Millisecond Forums

+x

 

Dave - 3/6/2026

+x

 

wmfang - 3/6/2026

In my data, I noticed that subjects have multiple participation sessions. There should only be 1 row per subject ID (1 group, date, time, LinkedInCondition). But, some have multiple rows with different group ID, date, time, and (for some) different conditions. In the screenshot, I am only presenting subject IDs with multiple rows. 

I took a closer look at their responses. Some subjects are stopping part-way through the study, then coming back and starting it from the beginning at a later date and time. Some are on the same day and only minutes or an hour later (e.g., subject 98170). Other subjects go all the way through, then complete the study a second time at a later date (e.g., subject 82684).

I checked my script multiple times with collaborators, verifying that a group ID is randomly generated per session. 

Has anyone else run into and resolved this issue? What is going on? How do I prevent subjects from starting / completing the study multiple times? It's a very short study, only 5 - 7 minutes.

 

There isn't really an issue here, starting / completing a task multiple times is not prohibited, there is no good technical way to do it, and it is often desirable (somebody might get distracted and have to return later, there might be a technical issue on the first attempt, and many studies are multi-session / repeated measurement, where people are expected to complete a task multiple times or different tasks at different time points).

A new group ID is only generated when the launch page is reloaded. If it isn't, then the group ID stays the same, as is the case for e.g. subjects 79462 and 83395. For both of these, it seems unlikely that they actually completed the study twice, since the entries are only a minute apart. Maybe a problem on first launch, but you don't show the full log, so it's impossible to say.

For people who actually fully completed the task multiple times (on same or different dates), you could simply discard their later data.

+x

 

Dave - 3/6/2026

+x

 

Dave - 3/6/2026

+x

 

wmfang - 3/6/2026

In my data, I noticed that subjects have multiple participation sessions. There should only be 1 row per subject ID (1 group, date, time, LinkedInCondition). But, some have multiple rows with different group ID, date, time, and (for some) different conditions. In the screenshot, I am only presenting subject IDs with multiple rows. 

I took a closer look at their responses. Some subjects are stopping part-way through the study, then coming back and starting it from the beginning at a later date and time. Some are on the same day and only minutes or an hour later (e.g., subject 98170). Other subjects go all the way through, then complete the study a second time at a later date (e.g., subject 82684).

I checked my script multiple times with collaborators, verifying that a group ID is randomly generated per session. 

Has anyone else run into and resolved this issue? What is going on? How do I prevent subjects from starting / completing the study multiple times? It's a very short study, only 5 - 7 minutes.

 

There isn't really an issue here, starting / completing a task multiple times is not prohibited, there is no good technical way to do it, and it is often desirable (somebody might get distracted and have to return later, there might be a technical issue on the first attempt, and many studies are multi-session / repeated measurement, where people are expected to complete a task multiple times or different tasks at different time points).

A new group ID is only generated when the launch page is reloaded. If it isn't, then the group ID stays the same, as is the case for e.g. subjects 79462 and 83395. For both of these, it seems unlikely that they actually completed the study twice, since the entries are only a minute apart. Maybe a problem on first launch, but you don't show the full log, so it's impossible to say.

For people who actually fully completed the task multiple times (on same or different dates), you could simply discard their later data.

Adding: If you really want to prevent people from being able to start again, you can do something like

(IQJS)

<expt yourScriptsExptElement>
/ onExptBegin = {
    // terminate if session number is greater 1
    if (script.sessionId > 1) {
        return script.abort(true)
    }
}
...
</expt>

(IQX)

<expt yourScriptsExptElement>
/ onExptBegin = [
    // terminate if session number is greater 1
    if (script.sessionId > 1) {
        return script.abort(true)
    }
]
...
</expt>

Note that this would also bar people from restarting who may have a legitimate reason to (e.g. a technical issue or some interruption early on that prevented them from completing the task at that time).

I don't know your task and how it's structured, but one reason why some people might quit early and restart shortly after is having missed reading the instructions thoroughly and having no way to re-read them in-task. If that applies to your script, giving people a simple way to re-read instructions and/or repeat any practice portion may be worth considering.

Thank you very much for your help! I'll think about implementing this. 

