Support #1978

Potential issue with coding of wkaut1 wkaut2 wkaut3 wkaut4 and wkaut5 variables

Added by Tim Ballard 5 months ago. Updated 3 months ago.

Data inconsistency
Start date:
% Done:



Hi understanding society team,

I'm writing to query the coding of the items wkaut1, wkaut2, wkaut3, wkaut4, and wkaut5 in the understanding society survey. These items assess job autonomy, and the documentation suggests that, for they are coded such that 4 = none, 3 = a little, 2 = some, and 1 = a lot.

I found this information through the link below...

I'm wondering if there might actually be an error here. Theoretically autonomy should be positively correlated with mental health (which means that these items should be negatively correlated with mental health). However, I'm finding quite robust positive correlations between each of these items and the scghq1_dv variable, which captures mental health (around 0.1). This finding is a striking contrast to the job autonomy--mental health relationship that's been established in the literature (and that's demonstrated in other large scale panel surveys). So I'm wondering if there's an error in the documentation it's actually the case that 4 = a lot and 1 = none.

You're help is much appreciated!


reg scghq1_dv wkaut1.jpg (37.4 KB) reg scghq1_dv wkaut1.jpg Understanding Society User Support Team, 09/27/2023 10:29 AM

Updated by Understanding Society User Support Team 5 months ago

Hello Tim

The variable "w_scghq1_dv" consolidates valid responses to 12 questions from the General Health Questionnaire (GHQ) into a single scale. The resulting scale runs from 0 (indicating the least distress) to 36 (indicating the most distress). Additionally, variables measuring job autonomy “w_wkaut*” (*=1 to 5) run from 1 (a lot of autonomy) to 4 (no autonomy).

As you can see both variables are on a negative scale, meaning that higher values correspond to more levels of distress and less autonomy in the job. This likely explains the positive correlation you observed.

For illustration, consider the example of reversing the scale of "w_scghq1_dv" in Wave 2, where the correlation turns negative.
gen scghq1_dv = 36 - b_scghq1_dv if b_scghq1_dv >=0
reg scghq1_dv wkaut1 if scghq1_dv>=0 & wkaut1>0, r

I hope this information is helpful.

Best wishes,
Roberto Cavazos
Understanding Society User Support Team


Updated by Understanding Society User Support Team 3 months ago

  • Status changed from Feedback to Resolved
  • % Done changed from 50 to 100

Also available in: Atom PDF