Table of Contents | Previous | Next |
PREDICTING EARLY HEAD START PROGRAM USE AND ACCEPTANCE BY PARENTS
JoAnn L. Robinson, Sheridan Green, Nancy Song and Robert
Emde
University of Colorado
Jon Korfmacher
Erikson Institute
Rebecca Soden
Clayton/Mile High Family Futures Early Head Start
Acknowledgements. This work was funded by the Clayton/Mile-High Early Head Start Local Research Partner grant (#ACYF90YF0009/04) and by the Jay and Rose Phillips Family Foundation grant to the last author, who served as Principal Investigator of the project. We extend our appreciation to the families who participated in this study and the research assistants and Clayton program staff who diligently worked to gather this information. We also extend special recognition to Mitzi Kennedy, Charmaine Lewis, and Chris Sciarrino for their appreciation of the importance of the research process. This report was extracted from a paper by the same authors published in the journal, NHSA Dialogue, under the title “Predicting Program Use and Acceptance by Parents Enrolled in Early Head Start.”
The research partnership between the Program for Early Developmental Studies in the Dept. of Psychiatry at the University of Colorado Health Sciences Center and the Clayton/Mile High Family Futures Early Head Start (CMHFFEHS) included as a principal goal the documentation of client engagement with the program. We sought to more fully understand how families respond differently to the home visiting intervention and to see what works best for whom under what circumstances. The research agenda included acquiring psychologically meaningful information about clients during the enrollment process. Using initial client assessments might assist home visitors during this relationship formation phase, not as a replacement for their own judgment of families, but as a supplementary tool to help meet client needs.
In this paper, we will present how these issues were addressed in our collaboration. Our collaborative goal was to explore whether client characteristics were predictive of patterns of participation in home visitation. If they were, we could then discuss whether the results might lead to a refinement of practices. We will begin with a brief review of prior efforts in the field to investigate determinants of program participation.
Typically, randomized trials of intervention investigate the intervention’s effectiveness, asking the question, “Does the program work?” In such “intent-to-treat” approaches, there is an assumption of a monolithic treatment that is uniformly delivered, no matter how families may vary in amount of services delivered. Participants also experience an intervention in different ways and this will likely influence the benefits they receive. Previous studies have shown that variations in a mother’s engagement with a program is related to program outcomes (Barnard et al., 1986; Korfmacher, Adem, Ogawa & Egeland, 1997; Lieberman, Weston & Pawl, 1991).
What predicts engagement in an intervention? How individuals use services and respond to an intervention depends, in part, on their personal characteristics and on the amount and type of support they receive from family, partners, and friends. Such links have been demonstrated for years in psychotherapy literature (Bergin & Garfield, 1994; Roth & Fonagy, 1996). Research suggests that mothers with who are at increased risk for parenting difficulties can be particularly helped by early interventions. For example, home visiting programs may be more successful among mothers with mental health difficulties such as depression (Lyons-Ruth, Connell, Grunebaum & Botein, 1991).
It is also possible that level of social support or interpersonal relatedness influence program use and engagement, although findings are inconsistent. Some studies show social support related to increased use and commitment to an intervention, while others demonstrate lower participation as social support increases (Birkel & Reppucci, 1983; Dunst, Lee & Trivette, 1988; Powell, 1988; Unger & Wandersman, 1988).
A study by Korfmacher and colleagues (1997) illustrates in particular the complexities of examining this issue. The authors found that insecure/dismissive mothers (with respect to their memories for past caregiving experiences) had an equally high level of contact with their home visitors as mothers with more secure representations. Qualitative ratings by the home visitors, however, suggested that their emotional involvement in the sessions was fairly shallow. In other words, although mothers were often available for visits, they seemed to keep their home visitors at an emotional distance, paralleling their avoidance of emotions when reflecting on their relationship history. This split between amount of contact and qualitative features of the contact suggests that any examination of the relationship between participant characteristics and program use should examine multiple dimensions of program participation.
Methods
Sample. One hundred sixty-two low income women who were either pregnant or had a child under the age of 12 months were recruited by program staff to participate in the national study of Early Head Start (EHS). They signed an informed consent form with the understanding that approximately half would be randomly selected to receive EHS program services plus developmental screenings and half to receive developmental screenings only. Eighty-three women were randomly assigned to the EHS services group.