+x

 

Dave - 3/6/2026

+x

 

wmfang - 3/6/2026

In my data, I noticed that subjects have multiple participation sessions. There should only be 1 row per subject ID (1 group, date, time, LinkedInCondition). But, some have multiple rows with different group ID, date, time, and (for some) different conditions. In the screenshot, I am only presenting subject IDs with multiple rows. 

I took a closer look at their responses. Some subjects are stopping part-way through the study, then coming back and starting it from the beginning at a later date and time. Some are on the same day and only minutes or an hour later (e.g., subject 98170). Other subjects go all the way through, then complete the study a second time at a later date (e.g., subject 82684).

I checked my script multiple times with collaborators, verifying that a group ID is randomly generated per session. 

Has anyone else run into and resolved this issue? What is going on? How do I prevent subjects from starting / completing the study multiple times? It's a very short study, only 5 - 7 minutes.

 

There isn't really an issue here, starting / completing a task multiple times is not prohibited, there is no good technical way to do it, and it is often desirable (somebody might get distracted and have to return later, there might be a technical issue on the first attempt, and many studies are multi-session / repeated measurement, where people are expected to complete a task multiple times or different tasks at different time points).

A new group ID is only generated when the launch page is reloaded. If it isn't, then the group ID stays the same, as is the case for e.g. subjects 79462 and 83395. For both of these, it seems unlikely that they actually completed the study twice, since the entries are only a minute apart. Maybe a problem on first launch, but you don't show the full log, so it's impossible to say.

For people who actually fully completed the task multiple times (on same or different dates), you could simply discard their later data.

Adding: If you really want to prevent people from being able to start again, you can do something like

(IQJS)


(IQX)


Note that this would also bar people from restarting who may have a legitimate reason to (e.g. a technical issue or some interruption early on that prevented them from completing the task at that time).

I don't know your task and how it's structured, but one reason why some people might quit early and restart shortly after is having missed reading the instructions thoroughly and having no way to re-read them in-task. If that applies to your script, giving people a simple way to re-read instructions and/or repeat any practice portion may be worth considering.

+x

 

wmfang - 3/6/2026

In my data, I noticed that subjects have multiple participation sessions. There should only be 1 row per subject ID (1 group, date, time, LinkedInCondition). But, some have multiple rows with different group ID, date, time, and (for some) different conditions. In the screenshot, I am only presenting subject IDs with multiple rows. 

I took a closer look at their responses. Some subjects are stopping part-way through the study, then coming back and starting it from the beginning at a later date and time. Some are on the same day and only minutes or an hour later (e.g., subject 98170). Other subjects go all the way through, then complete the study a second time at a later date (e.g., subject 82684).

I checked my script multiple times with collaborators, verifying that a group ID is randomly generated per session. 

Has anyone else run into and resolved this issue? What is going on? How do I prevent subjects from starting / completing the study multiple times? It's a very short study, only 5 - 7 minutes.

 

There isn't really an issue here, starting / completing a task multiple times is not prohibited, there is no good technical way to do it, and it is often desirable (somebody might get distracted and have to return later, there might be a technical issue on the first attempt, and many studies are multi-session / repeated measurement, where people are expected to complete a task multiple times or different tasks at different time points).

A new group ID is only generated when the launch page is reloaded. If it isn't, then the group ID stays the same, as is the case for e.g. subjects 79462 and 83395. For both of these, it seems unlikely that they actually completed the study twice, since the entries are only a minute apart. Maybe a problem on first launch, but you don't show the full log, so it's impossible to say.

For people who actually fully completed the task multiple times (on same or different dates), you could simply discard their later data.

In my data, I noticed that subjects have multiple participation sessions. There should only be 1 row per subject ID (1 group, date, time, LinkedInCondition). But, some have multiple rows with different group ID, date, time, and (for some) different conditions. In the screenshot, I am only presenting subject IDs with multiple rows. 

I took a closer look at their responses. Some subjects are stopping part-way through the study, then coming back and starting it from the beginning at a later date and time. Some are on the same day and only minutes or an hour later (e.g., subject 98170). Other subjects go all the way through, then complete the study a second time at a later date (e.g., subject 82684).

I checked my script multiple times with collaborators, verifying that a group ID is randomly generated per session. 

Has anyone else run into and resolved this issue? What is going on? How do I prevent subjects from starting / completing the study multiple times? It's a very short study, only 5 - 7 minutes.