Table of Contents | Previous | Next |
Appendix 2.1: Comparison of Head Start Grantees/Delegate Agencies and Centers in Saturated and Non-Saturated Communities
As discussed in Chapter 2, there is potential for undercoverage bias due to the exclusion from the sampling frame of Head Start grantees/delegate agencies and centers in communities saturated by the program, i.e., communities with too few extra families interested in Head Start (beyond those the program can accommodate) to provide a randomly selected non-Head Start group for the study. Newly entering Head Start children in these saturated communities had no chance of selection and therefore are not represented by our sample. Consequently, the potential for bias arises if the saturated grantees/delegate agencies and centers are systematically different from the non-saturated grantees/delegate agencies and centers we retained in the sampling frame and if the characteristics on which they differ are correlated with the outcome measures for and impact estimates on the children they enroll. However, if the children in these excluded grantees/delegate agencies and centers represent only a small percentage of the Head Start population, then the potential for bias is much less. Based on the sample coverage rate reported in Chapter 2, 15.5 percent of the children served by Head Start nationally are omitted from the study. This noncoverage rate is based on grantees and centers identified in the sample frame and samples that were excluded due to saturation. It equals 1 minus the product of four coverage rates: program frame x program sample x center frame x center sample. Mathematically, this equates to 1-(0.962 x 0.975 x 0.952 x 0.947) = 1-0.845 = 0.155.
Head Start Grantees/Delegate Agencies
Exhibits A.2.1.1 and A.2.1.2 compare saturated and non-saturated grantees/delegate agencies by a few qualitative characteristics and enrollment variables available on the Head Start Program Information Report (PIR) database (and, for newly entering enrollment, telephone screening confirmation calls to grantees and delegate agencies prior to sampling). The grantees/delegate agencies were weighted to account for sampling of broad geographic areas (i.e., PSUs) and for the subsampling of grantees/delegate agencies in three large urban cities prior to the telephone screening (see Chapter 1). This is necessary to draw conclusions about the entire population of children served by Head Start and not merely the children served by grantees/delegate agencies in the 25 sampled PSUs that were screened to determine saturation. Tests of statistical significance were performed to reduce the possibility of drawing false conclusions from differences that may be due to sampling error. The hypothesis testing was done in WesVar using jackknife replicate weights to account for the study’s complex sample design.
As shown in these tables, the saturated grantees/delegate agencies are much smaller, much more likely to be school-based, and have smaller percentages of Hispanic enrollment than the non-saturated grantees/delegate agencies. Although they appear to be more often located in the midwest, differences in the distribution of saturated vs. non-saturated grantees/delegate agencies by Head Start regions are not statistically significant. A cautionary note is that variances at the program level are not very stable because the number of saturated grantees/delegate agencies is small. In addition, variances do not include the between-PSU component of variance due to sampling PSUs; thus, they are underestimates, and the p-values may be slightly overstating the significance of the differences.
| Enrollment Variable | Saturated Programs | Non-Saturated Programs | P-Value (t-Test of Difference) |
|---|---|---|---|
| Percent Hispanic Enrollment | 9% | 26% | 0.001 |
| Percent Black Enrollment | 20% | 33% | 0.134 |
| Age 3 Enrollment as Percent of Total Enrollment | 52% | 49% | 0.535 |
| Average Total Enrollment | 188 | 571 | <0.001 |
| Average Newly Entering Enrollment | 113 | 388 | <0.001 |
| Characteristics | Saturated Programs | Non-Saturated Programs | P-Value (Chi-Square Test of Association) | |
|---|---|---|---|---|
| School-based | 0.018 | |||
| Yes | 66% | 21% | ||
| No | 34% | 79% | ||
| Metro Status | 0.91 | |||
| MSA | 66% | 68% | ||
| Non-MSA | 34% | 32% | ||
| Level of Pre-K Services in State | 0.60 | |||
| Similar to Head Start | 35% | 25% | ||
| Some Head Start-Like | 27% | 20% | ||
| Remaining States | 38% | 55% | ||
| Head Start Region | 0.15 | |||
| Northeast | 24% | 25% | ||
| Midwest | 48% | 24% | ||
| South | 28% | 39% | ||
| Plains | 0% | 4% | ||
| West | 0% | 8% | ||
Head Start Centers
Exhibits A.2.1.3 and A.2.1.4 compare saturated and non-saturated centers by various qualitative characteristics and enrollment variables available from Center Information Forms (CIFs) completed by all centers in the sampled grantees and delegate agencies. All hypothesis testing was again done in WesVar using jackknife replicate weights to account for the study sample design. The replicate weights do not include the between-PSU variance component, therefore the p-values in these tables may slightly overstate the significance of the difference. In Exhibit A.2.1.3 the chi-square test was not able to detect a significant difference for type of program option offered, whether staff are school employees, metro status, region, or level of Pre-K services available in the state. With respect to enrollment, Exhibit A.2.1.4 shows that the saturated centers are smaller, have fewer Hispanic children, and have a larger percentage of first year 3-year-olds than the non-saturated centers. As expected, these centers do not have waiting lists, a significant difference from non-saturated centers.
Two graphs follow Exhibit A.2.1.4 that show the percentage of centers that are saturated for each of the 84 grantees/delegate agencies with less than 100 percent saturation rate. The saturation rate was calculated two ways: as the percentage of centers in each program that are saturated and as the percentage of newly entering enrollment in saturated centers for each program. The average percentage of saturated centers is 16.6 percent while the average percentage of newly entering enrollment in saturated centers is 13.2 percent, another indication that the saturated centers tend to be smaller. The graphs show the extreme variation among grantees/delegate agencies in the share of centers operating in saturated communities and the share of newly entering children served by those centers.
| Characteristics | Saturated Centers | Non-Saturated Centers | P-Value (Chi-Square Test of Association) | |
|---|---|---|---|---|
| Program Option | 0.44 | |||
| Full-Day Only | 35% | 28% | ||
| Part-Day Only | 52% | 50% | ||
| Other | 13% | 22% | ||
| Staff Are School Employees | 0.249 | |||
| Yes | 17% | 11% | ||
| No | 83% | 89% | ||
| Metro Status | 0.64 | |||
| MSA | 74% | 70% | ||
| Non-MSA | 26% | 30% | ||
| Head Start Region | 0.376 | |||
| Northeast | 32% | 27% | ||
| Midwest | 34% | 20% | ||
| South | 17% | 31% | ||
| Plains | 12% | 11% | ||
| West | 4% | 11% | ||
| Level of Pre-K Services in State | 0.212 | |||
| Similar to Head Start | 40% | 22% | ||
| Some Head Start-Like | 15% | 18% | ||
| Remaining States | 45% | 60% | ||
| Enrollment Characteristic | Saturated Centers | Non-Saturated Centers | P-Value (t-test of Difference) |
|---|---|---|---|
| Percent Hispanic Enrollment | 17% | 30% | 0.005 |
| Percent Black Enrollment | 38% | 26% | 0.204 |
| Percent Newly Entering Enrollment | 65% | 66% | 0.985 |
| Age 3 Enrollment as Percent of Newly Entering Enrollment | 54% | 47% | 0.037 |
| Number of Children on Waiting List as Percent of Total Enrollment | 0% | 15% | <0.001 |
| Average Number Funded Slots | 37 | 48 | 0.036 |
| Average Total Enrollment | 26 | 47 | <0.001 |
| Average Newly Entering Enrollment | 16 | 31 | <0.001 |
| Average Number on Waiting List | 0 | 9 | <0.001 |
[D] |
| Table of Contents | Previous | Next |