Millisecond Forums

+x

 

Selene_MQ - 4/28/2020

Good morning,

I am very new to Inquisit and still not familiar with how it works. I have been trying to adapt the existing Emotional-n-back task that I got from the library, to create a simple 1-back task using 56 different images. The idea is for participants to implicitely learn these images by doing a 1-back task. So far, the script seems to be doing approximately what I want, except that when I calculated at the end the number of times each image was presented, it varied greatly (with 8 blocks, and 64 images per block, I would expect each of my 56 images to appear about 9 times, but in reality it varied between 2 and 23). I really don't know how to control this so that each image appears the same number of times across the experiment.

I have attached the script to this post, and would greatly appreciate any help you may provide,

Thank you so much in advance,
Selene

+x

 

Dave - 4/28/2020

+x

 

Selene_MQ - 4/28/2020

Good morning,

I am very new to Inquisit and still not familiar with how it works. I have been trying to adapt the existing Emotional-n-back task that I got from the library, to create a simple 1-back task using 56 different images. The idea is for participants to implicitely learn these images by doing a 1-back task. So far, the script seems to be doing approximately what I want, except that when I calculated at the end the number of times each image was presented, it varied greatly (with 8 blocks, and 64 images per block, I would expect each of my 56 images to appear about 9 times, but in reality it varied between 2 and 23). I really don't know how to control this so that each image appears the same number of times across the experiment.

I have attached the script to this post, and would greatly appreciate any help you may provide,

Thank you so much in advance,
Selene

<list notargetvalue>
/ items = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56)
/ not = (values.currenttarget)
/ replace = true
</list>

You're sampling items randomly *with* replacement. If you want a more even distribution, sample without replacement.

+x

 

Selene_MQ - 4/29/2020

+x

 

Dave - 4/28/2020

+x

 

Selene_MQ - 4/28/2020

Good morning,

I am very new to Inquisit and still not familiar with how it works. I have been trying to adapt the existing Emotional-n-back task that I got from the library, to create a simple 1-back task using 56 different images. The idea is for participants to implicitely learn these images by doing a 1-back task. So far, the script seems to be doing approximately what I want, except that when I calculated at the end the number of times each image was presented, it varied greatly (with 8 blocks, and 64 images per block, I would expect each of my 56 images to appear about 9 times, but in reality it varied between 2 and 23). I really don't know how to control this so that each image appears the same number of times across the experiment.

I have attached the script to this post, and would greatly appreciate any help you may provide,

Thank you so much in advance,
Selene

<list notargetvalue>
/ items = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56)
/ not = (values.currenttarget)
/ replace = true
</list>

You're sampling items randomly *with* replacement. If you want a more even distribution, sample without replacement.

Thanks a  lot for that. It does make it slightly better, but there is still a large variation in the number of times each image is seen (from 7 to 18 times instead of having them all presented 9 times). Is there another way to make it more even, that doesn't require massive changes in the script?

Thank you so much!
Selene

+x

 

Dave - 4/29/2020

+x

 

Selene_MQ - 4/29/2020

+x

 

Dave - 4/28/2020

+x

 

Selene_MQ - 4/28/2020

Good morning,

I am very new to Inquisit and still not familiar with how it works. I have been trying to adapt the existing Emotional-n-back task that I got from the library, to create a simple 1-back task using 56 different images. The idea is for participants to implicitely learn these images by doing a 1-back task. So far, the script seems to be doing approximately what I want, except that when I calculated at the end the number of times each image was presented, it varied greatly (with 8 blocks, and 64 images per block, I would expect each of my 56 images to appear about 9 times, but in reality it varied between 2 and 23). I really don't know how to control this so that each image appears the same number of times across the experiment.

I have attached the script to this post, and would greatly appreciate any help you may provide,

Thank you so much in advance,
Selene

<list notargetvalue>
/ items = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56)
/ not = (values.currenttarget)
/ replace = true
</list>

You're sampling items randomly *with* replacement. If you want a more even distribution, sample without replacement.

Thanks a  lot for that. It does make it slightly better, but there is still a large variation in the number of times each image is seen (from 7 to 18 times instead of having them all presented 9 times). Is there another way to make it more even, that doesn't require massive changes in the script?

Thank you so much!
Selene

Well, it's an n-back task. I.e. some images *have* to be repeated more than others, because they're target images, i.e. the same image as than n (here: 1) trials before. Other images aren't targets, so will ultimately be displayed less often, because -- among other things -- there are fewer target trials than there are non-target trials. In short, no, I don't think there's any easy way to force each image to be displayed exactly (or close to) 9 times over the course of the task. In fact, I don't know that there's any way at all.

