Millisecond Forums

+x

 

Kevin_Bague - 1/30/2020

Hello, I'm coding an experiment in which I would like include a random allocation of participants. I have 3 conditions. So I created one block for each condition and used expt element with noreplace function to random block apparition. But I would like the same sample size in each condition, and so in each block. Do someone know how do this ? 

Thank you for your answers

Set up three <expt> elements like so

<expt>
/ subjects = (1 of 3)
/ groupassignment = groupnumber
/ blocks = [1=condition1;]
</expt>

<expt>
/ subjects = (2 of 3)
/ groupassignment = groupnumber
/ blocks = [1=condition2;]
</expt>

<expt>
/ subjects = (3 of 3)
/ groupassignment = groupnumber
/ blocks = [1=condition3;]
</expt>

If you're running this on the web, indicate the number of groups as 3 in your web experiment settings and select either sequential or random (without replacement) group ID generation. If you're running this in Inquisit Lab, the assignment is fully under your control, just enter the group number corresponding to the condition you want a given participant to receive at the start.

+x

 

Dave - 1/30/2020

+x

 

Kevin_Bague - 1/30/2020

Hello, I'm coding an experiment in which I would like include a random allocation of participants. I have 3 conditions. So I created one block for each condition and used expt element with noreplace function to random block apparition. But I would like the same sample size in each condition, and so in each block. Do someone know how do this ? 

Thank you for your answers

Set up three <expt> elements like so

<expt>
/ subjects = (1 of 3)
/ groupassignment = groupnumber
/ blocks = [1=condition1;]
</expt>

<expt>
/ subjects = (2 of 3)
/ groupassignment = groupnumber
/ blocks = [1=condition2;]
</expt>

<expt>
/ subjects = (3 of 3)
/ groupassignment = groupnumber
/ blocks = [1=condition3;]
</expt>

If you're running this on the web, indicate the number of groups as 3 in your web experiment settings and select either sequential or random (without replacement) group ID generation. If you're running this in Inquisit Lab, the assignment is fully under your control, just enter the group number corresponding to the condition you want a given participant to receive at the start.

Thank you very  much for your answer. I'm running code in Inquisit Lab. I have thought to use this method, but I can't make a random allocation, because I must choose the  participant group.... 

+x

 

Kevin_Bague - 1/31/2020

+x

 

Dave - 1/30/2020

+x

 

Kevin_Bague - 1/30/2020

Hello, I'm coding an experiment in which I would like include a random allocation of participants. I have 3 conditions. So I created one block for each condition and used expt element with noreplace function to random block apparition. But I would like the same sample size in each condition, and so in each block. Do someone know how do this ? 

Thank you for your answers

Set up three <expt> elements like so

<expt>
/ subjects = (1 of 3)
/ groupassignment = groupnumber
/ blocks = [1=condition1;]
</expt>

<expt>
/ subjects = (2 of 3)
/ groupassignment = groupnumber
/ blocks = [1=condition2;]
</expt>

<expt>
/ subjects = (3 of 3)
/ groupassignment = groupnumber
/ blocks = [1=condition3;]
</expt>

If you're running this on the web, indicate the number of groups as 3 in your web experiment settings and select either sequential or random (without replacement) group ID generation. If you're running this in Inquisit Lab, the assignment is fully under your control, just enter the group number corresponding to the condition you want a given participant to receive at the start.

Thank you very  much for your answer. I'm running code in Inquisit Lab. I have thought to use this method, but I can't make a random allocation, because I must choose the  participant group.... 

Then you need to script something outside of Inquisit that performs the random allocation and then have that script launch Inquisit from the command line accordingly.

https://www.millisecond.com/support/docs/v5/html/howto/howtocommandline.htm