Project

General

Profile

Support #1012

Inconsistency on Wave 2 related to Raceldv compared to other waves

Added by Felippe Oliveira over 5 years ago. Updated over 5 years ago.

Status:
Resolved
Priority:
Normal
Assignee:
Felippe Oliveira
Category:
Data inconsistency
Start date:
08/02/2018
% Done:

100%

Description

Hello folks,

First of all, congratulation for the fantastic database and space for discussion.

I have been working on an SEM using Ethnicity, Socioeconomic Status, and Health.

The earlier observed variable (Race) seems strange compared to other waves. When I restrict my analysis to employees above 60 years old ( b_dvage < 60 ) receiving a salary(b_payg), the racial distribution(b_racel) of white respondents drops significantly when compared to waves 1 and 3 to 5.

If you want to mimic, I further restrict it to non-proxy interviewees and thoroughly answered questionnaires.

ANOTHER QUESTION
I checked across all waves a model that worked flawlessly from wave 1 to 5. Was there a change in methodology for waves 6 and 7?

Once I finish my work I am happy to share it to you, if it interests.

Best wishes,

Felippe

Also available in: Atom PDF