Content uploaded by Hilary F Byrnes
Author content
All content in this area was uploaded by Hilary F Byrnes
Content may be subject to copyright.
The Relationship between Neighborhood Characteristics and
Recruitment into Adolescent Family-Based Substance Use
Prevention Programs
Hilary F. Byrnes, Ph.D.,
Prevention Research Center, Pacific Institute for Research and Evaluation, Berkeley, CA
Brenda A. Miller, Ph.D.,
Prevention Research Center, Pacific Institute for Research and Evaluation, Berkeley, CA
Annette E. Aalborg, DrPH, and
Kaiser Permanente, Division of Research, Oakland, CA
Carolyn D. Keagy, MA
Prevention Research Center, Pacific Institute for Research and Evaluation, Berkeley, CA
Abstract
Youth in disadvantaged neighborhoods are at risk for poor health outcomes. Characteristics of
these neighborhoods may translate into intensified risk due to barriers utilizing preventive care,
such as substance use prevention programs. While family-level risks affect recruitment into
prevention programs, few studies have addressed the influence of neighborhood risks. This study
consists of 744 families with an 11-12 year old child recruited for a family-based substance use
prevention program. Using U.S. Census data, logistic regressions showed neighborhoods were
related to recruitment, beyond individual characteristics. Greater neighborhood unemployment
was related to decreased agreement to participate in the study and lower rates of high school
graduation were related to lower levels of actual enrollment. Conversely, higher rates of single-
female headed households were related to increased agreement. Recruitment procedures may need
to recognize the variety of barriers and enabling forces within the neighborhood in developing
different strategies for the recruitment of youth and their families.
Keywords
neighborhoods; prevention program recruitment; adolescent substance use; health care utilization
Beyond the effects of individual background characteristics, certain neighborhood
characteristics can place adolescents at increased risk for poorer health outcomes, such as
delinquency and substance use.1-4 Some neighborhood characteristics may make access to
health care difficult, intensifying risk for problem behaviors. Neighborhood characteristics
may impact not only utilization of treatment services, but also for prevention of disease.
Hilary F. Byrnes, Ph.D., Associate Research Scientist, Prevention Research Center, Pacific Institute for Research and Evaluation,
1995 University Ave., Suite 450, Berkeley, CA 94704; phone: 510-708-2215; fax: 510-644-0594; email: hbyrnes@prev.org.
Brenda A. Miller, Ph.D., Senior Research Scientist, Prevention Research Center, Pacific Institute for Research and Evaluation, 1995
University Ave., Suite 450, Berkeley, CA 94704; phone: 510-883-5768; fax: 510-644-0594; email: bmiller@prev.org
Annette E. Aalborg, DrPH, Research Scientist, Kaiser Permanente, Division of Research, 2000 Broadway, Oakland, CA 94612;
phone: 510-891-3498; email: Annette.E.Aalborg@nsmtp.kp.org
Carolyn D. Keagy, MA, Sociology Ph.D. Student, University of California, San Francisco, Department of Social and Behavioral
Sciences, 3333 California St., #LHts-455, San Francisco, CA 94118; phone: 415-713-8011; email: Carolyn.Keagy@UCSF.edu
Conflict of Interest Statement: The authors do not have any conflicts of interest.
NIH Public Access
Author Manuscript
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
Published in final edited form as:
J Behav Health Serv Res
. 2012 April ; 39(2): 174–189. doi:10.1007/s11414-011-9260-0.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Ability to utilize prevention services is critical because of the potential to decrease costs for
individuals and communities from treatment and delinquent activity.
There are several plausible explanations for the association between neighborhood
characteristics and poor youth outcomes. According to theoretical models, social
disorganization, a characteristic of the neighborhood context, makes it difficult for residents
to control their environments.5 Both physical (e.g., abandoned buildings, vandalism) and
social (e.g., public drinking, unsupervised children) characteristics of neighborhoods have
previously been used to indicate neighborhood disorganization.6 Census variables linked to
families' addresses through geocoding are also often used to indicate disorganization,
because demographic characteristics like ethnic heterogeneity may interfere with residents'
abilities to create stable social networks and impose shared norms in their community.7
Disorganization can disrupt healthy behavior for both adults and youth in the neighborhood.
