A Study of Engagement in Neighborhood-Based

Child Welfare Services

Julie Cooper Altman

Adelphi University

This article reports the results of a mixed-method study that examined processes and outcomes of parent-worker engage-
ment in child welfare. Knowledge gained from a qualitative exploration of engagement at one neighborhood-based child
welfare agency informed the gathering of quantitative data from 74 different parent-worker dyads in this sequential
exploratory design. Seven themes instrumental to engagement emerged: (a) clear, collaborative goal setting; (b) hope-
fulness; (c) parent acknowledgment of their situations; (d) motivation; (e) respect for cultural differences; (f) honest and
straightforward communication; and (g) persistent and timely efforts by all. Quantitative analyses yielded little rela-
tionship between engagement and either visitation rate or case disposition by 9 months post placement. Although these
data provide support for the clinical benefits of working to improve parent-worker engagement in child welfare services,
they fail to provide evidence of a relationship between engagement and improved case outcomes.

Keywords: client engagement; child welfare service delivery; working alliance

Engaging clients has long been considered a fundamen- client engagement was conducted. Through interviews, the
tal step in the interpersonal helping process, a key part experiences and perspectives of parents, foster parents,
of the process of client change (Compton & Galaway, workers, and supervisors in one child welfare agency
1994; Hepworth, Rooney, Dewberry-Rooney, Strom- were gathered and analyzed. Then, and based on those
Gottfried, & Larsen, 2007; Kadushin & Kadushin, data, engagement was operationalized, measured, and
1997; Woods & Hollis, 2000). Yet, for all its importance, related to certain outcomes for 74 different parent-worker
we know very little about how it happens, when it hap- dyads.
pens, why it happens, and how it may be related to fur-
ther change processes. In the field of child welfare, the
Adoption and Safe Families Act (P.L. 105-89) has short- BACKGROUND AND SIGNIFICANCE
ened the timeline child welfare workers have to work
toward permanency with the families of children who The need for child welfare workers to successfully
have been placed in out-of-home care. Knowledge of engage clients in the helping process has long been realized
how best to engage and move quickly toward change but seldom addressed in the empirical literature (Altman,
with these families is now critical. 2005; Dawson & Berry, 2002; Pecora, 1989; Rooney, 1992;
This article reports the findings of a sequential Yatchmenoff, 2005). Complicating efforts are the disparate
exploratory mixed-method study of the processes and definitions and competing, unclear, or incomplete concep-
outcomes of client engagement in the New York City tualizations of engagement itself. Indeed, engagement can
neighborhood-based child welfare system. Initially, a be seen as a process or a product, an intra- or interpersonal
qualitative exploration of the concept and process of effort, worker or client driven, and both a cause of partic-
ipation and its result.
From one perspective, engagement is viewed as an
Author’s Note: This article was accepted under the guest editorship of interactional, interpersonal process, beginning when work-
Dr. Anne E. (Ricky) Fortune. The research study on which this article is based
was funded through a grant from the New York Community Trust. Portions
ers establish communication with a potential client and
of this article were previously presented at the International Practice ending when there is a preliminary agreement to work
Research Symposium, June 1-2, 2005, sponsored by the School of Social together (Compton & Galaway, 1994). It can be seen as the
Welfare at the State University of New York at Albany. Address correspon-
process whereby the social worker creates an environment
dence to Dr. Julie Altman, Adelphi University School of Social Work, One
South Avenue, Garden City, NY 11530; e-mail: [email protected]. of warmth, empathy, and genuineness that enables a client
to enter into a helping relationship and actively work
Research on Social Work Practice, Vol. 18 No. 6, November 2008 555-564
DOI: 10.1177/1049731507309825 toward change. It can also be seen as the degree to which a
© 2008 Sage Publications given client is committed to collaboratively working with
555
556 RESEARCH ON SOCIAL WORK PRACTICE

