Why does exporting 5-point Likert scale responses result in values outside the scale? | XM Community
Skip to main content
Hi,



I have a 5-point Likert scale question in my survey. When I export the data and analyze the outputs of this question I am seeing values of 6. Does anyone understand why Qualtrics is outputting a higher value than the scale?



Thanks
Hi,



In my survey I have a 5-point Likert scale question. Somehow this scale outputs a value of 6 sometimes when outputting the data file. This seems to happen when selecting the highest option on the scale. Any ideas why this might be happening?



Thanks
Check the survey question and how it is coded. In the survey view, you can click on the grey gear or cog to the left of the question and select "recode values." This will show how the scale is currently coded and you can make adjustments. Occasionally the numbers of my scaled questions get funky like starting with 11 and go to 15 for a 5 point scale. Any edits you make on the survey will be updated in the data & analysis tab and when you export your data.

I realize this question is 4 years old but I just had the same thing happen to a 5 point likert scale. Qualtrics coded my 1-5 as 1,3,5,6,7. I tried recoding as suggested but it didn’t hold and still exported the weird coding. Had to recode in SPSS. Anyway, my question is why does Qualtrics do this? By default it should code a 5 point likert scale as 1-5. It’s really annoying.

The skip in id numbers is caused by editing (i.e., adding/deleting scale points). Qualtrics does this to prevent you from erroneous data (i.e., different answers with the same id).  Recoding should have worked as long as you published after you recoded.

I had the same problem. The output included a number not on the scale.  The scale was  2,3,4,5,6,10 (not sure why), but there are many 8s in the data.

When I tried to recode in the survey the output remained.

 

I had the same problem. The output included a number not on the scale.  The scale was  2,3,4,5,6,10 (not sure why), but there are many 8s in the data.

When I tried to recode in the survey the output remained.

 

You need to republish after you recode.

thanks!

I actually corrected it in spss. But also re-published and it seems to have worked.

 

Leave a Reply


Terms & Conditions