In contrast, neighborhoods with high social organization are more likely to have stronger
social ties and informal social controls, which help spread and healthy behavior among
residents.8,9
Disorganization may lead to negative adolescent outcomes by interfering with neighbors'
ability to develop stable social networks and to enforce shared values,7 as well as higher
crime rates and lack of positive role models.10 Further, Sampson, et al.8 suggest that the
potential for shared norm enforcement is less available to residents through lower levels of
social capital in such neighborhoods. For example, neighborhoods with greater concentrated
disadvantage had less child-centered social control,8 which puts more of a burden on
individual parents.
High levels of disorganization and low levels of social capital in a neighborhood could
intensify the risk of adolescent problem behaviors by making health care utilization
particularly difficult. Disorganization may create barriers through lowered availability of
medical facilities, less knowledge or means to use available services,11,12 and stressors that
may require families to only deal with immediate needs rather than less urgent needs, such
as prevention. Low social capital in these neighborhoods could also affect utilization of
preventive services, as residents would not be able to use social ties and shared norms
available through social capital to spread information and values for healthy behavior and
use of available services. This may be intensified for low-income families who are already
facing greater difficulties in utilizing health care.
Neighborhood Characteristics and Utilization of Treatment Health Care
Services
Prior neighborhood studies have focused on utilization of treatment services, or medical
services in general, rather than specifically on preventive services. However, these studies
are informative for understanding how neighborhood characteristics may pose particular
barriers to utilizing preventive health care. Residents of disorganized neighborhoods are less
likely to use treatment and medical services,13 independent of residents' own demographic
characteristics. Specifically, areas with higher rates of unemployment are related to less
health care utilization, while areas with higher income and more health centers are related to
increased utilization.14
There are several barriers to utilization of health care services in disorganized
neighborhoods. For example, neighborhood characteristics are associated with availability of
health care services and the lack of knowledge or means to use such services.11,12 This is
demonstrated in the history of lower supplies of medical services in low-income and inner-
city neighborhoods,15,16 although typically a greater demand exists in these areas due to
Byrnes et al. Page 2
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
higher density populations and greater health problems.12,17 These difficulties often
translate into longer waits for patients and poorer quality care in disorganized
neighborhoods.18-20 Some families in these neighborhoods resort to the use of the
emergency room for non-emergency health care, adding to their likelihood of long waits and
lack of preventive care.21 For families wishing to avoid these hassles and receive quality
care, they must travel outside of their neighborhoods; however transportation can be a
problem for residents of these areas, making location extremely important to health care
utilization for these residents.12
In addition to difficulties utilizing health care, stressors in disorganized neighborhoods may
force families to focus on immediate needs (e.g., food, clothes, shelter) rather than on non-
emergency health care, such as prevention services. For these families, expenses for regular
preventive medical care may be seen as appropriate only when there is discretionary income.