a worker toward change. Others conceptualize it for New York City’s model of neighborhood-based child
research purposes as an outcome, usually measured by rates welfare services was designed to meet some of these
of client participation (Littell, Alexander, & Reynolds, challenges, emphasizing community service networks; a
2001; Littell & Tajima, 2000) or attrition (Daro, McCurdy, culturally sensitive, family-centered philosophical ori-
Falconnier, & Stojanovic, 2003). Prinz and Miller (1996) entation; and a reconceptualization of the role of birth
defined engagement as “the participation necessary to and foster parents as integral collaborators. The model
obtain optimal benefits from an intervention” (p. 163). builds on research suggesting the benefits of service par-
Yatchmenoff (2005) was one of the first to attempt to ticipation for families: Parents who cooperate are less
define and measure engagement as a product in the child likely to be referred to court (Karski, 1999), uncooperative
welfare context. She defined it fluidly as “positive involve- parents may not be offered needed services (Jones, 1993),
ment in the helping process” (p. 86) and identified within active participation in services reduces the chances parents
that construct five factors: receptivity, expectancy, invest- will lose custody of children (Atkinson & Butler, 1996),
ment, mistrust, and a working relationship. and parents more able to visit their children are more
For the purposes of this study, engagement was under- likely to be reunited (Davis, Landsverk, Newton, &
stood as Yatchmenoff defined it and conceptualized as a Granger, 1996). Greater parental involvement in treatment
dynamic, complex, and multi-level phenomenon influ- planning results in fewer subsequent reports of child mal-
enced by and composed of a number of interwoven fac- treatment (Littell, 1997).
tors (to be better understood and determined) at the Despite these improvements, a recent advisory panel
worker, client, agency, and larger social environment continued to note that “the most significant challenge fac-
levels. This conceptualization was used merely to provide ing (New York City’s Child Welfare) system is to make
a guiding framework for the start of the qualitative further, critically needed improvements with regard to per-
inquiry, and as a beginning heuristic for analysis. manency,” the heart of which should be a “re-thinking
Interest in a better understanding of the processes and out- of the role of parents, around the primary themes of
comes of engagement in child welfare services is growing enhanced respect, engagement and partnership” (Special
(Altman, 2005; Daro, McCurdy, & Nelson, 2005; Dawson Child Welfare Advisory Panel, 2001). More knowledge of
& Berry, 2002; Dore & Alexander, 1996; Lee & Ayon, 2004; how engagement and partnership unfold in child welfare
Littell, 2001; Littell & Tajima, 2000; Okamoto, 2001; Petras, practice as well as their importance to the ongoing perma-
Massat, & Essex, 2002; Shonfeld-Ringel, 2001; Smithgall, nency planning for children in the child welfare system
2006; Yatchmenoff, 2005). A number of issues, however, were the goals of this investigation.
compound both the complexity and importance of engaging
clients in the helping process.
First, the 1997 Adoption and Safe Families Act METHOD
requires that decisions about termination of parental
rights be made within 12 months of a child entering out- Research Questions
of-home care, thus reducing the window open to families
to engage in services aimed at changing the conditions Three research questions were addressed in this
that led to placement. Second, most parents in the child study: What is the process of parent-worker engagement
welfare system can be considered involuntary or nonvol- in neighborhood-based child welfare services, and how
untary clients (Rooney, 1992), and either have not asked is it influenced? Is parent-worker engagement related to
for or, initially at least, do not see the need for the services visitation between parent and child in neighborhood-
offered. Guilt, anger, fear, shame, confusion, hostility, based child welfare services? Is parent-worker engage-
suspicion, or depression may lead the clients to appear ment related to reunification in neighborhood-based
reluctant, unmotivated, or unwilling, compounding the child welfare services?
difficulties workers may have in engaging them in ser-
vices (Blythe, Ivanoff, & Tripodi, 1994). Third, the high Research Design
rate of child welfare worker turnover (Cyphers, 2001)
compounds the difficulty parents have in relating to yet A sequential exploratory mixed-method design was used
another worker, one often either inexperienced or over- to answer the research questions. Qualitative interviews
burdened. Further complicating this is the inherently were held with participants of one neighborhood-based
unequal relationship between worker and client, where a child welfare agency to better understand the concept and
collaborative ideal in treatment planning is difficult to processes of engagement from their perspective. Based on
achieve (Alexander & Dore, 1999; Rooney, 2000). this knowledge, engagement was then operationalized,
 Altman / NEIGHBORHOOD-BASED CHILD WELFARE SERVICES 557

TABLE 1: Demographics of Sample

Qualitative Quantitative

Variable Parent (n = 16) Foster Parent (n = 7) Worker (n = 9) Supervisor (n = 4) Parent (n = 74) Worker (n =
74)

Gender (% women) 94 86 89 75 72 92
Age M (SD) 37 52 34 40 33.2 (9.4) 34.2 (10.2)
Caseload M (SD) — — n/a n/a — 24.8 (7.9)
Education (%)
< High school n.a. n.a. n.a. n.a. 44.4 —
High school 37.8 —
Some college 8.2 1.4
College 9.6 87.3
Graduate degree — 11.3
Months of child welfare — 144 60 144 — 37.7 (32.9)
experience M (SD)
Ethnicity (%)
African American 12.5 29 11 25 64.9 53.5
Latina/o 75 71 78 50 18.9 33.8
Asian American — — 11 — — 2.8
White 12.5 — — 25 8.1 2.8
Other — — — — 8.1 7
Parental Status (%)
Parent 100 100 50 75 100 58
Not a parent 0 0 50 25 0 42

and measured, so that a longitudinal analysis of the rela- who had a foster care case opened within the past 30
tionship of engagement to certain case outcomes could be days was asked for informed consent to participate in
conducted in the second and third years of the study. This the research. Of 155 parents eligible, 29 were unable to
two-phase mixed-method research design thus allowed be found, 49 declined, and 77 participated, for a 61%
for much needed dynamic, longitudinal analyses of the response rate. (Only 74 dyads were used in the analysis
processes and outcomes of engagement in child welfare due to missing data.) Once the parent agreed to partici-
services. Approval for the design and implementation of pate, his or her worker was asked to participate in the
this study was given by the researcher’s university institu- study. All workers who were asked gave their informed
tional review board (IRB), the child welfare agency’s IRB, consent, a 100% response rate. Demographics of the
and both the city- and state-level child welfare authority. sample can be found in Table 1.

Sample Data Collection

Participants in the first, qualitative phase of the In the qualitative research phase, informed consent
research project consisted of 36 persons, all from one of was obtained and interviews were conducted in loca-
the neighborhood-based family service centers operat- tions and in the language of the participants’ choosing.
ing under the auspices of a large New York City–based A qualitative interview guide was developed and used
child welfare agency. The 16 parent and 7 foster parent by the research team, covering the following domains:
participants were a purposive sample, suggested by conceptualization of engagement, its attributes, differ-
agency workers and administrators for their willingness ential aspects, development, meaning, related factors,
to share their experiences and perspectives of positive ways to promote, and perceived benefits of engagement.
engagement in neighborhood-based child welfare ser- Each participant was interviewed at least once, many
vices. The entire staff of nine workers and four supervi- twice, and all received $20 stipends for their participa-
sors from the same center were asked and agreed to tion. Each interview lasted between 1 and 3 hours, and
participate. was tape recorded.
Participants in the second, quantitative phase of the Once informed consent was obtained from subjects in
research project came from a different sample of work- the quantitative phase of the study, interviews were sched-
ers and parents from three other neighborhood service uled to collect data in a language and at a site and time
centers of the aforementioned child welfare agency. In chosen by subjects. Most were completed at the agency
this rolling, yearlong data collection phase, every parent site, and took between 1 and 2 hours. Data collection from
558 RESEARCH ON SOCIAL WORK PRACTICE

