The Future of Children, Princeton - Brookings: Providing research and analysis to promote effective policies and programs for children.

Existing studies of the effectiveness of PAT and home visiting in general have tended to consider the intervention as a "black box." Little is known about variations in the ways families experience home visiting programs and the differences those variations make in the outcomes children and families achieve.

One clear difference in families' experiences of PAT was the degree to which they were exposed to the PAT curriculum through home visits. In any program, some percentage of enrollees discontinue their participation before the program is completed. As discussed earlier, 43% of families that were assigned to the participant group in the Salinas Valley PAT Demonstration dropped out of the intervention before the children reached their third birthdays, and 57% of the teens in the three intervention groups (PAT, case management, and PAT plus case management) in the Teen PAT Demonstration dropped out before the children's second birthdays.

Even among those who persisted in the program, other variations in exposure occurred because families were not always able to keep their home visit appointments. For example, 11% of attempted home visits by Teen PAT parent educators were "no-shows," that is, the teen was not home when the parent educator arrived for the arranged visit. Case managers had about 35% of their attempted home contacts end as no-shows for the case management and combined intervention groups. More than 40% of Salinas Valley PAT families that persisted through the children's third birthdays had gaps of four months or more in their home visits, largely because many families returned to Mexico for extended periods each winter.

These kinds of variations resulted in considerably reduced exposure to the program for all groups when compared to the monthly home visits intended by program planners at these sites. Salinas Valley PAT families as a whole averaged almost 20 home visits in three years, with those that completed the program (n=169) averaging 28 visits, as compared with 8 visits for dropouts (n=129). Teen PAT families averaged 10 home visits and/or in-person case management contacts. Those that persisted through the children's second birthdays (n=76 for the PAT group, 64 for the case management group, and 84 for the combined intervention) averaged about three times the exposure of both home visits (16) and case management contacts (17 and 14 contacts, respectively, for the case management and combined intervention groups) as did dropouts (n=92 for the PAT group, 105 for the case management group, and 89 for the combined intervention group), with an average of five home visits and/or case management contacts. Given the high proportion of families that dropped out of their programs and the relatively low level of program service that dropouts experienced, it is important to assess whether PAT achieved more substantial impacts when experienced at a higher level of exposure.

In the Salinas Valley PAT Demonstration, parent and child outcomes for families that had dropped out of the program and families that persisted in the program were analyzed separately to determine whether each group had significantly better outcomes than control group members. A more fine-grained analysis was conducted to determine whether the number of home visits that families received contributed significantly to parent and child outcomes relative to no home visits (that is, the control group).

In the Teen PAT Demonstration, the effect on parent and child outcomes of receiving the intended level of program service as compared with not receiving the intended level of program service was examined. For the PAT group, the intended level of program service as defined by the Parents as Teachers National Center (PATNC) was 17 home visits for each family through the children's second birthdays.³⁷ For the case management group, four personal contacts per year was the minimum acceptable level of service as determined by the state funding agency for those services. For the combined intervention group, both thresholds needed to be met for a family to be considered to have received the intended level of program service.

Families that participated more in the programs may have differed from other families in ways other than the level of services they received. Because these differences may have influenced outcomes, multivariate analyses that controlled for a variety of individual characteristics³⁸ were conducted. However, even though the influence of important measurable differences between families that may have affected their outcomes was taken into consideration, it is possible that unmeasured factors may have distinguished high-exposure families and influenced the relationship between program exposure and outcomes.

Neither demonstration found that greater exposure to program services was related to greater parenting impacts. Although virtually all relationships between having greater exposure to a program and parenting outcomes were positive, in no case did they reach statistical significance, controlling for other factors.

However, Table 8 illustrates that the level of service participants received did relate to the level of benefit to children. In the Salinas Valley Demonstration, persisting in PAT through the children's third birthdays was significantly related to higher DPII scores in the cognitive, social, and self-help domains. Persisting to the end of the intervention (rather than being in the control group) was associated with a developmental advantage of 1.7 to 2.6 months across these domains, controlling for other factors. Similarly, each home visit was associated with a significant benefit in these domains. Home visits were associated with a one-month developmental advantage per 10 home visits in the cognitive, social, and self-help domains for children in the PAT participant group in the Salinas Valley study.

Findings regarding exposure to program services were less consistent in the Teen PAT Demonstration. Only in the case of cognitive development as measured by the DPII was the impact of receiving the intended level of services statistically significant, controlling for other factors in the analyses. Receiving the intended level of exposure was associated with a 1.6-month developmental advantage for children assigned to the case management group (p<.05) and a 1.7-month developmental advantage for children assigned to the combined PAT-plus-case-management group (p<.10). No impact was noted for children in the PAT-alone group, even when the intended number of home visits was made.

The inconsistent findings regarding the influence of the level of program exposure on outcomes mirror the inconclusiveness of other home visiting and family support research, in which intensity has been found to contribute to greater impact in some studies³⁹ but not in others.⁴⁰