Preventive care and non-emergency care are often delayed until a problem becomes so
severe that it requires hospitalization.17,22
Neighborhood Influences on Utilization of Preventive Health Care
Although studies have not examined the relationship between disorganized neighborhoods
and utilization of preventive health services for adolescent substance use in particular, data
suggest that families in disorganized neighborhoods are less likely to utilize services to
prevent a variety of health-related problems. For example, residents of disorganized
neighborhoods are less likely to use preventive services for themselves, such as dental
services 23,24 and prenatal care,25 controlling for individual characteristics. Similarly,
residents of areas with higher median education are about 1-½ to 2 times more likely to have
received a mammogram, while residents of areas with more Hispanic residents, low income,
higher rates of poverty, and higher proportions of immigrants are about 1-½ times less likely
to seek mammograms, breast exams, or pap smears.26 Residents of disorganized
neighborhoods are also less likely to use preventive services for their children, such as
immunizations.23 Lower levels of preventive care utilization could be due to less availability
in inner-city doctor's offices,15 long waiting times, transportation barriers,23,25 and
competing priorities.27 Disorganized neighborhoods may also have fewer preventive
resources and attract less well-educated doctors that may not recommend preventive care.20
Family-based Prevention Programs
Most youth substance use prevention programs solely target the individual youth, such as in
school-based approaches.28 While youth-only approaches are effective, effect sizes are
generally very small. Since substance use is influenced by family factors,29 substance use
prevention tends to be most effective when programs focus on family strengthening
activities.28 On average, effect sizes for family-based prevention strategies have been
reported as two to nine times larger than child-only approaches.30,31
As family-based approaches have been shown to be more effective than other types of
approaches, it is important to determine factors that are related to participant engagement,
including recruitment, into such programs. This is especially important to examine, as
engagement into family-based prevention programs is very challenging, particularly when
the target population is not “captive”, as with school-based programs.32-34
Although neighborhood-level factors have been associated with less use of other medical
preventive health care services,26,35 few studies have examined the impact of neighborhood-
level risk factors in family-based prevention programs for addressing adolescent risky
behaviors such as alcohol and other drug use. Neighborhood disorganization could impact
the success of family-based substance use prevention programs indirectly by affecting who
Byrnes et al. Page 3
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
is recruited. Although studies have not examined neighborhood level risks for utilization of
family-based prevention, family-level risk factors have been shown to affect recruitment.34
Specifically, for family components of multi-component prevention programs, families that
are two-parent, have higher SES, or are White or Hispanic, are more likely to attend.36-39
However, findings from these multi-component interventions may not generalize to
freestanding family-based interventions, as the length and organization of the two types of
interventions may be quite different.34 Findings from freestanding family-based prevention
programs show that parent education is related to participation,40 although family income is
generally not related.34 Despite these studies demonstrating family-level effects on
participation in family-based prevention programs, studies have not examined the role of
neighborhood-level risks not accounted for by these family-level effects.
Hypotheses
The goals of this study were to examine the association between neighborhood
disorganization (i.e., low socioeconomic status and residential instability) and recruitment,
as measured by agreement to participate and actual enrollment, into family-based adolescent
prevention programs focusing on substance use and the promotion of healthy behaviors.
Consistent with neighborhood disorganization theory 5 and prior studies showing that
neighborhood disorganization is associated with less use of treatment and preventive health
care services,13,35 hypotheses were that families residing in disorganized neighborhoods
would be less likely to be recruited into family-based prevention programs focused on
adolescent alcohol and other drug use (AOD). These relationships were examined while
controlling for individual-level factors in order to determine the relationships of
neighborhood disorganization above and beyond individual characteristics. Researchers
suggest that it is the concentration of individual characteristics (e.g., low income
neighborhood) that causes problems for residents outside of effects of individual or family
characteristics (e.g., low-income household). For example, Wilson41 postulated that the
concentration of disadvantage (e.g., poverty, unemployment in a neighborhood) resulted
from fewer unskilled job opportunities and the middle-class leaving the inner-city after
World War II; these changes led to social isolation and changes in neighborhood norms and
behavior for those remaining.
Methods
Sample and Procedures
This study utilizes data from the first wave of an ongoing longitudinal study designed to
examine the impact of choice on prevention program recruitment and participation as well as
adolescent outcomes. For this paper, families identified from two Kaiser Permanente (KP)
medical centers (Oakland and Vallejo) in the San Francisco Bay Area were included in
analyses. Families from these centers represent a diversity of socioeconomic statuses,
ethnicities, and neighborhoods. Study procedures were approved by the Institutional Review
Boards at the Pacific Institute for Research and Evaluation (PIRE), Kaiser Foundation
Research Institute (KFRI), and University of California, Berkeley.