parent-worker dyads began as soon as possible after variables in this quantitative phase of the study: hope-
placement of the target parent’s child(ren), but no more fulness and satisfaction. Parent and worker hopefulness
than 3 months post placement, and every 3 months was measured on a 5-point Likert-type scale, asking,
thereafter for 9 months. Worker changes, inability to “How hopeful are you that you and your child/ren (the
find parents, and/or case reunification led to only 42 child and their birth family) will be reunited?” Responses
dyads interviewed at Time 2, and 26 dyads at Time 3. At categories ranged from 1 (not very hopeful) to 5 (very
each time point, data from each member of the dyad hopeful). Regarding satisfaction, parents were asked,
were gathered within a week of the other. Participants on a scale from 1 (very unsatisfied) to 5 (very satisfied),
received $20 stipends for each interview. how satisfied they were with their current visitation
plan, their current agency worker, and the agency gen-
Instrumentation erally. Workers were asked, using the same scale, how
satisfied they were with their work with this family cur-
Data collected in the second, quantitative phase of the rently, and with their supervision and work at the agency
study were based on that which was discovered regarding currently, and how supported they feel in the work they
the conceptualization and processes of engagement in do at the agency.
neighborhood-based child welfare services from the first, After the interviews, data were collected from case
qualitative phase of the study. Engagement was viewed as records on demographics, case risk, casework contact,
the degree to which client and workers were invested in visitation rate, and case disposition. Case risk was taken
collaboratively working together to address the needs that from the Uniform Case Record (UCR) case risk rating
led the client family to the agency. Because this way of form, where 1 was low and 5 was high. The casework con-
viewing engagement was syntonic, conceptually, with the tact rate was calculated using a ratio of UCR-documented
idea of the working alliance (Bordin, 1979), and because home visits and phone calls and in-person office appoint-
the working alliance construct has been used increasingly ments that workers made to parents divided by weeks for
in child welfare research (Alexander, Littell, Girvin, & the interval period reported. Visitation rate was calculated
Reynolds, 2001), engagement in the quantitative part of using a simple ratio of completed parent-child visits, as
the study was measured using worker and client versions documented in the UCR, divided by weeks for the inter-
of the Working Alliance Inventory (WAI), adapted for val period reported. Case disposition was taken from the
child welfare use. latest UCR in the casefile. It was a nominal-level vari-
An alliance between client and worker that influences able in which the child was (a) returned home, (b) still in
the process of change has long been assumed an essential care with a primary goal of return home, or (c) still in care
component of any social work intervention. The thera- with a primary goal of adoption.
peutic, helping, or working alliance in psychotherapy
(Bordin, 1979; Frank, 1991; Luborsky, 1976) has been Data Analysis
used synonymously with engagement in social work
(Poulin & Young, 1997; Pritchard, Cotton, Bowen, & Analysis of the qualitative data was completed accord-
Williams, 1998) and is thought to be a key process vari- ing to Spradley’s (1979) developmental research sequence
able in child welfare research (Dore & Alexander, 1996). (DRS). Members of the research team read interview tran-
The WAI is a 36-item inventory with three subscales scripts and field notes independently, then met for joint
(Goal, Bond, and Task) and good psychometric proper- coding and consensual data analysis. To start, data were
ties (Horvath & Greenberg, 1994). Sixty-seven internal analyzed by four system levels (parent, worker, agency,
consistency estimates, six interrater reliability estimates, and the larger system) and by participant category (parent,
and four study characteristics were recently analyzed; reli- foster parent, worker, and supervisor). According to the
ability estimates of WAI scale scores appear to be robust. DRS, three levels of analyses then took place.
Mean reliability estimates ranged, in this sample of stud- First, a variety of semantic relationships were devel-
ies, from .79 to .97, with a modal estimate of .92 (Hanson, oped and fitted from coded data in order to produce initial
Curry, & Bandalos, 2002). WAI scores range from 36 domains (e.g., all codes that fit the semantic relationship
(weak alliance) to 252 (strong). “X is a cause of good engagement” comprised one ana-
Two other related constructs, both found to be influ- lytic domain). Taxonomic analysis of each domain was
ential to the processes of engagement in the first phase done next (e.g., in the parent participant domain, “causes
of the study and, increasingly, across the research liter- for good engagement”; and at the worker system level,
ature (Chapman, Gibbons, Barth, & McCrae, 2003; Lee four subcategories emerged: interpersonal skills, case-
& Ayon, 2004; Mahoney, 2002), were also measured as related interventive skills, personality, and attribute).
 Altman / NEIGHBORHOOD-BASED CHILD WELFARE SERVICES 559

Taxonomies of each of these subcategories were then Parents asked that workers help in setting very clear
completed. Next, organized tables of data (matrices) directions to get where they needed for reunification.
were developed from these, again according to both Agency staff and foster parents suggested that participa-
system level and participant category. Themes in the tion in goal setting helped empower parents:
data were then identified through the comparing and
contrasting of taxonomies, overlaying of the matrices, Worker: For them to accept the service plan, it’s very, very, very
important that they participate in formulating the goals and
and writing of analytic memos.
the service plan. Their participation in that is very, very, very
Quantitative data collected from the interviews and important, and it helps them, you know, to get some sense of
case records of each parent-worker dyad were entered control over the whole situation.
into SPSS for analysis. Univariate and bivariate analyses
were conducted in order to generate frequencies and Hopefulness
measures of central tendency and dispersion for all rel- Workers considered hopefulness an important moti-
evant variables. Pearson r correlations were calculated vating force; parents saw it as affirmation for their
to determine relationships between groups and selected capacity as caregivers. All saw it as an expectation that
variables, and t tests were calculated to measure differ- future experiences with each other, and with the change
ences between selected groups and variables. process, would be positive and beneficial:

Supervisor: [You need to] send that message . . . that there is

FINDINGS hope regardless of what happens, you’re with us and we want
to do something, we want to help you, we want to work with
Qualitative Findings you to get you to where you want to be.

