[Part 5: How Many Children Have Special Needs?]
[Part 7: Parents' Reasons for Participating in Even Start]
Chapter 3
The law requires Even Start projects to recruit and serve families "most in need" of Even Start services; the law indicates that families "most in need" are those with a low level of income, a low level of adult literacy or English language proficiency, and "other need-related indicators." However, the law does not provide specific criteria of "family need," thus creating a major challenge in answering the question: "Do projects serve families most in need?" The type and extent of family needs vary in different communities. To answer this question, we would need to compare the characteristics of Even Start families with all families living in the project's service area—an approach not possible within the scope of this evaluation. 39 Thus, for the 1994-95 evaluation report we developed a data-based working definition of "the neediest Even Start families." The working definition was based on (1) types of economic, educational, and social disadvantages relevant to the Even Start program and (2) distributions of Even Start participants on these characteristics. The working definition was further refined in 1995-96 and 1996-97.40
The following seven features of families were used as indicators of family need level:41
We assigned each family a need index value based on the number of the seven characteristics on which the family matched our working definition of "needy." The need index that resulted represents a combination of multiple disadvantages including extreme poverty; limited educational experiences; limited English proficiency; problems associated with single-parent family status; the difficulty of raising multiple children given limited income and earning capacity; and having a child(ren) with disabilities. Families' need index scores could range from zero to seven, where zero indicated a low level of need relative to other Even Start families, and seven indicated the presence of all seven disadvantages.46
For the approximately 31,500 families who participated in 1996-97 (and had data for the needy-family analyses), the average need index was 3.2.47 Thus, on average, Even Start families had about three of the seven disadvantages listed above. Forty-five percent of families had four or more disadvantages, 18 percent had five or more, and about 3 percent had none of the seven disadvantages.48 In the remainder of this report, families with four or more need indicators (or disadvantages) are referred to as the "very needy families."
For selected analyses, we grouped projects by the percent of very needy families according to our working definition. This approach does not mean that some projects fail to serve families who qualify as "most in need" in their communities. Our working definition indicates a family's level of need (or extent of disadvantage) in relation to all other families participating in Even Start. As discussed earlier in this chapter, Even Start families as a whole are disadvantaged in many areas of functioning relative to the general population. In applying the working definition of "very needy families" and focusing our analysis on families who are needier than others, we must keep in mind that on the whole, Even Start projects are recruiting and serving needy families.
The average need index varied little across regions. However, as shown in Exhibit 3.19, families living in urban areas reported slightly higher levels of need than families living in rural areas, except in the South where the need levels were consistent across urban and rural communities.
Exhibit 3.19: Average Need Index, by Region and Type of Community: 1996-97 Participants
|
Region |
|||||
|
Total |
Northeast |
South |
Midwest |
West |
|
|
Rural |
3.1 (16,647) |
3.0 (3,418) |
3.3 (7,344) |
2.9 (2,808) |
3.2 (3,077) |
|
Mixed |
3.2 (6,406) |
3.4 (770) |
3.2 (3,598) |
2.9 (1,096) |
3.1 (942) |
|
Urban |
3.4 (14,230) |
3.4 (2,551) |
3.3 (5,260) |
3.3 (3,218) |
3.6 (3,201) |
|
Total |
3.2 (37,283) |
3.2 (6,739) |
3.3 (16,202) |
3.1 (7,122) |
3.4 (7,220) |
|
Note: The numbers of 1996-97 families included in this analysis are indicated in parentheses. Exhibit reads: In 1996-97, Even Start families in the rural areas of the Northeast had an average need index of 3.0 on a scale of zero to seven. |
The average need levels were greater among some racial/ethnic groups. African American, Asian, and Hispanic families experienced an average of three to four needs; American Indian and Caucasian families experienced, on average, slightly fewer than three needs (Exhibit 3.20).
Exhibit 3.20: Average Need Index, by Parent Race/Ethnicity and Age: 1996-97 Participants
|
Parent Age |
|||||
|
Total |
Less than 20 Years |
20-29 Years |
30-39 Years |
40 Years or Older |
|
|
Hispanic |
3.5 (13,643) |
3.4 (852) |
3.4 (6,241) |
3.6 (5,184) |
3.8 (1,366) |
|
Asian |
3.4 (1,283) |
3.4 (27) |
3.3 (342) |
3.2 (618) |
4.2 (296) |
|
African American |
3.4 (7,449) |
3.1 (1,735) |
3.5 (3,728) |
3.5 (1,482) |
3.1 (504) |
|
American Indian |
2.9 (1,217) |
2.8 (203) |
2.8 (590) |
3.2 (315) |
3.2 (109) |
|
Caucasian |
2.8 (11,400) |
2.5 (1,920) |
2.9 (6,057) |
2.9 (2,735) |
2.8 (688) |
|
Total |
3.2 (34,992) |
2.9 (4,737) |
3.2 (16,958) |
3.4 (10,334) |
3.5 (2,963) |
|
Note: The numbers of 1996-97 participating parents included in this analysis are indicated in parentheses. Exhibit reads: On average, families with Hispanic teen parents had 3.4 of the seven need indicators in 1996-97. |
Due to the small range of the need index (zero to seven), most average need scores differed only by decimal points. However, some differences (e.g., between 4.2 and 2.5) would translate into a typical family in one group experiencing two additional types of economic, educational, or social disadvantage compared to a typical family in the other group.