Eligibility criteria were that at least one family member must be a member of KP at the time
the sample was drawn, and that there must be an 11-12 year old child. Families with an
11-12 year old were targeted because the programs are designed for the prevention of
adolescent substance use and the promotion of healthy behaviors, prior to the age which
most adolescents have initiated use. A large increase in alcohol use is seen after this age,
from approximately 3.4% of 12-13 year olds currently drinking alcohol (past month use) to
13.1% among 14-15 year olds.42
Byrnes et al. Page 4
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
KP provided a list of families who met these eligibility criteria (N = 1,983). Then families
were assigned to one of two conditions: 1) Random Control Trial (RCT), in which families
were randomly assigned to one of two programs or a control condition and 2) Choice
condition, in which families chose between the two programs. Families were sent letters on
KP letterhead inviting participation in the study, signed by the Chief of Pediatrics. Families
were invited to call to schedule an enrollment interview, and also given the option to “opt-
out” of the study by calling a number to be placed on a no-contact list. Trained recruiters,
representing diverse ethnic backgrounds, then called families to confirm eligibility (families
were excluded if the child did not live with his/her parents, if the family did not speak
English, or if the child was currently participating in ATOD treatment) and to schedule
enrollment interviews if the family agreed to participate in the study. The scripts used during
these recruitment calls are available upon request. If contact information was incorrect, KP
was usually able to provide corrected information. However, for a few cases (5% of
attempted contacts) where address information was correct but phone numbers were not, KP
was not able to provide new telephone contact information. In these cases, correspondence
was conducted via mail. Recruiter contacts confirmed that 823 families were eligible. Of the
823 eligible families, the addresses of 79 families were not able to be geocoded to allow for
the collection of archival neighborhood data from the census, either due to incorrect
addresses or due to residences in new subdivisions/streets that were not reflected in current
maps.
Of the remaining 744 families, 351 families agreed to participate in the study. Of these, 214
families came to the health care facility to enroll in the study and completed the first
interview (mothers/female caregivers and youth completed separate face-to-face baseline
enrollment interviews). Parent and youth each received $30 for interview completion.
Age and gender data were available for the entire sample. About half (47.4%) of the youth
were female. Per study criteria, youth were 11 or 12 years of age at recruitment (x̄= 11.5;
SD
= .50). Mothers' ages ranged from 24 to 69 (
M
= 41.93;
SD
= 7.05). Ethnicity and education
information was available for a smaller sample. Mothers provided ethnicity information, and
were allowed to endorse multiple ethnicities, resulting in the following racial breakdown:
37.8% White, 32.3% African-American, 17.1% Hispanic, 16.8% Asian, 3.1% American
Indian/Alaskan Native, and 1.0% Pacific Islander. Nearly half (45.2%) of mothers had
graduated from college.
Programs
The programs offered in the study were: Strengthening Families Program: For Parents and
Youth 10-14 (SFP) 43 and Family Matters (FM).44 Both are theory-based universal
programs, not targeted at specific groups, and focus on similar risk and protective factors for
adolescent substance use and the promotion of healthy behaviors. Both have been evaluated
in rigorous randomized experimental designs and shown to be effective in preventing
adolescent ATOD.45-48 The programs have substantial differences in their structure and time
investment for families.
SFP—SFP is a group program led by health educators in seven weekly sessions at KP
medical centers. The first hour involves separate groups for parents and adolescents, while
the second hour involves a combined session for families to practice skills learned in the
first hour. Based on the biopsychosocial model, SFP targets family risk and protective
factors that are related to adolescent problem behavior.4950,51 For example, risk factors such
as poor parental discipline and poor parent-adolescent relations and protective factors
including parental empathy and parent-adolescent bonding are emphasized in the
program.52,53
Byrnes et al. Page 5
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
FM—FM is a program that parents lead for their family at home using four booklets mailed
to them one at a time. Health educators call to provide encouragement and discuss any
issues parents have with the program. FM development was guided by key concepts in
public health and health promotion practice, including the targeting of environmental or
ecological risk and protective factors.54,55
Measures
Neighborhood Disorganization—Census data was gathered from the 2000
Census of
Population and Housing
, publicly available from the U.S. Census Bureau to assess
neighborhood characteristics for the census block groups in which the participating families
live. Families' addresses were geocoded to determine block group membership so that
families could be linked to census data for their block group. All census data are presented
in proportions (i.e., the number of people who make up each characteristic divided by the
total population of the neighborhood block group). Items were standardized prior to
analyses.
