

You only have to rely on your participants; that the remember their first choice.

By the way: How do you merge the data of the two waves?

Joffm

I try to find the plugin called "updateTokenByRespons" seems it not default install in the LS, and I didn't find the official LS link to download it. Can you tell me the official link of updateTokenByRespons?

Joffm already mentioned that in your case update token by repsonse won't help, as you can't use tokens.

The plugin is a third party plugin provided by Denis (a forum user and core developer). So there won't be any official link from Limesurvey.

By the way: How do you merge the data of the two waves?

That's a good question, I have to radomly pick a same as BEFORE case , and update to the NEW data manually.

Luckily, There just small part of the participant change their idea. But I am alos consider how to achieve this kind of suevey in a anonymously way. Any possible way I would love to try.

Oh Thanks!
I have downloaed the plugin as you suggested. But it can not install on LS6.0.6+230508, So I can not test it

I can install the plugin but it not compatible with LS6.0.6+230508.

I have an idea suddently, Could it be a way to generate a token for participants after submition. And the token will became the second survey key if they join in again, for example when they answer the question "Have you participated this survey before", Yes→then they can input the token. Not a good UX but the token can be a gift code later for reward.

I can install the plugin but it not compatible with LS6.0.6+230508.

I have an idea suddently, Could it be a way to generate a token for participants after submition. And the token will became the second survey key if they join in again, for example when they answer the question "Have you participated this survey before", Yes→then they can input the token. Not a good UX but the token can be a gift code later for reward.

Did you install it correctly?
How exactly did you do it?
You should have a folder "update TokenBy Response" in the "plugin" folder of your installation.
Maybe you have to adapt the <compatibility> in "config.xml"
Here it is in version 6.x.
 

Anyway:
We alreasy told you:
In an open survey  - if you only spread the generic link in your chat group - there is no token (participants) table.

And - of course - you have to find a way to merge the two surveys.
Thus is discussed many times with many different ideas.
One is: Let the respondent generate a code that he can remember and reproduce easily, like
First letter of your mother's first name,
Last letter of your name,
Two digits year of birth of your mother,
and whatever
Then you get a string like "BJ44JFU7"
Either you display the string at the end saying "Remember this, you will have to enter it in the second wave"
or you ask the questions again in the second wave.

Now you have a code to merge the waves.

Joffm