The average age of the mothers was 21.4 years, 74 percent were unmarried, 40 percent were Latina, 36 percent were African American, 17 percent were European American, and 7 percent represented other ethnic groups. Seventy-two percent of the women spoke English well, but 28 percent were more proficient in Spanish. Forty-seven percent did not complete high school, 24 percent completed high school, and 29 percent reported some technical or college-level education.
Data Gathering. Following the recruitment visit and before random assignment, mothers completed a 2-hour interview with research staff that included the Head Start Family Information System (HSFIS) enrollment questions (providing basic demographic information) plus selected standardized questionnaires of psychological constructs that included measures of: depression (using the Center for Epidemiological Studies-Depression scale [CES-D: Radloff, 1977]), sense of personal mastery (Pearlin & Schooler, 1967), attitudes toward relationships (Simpson, Rholes & Nelligan, 1992), violence in partnered relationships (10 items from the Conflict Tactics Scale, Straus & Gelles, 1990), and stressful life events (using a 20-item life events scale by Mathematica Policy Research, 1996). Women’s literacy was evaluated using the Letter-Word Identification subtest of the Woodcock-Johnson Tests of Achievement (Woodcock & Mather, 1989). Table 1 displays the sample means and standard deviations for these assessments. It is noteworthy that while 45 percent of women had depression scores above the CES-D cut-off, reported stressful life events and domestic violence were very low. The group means for personal mastery and difficult relationship attitudes were slightly about the mid-point for these scales.
| Item | Mean | Standard Deviation |
|---|---|---|
| Depression | 16.97 | 10.15 |
| Domestic Violence | 0.64 | 1.63 |
| Personal Mastery | 2.98 | 0.55 |
| Difficult Relationship Attitudes | 2.68 | 0.66 |
| Stressful Life Events | 0.27 | 0.13 |
| Literacy Level (grade equivalent) | 11 | 4.21 |
Home visitors tracked their contacts with families and recorded information about each visit on the collaboratively designed home visit tracking form. The average number of months women were in the program as of 12/31/99 was 21 months and ranged from less than one month to 37 months. The average number of visits over the three years was 45 (median = 44) but ranged from 0-131; this included 10 mothers who had no home visits after random assignment. The average length of visits was 80.9 minutes. Following each visit, home visitors rated mother’s level of involvement on a scale ranging from: (1) Not involved, inattentive, not participating to (5) Highly interested/involved, attentive. The distribution of these involvement ratings was positively skewed; the mean was 4.80 (SD = .27) and the median was 4.91. Average involvement was below the mean for 40 percent of the 83 women, including the ten women who completed no visits. Three years after enrollment began, at a point when recruitment was completed and families had potentially received services for two years or more, the research team undertook examination of the data and shared those results with program staff.
Data Analysis. In order to examine the association between maternal baseline characteristics and subsequent patterns of participation, we began by correlating the five psychological dimensions, plus maternal age and literacy level with the frequency of home visits and the average level of involvement rated by home visitors. As noted earlier, however, we realized that amount of visiting and involvement in the visits provides more information when considered together. To capture this, an aggregate measure containing classifications of participation was constructed. This was done by cross-tabulating information about frequency and involvement, dichotomizing each distribution at the mean.
This produced five groups; two clearly reflected lower participation. Never Engaged women completed no visits (N = 10) and Disengaged women completed fewer than average visits and were rated lower than average on involvement (N = 15). Two patterns reflected mixed participation in the program: Superficially Engaged women who completed greater than average visits but were rated lower than average on involvement (N = 10) and Sporadically Engaged women who completed fewer than average visits but were rated higher than average on involvement (N = 15). The fifth group included Highly Engaged women who completed greater than average visits and were rated higher than average on involvement (N = 33). Number of visits and average involvement were weakly, but significantly correlated (r = .29, p < .05). Using the GLM procedure from SAS, we tested differences between the least square means of these five groups on the five psychological dimensions, plus maternal age and literacy level.