+x

 

Selene_MQ - 4/30/2020

+x

 

Dave - 4/29/2020

+x

 

Selene_MQ - 4/29/2020

+x

 

Dave - 4/28/2020

+x

 

Selene_MQ - 4/28/2020

Good morning,

I am very new to Inquisit and still not familiar with how it works. I have been trying to adapt the existing Emotional-n-back task that I got from the library, to create a simple 1-back task using 56 different images. The idea is for participants to implicitely learn these images by doing a 1-back task. So far, the script seems to be doing approximately what I want, except that when I calculated at the end the number of times each image was presented, it varied greatly (with 8 blocks, and 64 images per block, I would expect each of my 56 images to appear about 9 times, but in reality it varied between 2 and 23). I really don't know how to control this so that each image appears the same number of times across the experiment.

I have attached the script to this post, and would greatly appreciate any help you may provide,

Thank you so much in advance,
Selene

<list notargetvalue>
/ items = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56)
/ not = (values.currenttarget)
/ replace = true
</list>

You're sampling items randomly *with* replacement. If you want a more even distribution, sample without replacement.

Thanks a  lot for that. It does make it slightly better, but there is still a large variation in the number of times each image is seen (from 7 to 18 times instead of having them all presented 9 times). Is there another way to make it more even, that doesn't require massive changes in the script?

Thank you so much!
Selene

Well, it's an n-back task. I.e. some images *have* to be repeated more than others, because they're target images, i.e. the same image as than n (here: 1) trials before. Other images aren't targets, so will ultimately be displayed less often, because -- among other things -- there are fewer target trials than there are non-target trials. In short, no, I don't think there's any easy way to force each image to be displayed exactly (or close to) 9 times over the course of the task. In fact, I don't know that there's any way at all.

Thanks for your answer. I do understand that the targets have to be repeated, but because each image in the pool can be a target (i.e. there isn't a separate pool of images that constitute the non-target images), I think it's possible to think of a solution that will present each image exactly as many times as possible.

So my new question is: if I can figure out an order of presentation that works (for example: 4 6 8 8 5 9 1 12 13 13 etc... over the entire experiment), how easy is it to then hard-code the order in the script?

Thanks a lot!

+x

 

Dave - 4/30/2020

+x

 

Selene_MQ - 4/30/2020

+x

 

Dave - 4/29/2020

+x

 

Selene_MQ - 4/29/2020

+x

 

Dave - 4/28/2020

+x

 

Selene_MQ - 4/28/2020

Good morning,

I am very new to Inquisit and still not familiar with how it works. I have been trying to adapt the existing Emotional-n-back task that I got from the library, to create a simple 1-back task using 56 different images. The idea is for participants to implicitely learn these images by doing a 1-back task. So far, the script seems to be doing approximately what I want, except that when I calculated at the end the number of times each image was presented, it varied greatly (with 8 blocks, and 64 images per block, I would expect each of my 56 images to appear about 9 times, but in reality it varied between 2 and 23). I really don't know how to control this so that each image appears the same number of times across the experiment.

I have attached the script to this post, and would greatly appreciate any help you may provide,

Thank you so much in advance,
Selene

<list notargetvalue>
/ items = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56)
/ not = (values.currenttarget)
/ replace = true
</list>

You're sampling items randomly *with* replacement. If you want a more even distribution, sample without replacement.

Thanks a  lot for that. It does make it slightly better, but there is still a large variation in the number of times each image is seen (from 7 to 18 times instead of having them all presented 9 times). Is there another way to make it more even, that doesn't require massive changes in the script?

Thank you so much!
Selene

Well, it's an n-back task. I.e. some images *have* to be repeated more than others, because they're target images, i.e. the same image as than n (here: 1) trials before. Other images aren't targets, so will ultimately be displayed less often, because -- among other things -- there are fewer target trials than there are non-target trials. In short, no, I don't think there's any easy way to force each image to be displayed exactly (or close to) 9 times over the course of the task. In fact, I don't know that there's any way at all.

Thanks for your answer. I do understand that the targets have to be repeated, but because each image in the pool can be a target (i.e. there isn't a separate pool of images that constitute the non-target images), I think it's possible to think of a solution that will present each image exactly as many times as possible.

So my new question is: if I can figure out an order of presentation that works (for example: 4 6 8 8 5 9 1 12 13 13 etc... over the entire experiment), how easy is it to then hard-code the order in the script?

Thanks a lot!

You have to hardcode two things, then.

- The full item order in the non-target list, and set it to /selectionmode = sequence
- The order of target and non-target trials run by the blocks. That you'll then also have to switch to a list, and set it to sequence.

That's not necessarily hard, but fiddly and error-prone given the amount of trials.

