How do you randomize the order of entire blocks? | XM Community
Solved

How do you randomize the order of entire blocks?

  • 25 May 2018
  • 5 replies
  • 3297 views

Hi!
I have been asked to randomize the order in which 3 different blocks appear to participants.
Does anyone know how to do this?

Thank you
Jade 🙂
icon

Best answer by TomG 25 May 2018, 14:02

View original

5 replies

Userlevel 7
Badge +27
Put the blocks under a Randomizer in the survey flow and present all three.
Userlevel 7
Badge +38
@TomG is right. If you need more help check out the resources Qualtrics has on this: https://www.qualtrics.com/support/survey-platform/survey-module/survey-flow/standard-elements/randomizer/
That page says: " You have Evenly Present Elements selected. You have been randomly assigning respondents 1 of 3 blocks for about a week. 15 people have responded, meaning each block was presented 5 times. You decide to add a new, fourth block. The fourth block may appear 5 times in a row to equalize it with the other blocks. After this, the assignment will be random again."

If I have 20 blocks and select choose to randomly present "20" of the following elements with Evenly Present Elements checked, then is it possible that a survey respondent will see the same block twice (for example, seeing block #20 twice and never seeing block #1) because two people have started the survey at the same time and block #1 was already presented to them (so the count for block #1 is higher than block #20)?
Userlevel 7
Badge +27
> @jamyli said:
> That page says: " You have Evenly Present Elements selected. You have been randomly assigning respondents 1 of 3 blocks for about a week. 15 people have responded, meaning each block was presented 5 times. You decide to add a new, fourth block. The fourth block may appear 5 times in a row to equalize it with the other blocks. After this, the assignment will be random again."
>
> If I have 20 blocks and select choose to randomly present "20" of the following elements with Evenly Present Elements checked, then is it possible that a survey respondent will see the same block twice (for example, seeing block #20 twice and never seeing block #1) because two people have started the survey at the same time and block #1 was already presented to them (so the count for block #1 is higher than block #20)?

No. A respondent only sees any block once. Evenly Present is only applicable if respondents aren’t seeing all the blocks.
> @TomG said:
> > @jamyli said:
> > That page says: " You have Evenly Present Elements selected. You have been randomly assigning respondents 1 of 3 blocks for about a week. 15 people have responded, meaning each block was presented 5 times. You decide to add a new, fourth block. The fourth block may appear 5 times in a row to equalize it with the other blocks. After this, the assignment will be random again."
> >
> > If I have 20 blocks and select choose to randomly present "20" of the following elements with Evenly Present Elements checked, then is it possible that a survey respondent will see the same block twice (for example, seeing block #20 twice and never seeing block #1) because two people have started the survey at the same time and block #1 was already presented to them (so the count for block #1 is higher than block #20)?
>
> No. A respondent only sees any block once. Evenly Present is only applicable if respondents aren’t seeing all the blocks.

Is the following correct?

Randomiser with 20 items (blocks/groups) underneath it

1. randomly present *19* of the following. evenly present is *unchecked*
> for 1:19, randomly select block (1-20) to display

2. randomly present *19* of the following. evenly present is *checked*
> for 1:19, randomly select block from the blocks with the lowest presentations across all respondents with no repeats

3. randomly present *20* of the following. evenly present is *unchecked*
> for 1:20, randomly select block (1-20)

4. randomly present *20* of the following. evenly present is *checked*
> for 1:20, randomly select block from the blocks with the lowest presentations across all respondents with no repeats

Okay. I was thinking that how 2 and 4 are implemented is functionally difference, but it's not, since the 'with no repeats' means that the same algorithm can work with both cases. Even if the count is 20 for block #1 and 0 for all other blocks, participants will still see block #1 with evenly present elements checked, since there is the condition of 'no repeats'.

To me it wasn't clear that 'evenly present elements' meant 'evenly present elements without any repetition'. If participants aren't seeing all the blocks, they will still see only 1 of each block if 'evenly present elements' is checked I guess.

Leave a Reply