Results
Examining associations between the variables assessed at program entry (the five psychological constructs, maternal age and literacy level) with the later frequency of home visits and mother’s level of involvement in visits, yielded no significant correlations (see Table 2). In other words, program participation variables, when considered separately as single constructs, were not predicted by baseline characteristics.
| Item | Number of Visits (n = 83) |
Average Involvement (n = 73) |
|---|---|---|
| Depression | 0.02 | -0.14 |
| Domestic Violence | 0.07 | -0.02 |
| Personal Mastery | -0.02 | 0.21 |
| Difficult Relationship Attitudes | 0.19 | 0 |
| Stressful Life Events | -0.06 | -0.14 |
| Literacy Level | 0.07 | 0.16 |
| Maternal Age | -0.04 | 0.11 |
The story was different, however, when program participation dimensions were combined into classifications. We examined the associations between the five classifications of participation and these same baseline variables (see Table 3). As can be seen, significant differences were observed primarily between the Superficially Engaged women and the women with higher involvement scores (sporadically engaged and highly engaged).
| Depression | Domestic Violence |
Mastery | Relationship Attitudes |
Stressful Life Events |
Literacy Level |
Maternal Age |
|
|---|---|---|---|---|---|---|---|
| Never Engaged |
16.9 | 0.01 | 3.1 | 2.34 | 0.33 | 12.32 | 24.60b,c |
| Disengaged | 18.6 | 0.06 | 2.87 | 2.55 | 0.25 | 10.07 | 21.13 |
| Superficially Engaged |
20.2 | 0.05 | 2.54a,d,e | 3.24a,b,d,e | 0.36d,e | 10.25 | 19.9 |
| Sporadically Engaged |
16.3 | 0.03 | 3.05 | 2.7 | 0.24 | 10.69 | 19.8 |
| Highly Engaged |
17 | 0.09 | 3.09 | 2.67 | 0.26 | 11.4 | 21.91 |
| (a) p < .05 different from Never Engaged (d)
p < .05 different from Sporadically Engaged (b) p < .05 different from Disengaged (e) p < .05 different from Highly Engaged (c) p < .05 different from Superficially Engaged |
Superficially Engaged women, compared to sporadically engaged women, reported lower personal mastery (T1,82 = 2.35, p < .05), more difficult relationship attitudes (T1,82 = 2.06, p < .05), and more stressful life events (T1,82 = 2.11, p < .05). Superficially Engaged women, compared to Highly Engaged women, reported a lower sense of personal mastery (T1,82 = 2.82, p < .01), difficult attitudes toward relationships (T1,82 = 2.51, p < .02), and stressful life events (T1,82 = 2.05, p < .05). In addition, Superficially Engaged women also reported lower mastery and more difficult attitudes toward relationships than the Never Engaged women (T1,82 = 2.34, p < .05) and their attitudes toward relationships were also more difficult than the Disengaged group (T1,82 = 2.65, p < .01). Finally, maternal age also differed across groups: the Never Engaged women were significantly older at the time of program enrollment than either the Superficially Engaged (T1,82 = 2.36, p < .05) or Sporadically Engaged women (T1,82 = 2.64, p = .01). In addition, there were non-significant trends for Never Engaged women to be older than Disengaged and Highly Engaged women.
Discussion
Striking differences were found between the Superficially Engaged women and women rated by their home visitors as more involved during their meetings. This pattern of participation seemed to delimit a meaningful subgroup of mothers with particular psychological characteristics at program entry. Superficially Engaged women had a lower sense of personal mastery, more difficult attitudes toward relationships, and greater life stress at enrollment than women who became more involved in the program. Participating in a superficial manner meant that although clients were often home and available for visits, visitors experienced these women as less attentive and inquisitive during their meetings. This fits the greater reticence reflected in their attitudes toward relationships reported at enrollment, a finding similar to results reported by Korfmacher and colleagues (1997) with a different home visiting model. The lower sense of personal mastery and high levels of stress reflected in the number of recent life events reported at enrollment suggest that their relatively high attendance may be a reflection of some disorganization or lack of mobilization that kept them home and available for visits.
With respect to these psychological characteristics, Superficially Engaged women appear to have greater risks than the other groups and may stand to benefit most from the home visit services compared to their control group counterparts. That they appeared less inclined to engage with visitors is one of the challenges of relationship-based practice. Other programs have demonstrated the benefits of home visitor persistence in meeting with families in need (e.g., Olds & Korfmacher, 1998), even when these families were very passive or ambivalent in their involvement (e.g., Greenspan et al., 1987).
Both research and program staff were surprised that Never Engaged or Disengaged women did not differ from women who had patterns of greater program participation. We jointly considered how the Never Engaged group might have been an artifact of the random assignment process. The women in this group were older than the other groups, had the highest sense of personal mastery, and the most positive attitudes toward relationships. Therefore, they may have had lower perceived needs for the structured relationship-based approach of the program.