Analysis of the qualitative data pointed to a common Acknowledgment

conceptual understanding of engagement among partic- If engagement in services is to occur, some level of
ipants. In the context of this neighborhood-based child acknowledgment on the part of the parents that they are
welfare agency, engagement was viewed as the degree aware of and responsible for the situation they are in helps.
to which clients and workers are invested in collabora- Both foster parents and workers found it much easier to
tively working together to address the needs that led the engage with a client who had some capacity for insight and
client family to the agency. In this way, engagement is who confessed some responsibility for their plight, and
seen as a dynamic process that involves the establish- parents agreed:
ment of a helping relationship between worker and
client so that active work toward change can begin, a Parent: So it’s like wake up and smell the coffee; you did some-
conceptualization similar to the notion of a working thing wrong, now make it right.
relationship (Horvath & Greenberg, 1994). Motivation
This process of engagement in neighborhood-based
child welfare services was further found to be influenced Motivation as a force for engagement in services was
by seven main themes, as revealed in the qualitative data viewed partly as compliance, partly as true desire for reuni-
common across all participant groups: (a) the need for fication. Both parents and workers found it a source of
families and workers to set common and clear goals energy to propel their work together, and related it to the
together; (b) the need for all involved to maintain a sense earlier theme of hopefulness. Participants agreed motiva-
of hopefulness during the change process; (c) the need for tion was central to the perseverance needed to engage
parents to be aware of, acknowledge, and understand their together in the work toward the end goal. When parents
situations accurately; (d) the need for all parties to be con- become unmotivated, difficulties emerged:
sistently motivated in their change efforts; (e) the need for
Foster parent: I think she got comfortable knowing that my
workers to identify, understand, and respect cultural differ-
husband and I were taking very good care of him, and she
ences in their relationships with families; (f) the need for kind of slipped . . . she just got into that comfortable mode.
truthful, honest, and respectful communication; and (g) the
need for persistent, diligent, and timely work by all parties. Respect for Cultural Differences
The importance of workers demonstrating cultural
Clear Collaborative Goals
awareness, respect, and understanding was a prominent
All stakeholders communicated the importance of theme. Parents admitted that cultural similarity to the
collaborative goal setting in effective engagement. worker and foster parent also helped their relationship:
560 RESEARCH ON SOCIAL WORK PRACTICE

TABLE 2: Parent-Worker Working Alliance Inventory (WAI) Scores Over Time

WAI Total Score and Subscale Time 1: M (SD) Time 2: M (SD) Time 3: M (SD)

Average WAI—parent 187.92 (54.39) 185.16 (56.08) 187.00 (61.35)

Bond subscale 60.4 (15.6) 59.1 (17.0) 61.1 (18.0)
Task subscale 65.2 (16.6) 64.4 (17.5) 65.7 (20.0)
Goal subscale 61.6 (18.9) 62.0 (17.9) 62.5 (20.0)
Average WAI—worker 195.39 (42.04) 195.12 (37.11) 207.81**
(29.67)
Bond subscale 69.1 (10.1) 69.6 (9.8) 68.0 (9.3)
Task subscale 66.9 (12.2) 65.3 (18.5) 68.9 (10.8)
Goal subscale 65.1 (13.9) 63.0 (13.7) 67.8 (12.7)

Parent: Because an Asian worker, where they come from or how they more positive working alliances than did parents (see
live is totally different from how I live. . . . They may say no it’s Table 2). Worker and parent scores were significantly cor-
not, because she breathes and walks the same. No, she does not.
related at Time 2 (r = .494, p < .01). Workers’WAI scores
She may not watch the same shows we watch . . . like a black
worker that I may have, we can sit down and have stuff in common. increased significantly between Time 2 and Time 3 (t =
We don’t have nothing in common. Nothing, nothing, nothing. –2.76, df = 19, p = .01). Of the three subscales, Bond,
Task and Goal, parents scored, on average, highest on the
Communication Task subscale. Workers scored, on average, highest on the
The importance of good communication at all system Bond subscale. WAI scores did not vary significantly
levels was seen as key to effective engagement. Although, across race or gender. They did vary significantly accord-
however, parents asked for assertive, honest, clear, even ing to worker’s parental status. Client parents rated their
urgent messages about what they should be doing to ensure initial alliance significantly more positively when their
their children’s return, workers reported a hesitancy to do so: workers were parents (t = 2.15, df = 67, p = .03).