+x

 

Selene_MQ - 5/4/2020

+x

 

Dave - 4/30/2020

+x

 

Selene_MQ - 4/30/2020

+x

 

Dave - 4/29/2020

+x

 

Selene_MQ - 4/29/2020

+x

 

Dave - 4/28/2020

+x

 

Selene_MQ - 4/28/2020

Good morning,

I am very new to Inquisit and still not familiar with how it works. I have been trying to adapt the existing Emotional-n-back task that I got from the library, to create a simple 1-back task using 56 different images. The idea is for participants to implicitely learn these images by doing a 1-back task. So far, the script seems to be doing approximately what I want, except that when I calculated at the end the number of times each image was presented, it varied greatly (with 8 blocks, and 64 images per block, I would expect each of my 56 images to appear about 9 times, but in reality it varied between 2 and 23). I really don't know how to control this so that each image appears the same number of times across the experiment.

I have attached the script to this post, and would greatly appreciate any help you may provide,

Thank you so much in advance,
Selene

<list notargetvalue>
/ items = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56)
/ not = (values.currenttarget)
/ replace = true
</list>

You're sampling items randomly *with* replacement. If you want a more even distribution, sample without replacement.

Thanks a  lot for that. It does make it slightly better, but there is still a large variation in the number of times each image is seen (from 7 to 18 times instead of having them all presented 9 times). Is there another way to make it more even, that doesn't require massive changes in the script?

Thank you so much!
Selene

Well, it's an n-back task. I.e. some images *have* to be repeated more than others, because they're target images, i.e. the same image as than n (here: 1) trials before. Other images aren't targets, so will ultimately be displayed less often, because -- among other things -- there are fewer target trials than there are non-target trials. In short, no, I don't think there's any easy way to force each image to be displayed exactly (or close to) 9 times over the course of the task. In fact, I don't know that there's any way at all.

Thanks for your answer. I do understand that the targets have to be repeated, but because each image in the pool can be a target (i.e. there isn't a separate pool of images that constitute the non-target images), I think it's possible to think of a solution that will present each image exactly as many times as possible.

So my new question is: if I can figure out an order of presentation that works (for example: 4 6 8 8 5 9 1 12 13 13 etc... over the entire experiment), how easy is it to then hard-code the order in the script?

Thanks a lot!

You have to hardcode two things, then.

- The full item order in the non-target list, and set it to /selectionmode = sequence
- The order of target and non-target trials run by the blocks. That you'll then also have to switch to a list, and set it to sequence.

That's not necessarily hard, but fiddly and error-prone given the amount of trials.

Thank you very much for your answer, Dave.
I generated, for each of the 8 blocks, the sequence that I want, and have also generated for each block a vector with the order of Target/Non-target that matches my sequence.
I am not sure how to implement it, because each of the 8 blocks will have a different sequence... Would you please be able to help with this? I have attached the script, and under the "LIST" tab, you can see the orders that I generated.

Thank you so much for your help,
Selene

+x

 

Dave - 5/4/2020

+x

 

Selene_MQ - 5/4/2020

+x

 

Dave - 4/30/2020

+x

 

Selene_MQ - 4/30/2020

+x

 

Dave - 4/29/2020

+x

 

Selene_MQ - 4/29/2020

+x

 

Dave - 4/28/2020

+x

 

Selene_MQ - 4/28/2020

Good morning,

I am very new to Inquisit and still not familiar with how it works. I have been trying to adapt the existing Emotional-n-back task that I got from the library, to create a simple 1-back task using 56 different images. The idea is for participants to implicitely learn these images by doing a 1-back task. So far, the script seems to be doing approximately what I want, except that when I calculated at the end the number of times each image was presented, it varied greatly (with 8 blocks, and 64 images per block, I would expect each of my 56 images to appear about 9 times, but in reality it varied between 2 and 23). I really don't know how to control this so that each image appears the same number of times across the experiment.

I have attached the script to this post, and would greatly appreciate any help you may provide,

Thank you so much in advance,
Selene

<list notargetvalue>
/ items = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56)
/ not = (values.currenttarget)
/ replace = true
</list>

You're sampling items randomly *with* replacement. If you want a more even distribution, sample without replacement.

Thanks a  lot for that. It does make it slightly better, but there is still a large variation in the number of times each image is seen (from 7 to 18 times instead of having them all presented 9 times). Is there another way to make it more even, that doesn't require massive changes in the script?

Thank you so much!
Selene

Well, it's an n-back task. I.e. some images *have* to be repeated more than others, because they're target images, i.e. the same image as than n (here: 1) trials before. Other images aren't targets, so will ultimately be displayed less often, because -- among other things -- there are fewer target trials than there are non-target trials. In short, no, I don't think there's any easy way to force each image to be displayed exactly (or close to) 9 times over the course of the task. In fact, I don't know that there's any way at all.