Disengaged women, who gradually withdrew from program participation, on the other hand, did not stand out in any way. Their low engagement with the program may have been occasioned by an unexpected move from the neighborhood or other factors unrelated to their initial reported characteristics. Perceived passivity and uninvolvement may also have discouraged visitors from pursuing these women. The enrollment assessment measures, however, did not provide insight about them.
As a result of these partnership discussions, CMHFFEHS decided that supplementing their enrollment process with brief assessments of psychological characteristics of mothers would be a valuable addition to their regular protocol for two reasons. First, in the spirit of continuous learning and improvement, they expressed a desire to actively collect such information themselves and see if they could respond better to individual needs and improve engagement levels. Second, they developed a strong appreciation for the role of assessment in establishing client-home visitor relationships.
Finally, we should consider what resources a program might need to implement a systematic assessment process at program entry. A trained staff person who is able to administer standardized questionnaires as part of an enrollment process is essential. Also essential is an information management system that will permit timely data basing along with a plan for periodically summarizing information so as to examine distributions and associations between participation and initial characteristics. A continuous improvement partner working with the program may provide these data skills. Most importantly, supervisors and visitors need training and on-going support to interpret the profiles that emerge from these assessments. Such expertise is crucial for information of this kind to be used appropriately as a supplement to the judgments of program staff regarding how to work successfully with the families they serve.
References
Barnard, K.E., Magyary, D., Sumner, G., Booth, C.L., Mitchell, S.K., & Spieker, S. (1988). Prevention of parenting alterations for women of low social support. Psychiatry, 51, 248-253.
Bergin, A. E., & Garfield, S. L. (1994). Handbook of psychotherapy and behavior change. New York: John Wiley & Sons, Inc.
Birkel, R. C., & Reppucci, N. D. (1983). Social networks, information-seeking, and the utilization of services. American Journal of Community Psychology, 11, 185-205.
Dunst, C. J., Lee, H. E., & Trivette, C. M. (1988). Family resources, personal well-being, and early intervention. Journal of Special Education, 22, 108-116.
Greenspan, S., Wieder, S. Lieberman, A. F., Nover, R., Robinson, M. & Lourie, R. (Eds.) (1987). Infants in multirisk families. Madison, CT: International Universities Press.
Korfmacher, J., Adam, E., Ogawa, J., & Egeland, B. (1997). Adult attachment: Implications for the therapeutic process in a home visitation intervention. Applied Developmental Science, 1, 43-52.
Lieberman, A. F., Weston, D., & Pawl, J. H. (1991). Preventive intervention and outcome with anxiously attached dyads. Child Development, 62, 199-209.
Lyons-Ruth, K., Connell, D. B., Grunebaum, J. U., & Botein, S. (1990). Infants at social risk: Maternal depression and family support services as mediators of infant development and security of attachment. Child Development, 61, 85-98.
Olds, D. L., & Korfmacher, J. (1998). Maternal psychological characteristics as influences on home visitation contact. Journal of Community Psychology, 26, 23-36.
Pearlin LI, Schooler C. (1967). The structure of coping. Journal of Health and Social Behavior, 19, 2-21.
Powell, D. R. (1988). Toward an understanding of the program variable in comprehensive parent support programs. H. B. Weiss, & F. H. Jacobs (Eds.), Evaluating family programs (pp. 267-286). New York: Aldine de Gruyter.
Radloff, L. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385-401.
Roth, A, & Fonagy P. (1996). What works for whom? A critical review of psychotherapy research. New York: The Guilford Press.
Simpson, J., Rholes, W., & Nelligan, J. (1992). Support seeking and support giving within couples in an anxiety-provoking situation: The role of attachment styles. Journal of Personality and Social Psychology, 62, 4343-446. Straus, M. A. & Gelles, R. J. (1990). Physical Violence in American Families: Risk Factors and Adaptations to Violence in 8,145 Families. New Brunswick, NJ: Transaction.Unger, D. G., & Wandersman, L. P. (1988). The relation of family and partner support to the adjustment of adolescent mothers. Child Development, 59, 1056-1060.
Woodcock, R. W., & Mather, N. (1989). Woodcock Johnson-Revised test of achievement: Examiner's manual. In R.W. Woodcock & M.B. Johnson, Woodcock-Johnson Psycho-Educational Battery - Revised. Chicago: Riverside.
| Table of Contents | Previous | Next |