Millisecond Forums » Millisecond Forums » Inquisit 6 » Joining two part experiment with subject ID

Joining two part experiment with subject ID

InquisitJK — Thu, 07 Jan 2021 20:12:11 GMT

Hi all,

I have an experiment that is split into two-parts, subjects will complete part 1 and then a week later part 2 via the millisecond platform. Due to counterbalancing, the experiment has three different formats which I assign to subjects via a "batch" script. Here is my question: if the subject is assigned format 1 for part 1 of the experiment how do I ensure that they will be assigned to format 1 for part 2 of the experiment?
As of right now subjects just click the link for each part of the experiment and so there is currently no link between the experiment parts.

Thanks,
JK

RE: Joining two part experiment with subject ID

Dave — Thu, 07 Jan 2021 20:12:11 GMT

+x[quote]

[b]InquisitJK - 1/7/2021[/b]
+x[quote]

[b]Dave - 12/21/2020[/b]
+x[quote]

[b]InquisitJK - 12/21/2020[/b]
Hi all,

I have an experiment that is split into two-parts, subjects will complete part 1 and then a week later part 2 via the millisecond platform. Due to counterbalancing, the experiment has three different formats which I assign to subjects via a "batch" script. Here is my question: if the subject is assigned format 1 for part 1 of the experiment how do I ensure that they will be assigned to format 1 for part 2 of the experiment?
As of right now subjects just click the link for each part of the experiment and so there is currently no link between the experiment parts.

Thanks,
JK
[/quote]

You need to provide more details. It's entirely unclear from your description how you currently assign participants to one of the three part 1 counterbalancing conditions, and it's not clear what "right now subjects just click the link for each part of the experiment" means.
[/quote]

Hi,
Sorry for the late reply.
The experiment has a one week delay and so it has been set up on Inquisit Millisecond into Part 1 and Part 2, and there are separate batch scripts for each Part. Within each batch script, there are 3 different formats for counterbalancing purposes. The two html links provided by Inquisit Millisecond were then input into SONA where there is a Part 1 and Part 2 session (so obviously there is one html link per session).

The problem lies in the fact that participant, let's call him Jack, may be randomly assigned to Format 1 for Part 1 (also known as Session 1) due to the batch script, and then Format 2 for Part 2 (also known as Session 2). In order for the data to make sense it is critical that the Formats match across days.

I require the use of a Batch script to randomize subjects because I want this study to be conducted remotely and so I do not want to be contacting participants what group and subject number they are every single time.

Thank you for your time and if you need more clarity please let me know,
Paige

[/quote]

The way you've set this up, these are entirely separate and unconnected experiments from Inquisit's perspective. There is no way for a experiment B ("session 2") to figure out whatever condition a given participant was randomly assigned to in a previous experiment A ("session 1"), so this is not going to work.

What you could have done is pre-assign each participant to a given condition and then give them correponding links for parts A and B with their condition already included as URL parameter.

If you've already had your participants complete session 1, then the only thing you can do now is to look up each participant's condition in that first session (you have that information in your data files and logs) and then send each participant a link for session 2 with the corresponding condition included as URL parameter.

RE: Joining two part experiment with subject ID

InquisitJK — Thu, 07 Jan 2021 18:03:06 GMT

+x[quote]

[b]Dave - 12/21/2020[/b]
+x[quote]

[b]InquisitJK - 12/21/2020[/b]
Hi all,

I have an experiment that is split into two-parts, subjects will complete part 1 and then a week later part 2 via the millisecond platform. Due to counterbalancing, the experiment has three different formats which I assign to subjects via a "batch" script. Here is my question: if the subject is assigned format 1 for part 1 of the experiment how do I ensure that they will be assigned to format 1 for part 2 of the experiment?
As of right now subjects just click the link for each part of the experiment and so there is currently no link between the experiment parts.

Thanks,
JK
[/quote]

You need to provide more details. It's entirely unclear from your description how you currently assign participants to one of the three part 1 counterbalancing conditions, and it's not clear what "right now subjects just click the link for each part of the experiment" means.
[/quote]

Hi,
Sorry for the late reply.
The experiment has a one week delay and so it has been set up on Inquisit Millisecond into Part 1 and Part 2, and there are separate batch scripts for each Part. Within each batch script, there are 3 different formats for counterbalancing purposes. The two html links provided by Inquisit Millisecond were then input into SONA where there is a Part 1 and Part 2 session (so obviously there is one html link per session).

The problem lies in the fact that participant, let's call him Jack, may be randomly assigned to Format 1 for Part 1 (also known as Session 1) due to the batch script, and then Format 2 for Part 2 (also known as Session 2). In order for the data to make sense it is critical that the Formats match across days.

I require the use of a Batch script to randomize subjects because I want this study to be conducted remotely and so I do not want to be contacting participants what group and subject number they are every single time.

Thank you for your time and if you need more clarity please let me know,
Paige

RE: Joining two part experiment with subject ID

Dave — Mon, 21 Dec 2020 12:51:05 GMT

+x[quote]

[b]InquisitJK - 12/21/2020[/b]
Hi all,

I have an experiment that is split into two-parts, subjects will complete part 1 and then a week later part 2 via the millisecond platform. Due to counterbalancing, the experiment has three different formats which I assign to subjects via a "batch" script. Here is my question: if the subject is assigned format 1 for part 1 of the experiment how do I ensure that they will be assigned to format 1 for part 2 of the experiment?
As of right now subjects just click the link for each part of the experiment and so there is currently no link between the experiment parts.

Thanks,
JK
[/quote]

You need to provide more details. It's entirely unclear from your description how you currently assign participants to one of the three part 1 counterbalancing conditions, and it's not clear what "right now subjects just click the link for each part of the experiment" means.