Custom conditions for moving to the next page

[saudiwin]

Description

At present it's impossible to add custom validation conditions before moving to the next page. I can write an observer on the next button but I can't prevent the page ID to be set to the new value because that's done in the sd_server function.

My proposal is that sd_server takes a named list of boolean reactive values with one value for each page ID; before moving to the next page, the relevant code in the sd_server function would check for TRUE for that reactive value for that page ID. That way custom validation functions could be written for each page that would go beyond required answers (especially as, see my issue, sliders don't really fit into required questions etc.)--without adding unnecessary complexity to the underlying code base.

[jhelvy]

Perhaps you are looking for https://surveydown.org/docs/conditional-logic#conditional-forward-page-skipping

[jhelvy]

We also documented other �ways to control navigation here

[saudiwin]

OK, but I'm looking to prevent people from moving forward. So could I set the page target as the page I don't want the user to be able to leave until they meet the condition?

…

On Mon, Jun 2, 2025 at 11:50 AM John Helveston ***@***.***> wrote: Perhaps you are looking for https://surveydown.org/docs/conditional-logic#conditional-forward-page-skipping — Reply to this email directly, view it on GitHub <#202 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACTVVOMEFDKOOJC36NNO3N33BRXDTAVCNFSM6AAAAAB6NKTMC2VHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGMZUGU2TANI> . You are receiving this because you authored the thread.Message ID: ***@***.*** com>

[LarryVincent]

Seconding the need to validate data before respondent can advance to next question. It's a frequent use case in my survey work, especially when trying to capture certain kinds of specific data (i.e., valid phone number, four-digit year, etc.). After the basic functionality is created, at some point it would also be great to have some convenience validation handlers sd_validate_phone, sd_validate_numeric, sd_validate_text, etc. Those would be "lucky strike extras" but could save a lot of time. If you can also link those to custom javascript that highlights the field with an error message, that would raise the bar considerably.

Larry Vincent
Professor of the Practice, Marketing
USC Marshall School of Business

[jhelvy]

Thanks for bringing this back up. Since this discussion thread has several different topics in it, I just made a new issue to address the specific request for custom validation. I agree this is a necessary addition, and it shouldn't be too hard to implement. We'll work on it. The additional sd_validate_*() functions you proposed may be part of the solution.

[jhelvy]

Ah okay so this is like the opposite of sd_skip_foward(). With skipping we send someone to a forward page based on a condition being met. In your case, you want to prevent someone from being able to move foward until some condition is met. We don't have that natively supported yet, but we can probably add it following a similar to how we have designed sd_skip_foward(), where you would give a condition ~ target pattern. In this case, the target would be the page you're not allowed to leave until the condition is met.

We have internal logic for this kind of prevention in the sd_server() already with required questions. This prevents the respondent from going forward until the question is answered. So you're just asking for the ability to set some other condition. I think we can figure this out.

The only question I have is what the function name should be. Maybe sd_require_condition() or sd_require_question()? Like:

sd_require_condition(
  condition ~ page_id
)

I'm thinking of this name because we currently use required_questions as the argument name in sd_server() to require a question be answered.

[saudiwin]

Yes it would be something like sd_require_condition. Adding this in would make the platform remarkably flexible because any type of arbitrary validation could be added to a page. The only other thing is perhaps it would be good to pass a string for a pop-up if the condition fails. This isn’t as necessary for skipping forward, but if the user gets stuck failing a validation check, it’d be great to inform them. Could be done separately by adding a reactive function to the next button, too (but involves additional coding etc). Robert Kubinec Assistant Professor of Political Science Department of Political Science University of South Carolina

…

On Jun 2, 2025, at 1:26 PM, John Helveston ***@***.***> wrote: Ah okay so this is like the opposite of sd_skip_foward(). With skipping we send someone to a forward page based on a condition being met. In your case, you want to prevent someone from being able to move foward until some condition is met. We don't have that natively supported yet, but we can probably add it following a similar to how we have designed sd_skip_foward(), where you would give a condition ~ target pattern. In this case, the target would be the page you're not allowed to leave until the condition is met. We have internal logic for this kind of prevention in the sd_server() already with required questions <https://surveydown.org/docs/server-options.html#required-questions>. This prevents the respondent from going forward until the question is answered. So you're just asking for the ability to set some other condition. I think we can figure this out. The only question I have is what the function name should be. Maybe sd_require_condition() or sd_require_question()? Like: sd_require_condition( condition ~ page_id ) I'm thinking of this name because we currently use required_questions as the argument name in sd_server() to require a question be answered. — Reply to this email directly, view it on GitHub <#202 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACTVVOMO4CHZFUGR4YTQ6B33BSCKXAVCNFSM6AAAAAB6NKTMC2VHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGMZUGYZTONA>. You are receiving this because you authored the thread.

