How to connect the "age" factor with an external database, depending on the age of each participant?
19 views
Skip to first unread message
Marco A. López
Dec 17, 2017, 12:57:07 PM12/17/17
to E-Prime
Hello everyone,
I have an experimental design in which the participants listen randomly to several songs included in a database. For it, I have created a list with several nested sublists. The idea would be that each participant listens a sequence of songs based on the years in each one was between 15 and 30 years old. For example, a person with 35 years would listen to songs between 1993 and 2008. My problem is to create an inline or instruction, so that automatically the program assigns each participant the songs corresponding to their age...How could I do that?
On the other hand, I have a task in which the participants listen to a sequence of songs classified by decades, in 10 blocks. The idea would be that presentation order was randomized, with the constraint that one song from each block must be sampled before a second song could be drawn from a particular block. In this case, what instructions would be the ideal? Could it be done only with objects?
Thanks in advance for any help.
Marco
David McFarlane
Dec 18, 2017, 2:34:48 PM12/18/17
to e-p...@googlegroups.com
Marco,
Addressing only your second question regarding a specific case of
"randomization with constraints" ...
For this case, I would make a List or nested List with 10 levels (rows),
each row set to Weight 1, and set the List to Random order -- let's call
this TrialList. Each row would have one nested List for one decade, and
each of these decade Lists would be set to Random order. So again, one
TrialList with 10 rows, and one nested decade List for each row. If I
wanted to run more than 10 trials with TrialList, I would set its Exit
properties as needed in order to run through the 10 rows as many times
as needed.
With this setup, E-Prime would shuffle the 10 rows of TrialList, and
shuffle each nested List. It would then run each of the 10 trials in
the shuffled order, picking one member of each nested decade List as it
runs. At the end of 10 trials it would reshuffle TrialList, and more
trials in the new shuffled order, picking new values wherever it left
off from the previously shuffled nested Lists, and if you set "No Repeat
After" to Yes in TrialList then EP will make sure that it does not
repeat the same row at the start of each reshuffle of TrialList.
For anything more complicated, you might have to generate a randomized
order outside of E-Prime, and then set up all your Lists to run that
order as Sequential. Search this Group for terms such as "random",
"pseudorandom", "pseudo-random", "constrain", or "constraint" to find
many more discussions of this topic.
-- David McFarlane
On 2017-12-17 12:57 PM, Marco A. López wrote:
> Hello everyone,
>
> I have an experimental design in which the participants listen randomly to
> several songs included in a database. For it, I have created a list with
> several nested sublists. The idea would be that each participant listens a
> sequence of songs based on the years in each one was between 15 and 30
> years old. For example, a person with 35 years would listen to songs
> between 1993 and 2008. My problem is to create an inline or instruction, so
> that automatically the program assigns each participant the songs
> corresponding to their age...How could I do that?
>
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__lh3.googleusercontent.com_-2DWw5mRmkdVZE_WjasfsfrZhI_AAAAAAAAAbM_bCXS2CbTntcmk5lkj6hr-2DJ4-2D5ha2TGtuQCLcBGAs_s1600_Music-5FList.png&d=DwIBaQ&c=nE__W8dFE-shTxStwXtp0A&r=rVCe0bDlnqQHHAg2cizkHg&m=56s2gQwV_ijgqQcxiX60daThZB6MlWiS3h91yke7e_U&s=fioLHTGL32gHtkYiQQ-yWVm2cyJQlVfpllsOi91kLMA&e=>
Addressing only your second question regarding a specific case of
"randomization with constraints" ...
For this case, I would make a List or nested List with 10 levels (rows),
each row set to Weight 1, and set the List to Random order -- let's call
this TrialList. Each row would have one nested List for one decade, and
each of these decade Lists would be set to Random order. So again, one
TrialList with 10 rows, and one nested decade List for each row. If I
wanted to run more than 10 trials with TrialList, I would set its Exit
properties as needed in order to run through the 10 rows as many times
as needed.
With this setup, E-Prime would shuffle the 10 rows of TrialList, and
shuffle each nested List. It would then run each of the 10 trials in
the shuffled order, picking one member of each nested decade List as it
runs. At the end of 10 trials it would reshuffle TrialList, and more
trials in the new shuffled order, picking new values wherever it left
off from the previously shuffled nested Lists, and if you set "No Repeat
After" to Yes in TrialList then EP will make sure that it does not
repeat the same row at the start of each reshuffle of TrialList.
For anything more complicated, you might have to generate a randomized
order outside of E-Prime, and then set up all your Lists to run that
order as Sequential. Search this Group for terms such as "random",
"pseudorandom", "pseudo-random", "constrain", or "constraint" to find
many more discussions of this topic.
-- David McFarlane
On 2017-12-17 12:57 PM, Marco A. López wrote:
> Hello everyone,
>
> I have an experimental design in which the participants listen randomly to
> several songs included in a database. For it, I have created a list with
> several nested sublists. The idea would be that each participant listens a
> sequence of songs based on the years in each one was between 15 and 30
> years old. For example, a person with 35 years would listen to songs
> between 1993 and 2008. My problem is to create an inline or instruction, so
> that automatically the program assigns each participant the songs
> corresponding to their age...How could I do that?
>
Marco A. López
Dec 26, 2017, 3:00:59 PM12/26/17
to E-Prime
Hi David, thank you very much for the information and apologize for the delay in the answer, I had not been able to check it until now.Following your advice I have created a general list with several nested lists by decade. In both the selection is in random order. As conditions in the general list I have established to pass all the samples, and in the nested ones “exit list” after each example, it works ...
All best
Marco
Reply all
Reply to author
Forward
0 new messages