hispanic |
Hispanic |
SIPP Synthetic Beta v6 |
|
numeric |
hispanic |
In the 1984 & 1990-1993 SIPP panels, a value for ethnicity is included on each wave
file. Thus, there are actually as many ethnicity variables as there are waves of
the survey and some changes occur across waves as a result of data collection error.
Ethnicity is chosen by creating an array of variables ethncty1-ethncty{max number
of waves} and choosing the first non-missing value. Thus, ethnicity comes from the
first wave in which the individual was interviewed instead of from a fixed point in
the survey. Respondents are coded as Hispanic if they have an ethnicity code between
14 and 20. In the 1996-2008 panels, the longitudinally-edited version contains only
one value for ethnicity across all waves (eorigin) and this value is used. Respondents
are coded as Hispanic if they have an ethnicity code between 20 and 28 in 1996 and
2001, or if they have an ethnicity code of 1 in 2004 & 2008.