Parent: They [the workers] don’t get to the point. They don’t tell Satisfaction
you what you have to do right at the first meeting. It’d be the Only about half of the parents reported being satisfied,
second meeting, and then they tell you. . . . They don’t be
straightforward with you.
somewhat satisfied, or very satisfied with the visitation plan
Worker: Personally, I feel a little afraid of confrontation and say- they agreed to with the agency at Time 1; that grew by
ing anything negative. . . . So sometimes I kind of avoid it, you Time 3, when more were getting unsupervised home visits
know, I kind of waited to confront her, you know, so I think with their children, to 63.7%. Most parents reported being
that’s something that I need to deal with. satisfied, somewhat satisfied, or very satisfied with their
workers at all three time points (73% at Time 1, 76.2% at
Persistence and Timeliness
Time 2, and 75% at Time 3). In terms of overall satisfac-
All believed persistence and timeliness to be key ingre- tion with the agency, 64.3% of parents reported being sat-
dients of effective engagement in child welfare services. isfied, somewhat satisfied, or very satisfied at Time 1, and
Stakeholders felt that neither parents, nor workers, nor the this varied little over time (to 66.6% and then 73.1%). Most
institutions involved should lose the importance of urgency workers reported being satisfied, somewhat satisfied, or
and determination when striving toward permanence for very satisfied with the work they were doing with the index
children in out-of-home care. family at Time 1 (71.9%), and this impression grew over
time (to 87.8% and then 92.5%). They also reported satis-
Parent: You know and she don’t seem to get the urgency. She faction with their work at the agency (85.9%, 80.5%, and
don’t feel the urgency I feel. I need to do this now. I don’t have
85.2%) and with their supervision (80%, 80.5%, and
time for you to wait for these people to get back to you. She
keep telling me they didn’t get back to me. What kind of shit 85.1%). Most reported feeling quite supported in their
is that? What kind of shit is that? work across time (78.9%, 71.4%, and 81.4%) (see Table 3).

Hopefulness
Quantitative Findings
Parents consistently ranked themselves far more
Working Alliance
hopeful that their child(ren) would be returned to their
Working alliance inventory (WAI) scores were quite care than did their workers. At Time 1, 94.5% of parents
high overall, and although not significantly different at any reported being hopeful, mostly hopeful, or very hopeful
of the three time points, workers consistently reported that they would be reunited with their children, whereas
 Altman / NEIGHBORHOOD-BASED CHILD WELFARE SERVICES 561

TABLE 3: Parent-Worker Satisfaction Over Time

Satisfaction Time 1: %,a M, and (SD) Time 2: %, M, and (SD) Time 3: %, M, and
(SD)

Parent
With visitation plan 49.3 2.62 (1.48) 42.9 2.57 (1.54) 63.7 3.05 (1.46)
With worker 73.0 3.43 (1.44) 76.2 3.42 (1.39) 75.0 3.29 (1.37)
With agency 64.3 2.97 (1.23) 66.6 3.21 (1.44) 73.1 3.19 (1.39)
Worker
With family’s casework 72.4 3.33 (1.2) 93.5 3.58 (1.04) 92.6 3.85 (.99)
With agency 85.9 3.45 (.87) 80.5 3.19 (1.01) 85.2 3.19 (.68)
With supervision 80.0 3.39 (1.09) 80.5 3.27 (1.28) 85.2 3.26 (.94)
With level of felt support 78.9 3.54 (1.06) 71.5 3.21 (1.24) 81.5 3.41 (1.04)

TABLE 4: Parent-Worker Hopefulness Over Time first 6 months of each case, one contact with parents every
Time 1: %, a
Time 2: %, Time 3: %, 3 weeks. Casework contact was significantly negatively
Hopefulness M, and (SD) M, and (SD) M, and (SD) related to worker hopefulness regarding reunification at
Time 1 (r = –.319, p < .05) and Time 2 (r = –.493, p <
Parent 94.5 4.66 (.87) 95.1 4.68 (.93) 95.2 4.67
(.97) .05), and worker satisfaction with supervision at Time1
Worker 74.9 3.69 (1.54) 81.0 3.64 (1.3) 88.9 3.93 (r = –.389, p < .01) and Time 2 (r = –.460, p < .01).
(1.3) Casework contact was not related to WAI scores for
a. Percentage reporting hopeful, somewhat hopeful, or very hopeful. parents or workers at any time point.
only 74.9% of workers did. Parents’ hopefulness about Visitation rate averaged one visit per week for the
the ultimate case outcome did not diminish over time; first 6 months of the case. Visitation rate was not related
workers’ hopefulness that families would be reunited to WAI scores of parents or workers at any time point.
increased over time (see Table 4). Visitation rate was also not related to case disposition.
At no time point were WAI scores related to case dis-
WAI, Case Risk, Satisfaction, and Hopefulness position. Neither was visitation rate related to case dispo-
sition. Case risk rating (r = .247, p < .01) and casework
At Time 1, WAI scores of parents were significantly
contact with parents (r = –.274, p < .05) were the only
lower as the family risk rating was higher (r = –.343, p <
two variables related to case disposition.
.01), but not at Time 2 or Time 3. Parent WAI scores were
significantly related to their self-reported satisfaction with
Limitations
their visitation plan at Time 1 (r = .329, p < .01) and
Time 2 (r = .473, p < .01); with their worker at Time 1 These findings are limited by a number of factors. First,
(r = .520, p < .01), Time 2 (r = .581, p < .01), and Time although the decision to limit this investigation to a single
3 (r = .667, p < .01); and with the agency at Time 1 (r = neighborhood-based child welfare agency increased the
.4, p < .01) and Time 3 (r = .523, p < .01). Parents’ WAI trustworthiness of the qualitative portion of the mixed-
scores at Time 2 were also related to their hopefulness method study, it could be seen as limiting the generaliz-
regarding future reunification (r = .473, p < .01). ability of the quantitative findings. Furthermore, given the
WAI scores for workers were significantly associated longitudinal nature of the quantitative design, attrition
with their satisfaction with the family’s casework at of the nested worker-client pairs limited the capacity to
Time 1 (r = .527, p < .01), Time 2 (r = .618, p < .01), analyze the paired data. Had a larger sample allowed,
and Time 3 (r = .550, p < .01), and their hopefulness sophisticated hierarchical models could have yielded a
that reunification would occur for this family at Time 1 more complete understanding of the relationship between
(r = .345, p < .01) and Time 2 (r = .523, p < .01). engagement and key outcomes.
Although data collection took place using the first
Casework Contact, WAI, Visitation, and Case Disposition
language of participants, some loss of meaning in the
Approximately 30% (n = 22) of families were reunited translation from Spanish to English could have affected
by the end of the study. Eight percent (n = 6) moved toward the qualitative data, and although the WAI has estab-
permanency, with the goal changed to adoption. The lished psychometric properties, those were established
majority of the cases (62%, n = 46) were continuing their only for English versions of the instrument.
work toward reunification at the study’s end. Person-to- The use of administrative data for key child welfare
person casework contact with parents averaged, for the outcomes might also have affected the results of this
562 RESEARCH ON SOCIAL WORK PRACTICE

