Individual-Level Concepts & Issues...
"Code 2 Individuals" in 1994 and 2000
"Code 2 Individuals" are individuals in the 1994 and 2000 household
surveys who
may have been listed on two different household
rosters. These indivdiuals are coded '2' on Question 1.1 in
the household where they resided in the previous round of data collection, a.k.a. their
origin household.
The '2' code means that since the
last round of data collection this individual had moved from their
origin household, for 2 months or more, to either another household in the village or to a
temple in the
village. If the
individual moved to another household in the village, then the other
household information was noted in the questionnaire and later
verified by the field supervisors. If the information was
verified and the
individual indeed moved to the other household, then this other
household was considered the code 2 individual's
destination household.
So, code 2 individuals with a verified destination household are
listed on two household rosters - their origin household and their
destination household; while code 2
individuals who do not
have a verified destination household or who moved
to a temple are listed only on their origin
household's roster. In the 1994 survey, there
are 2,516 code 2 individuals who appear on two rosters, and 124 who appear on one roster. In the 2000 survey,
there are 2,511 code 2 individuals on two rosters, and 83 who appear on roster. See the
Code 2 Individuals document for various examples from the 1994 and 2000 surveys.
"Code 5 Individuals" in 1994 only
"Code 5 Individuals" are individuals in the 1994 household
survey who were coded as migrants (Q1=3) during the
fieldwork. However, it was determined years later from
preparation work for the 2000 data collection that the same individuals
were listed on a second household roster in 1994. This
created a situation very similar to the tracked code 2 individuals in
that the one individual is listed on two household rosters - an origin
household and a destination household. The only difference is that the household reported that the individual was
a migrant and not a code 2 person.
So, rather than code these individuals with a '3' or a '2' on Question
1.1, a new code of '5' (or '3' plus '2') was created. There are
only 36 individuals that fit this scenario. Please see the
Code 5 Individuals document for a couple of examples.
"Code 2 Migrants" in 1994 and 2000
"Code 2 Migrants" are individuals in the 1994 and 2000 household
surveys who were initially "code 2 individuals" during the
fieldwork. However, it was determined from the destination
household
verification process (interviewer notes
on the Code 2 Tracking Form in 1994 and supervisor verification in
2000) that the individual had migrated from
their destination household. Once the individual was
determined to be a code 2 migrant, their status code for Question 1.1
was changed from a '2' to a '3,' and they were assigned a code of '1' on the flag variable,
CODE2MIG. Below are some examples of Code 2 Migrants from the 1994 and 2000 Surveys:
1994 Survey
|
2000 Survey
|
|
CODE2-
MIG
|
HHID94 | CEP94 | Q1 | Q12 | MIPRO | MIDIS |
1
|
20010201 | 105 | 3 | 20422001 | 1 | 9999 |
1
|
20130351 | 104 | 3 | 21502013 | 9 | 910 |
1
|
20160051
|
202
|
3
|
20912016
|
9999
|
9999
|
1
|
20290311 | 103 | 3 | 21212029 | 28 | 9999 |
1
|
20300171 | 103 | 3 | 21572030 | 1 | 9999 |
1
|
20340211 | 103 | 3 | 21002034 | 27 | 2712 |
1
|
20340671 | 104 | 3 | 20772034 | 50 | 9999 |
1
|
20470471 | 105 | 3 | 21452047 | 59 | 9999 |
1
|
20490831 | 103 | 3 | 21042049 | 52 | 9999 |
1
|
20500181 | 105 | 3 | 20442050 | 1 | 9999 |
|
CODE2-
MIG
|
HHID00 | CEP00 | X1 | DHHID00 | DCEP00 |
1
|
020192029 | 06 | 3 | 020192016 | 96 |
1
|
020312057 | 04 | 3 | 020312043 | 96 |
1
|
020510096 | 06 | 3 | 020506036 | 96 |
1
|
020614006 | 07 | 3 | 020614005 | 96 |
1
|
020705067 | 06 | 3 | 020705015 | 96 |
1
|
021310039 | 09 | 3 | 021307079 | 96 |
1
|
021407018 | 05 | 3 | 021407002 | 96 |
1
|
170204003 | 04 | 3 | 170206003 | 96 |
1
|
180101024 | 09 | 3 | 180101050 | 96 |
1
|
180103036 | 04 | 3 | 180103117 | 96 |
|
- The Q12 variable was re-coded to point to the individual's initial destination household (2
+ ban lek ti + village number) before s/he migrated.
- The migrant destination
information, MIPRO and MIDIS, came from the interviewer notes on the Code 2 Tracking Form.
|
- The DHHID00 variable shows the code 2 migrant's initial destination household before s/he migrated.