[jhelvy]

We already have a pop up message for required questions, so we would probably want to make this a required argument for something like sd_require_condition() so that respondents know why they've been blocked from proceeding. We could have a generic default message.

[saudiwin]

Also unrelated (perhaps I could do a different feature request), but I think it’d be really useful to have some kind of reactive div or counter for the user answering questions. As in when a page loads, there’s a black box around each unanswered question, then when they select an answer, the black box disappears. That way they’d know when it was answered or not. Robert Kubinec Assistant Professor of Political Science Department of Political Science University of South Carolina

…

On Jun 2, 2025, at 1:30 PM, Bob K ***@***.***> wrote: Yes it would be something like sd_require_condition. Adding this in would make the platform remarkably flexible because any type of arbitrary validation could be added to a page. The only other thing is perhaps it would be good to pass a string for a pop-up if the condition fails. This isn’t as necessary for skipping forward, but if the user gets stuck failing a validation check, it’d be great to inform them. Could be done separately by adding a reactive function to the next button, too (but involves additional coding etc). Robert Kubinec Assistant Professor of Political Science Department of Political Science University of South Carolina > On Jun 2, 2025, at 1:26 PM, John Helveston ***@***.***> wrote: > > > Ah okay so this is like the opposite of sd_skip_foward(). With skipping we send someone to a forward page based on a condition being met. In your case, you want to prevent someone from being able to move foward until some condition is met. We don't have that natively supported yet, but we can probably add it following a similar to how we have designed sd_skip_foward(), where you would give a condition ~ target pattern. In this case, the target would be the page you're not allowed to leave until the condition is met. > > We have internal logic for this kind of prevention in the sd_server() already with required questions <https://surveydown.org/docs/server-options.html#required-questions>. This prevents the respondent from going forward until the question is answered. So you're just asking for the ability to set some other condition. I think we can figure this out. > > The only question I have is what the function name should be. Maybe sd_require_condition() or sd_require_question()? Like: > > sd_require_condition( > condition ~ page_id > ) > I'm thinking of this name because we currently use required_questions as the argument name in sd_server() to require a question be answered. > > — > Reply to this email directly, view it on GitHub <#202 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACTVVOMO4CHZFUGR4YTQ6B33BSCKXAVCNFSM6AAAAAB6NKTMC2VHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGMZUGYZTONA>. > You are receiving this because you authored the thread. >

[jhelvy]

This is an interesting idea. We currently don't provide any sort of user feedback that a question has been marked as answered. Could you open that as a feature request in a new issue?

[jhelvy]

@pingfan-hu I wonder if you could use a similar approach to what you did with the recent change for required questions? In the case of required questions, you get a red box around each required question that wasn't answered. But maybe you could have a thin black line around any unanswered question that disappears once it's answered? It could also be an option in sd_server() so users can turn it off if they want.

[saudiwin]

I think this could be a real game-changer as "survey satisficing" is a big open problem in survey methodology, and Qualtrics has fairly limited "nudges". In fact, this could be a cool research paper -- run an online survey where we randomize these features and see to what extent we can lower non-response rates.

[pingfan-hu]

I think this is a great idea. It's not hard to achieve since I only need an inverse logic to check for the "answered" status. I'll work out a smooth UI and see how it feels as a survey participant.

[cwilligv]

Hi @jhelvy , I have just open a new discussion around this same topic. I need to validate the answer of a text-based question before letting the user continue to the next question. I couldn't make it work with conditionals so this would be very helpful. In addition to the required answer, a condition that can be provided as a custom function.
here is the discussion I opened: https://github.com/orgs/surveydown-dev/discussions/206

[jhelvy]

I just replied on that thread, this might be useful.

[pingfan-hu]

Hi folks, we just bumped surveydown to v0.13.2 adding the new feature of sd_stop_if(), which serves as a conditional logic for "stopping navigation if a condition is TRUE.

It's not on CRAN yet so please install the latest version using:

pak::pak("surveydown-dev/surveydown", ask = FALSE)

A more inclusive discussion and reasoning flow can be accessed in issue #216 .

Feel free to check our document and templates for this updated conditional logic.