study. As with most secondary data, these data might Workers must also listen carefully to the parent’s story
include certain misinformation or nonsystematic record- and acknowledge, where appropriate, factors contribut-
keeping errors. ing to the child’s placement for which the parent is not
responsible.
Establishing a relationship with clients cross-culturally
DISCUSSION AND APPLICATIONS requires careful attention to traditional notions and impacts
TO PRACTICE of empathy, mutuality, power and authority, use of self,
and communication (Abney, 2000; Shonfeld-Ringel,
This study yields some knowledge about the complex- 2001). These findings and others (McCurdy, Gannon, &
ity involved in engaging nonvoluntary clients in working Daro, 2003) support the need for cultural awareness and
relationships quickly and at times of crisis. How important competence on the part of workers to be taken seriously
attempting to form these relationships is in child welfare at the agency level, especially given that no group of
services remains unclear. Data from this study failed to children in the United States is growing faster than
support the premise that good working relationships children in immigrant families (Brandon, 2002).
between parents and workers lead to better visitation or Helping workers increase their capacity for persistent
reunification rates, at least for the shorter term. and timely intervention needs to be supported at a
Even so, a number of strategies regarding how to pro- number of levels. Central to this is enhanced capacity
mote engagement in neighborhood-based child welfare for workers to offer truthful, honest, respectful, coher-
services were uncovered in this research. The impor- ent, and strong messages to clients. Parents in this study
tance of worker-parent collaboration in the setting of wanted a more assertive approach from their workers,
goals should not be minimized, especially with clients and expect timely action on their worker’s part in ful-
who may be seen as hostile or indifferent. Starting filling responsibilities. Workers could be helped to
where that client is and setting realistic, workable, flex- deliver it through enhanced supervision, increased train-
ible goals are considered vital in the efforts made to ing, and/or opportunities for case-based clinical consul-
engage vulnerable families. Clients expect workers to tation (Strand & Badger, 2005).
also have goals for which they are accountable. This In this study, casework contact was related to case dis-
makes the partnership a reality rather than an empty bit position. Further dialogue around who the primary client
of rhetoric (Alexander & Dore, 1999). is in casework practice in child welfare is clearly needed.
The importance of keeping a positive, hopeful outlook As one foster parent said, as she was describing the diffi-
toward the future of their work together was seen in this culties parents often encounter in working to reunite with
study to be a valuable link toward effective engagement and their children, “But who works on the birth mother’s
positive case disposition. Helping workers build skills for behalf? I wonder if there’s someone to help them.”
motivating a client to change continues, however, to remain We remain uncertain as to the importance of working
a formidable challenge (Pecora, 1989). The power of hope in all these ways to enhance the engagement process in
and positive expectancy is seldom harnessed for use in child welfare, a finding shared in recent work by others
social work intervention, yet empirical findings increasingly (Barber & Delfabbro, 2004). In our study, parent-worker
point to its efficacy (Mahoney, 2002). This study adds to the engagement, as measured by the Working Alliance
evidence that suggests we should understand its role better. Inventory, was not significantly related to important
Of particular note is how much more hopeful parents were case outcomes such as rate of parent-child visitation or
than workers that reunification would occur. Capitalizing on reunification. This finding may be due to several limita-
the early hopefulness parents express as a strengths-based tions of the study itself: a small sample size, and the
intervention is worthy of further study. potential that the WAI was not a valid and reliable mea-
Consensus around the need for clients to realistically sure of parent-worker engagement. Conceptually, the
appraise and acknowledge their current situations idea that engagement is equivalent to a working rela-
should also be further examined. Workers need to clar- tionship may be a faulty one. Or working relationships in
ify to what extent it is their need for the parents to con- child welfare practice may not share the same conceptual
fess to their poor parenting practices, and to what extent validity as working relationships in voluntary practice con-
it truly is part of a clear and accurate attempt to help texts when measured by the WAI. Replicating the research
parents understand the dynamics that led to their child’s presented here using other ways of measuring engage-
removal. Workers should see this as an essential part of ment should be considered.
the work they do with clients, not as a test that must be It also may be that parent-child visitation is not the
passed prior to the beginning of real work together. durable factor it once was as an intermediate case outcome
 Altman / NEIGHBORHOOD-BASED CHILD WELFARE SERVICES 563