Low Socioeconomic status (SES)—Low SES variables are the most commonly
examined census variables used in neighborhood studies.2 Five items commonly used to
reflect low SES were used, including the rates of overall unemployment, persons below the
poverty line, households receiving public assistance, high school dropouts, and female-
headed households. Items were standardized and high scores indicate the existence of a
greater proportion of low-income residents.
Residential Instability—Residential instability is also a frequently used indicator of
neighborhood characteristics.2 The proportion of residents who have moved in the past 5
years was used to indicate residential instability. This item was standardized and high scores
indicate higher levels of residential instability.
Recruitment—Two variables were used as indicators of recruitment into the study: 1)
whether the mother agreed to participate in the study (indicated by scheduling an
appointment for a baseline enrollment interview), and 2) whether the family actually
enrolled in the study, indicated by signing the consent forms and completing the face-to-face
baseline interview at the KP Medical Center.
Individual variables—Background variables include mothers' reports of her ethnicity
(White = 1) and family SES, as indicated by mothers' education level (college graduate = 1).
Mothers' age and youths' gender (male = 1) were also obtained through KP patient records.
Data Analysis—Descriptive analyses provided an overview of neighborhood
characteristics and average recruitment rates. To determine if non-independence among
observations sampled from the same block group might bias statistical tests, an intraclass
correlation coefficient (ICC) was computed for agreeing to participate and enrolling in the
study. Results showed low ICC values (ρ = 0.02 and 0.03, respectively) indicating that non-
independence at the neighborhood level was negligible. It is also worth noting that, on
average, each block group had fewer than 2 families each (1.58 families per block group).
Using the formula provided by Kish,56 the design effect was estimated at 1.01, resulting in
an effective N for these analyses of 737, which differs only trivially from the total sample of
744. Given that the observations were essentially independent and that a large proportion of
the block groups had only 1 family, family was treated as the highest level unit of analysis.
Accordingly, standard logistic regression analyses were used to examine relationships
between neighborhood characteristics and recruitment rather than a multilevel regression
approach
Byrnes et al. Page 6
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Agreeing to Participate—First we examined predictors of agreeing to participate in the
study. Within this set of analyses, logistic regressions were conducted for two groups: 1)
Sample 1, which was the full sample of 744 families, and 2) Sample 2 (N = 456), which
includes only those families from Sample 1 providing complete demographic information in
order to control for individual SES and ethnicity in the models. Even parents who refused to
participate in the study were asked for demographic information at the initial contact for
recruitment. Missing demographic data was not refused to be provided, however, many of
the families who did not agree to participate ended the call before demographic questions
were asked and so did not provide complete demographic information (N = 288). In these
cases, responses were supplemented with two variables considered to be non-identifying,
parent age (not date of birth) and child gender, obtained through KP records. KP did not
provide information regarding any other variables. At the time of enrollment into KP,
members were informed that their demographic information may be used in research studies.
However, members could request to be placed on a no-contact list. This list was used to
exclude some individuals from any contact for our study.
For analyses with Sample 1, controls were the available demographic variables, parent age
and child gender. Sample 2 was included in analyses so that it would be possible to
determine neighborhood effects above and beyond any effects of similar characteristics at
the individual level. For example, a low-SES family might experience very different
outcomes living in a low-SES neighborhood as compared to a high-SES neighborhood based
on differing resources in the neighborhood.
Enrolling in the Study—The second set of analyses examined neighborhood predictors
of enrolling in the study. Again, within this set of analyses, logistic regressions were
conducted for two groups: 1) Sample 3, including all of the families who agreed to
participate in the study (N = 351), and 2) Sample 4 (N = 341), including only those families
who agreed to participate that provided complete demographic information in order to
control for individual SES and ethnicity in the models.