Thanks for your answer. I do understand that the targets have to be repeated, but because each image in the pool can be a target (i.e. there isn't a separate pool of images that constitute the non-target images), I think it's possible to think of a solution that will present each image exactly as many times as possible.

So my new question is: if I can figure out an order of presentation that works (for example: 4 6 8 8 5 9 1 12 13 13 etc... over the entire experiment), how easy is it to then hard-code the order in the script?

Thanks a lot!

You have to hardcode two things, then.

- The full item order in the non-target list, and set it to /selectionmode = sequence
- The order of target and non-target trials run by the blocks. That you'll then also have to switch to a list, and set it to sequence.

That's not necessarily hard, but fiddly and error-prone given the amount of trials.

Thank you very much for your answer, Dave.
I generated, for each of the 8 blocks, the sequence that I want, and have also generated for each block a vector with the order of Target/Non-target that matches my sequence.
I am not sure how to implement it, because each of the 8 blocks will have a different sequence... Would you please be able to help with this? I have attached the script, and under the "LIST" tab, you can see the orders that I generated.

Thank you so much for your help,
Selene

You can creat lists of lists and select the list you need /onblockbegin. See e.g. https://www.millisecond.com/forums/Topic16652.aspx#16717, https://www.millisecond.com/forums/FindPost20679.aspx

+x

 

Selene_MQ - 5/5/2020

+x

 

Dave - 5/4/2020

+x

 

Selene_MQ - 5/4/2020

+x

 

Dave - 4/30/2020

+x

 

Selene_MQ - 4/30/2020

+x

 

Dave - 4/29/2020

+x

 

Selene_MQ - 4/29/2020

+x

 

Dave - 4/28/2020

+x

 

Selene_MQ - 4/28/2020

Good morning,

I am very new to Inquisit and still not familiar with how it works. I have been trying to adapt the existing Emotional-n-back task that I got from the library, to create a simple 1-back task using 56 different images. The idea is for participants to implicitely learn these images by doing a 1-back task. So far, the script seems to be doing approximately what I want, except that when I calculated at the end the number of times each image was presented, it varied greatly (with 8 blocks, and 64 images per block, I would expect each of my 56 images to appear about 9 times, but in reality it varied between 2 and 23). I really don't know how to control this so that each image appears the same number of times across the experiment.

I have attached the script to this post, and would greatly appreciate any help you may provide,

Thank you so much in advance,
Selene

<list notargetvalue>
/ items = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56)
/ not = (values.currenttarget)
/ replace = true
</list>

You're sampling items randomly *with* replacement. If you want a more even distribution, sample without replacement.

Thanks a  lot for that. It does make it slightly better, but there is still a large variation in the number of times each image is seen (from 7 to 18 times instead of having them all presented 9 times). Is there another way to make it more even, that doesn't require massive changes in the script?

Thank you so much!
Selene

Well, it's an n-back task. I.e. some images *have* to be repeated more than others, because they're target images, i.e. the same image as than n (here: 1) trials before. Other images aren't targets, so will ultimately be displayed less often, because -- among other things -- there are fewer target trials than there are non-target trials. In short, no, I don't think there's any easy way to force each image to be displayed exactly (or close to) 9 times over the course of the task. In fact, I don't know that there's any way at all.

Thanks for your answer. I do understand that the targets have to be repeated, but because each image in the pool can be a target (i.e. there isn't a separate pool of images that constitute the non-target images), I think it's possible to think of a solution that will present each image exactly as many times as possible.

So my new question is: if I can figure out an order of presentation that works (for example: 4 6 8 8 5 9 1 12 13 13 etc... over the entire experiment), how easy is it to then hard-code the order in the script?

Thanks a lot!

You have to hardcode two things, then.

- The full item order in the non-target list, and set it to /selectionmode = sequence
- The order of target and non-target trials run by the blocks. That you'll then also have to switch to a list, and set it to sequence.

That's not necessarily hard, but fiddly and error-prone given the amount of trials.

Thank you very much for your answer, Dave.
I generated, for each of the 8 blocks, the sequence that I want, and have also generated for each block a vector with the order of Target/Non-target that matches my sequence.
I am not sure how to implement it, because each of the 8 blocks will have a different sequence... Would you please be able to help with this? I have attached the script, and under the "LIST" tab, you can see the orders that I generated.

Thank you so much for your help,
Selene

You can creat lists of lists and select the list you need /onblockbegin. See e.g. https://www.millisecond.com/forums/Topic16652.aspx#16717, https://www.millisecond.com/forums/FindPost20679.aspx

Thanks for your answer Dace. I had a look at these posts and try to adapt it to my code. I think I'm not too far off, but it just won't start. Now it throws an error when it reaches the blocks : "Could not locate element ". " and "trial 2 is unassigned" for each of the blocks.

Any idea where I sent wrong?

Thanks so much!