- The DCEP00 variable is automatically coded as a '96' for all code 2 migrants.
|
Old Individuals treated as New Individuals in 1994 and 2000
There is a small group of individuals in the 1994 survey
(n=58) and in the 2000 survey (n=15) who were initially treated as new
individuals in the village. However, it was determined
long after the fieldwork, that the
individuals were actually "old individuals" or individuals who were in
the previous round(s) of data collection. Upon further
investigation, two whole 1984 households are a subset of these
indivduals in the 1994 survey, while one whole 1994
household is a subset of those indivduals in the 2000 survey. All
of these old indivduals have been flagged using the
FORM94 and
FORM00
variables. If the individual is listed on Form 2, then s/he is
coded as "F2-OLD". If the individual is listed on Form 3, then
s/he is coded as "F3-OLD". Below are the whole household examples in the
1994 and 2000 Surveys:
Two Old 1984 Households treated as New in the 1994 Survey
The first old 1984 household (VILL84='07' and HOUSE84='019') is a
one-person household, who was the last person listed on the roster
(CEP94='505') of a New
Household (HHID94=20070722) in the 1994 survey. The new household is below:
CEP94
|
FORM
|
Q1
|
Q2
|
Q3
|
FORM94
|
VILL84
|
HOUSE84
|
CEP84
|
HHTYPE94
|
| 501 | 3 | 1 | 37 | 1 | F3 |
|
|
| 2 |
| 502 | 3 | 1 | 28 | 2 | F1&F3&CODE2 | 07 | 038 | 104 | 2 |
| 503 | 3 | 1 | 10 | 2 | F3 |
|
|
| 2 |
| 504 | 3 | 1 | 1 | 1 | F3 |
|
|
| 2 |
| 505 | 3 | 1 | 68 | 2 | F3-OLD | 07 | 019 | 101 | 2 |
The second old 1984 household (VILL84='48' and HOUSE84='095') was
treated as a New Household (HHID94=20481102) in the 1994 survey.
The original 1984
household added two members (CEP94='504' & '505') since the 1984
survey:
CEP94
|
FORM
|
Q1
|
Q2
|
Q3
|
FORM94
|
VILL84
|
HOUSE84
|
CEP84
|
HHTYPE94
|
| 501 | 3 | 1 | 42 | 1 | F3-OLD | 48 | 095 | 101 | 2 |
| 502 | 3 | 1 | 40 | 2 | F3-OLD | 48 | 095 | 102 | 2 |
| 503 | 3 | 1 | 20 | 2 | F3-OLD | 48 | 095 | 103 | 2 |
| 504 | 3 | 1 | 26 | 1 | F3
|
|
|
| 2 |
| 505 | 3 | 1 | 2 | 2 | F3 |
|
|
| 2 |
| 506 | 3 | 1 | 17 | 1 | F3-OLD | 48 | 095 | 104 | 2 |
| 507 | 3 | 1 | 14 | 1 | F3-OLD | 48 | 095 | 105 | 2 |
| 508 | 3 | 1 | 74 | 2 | F3-OLD | 48 | 095 | 106 | 2 |
One Old 1994 Household treated as New in the 2000 Survey, BUT...
This old 1994 household (HHID94=20330892) was handled as a New
Household (HHID00='020908119') in the 2000 survey. But, it was
treated that way for a very good reason, as the household actually
moved from one study village to another study village between the 1994
and 2000 data collections. The four-person houshold is below:
CEP00
|
FORM
|
X1
|
X3
|
X4
|
FORM00
|
HHID94
|
CEP94
|
HHTYPE00
|
| 01 | 3 | 1 | 34 | 1 | F3-OLD | 20330892 | 501 | 2 |
| 02 | 3 | 1 | 35 | 2 | F3-OLD | 20330892 | 502 | 2 |
| 03 | 3 | 1 | 13 | 2 | F3-OLD | 20330892 | 503 | 2 |
| 04 | 3 | 1 | 10 | 1 | F3-OLD | 20330892 | 504 | 2 |
"Unaccounted For" Individuals
As one might expect with longitudinal data, there are several
individuals who are a part of the data in one round of data collection,
but then are
not a part of the data in the following round(s) of data
collection. For some unknown reason, the individuals have
disappeared from the data archive. Here are a few
possible scenarios as to why this situation occurred:
- During the household interview in 1994 and 2000, the interviewers
copied the names of individuals from a specific source to the
questionnaire. In 1994, the source was a copy of the 1984
household questionnaire, while in 2000 the source was a pre-printed
household roster. It is possible that the name of the
"unaccounted
for" person might have been skipped by mistake leaving the person off
the current household roster. In fact, this may be especially
true in the 1994 interview given that 33% of the unaccounted for individuals
were temporarily absent members on the 1984 questionnaire.
The temporarily absent members were listed separately from other household members and were sometimes listed on an additional
questionnaire page.
- The "unaccounted for" person could have been completely skipped
in the data entry phase.
- During the 1994 fieldwork only, one or more members of the old household roster were treated as new individuals (see the "Old
Individuals treated as New Individuals in 1994 and 2000"
above). In this case, if the "unaccounted for" person was
not part of the particular household being interviewed, then the
interviewers would have had no past record to go by to even know to add this person to the roster.
No matter the reason for the situation, these types of individuals
have been given the label of "unaccounted for" and have been placed in
two special data sets called
unacc94 and
unacc00, for the 1994 and 2000 survey data, respectively. In the 2000 survey data, a good portion of the "unaccounted for"
individuals are actually part of entire households that are unaccounted
for. The tables below show the numbers for both 1994 and
2000:
| 2000 Survey
|
1994 Survey |
| 1984 "Unaccounted For" |
| Individuals |
Whole Households |
| 279 |
39 |
| 1994 "Unaccounted For" |
| Individuals |
Whole Households |
| 252 |
43 |
|
| 1984 "Unaccounted For" |
| Individuals |
Whole Households |
| 171 |
0 |
|