Results
Overall Recruitment Rates
Of eligible families who were able to be contacted, about half (47.2%) agreed to participate
in the study, and more than half (61.0%) of those families enrolled in the study. Analyses
indicate few demographic differences between families who agreed to participate as
compared to those who refused, as well as between those who enrolled and those who did
not (See Tables I and II). Specifically, African-Americans were significantly more likely to
agree to participate, while Asians were less likely. Caucasians, Asians, college graduates,
and older parents were more likely to enroll, whereas African-Americans were less likely.
Neighborhood Disorganization
Neighborhoods of residence for families in the sample reflected a diversity of neighborhood
conditions. Table III presents descriptive statistics of neighborhood variables, including
means, standard deviations, and ranges.
Sample Differences
Analyses were conducted to determine demographic differences between the samples
providing all demographic information (Samples 2 and 4) and the remaining families
(Samples 1 and 3) on the variables available in all samples: child gender and parent age. No
significant differences between these sets of samples existed for child gender (χ2 = 0 .68,
p
= 0.41) or parent age (
t
= .44,
p
= 0.66).
Byrnes et al. Page 7
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Logistic Regressions
Results from logistic regression analyses are presented in Tables IV and V, and are
described separately below for agreeing to participate and enrolling in the study, and for
each sample. Logistic regression analyses allow for the calculation of odds ratios, which
provide information regarding the risk of a certain outcome occurring.57 Odds ratios equal to
one suggest that the outcome is as likely to happen as not. If the odds ratio is greater than
one, the outcome is more likely to happen than not, and if the odds ratio is less than one, the
outcome is less likely to happen than not.
Agreeing to Participate
Sample 1—For Sample 1, the individual variables included in the model were parent age
and child gender (see Table IV). Higher levels of neighborhood unemployment (OR = 0.73,
p
< .01) were related to decreased agreement to participate, such that the odds of agreeing
were decreased by 27% for each percentage point increase in neighborhood unemployment.
Sample 2—Table IV also presents findings for Sample 2. Among the neighborhood effects
tested, only higher rates of single-female headed households were related to increased
agreement to participate (OR = 1.43,
p
< .05). The odds of agreeing to participate were 43%
greater for each percentage point increase in neighborhood single-female headed
households.
Enrolling in the Study
Sample 3—As shown in Table V, for Sample 3, the odds of enrolling decreased by 44%
for each percentage point increase in neighborhood high school dropouts (OR = 0.56, p < .
01). Parent age was related to increased levels of enrollment (OR = 1.04,
p
< .05) so that the
odds of enrolling increased 4% for each year of parent age.
Sample 4—No neighborhood variables were related to enrollment for Sample 4 (see Table
V). However, two personal characteristics, ethnicity and education, were related to
enrollment, so that the odds of enrolling were about 100% higher for White parents than for
non-White parents (OR = 2.08,
p
< .05) as well as for college graduates than for those not
completing college (OR = 2.09,
p
< .01).
Discussion
Study findings suggest that families residing in disorganized neighborhoods may experience
both forces for and barriers against recruitment into family-based prevention programs
focused on substance use and promotion of healthy behaviors, independent of the effects of
individual characteristics. In contrast to prior research that has primarily emphasized the
barriers to utilizing health care, findings from this research suggest that there are some
neighborhood forces encouraging recruitment into prevention programs. Specifically,
families in neighborhoods with more single mothers were more likely to agree to participate
in the study, controlling for individual and other neighborhood characteristics.
Neighborhoods with many single mothers may be characterized by a lack of support and
resources for youth in particular, both within each family and by the lack of collective
supervision and resources available from other neighborhood adults to ease the burden on
individual families.58 These neighborhoods are characterized by families with only one
working adult (females), and may provide fewer neighborhood role models for youth, as
well as few resources remaining to help provide collective resources to other neighborhood
youth. Involvement in prevention programs may be seen as a way to gain the additional
parenting-specific resources lacking in the neighborhood.
