Besides school type, which had to be constructed from existing variables, most of the variables in this report were taken directly from the data files without manipulation. Some variables, such as locale type, were collapsed into fewer categories for the analysis. The definitions of important and unique variables are explained below.
School type From 1993 to 2003, the school type variable was a derived variable in the data files. In these years it was derived from the following questions:
Does (child) go to a public or a private school?
1 = Public
2 = Private
Is it (his/her) regularly assigned school or a school that you chose?
1 = Assigned
2 = Chosen
3 = Assigned school is school of choice (This response category was coded as a chosen school for this report)
Is the school affiliated with a religion? (1993)
1 = Yes
2 = No
Is the school church-related or not church-related? (1996, 1999, 2003)
1 = Church-related
2 = Not church-related
If the school was classified as public, it was further classified as either assigned or chosen. If the school was classified as private, it was further classified as either affiliated or not affiliated with a religion (1993) or church (1996, 1999, 2003). The response category "assigned school is school of choice" that appeared in all years was coded as a chosen school.
The school type variable in the 2007 data files was constructed differently than in previous years. In 2007, school type was constructed from three variables: SCHOICE, a parent-reported variable which comes from the same question asked in 1993, 1996,1999, and 2003; S07PBPV, a derived variable using data from both the 2004–05 CCD and the 2003–04 PSS; and S07TYPE, also a derived variable using data from the 2004–05 CCD and the 2003–04 PSS. Details about these variables follow:
Parents were asked the name and address of their children's schools and the schools were matched to the CCD and PSS data for some variables. The NHES 2007 datafile contains 7 variables whose data were obtained by matching the student's school to information on the CCD or PSS. These variables are: S07CHART (charter school status); S07NUMST (number of students in the school); S07PBPV (public or private control); S07SAMSX (coeducational status of school); S07TITL1 (Title 1 status of school); S07TYPE (sectarian status of school), and; SCHLGRAD (grades taught at school). Of these variables, two are used to create the school type variable: S07PBPV and S07TYPE. If the student's school was determined to be public because it was found in the CCD public school database (S07PBPV = 1), it was further classified as either assigned or chosen using response to SCHOICE, where the category "assigned school is school of choice" was coded as a chosen school. This report's 2007 estimates exclude cases where data for SCHOICE were missing. If the school was private because it was found in the PSS private school database, as determined by S07TYPE = 1, 2, or 3, then schools were classified as private. The analysis further distinguishes private schools as "Religious" if the S07TYPE indicated the school was "Catholic" or "Other religious," or as "Nonsectarian." Two cases in which data from the PSS were missing are excluded from the 2007 estimates.
In NHES reports based on school choice data from previous years, the categories for school type were "public, assigned"; "public, chosen"; "private, church-related"; and "private, not church-related." The private school categories "church-related" and "not church-related" reflect question wording from the 1996, 1999, and 2003 NHES questionnaires. The 2007 private school categories have been changed to "religious" and "nonsectarian" because they include cases where schools were considered religiously affiliated by the PSS but not necessarily church related. There is no reason to expect this change in wording to create notable differences in estimates between 2007 and previous years.
The poverty measures used in this report were developed by combining information about household composition and household income. Poor students were defined as those with household incomes below 100 percent of the poverty threshold; near-poor students as those with household incomes from 100 to 199 percent of the poverty threshold; and nonpoor students as those with household incomes at or above 200 percent of the poverty threshold. Information on exact incomes was not collected in every administration of the NHES, but categorical household income information was. To keep the measurement of poverty comparable across years, only categorical income information was used in this report. The poverty status measures used in this report were based on poverty thresholds published by the Bureau of the Census for 2006.3 Census poverty thresholds were rounded up or down to the nearest upper bound of an NHES income category. Thus, the poverty measures in this report are approximations of poverty. For example, in 2006, the Census poverty threshold for a four-person family was $20,614; this number was rounded to the nearest upper bound of an NHES income category, which was $20,000 (for the income category $15,001 to $20,000). So a four-person family making less than $20,000 a year would be counted as "poor" in this report. Similar calculations were performed to determine whether households were "near-poor" (with an income at or above the upper bound of the NHES income category closest to 100 percent of the Census poverty threshold, but below the upper bound of the NHES income category closest to 200 percent of the Census poverty threshold) or "nonpoor" (with an income at or above the upper bound of the NHES income category closest to 200 percent of the Census poverty threshold). For example, in 2006, 200 percent of the Census poverty threshold for a four-person family was $41,228; this number was rounded to the nearest upper bound of an NHES income category, which was $40,000 (for the income category $35,001 to $40,000). So a four-person family making $20,000 to $39,999 in 2006 was considered near-poor in this report. A four-person family making $40,000 or more in 2007 was considered nonpoor in this report. The definitions of poor, near-poor, and nonpoor differ across years because Census-defined weighted average poverty thresholds change somewhat from year to year to account for inflation, among other things. Poverty and 200 percent poverty thresholds used in this report are shown in table A-1 on the next page.
