📚️ Straightlining and Speeding | Basecamp Wednesdays | June 21st 2023 | XM Community
Skip to main content

What is Basecamp Wednesdays?

 

Every Wednesday we will look at different topics discussed on XM Basecamp and provide a brief description of the Basecamp resource.

Whether you’re just getting started on your XM journey or ready to take your program to the next level, this weekly series is meant to help all users find Basecamp resources. 

 

Basecamp Wednesdays | June 21st 2023
 

Topic: Straight Lining and Speeding

 

Issues such as straightlining and speeding can heavily impact the validity of collected data by adding bias or pulling averages in a certain direction. While there is no failsafe, there are steps survey writers can take to mitigate the effect these types of responses have on the data.

 

Straightlining is a problem typically scene with matrix table questions, where a respondent clicks the same answer for each row in the question. Speeding is when respondents attempt to complete the survey as fast as possible by randomly selecting answers. Respondents tend to do this when they’ve lost motivation to finish the survey or simply want to rush their way to the end. 

 

To reduce straightlining, it’s best to break up matrix tables into separate questions and limit the number of questions per page to one or two. An easy way to identify speeders is by comparing their survey duration to the average duration of your other responses. 

 

Increase your data’s validity by checking for these trends in your responses today!

 

Where to find this course:

  1. Go to XM Basecamp.
  2. Search for Minimizing Survey Fatigue and Bias.
  3. Click Register.
  4. Find the Understanding the Threat of Satisficing section.
  5. Select the Mitigating Straight Lining and Speeding video.

 

Basecamp Link: Mitigating Straight Lining and Speeding

Qualtrics Resources: 

Questions of the Week: How do you decide whether to include or exclude these types of respondents? Have these types of responses heavily impacted your data before?

I love this topic.  I would love to get others’ feedback on what they do on the following:

  • Is it a good idea to do this assessment for all surveys?  I typically only do it with panel data, where people might be getting rewarded for taking the survey and therefore not motivated to answer thoughtfully.  I typically don’t do it with our patient or employee surveys because if those folks are responding, it is typically because they have valid feedback to give.  
  • For straightlining, in practice, do you usually look at “did they give the same answer to every survey question” or just look at the responses in the matrices?  
  • Speeding - I typically look at what is 2-3 standard deviations below “average take time” and drop those out.  What do others do?

In practice, I have never seen a huge change from dropping the speeders/straightliners, but I definitely feel better about the cleanliness of the data!

I love this topic.  I would love to get others’ feedback on what they do on the following:

  • Is it a good idea to do this assessment for all surveys?  I typically only do it with panel data, where people might be getting rewarded for taking the survey and therefore not motivated to answer thoughtfully.  I typically don’t do it with our patient or employee surveys because if those folks are responding, it is typically because they have valid feedback to give.  
  • For straightlining, in practice, do you usually look at “did they give the same answer to every survey question” or just look at the responses in the matrices?  
  • Speeding - I typically look at what is 2-3 standard deviations below “average take time” and drop those out.  What do others do?

In practice, I have never seen a huge change from dropping the speeders/straightliners, but I definitely feel better about the cleanliness of the data!

@JessicaGregory_CHNw  I’m glad you enjoyed this week’s basecamp topic. 😄

We typically recommend to look at your data both with and without the speeders/straightliners, and see what, if any, impact they have on your data. Most of the time these responses simply add noise, so deleting them isn’t always necessary, however, I completely understand wanting to have your data as clean as possible! 

Leave a Reply


Terms & Conditions