Byrnes et al. Page 8
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Consistent with study hypotheses and neighborhood disorganization theory,5 neighborhood
barriers to recruitment were also found. Families in neighborhoods with higher rates of
general unemployment were less likely to agree to participate in the study, and those in
neighborhoods with more high school dropouts were less likely to enroll in the study. These
findings are consistent with prior studies showing that neighborhood disadvantage is
associated with less use of treatment services,13,14 suggesting that the effects of
neighborhoods on treatment seeking may parallel effects on prevention seeking. As such,
results are also consistent with studies finding that neighborhood disadvantage is related to
less use of other preventive health services.23,24,26 For example, Andersen and colleagues 14
found that in areas with higher rates of unemployment, both adults and children were less
likely to utilize health care services. Similarly, increased neighborhood levels of education
have been associated with increased use of mammograms.26
There are several possible explanations why neighborhood disorganization may make
recruitment into prevention programs less likely. Prior studies show that certain
characteristics of these neighborhoods may lower utilization of health services. For example,
families in more disorganized neighborhoods typically receive lower quality health services
and face more hassles such as long waits.15,18-20 Families experiencing these problems may
be less willing to participate in prevention programs, anticipating similar difficulties. In
addition to problems utilizing care, stressors in disorganized neighborhoods may force
families to concentrate on immediate pressures instead of on prevention.27,59 However,
determining the reason families do not participate in programs is difficult because many
families do not state a reason even if probed and may not explicitly think of neighborhood
influences as reasons. Future research should explore whether families see an association
between their neighborhood contexts and their ability to participate in family programs (e.g.,
fewer services available).
Results differed across samples. Specifically, findings using samples that included families
regardless of whether they provided demographic information (Samples 1 and 3) were
consistent with hypotheses that neighborhood disorganization appears to be a barrier to
recruitment, while findings using the samples including only those families providing
complete demographic information (Samples 2 and 4) pointed to features that encourage
families to participate in programs. Descriptive analyses showed that the families providing
complete demographic information (Samples 2 and 4) did not differ from the remaining
families (Samples 1 and 3) on the available demographic variables, child gender and parent
age. However, it is possible that differing findings may result from group differences on
other characteristics, such as a general tendency to seek help or problems the family may
already be experiencing, as these may be related to greater participation.
One sample difference is that when individual-level SES variables are added to the model
for Sample 4, neighborhood effects are no longer related to actual enrollment. Individual
characteristics may be more important in determining enrollment for this sample. White
parents and college graduates were much more likely to enroll in the program than were
nonwhites and those with less education. However, these groups were not any more likely to
agree to participate during recruitment, indicating that these variables may influence the
ability to follow through and actually participate, possibly due to greater resources and
fewer stressors allowing them the luxury to focus on prevention. However, regarding the
decision to agree to participate, with individual SES included, the neighborhood context still
remains an important positive influence. Specifically, families in neighborhoods with higher
rates of single female headed households were more likely to agree to participate, while
controlling for these individual variables. These findings point to the importance of
neighborhood features not accounted for by individual level characteristics.
Byrnes et al. Page 9
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Bivariate findings also showed that individual characteristics remain important in
influencing recruitment into such programs. Interestingly, African-American parents were
more likely to agree to participate but less likely to actually enroll than other groups, while
the opposite was found for Asian parents. It may be that African-Americans wanted to
participate, and so agreed, but barriers prevented them from actually enrolling in the study.
This is consistent with prior studies showing barriers to recruiting African-Americans, such
as underutilization of health services and distrust of the medical/scientific community.60,61
However, although Asian families are more likely to follow through with enrollment once
they agree to participate, they are less likely to agree in the first place than other groups.
Asian families might worry that the program will be culturally inappropriate or not relevant
to their needs.60 This might suggest that mail and phone recruitment alone may not be the
most effective recruitment strategy for Asian families. Some studies suggest that including
initial community involvement in recruitment efforts, such as involving health professionals
or community based organizations in the targeted ethnic communities, can increase
recruitment in minority populations.61 In addition, culturally competent interventions may
also be effective in attracting minority populations.
