What's the best way to match up a person's pre-test and post-test using different surveys?

You can use same unique number list in both contact files for pre test and post test. It will be easy for you to map respondent data.

If your pre test anonymous and you will capture the email id of these respondent in pre test and will target same respondent for post test survey. So while uploading the file you can upload their pre test respondent id as well to map the information.

You can map it through email id as well. But email ID is PII information, so we generally avoid it for using in any purpose except invite or reminder.

Hi Arun,

Thank you for responding to my post query about how to match pre and post-test surveys (What's the best way to match up a person's pre-test and post-test using different surveys?).

We are not collecting any identifiable information (i.e., email, name) for both pre-and post-surveys.

You stated that,
  “You can use same unique number list in both contact files for pre test and post test. It will be easy for you to map respondent data.

   If your pre test anonymous and you will capture the email id of these respondent in pre test and will target same respondent for post test survey. So while uploading the file you can upload their pre test respondent id as well to map the information.”

Also, you mentioned uploading the respondent’s pre-test id. When I exported a "test sample response" for both the pre-and post-test surveys and compared the respondent ids, the ids differed between both surveys.

However, when comparing the contact list for both pre-and post-test surveys, I found the contact ids were the same. However, I am not clear on how to map this data to match the surveys.

Can you direct me to an example on this platform of this process to identify the steps to map the pre- and post-test data points?

For surveys where we are not collecting personal data, yet want to match responses across time, we often add a question in the pre-survey that has the respondent create their own unique ID that no one else would know. 

For example, we may ask them to create an ID that represents the month of their birthday, the first two letters of their mom’s name, and then the last two digits of their zip code.  If someone has an August birthday, a mother named Mary, and lives in 55555 then their unique ID becomes 08MA55. 

You then prompt them with a question in the “post” survey to enter their unique ID, reminding them of the criteria used to create it.

Hope this helps in your scenario!

Hi @ivelisse, before responding to your query. Can you please confirm below few points.

1. do you have predefined list of target respondent or all these respondents are anonymous?
2. How are you planning to distribute your survey if you do not have email address?

I think, then I might be able to give you proper solution on that.

Thank you, Jessica! Your response is very helpful!

Hi @ArunDubey,

We will have a predefined list of target respondents and the person recruiting them will email them a survey link. However, we are not having respondents enter their email as part of the surveys because it is considered PII data. Also, it’s not part of the study design. 

then you can add one unique number column in your first survey contact list for each respondent, care the same column name embedded field in the pre survey, it will be captured by itself. Then add the same number column for post survey use the same number in post survey list.

for example: contact list containing record First name, last name, external data reference, email etc. add one more column resp_number and generate 6 digit unique ID. create same resp_number in your pre and post survey both. once pre survey will be completed. it will be captured from contact list into your survey data. then when you target the same respondent in post survey. same key will be captured in that survey data too. you can use the same key to validate both survey.

I hope it’s clear now.

Hi Arun,

I am still unclear how to set up the process that you suggested. Are there written instructions of this example to reference? Or are you available to connect over Zoom for demo?

@ArunDubey Arun, thank you for assisting and helping me resolve my query.

Hello, I am an instructor wanting to do the same, compare results from pre and post tests. I want to see how the answers changed (increased, decreased, etc) for each question over time.  This will show me how student perceptions changed and if the course content influenced their answer. Is this something I can do within the Qualtrics platform? It sounds like I might need to download data and do it in an excel sheet?  I would also like a link to an example if possible.  Thanks

@ArunDubey Hi: This thread is helping me a lot. I need to do this same thing without using identifiable personal information. Goal: Compare pre- and post-test survey results by student, taken weeks apart,  anonymous.

However, students will access the survey using an anonymous link via an LMS web page.

Can you advise me how to set up this up using @JessicaGregory_CHNw ‘s idea for self-creating their own ID? How would that work across the two surveys? I want to avoid having them remember a randomly generated code and want to make it easy to d/l the reports and see the compared scores by student with anonymity.

Much appreciated! 
Jill

@jmeilander Were you able to figure out how to compare pre and post results within Qualtrics?  We are trying to build out a similar report to analyze pre and post surveys. We have a unique identifier that is entered into the surveys, can Qualtrics pair the unique identifiers and then analyze the responses across the pre and post for comparative analysis? 

Yes, I used  @JessicaGregory_CHNw ‘s workaround of having students create their own personalized ID (“...ask them to create an ID that represents the month of their birthday, the first two letters of their best friend’s first name [changed], and then the last two digits of their zip code. If someone has an August birthday, a friend named Mary, and lives in 55555 then their unique ID becomes 08MA55.”).

This was the easiest way to make it all happen anonymously while using an anonymous link.

Here are some resources I used:

https://community.qualtrics.com/cx-dashboards-8/comparing-two-surveys-in-one-dashboard-16117 

https://www.qualtrics.com/support/frontline-feedback/comparisons-tab/comparisons-collections/ 

https://community.qualtrics.com/employee-360-19/i-need-to-do-a-pre-post-survey-comparison-for-20-leaders-what-is-the-best-way-to-do-this-12563?tid=12563&fid=19 

https://community.qualtrics.com/survey-platform-54/what-s-the-best-way-to-match-up-a-person-s-pre-test-and-post-test-using-different-surveys-23399 

https://www.qualtrics.com/support/survey-platform/common-use-cases-rc/pulling-data-into-a-second-survey-longitudinal-surveys/ 

https://community.qualtrics.com/survey-platform-54/pre-test-and-post-test-analysis-help-24676 

https://www.qualtrics.com/support/survey-platform/survey-module/survey-flow/standard-elements/passing-information-through-query-strings/ 

Good luck!