In all survey years from 1993 to 2007, region was determined by the Census definition of regions. The following states and the District of Columbia are in each Census region:
Northeast: Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, Vermont
South: Alabama, Arkansas, Delaware, the District of Columbia, Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, North Carolina, Oklahoma, South Carolina, Tennessee, Texas, Virginia, West Virginia
Midwest: Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, Wisconsin
West: Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, Wyoming
From 1993 to 2003, locale was calculated using a different definition than those used in 2007. Therefore, data prior to 2007 are not available in this report. ZIPLOCL is a variable on the file that classifies the residential ZIP Code into a set of 12 community types and was derived using the respondent's Zip Code and Census data. This report recodes ZIPLOCL into 4 categories:
1 = City (large, midsize, or small)
2 = Suburb (large, midsize, or small)
3 = Town (fringe, distant, or remote)
4 = Rural (fringe, distant, or remote)
Data on disability status are based on parent reports. Parents were asked whether or not a health professional ever told the parent that the sampled child had any of several types of disabilities. Each disability was asked in a "yes/no" format. If the parent reported that the child had one or more of the disabilities asked about, the child was coded as having a disability. Disability status was not asked in 1993 and was not asked of grades 6 through 12 in 1996.
The number of parents living in the household determined the family structure for each case. Parents include birth, adoptive, step, or foster parents. If two such parents were in the household, the number of parents living in the household was two. If one such parent was in the household, the number of parents living in the household was one. If there were no such parents in the household, then a student was identified as living with non-parent guardians.
Parents' highest education level
This variable indicates the highest level of education attained for the students' parents or guardians who resided in the household. This measure is used in this report in the same format in which it appears on the data file. The categories are:
1 = Less than a high school diploma or its equivalent (GED)
2 = High school diploma or equivalent (GED)
3 = Some college or vocational/technical education after high school
4 = Bachelor's degree
5 = Graduate or professional school (with or without a degree)
Parents' highest level of education is the highest level of educational attainment between both parents or guardians in the household or the only parent or guardian in the household. This means, for example, that parents whose highest education level was a high school diploma or GED did not attend a community college, vocational or technical school, or college or university beyond high school completion.
Parent satisfaction with various aspects of the school
The NHES measures parent satisfaction by asking parents how satisfied they are with aspects of the students' school: the school; the teachers; the academic standards; and order and discipline. In 2007, a fifth aspect was added, which was satisfaction with staff interactions with parents. Parents were asked to rate their satisfaction as very satisfied, somewhat satisfied, somewhat dissatisfied or very dissatisfied.
Parent involvement at the school
Parent involvement was measured by a series of questions about parents' attendance at different parent/school activities since the beginning of the school year: A general meeting, such as an open house, a back-to-school night, or a Parent Teacher Association (PTA) meeting; a regularly scheduled parent-teacher conference; a school event, such as a play, dance, sports event, or science fair; and volunteering at the school or serving on a committee. The analysis considers each activity individually and measures attendance as a binary variable, "yes, attended" or "no, did not attend." In 2007, the questions "Have you attended a general school meeting, for example, an open house or back to school night?" and "Have you attended a meeting of the parent-teacher organization or association?" were combined to ascertain the number of parents attending a general school meeting or PTA meeting. Similarly the questions "Have you served as a volunteer in (CHILD's) classroom or elsewhere in the school?" and "Have you served on a school committee?" were combined to ascertain whether or not a parent volunteered in the child's school or served on a school committee.
Part-time homeschooled students
Data for part-time homeschoolers has been collected differently across some years of the NHES. Therefore, it is not possible to consistently classify part-time homeschoolers as enrolled or not for all survey years. Because part-time homeschoolers make-up less than one half of a percent of all students, the effect of this inconsistency on estimates of enrolled students is negligible. Note that the inconsistency only applies to estimates of enrolled students and does not affect the estimates of homeschooled students presented in the report.
3 For exact details on the poverty thresholds for the 1993, 1996, 1999, 2003, and 2007 NHES, please see the Census Bureau, Current Population Survey (http://www.census.gov/hhes/www/poverty/threshld.html).