or predictor of reunification (Fanshel, 1975), although randomly assigned but similar child welfare parent pop-
recent evidence of an inclusive practice model suggests ulations would be a significant next research step.
otherwise (Leathers, 2002). Agency policy now exists to Work on developing a theory of engagement in child
support visitation as an institutionalized part of case welfare services is also needed, and some significant
practice. Years ago, it may have served more as a proxy progress to that end is being made (Daro et al., 2005).
for parental motivation. Similarly, reunification today Findings exist that engagement may be related to
may be more likely the result of factors outside the case- clients’ past and current experience in services, their
work relationship or agency practice, such as comple- personal networks, or their readiness to change (Daro
tion of an independent drug treatment program. It et al., 2003). From qualitative inquiries, factors such as
should also be noted that only half the parents were sat- perceived choice, analysis of cost-benefit, their under-
isfied with their visitation plan early in their case, standing of their own needs, and the potential fit and
despite being satisfied with their worker and the agency. effectiveness of the services offered were also found to
Much more work in understanding the quality and affect a client’s engagement (Altman, 2003). Further
impact of visitations can and should be done. research that embraces data from both quantitative and
The import of a relationship between parents in the qualitative perspectives should help in the comprehen-
child welfare system and their agency workers should sive development of a theory.
perhaps be reconsidered. Many of the parents who enter Although we look forward to larger systemic changes
the child welfare system are individuals for whom the that have the potential to alter case practice (Smith &
capacity to engage in healthy relationships may be Donovan, 2003), we must work to understand what can
severely limited. Furthermore, they often enter child wel- be done at the other levels to facilitate permanence for
fare systems in which they may be asked to form rela- children currently in the public child welfare system.
tionships with many workers—the protective services These findings lend some direction to that work.
worker who may have worked with them prior to placing
the child(ren) in care, a foster care worker whose primary
responsibility is ongoing case management and perma- REFERENCES
nency planning, mental health or substance abuse coun-
selors who may be working with the parent(s) on some
Abney, V. D. (2000). What principles and approaches can I use to engage
specific therapeutic goals, and perhaps others. Asking clients across cultures? In H. Dubowitz & D. DePanfilis (Eds.),
anyone to engage with that many helping professionals is Handbook for child protection practice. Thousand Oaks, CA: Sage.
optimistic; for these families, it may be overwhelming or Alexander, L. B., & Dore, M. M. (1999). Making the parents as
impossible. If, as this study suggests, having a good partners principle a reality: The role of the alliance. Journal of
working relationship with their foster care worker does Child and Family Studies, 8, 255-270.
Alexander, L. B., Littell, J. H., Girvin, H., & Reynolds, W. (2001,
little to ensure earlier reunification with their children, the January). The predictive capacity of the working alliance in child
import placed on this as the primary helping relationship welfare: Results of a longitudinal investigation of home-based
may be unfounded. Research that teases out the relative services. Paper presented at the Society for Social Work and
importance of the numerous workers involved in the lives Research annual meeting. Atlanta, GA.
of families in the child welfare system would be a logical Altman, J. C. (2003). A qualitative examination of client participation
in agency-initiated services. Families in Society, 84, 471-479.
next step. An exploratory investigation using qualitative
Altman, J. C. (2005). Engagement in children, youth, and family
methods could help in understanding the presence and services: Current research and best practices. In G. Mallon &
significant effects of these relationships from at least the P. Hess (Eds.), Child welfare for the twenty-first century: A hand-
parent’s perspective. book of practices, policies and programs (pp. 72-86). New York:
For clients not asking for services in a cooperative Columbia University Press.
arrangement, a different model of client engagement Atkinson, L., & Butler, S. (1996). Court-ordered assessment: Impact
of maternal noncompliance in child maltreatment cases. Child
and problem solving may be needed. Increasingly, Abuse and Neglect, 20, 185-190.
parental views are being taken into consideration in the Barber, J. G., & Delfabbro, P. H. (2004). Children in foster care.
contemplation of new interventions (Chapman et al., London: Routledge.
2003; Ribner & Knei-Paz, 2002). A practice model is Blythe, B., Ivanoff, A., & Tripodi, T. (1994). Involuntary clients in
needed that is perhaps more direct and expeditious, and social work practice. New York: Aldine de Gruyter.
Bordin, E. S. (1979). The generalizability of the psychoanalytic
that sees the parent of the child in placement as the
concept of the working alliance. Psychotherapy, 16, 252-269.
primary client to engage in change efforts. Experimental Brandon, P. (2002). The childcare arrangements of preschool children in
studies comparing the effectiveness of model interven- immigrant families (Foundation for Child Development Working
tion programs to a services-as-usual control group of Paper Series). New York: Foundation for Child Development.
564 RESEARCH ON SOCIAL WORK PRACTICE

