1. Age of the reference person: The respondent’s
reconciled age was employed.
2. Household income: In order to reduce
heteroskedasticity (unequal variance of the
disturbances), the natural logarithm of the reconciled
annual total household income before taxes was used
(Montalto & Sung, 1996). Although the original data
set included negative income (e.g. business losses in
self-owned business), negative income is closer to a
measure of assets (debts) rather than income.
Therefore, for households whose income was
negative, the natural logarithm of income was
recoded as zero.a
3. Female headed household
4. Race-ethnicity: The respondent’s race was
categorized into Hispanics, African-American, other
nonwhites, and non-Hispanic whites (base).
5. Education: a set of dummies were included with high
school graduates or equivalent as the base: other
categories were less than high school, some college,
bachelor’s degree, and graduate degree.
6. Marital status: Three dummy variables, divorced or
separated, widowed, and never been married, were
included (married or living with a partner as base.)
7. Size of household: The number of people in
household was employed.
8. Region: 9-level Census Division code was employed
for region (Table 1).
1. Age of the reference person: The respondent’s
reconciled age was employed.
2. Household income: In order to reduce
heteroskedasticity (unequal variance of the
disturbances), the natural logarithm of the reconciled
annual total household income before taxes was used
(Montalto & Sung, 1996). Although the original data
set included negative income (e.g. business losses in
self-owned business), negative income is closer to a
measure of assets (debts) rather than income.
Therefore, for households whose income was
negative, the natural logarithm of income was
recoded as zero.a
3. Female headed household
4. Race-ethnicity: The respondent’s race was
categorized into Hispanics, African-American, other
nonwhites, and non-Hispanic whites (base).
5. Education: a set of dummies were included with high
school graduates or equivalent as the base: other
categories were less than high school, some college,
bachelor’s degree, and graduate degree.
6. Marital status: Three dummy variables, divorced or
separated, widowed, and never been married, were
included (married or living with a partner as base.)
7. Size of household: The number of people in
household was employed.
8. Region: 9-level Census Division code was employed
for region (Table 1).
การแปล กรุณารอสักครู่..