On average, families with parents ages 40 or older had somewhat greater needs (3.5) than families headed by teen parents (2.9). Asian parents age 40 or older (less than 1 percent of all parents) produced the highest need index. Compared to the Even Start average, this group was far more likely to have low education and limited English proficiency and somewhat more likely to be below the poverty level, receiving government assistance, and have four or more children.
Virtually all Even Start families have economic and educational needs qualifying them for receiving Even Start services. Further focusing our attention on families with the greatest needs, we calculated the percentage of very needy families for each project. The average percentage of very needy families across all projects in 1996-97 was 45 percent. However, the prevalence of very needy families varied widely across projects. Ten percent of projects had 19 percent or fewer families with four or more needs; in another 10 percent of projects, at least 69 percent of their caseload had four or more needs.49
In Chapters 5 and 6, we examine (1) how projects respond to high percentages of very needy families in their service design and delivery and (2) the extent to which the very needy families are able to participate in Even Start services despite disadvantages that may hinder their participation.
Footnotes:
39 Economic characteristics of Even Start families as a whole or by state could be compared to similar national or state statistics. For example, 11 percent of all households in the United States had incomes below the federal poverty level (compared to 90 percent of Even Start households); nationally, 18 percent of adults over age 25 had not graduated from high school (compared to 86 percent of all Even Start parents) (the population data are from the 1996 Current Population Survey, U.S. Bureau of the Census). The Even Start population clearly represents an extremely disadvantaged segment of the population economically and educationally. However, this type of comparison does not answer the question of whether Even Start projects serve families most in need of their services in their communities.
40 While the basic approach behind creating the composite need index has remained consistent since the 1994-95 analyses, the yearly refinements of this measure precludes comparing the extent of need among Even Start families over time. Yearly trends for each of the need indicator measures (e.g., income, educational levels) have been discussed in earlier sections of this chapter.
41 Most of the data used to derive the need index represent family characteristics that can change over time such as family income, family structure, and parents' English proficiency. These data were collected for this evaluation only at the time of families' initial enrollment in the program. Approximately 40 percent of the 1996-97 participants had enrolled in prior years, and some of their need-related information may have changed by 1996-97. This was a potential problem in assessing the level of family need for continuing families in this evaluation. Neither increases nor decreases in the level of family need after enrollment were captured by the ESIS. At the same time, the changes would have affected the need index analyses only if they raised the family above the need threshold (e.g., an increase in income above the federal poverty level for the family, a change from a single-parent to a two-parent family, or significant improvements in parents' English proficiency).
42 Because the ESIS asked for family incomes, it is possible that in large households (e.g., extended families), there may have been incomes besides the Even Start family's reported income.
43 Reported receipt of public assistance was used as an indicator of poverty in addition to low family income. To the extent that some eligible families do not receive welfare due to pride or lack of information, this index may underestimate the level of economic need for some families relative to those who chose to receive public assistance.
44 Membership in a single-parent family is used as an index of family need separate from family income or the number of people supported by the family income. This index is intended to represent the difficulties single parents face in parenting and managing all family responsibilities without help from a partner.
45 Earlier in this chapter, we stated that 11 percent of all children participating in 1996-97 were reported to have special needs. The 14 percent reported here refers to families in which at least one child participating in Even Start has special needs. The percentage difference indicates that, in families with multiple participating children, only one child may have special needs.
46. Correlations among the seven need indicators are presented in Appendix B, Exhibit B.4. The correlations were generally low, the highest being .42 between "single parent" and "receiving welfare." No correlation was high enough to suggest that two or more variables represent essentially the same family characteristic.
The need index was derived as a family characteristic. Some of the seven need indicators were based on parent- and child-level data. In families with multiple adults and/or children participating in Even Start, the families were classified as having the specific need if at least one parent or child reported the need as defined above.
For parent's limited education and English proficiency, we marked a family as having these needs if at least one participating parent met the criteria set for the working definition of need. Data on education and English proficiency were collected only for participating parents. Thus, we could not determine whether some families marked as having these needs may include other adults with higher levels of education and English abilities who could reduce the extent of limitations experienced by the families.
If a family had data for four or more, but not all, of the seven indicators, we computed a prorated need index for the family based on available data. Families with missing data for four or more need indicators were excluded from analyses involving the need index. Of approximately 32,000 families for whom we received at least some data for the 1996-97 evaluation, 1.6 percent were excluded from the analysis of needy families due to incomplete data.
47 The standard deviation of the need index was 1.4.
48 The 3 percent of families with none of the seven need indicators may be an error resulting from the very conservative method used to determine whether a family income was at or below the federal poverty level. Further, these families may have experienced great needs that were not represented by the seven indices used in this analysis.
49 The working definition of very needy families was limited to somewhat imprecise measures of a few aspects of a family. It was possible that a very needy family headed by a single-parent who is a native English-speaker and who chooses to raise three children without receiving public assistance could have been classified as less needy. The need index was not intended to assess the actual extent of need of individual families, but rather, to serve as a tool for analytic purposes.
-###-
[Part 5: How Many Children Have Special Needs?] |
|
[Part 7: Parents' Reasons for Participating in Even Start] |