Chapman, M., Gibbons, C., Barth, R., & McCrae, J. (2003). Parental Luborsky, L. (1976). Helping alliances in psychotherapy. In
views of in-home services: What predicts satisfaction with child J. Claghorn (Ed.), Successful psychotherapy (pp. 92-116). New
welfare workers? Child Welfare, 82, 571-596. York: Brunner/Mazel.
Compton, B., & Galaway, B. (1994). Social work processes. Pacific Mahoney, M. J. (2002). Constructivism and positive psychology. In
Grove, CA: Brooks/Cole. C. R. Snyder & S. J. Lopez (Eds.), Handbook of positive psy-
Cyphers, G. (2001). Report from the child welfare workforce survey: chology (pp. 745-750). New York: Oxford University Press.
State and county data and findings. Washington, DC: American McCurdy, K., Gannon, R. A., & Daro, D. (2003). Participation pat-
Public Human Services Association. terns in home-based family support programs: Ethnic variations.
Daro, D., McCurdy, K, Falconnier, L., & Stojanovic, D. (2003). Family Relations, 51, 3-111.
Sustaining new parents in home visitation services: Key participant Okamoto, S. K. (2001). Individual and agency factors related to
and program factors. Child Abuse and Neglect, 27, 1101-1125. engagement in a parent training program. Journal of Family
Daro, D., McCurdy, K., & Nelson, C. (2005). Engagement and reten- Social Work, 5, 39-50.
tion in voluntary new parent support programs: Final report. Pecora, P. J. (1989). Improving the quality of child welfare services:
Chicago: Chapin Hall. Needs assessment for staff training. Child Welfare, 68, 403-419.
Davis, I., Landsverk, J., Newton, R., & Granger, W. (1996). Parental Petras, D., Massat, C., & Essex, E. (2002). Overcoming hopeless-
visiting and foster care reunification. Children and Youth Services ness and social isolation: The ENGAGE model for working
Review, 18, 363-382. with neglecting families toward permanence. Child Welfare, 81,
Dawson, K., & Berry, M. (2002). Engaging families in child welfare 225-248.
services: An evidence-based approach to best practice. Child Poulin, J., & Young, T. (1997). Development of a Helping
Welfare, 81, 293-317. Relationship Inventory for social work practice. Research on
Dore, M., & Alexander, L. (1996). Preserving families at risk of Social Work Practice, 7(4), 463-489.
child abuse and neglect: The role of the helping alliance. Child Prinz, R. J., & Miller, G. E. (1996). Parental engagement in inter-
Abuse and Neglect, 20, 349-361. ventions for children at risk for conduct disorder. In R. Peters &
Fanshel, D. (1975). Parental visiting of children in foster care: Key R. McMahone (Eds.), Preventing childhood disorders, substance
to discharge? Social Service Review, 49, 493-514. abuse and delinquency (pp. 161-183). Thousand Oaks, CA: Sage.
Frank, A. F. (1991). The therapeutic alliances of borderline patients. Pritchard, C., Cotton, A., Bowen, D., & Williams, R. (1998). A con-
In J. Clarkin, E. Marziali, & H. Monroe-Blum (Eds.), Borderline sumer study of young people’s views on their educational social
personality disorders (pp. 220-247). New York: Guilford. worker: Engagement as a measure of an effective relationship.
Hanson, W. E., Curry, K. T., & Bandalos, D. L. (2002). Reliability British Journal of Social Work, 28, 915-938.
generalization of working alliance inventory scale scores. Ribner, D. S., & Knei-Paz, C. (2002). Client’s view of a successful
Educational and Psychological Measurement, 62, 669-673. helping relationship. Social Work, 47, 379-387.
Hepworth, D. H., Rooney, R., Dewberry-Rooney, G., Strom- Rooney, R. (1992). Strategies for work with involuntary clients. New
Gottfried, K., & Larsen, J. (2007). Direct social work practice: York: Columbia University Press.
Theory and skills. Pacific Grove, CA: Brooks/Cole. Rooney, R. (2000). How can I use authority effectively and engage
Horvath, A., & Greenberg, L. (1994). The working alliance: Theory, family members? In H. Dubowitz & D. DePanfilis (Eds.),
research and practice. New York: John Wiley. Handbook for child protection practice. Thousand Oaks, CA:
Jones, L. (1993). Decision-making in child welfare: A critical review Sage.
of the literature. Child and Adolescent Social Work, 10, 241-262. Shonfeld-Ringel, S. (2001). A reconceptualization of the working
Kadushin, A., & Kadushin, G. (1997). The social work interview: A alliance in cross cultural practice with non-western clients:
guide for human service professionals. New York: Columbia Integrating relational perspectives and multicultural theories.
University Press. Clinical Social Work Journal, 29, 53-63.
Karski, R. L. (1999). Key decisions in child protective services: Smith, B., & Donovan, S. (2003). Child welfare practice in organi-
Report investigation and court referral. Children and Youth zational and institutional contexts. Social Service Review, 77,
Services Review, 21, 643-656. 541-563.
Leathers, S. (2002). Parental visiting and family reunification: Could Smithgall, C. (2006, January). Assessing worker-client relationship
inclusive practice make a difference? Child Welfare, 81, 595-616. in child welfare and family preservation services. Presentation at
Lee, C. D., & Ayon, C. (2004). Is the client-worker relationship asso- the Society for Social Work and Research annual meeting. San
ciated with better outcomes in mandated child abuse cases? Antonio, TX.
Research on Social Work Practice, 14, 351-357. Special Child Welfare Advisory Panel, Administration for Children’s
Littell, J. (1997). Effects of the duration, intensity, and breadth of Services. (2000). Advisory report on front line and supervisory
family preservation services: A new analysis of data from the practice. New York: Author.
Illinois Family First experiment. Children and Youth Services Spradley, J. (1979). The ethnographic interview. New York: Holt,
Review, 19(1/2), 17-39. Rinehart & Winston.
Littell, J. (2001). Client participation and outcomes of intensive Strand, V., & Badger, L. (2005). Professionalizing child welfare: An
family preservation services. Social Work Research, 25, 103-114. evaluation of a clinical consultation model for supervisors.
Littell, J., Alexander, L. B., & Reynolds, W. (2001). Client partici- Children and Youth Services Review, 27, 865-880.
pation: Central and underinvestigated elements of intervention. Woods, M. E., & Hollis, F. (2000). Casework: A psychosocial ther-
Social Service Review, 75, 1-28. apy (5th ed.). Boston: McGraw-Hill.
Littell, J., & Tajima, E. (2000). A multilevel model of client partici- Yatchmenoff, D. (2005). Measuring client engagement from the
pation in intensive family preservation services. Social Service client’s perspective in nonvoluntary child protective services.
Review, 74, 405-435. Research on Social Work Practice, 15, 84-96.