Both the advantages and limitations of telephone recruitment over other recruitment
methods should be noted. A substantial advantage is its relative low cost and effectiveness in
contacting a large number of families,62 and its efficiency in determining the eligibility of
potential participants.63 The inability to include families without phones and difficulty
contacting families with missing or changed phone numbers present limitations.62 Although
refusal rates are often higher for phone recruitment as compared to in-person recruitment,64
in-person recruitment can be very expensive and time-consuming. In addition, phone
recruitment tends to have lower refusal rates than does mail recruitment.63 A combination of
methods is often recommended as more effective than any single method.62
One limitation of this study is the cross-sectional nature of the study, meaning that causality
cannot be determined. Additionally, census boundaries may not match residents' perceptions
of their own neighborhood boundaries,65 and perceptions of neighborhood features may be
important to residents' behaviors. For example, both adult and youth perceptions of lower
social cohesion are related to increased levels of neighborhood alcohol and drug use among
youth.66 In the current study, neighborhood perceptions were assessed, but only after
recruitment into the program, thus limiting the current analyses to census data only. Future
studies will examine the role of neighborhood perceptions on level of participation in
prevention programs among those recruited. It is also possible that other individual variables
not examined could be related to recruitment. However, since the purpose of the paper was
to examine the relation of neighborhood variables with recruitment, we focused on
individual variables that are commonly included in neighborhood studies as controls.
The low recruitment rate of the study also presents a limitation in generalizing results to a
wider population. However, the engagement of participants in prevention programs,
especially, universal family-based programs, is a considerable challenge.32-34 As such,
recruitment rates for programs that target both youth and parents are typically very low.32
Our recruitment rate of 47% is comparable to other family-based prevention programs, such
as that reported by Heinrichs et al. 32 for a universal family program (31%) and Bronstein et
al. 67 for a parent program (38%).
Another issue is that recruiters were not matched to families based on race/ethnicity. It is
possible that being recruited by an individual of a different race or ethnicity may lessen
families' interest in participating. However, as recruiters were from diverse racial/ethnic
backgrounds, and only spoke to families over the phone, it is unlikely that this posed a
substantial barrier to recruitment.
Byrnes et al. Page 10
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Implications for Behavioral Health
This study indicates the importance of neighborhood characteristics for recruitment into
family-based prevention programs. Findings suggest that recruitment procedures may need
to emphasize different strategies according to specific neighborhood conditions to ensure
recruitment of families who are at risk. For example, programs provided in disorganized
areas may need to be altered in order to make them more accessible, such as providing
transportation vouchers, childcare, or meals,68 and emphasizing these additional features
during recruitment. These strategies may be less important in more socially organized
neighborhoods. In addition, when recruiting in neighborhoods with high rates of single-
female headed households, recruitment may be boosted by using protocols that emphasize
the potential of the program to help mothers gain parenting-specific resources and support of
other community adults.
Overall, findings from this study contribute a better understanding of neighborhood
characteristics that may indirectly put adolescents at risk for poor health outcomes, through
lowered participation in family-based prevention programs. As family-based programs have
been shown to be effective in reducing problem behaviors and strengthening family
protective factors,30,31 it is important to determine specific challenges to the engagement of
families in such programs. For example, transportation challenges may be addressed through
providing transportation vouchers. Addressing the challenge of busy schedules may be
addressed by providing meals so that families do not have to fit in mealtime around program
attendance. Future studies should extend these findings to adolescent outcomes in order to
confirm whether neighborhoods indirectly affect adolescent substance use and other
problem behaviors by first affecting program recruitment and participation. Findings from
this work have important implications for the development of prevention programs in the
field of adolescent substance abuse and other problem behaviors and help determine whether
health promotion messages are reaching families in at-risk areas.
Acknowledgments
Thanks are extended to Michael Todd and Joel Grube for their statistical guidance. Research for and preparation of
this manuscript were supported by NIAAA “Adolescent Family-Based Alcohol Prevention” R01-AA015323-01,
2005-2010, Brenda A. Miller, PI and NIAAA “Prevention Science Research Training Program Grant” T32
AA014125, 2004-2009, Genevieve Ames, PI. The content is solely the responsibility of the authors and does not
necessarily represent the official views of the National Institute on Alcohol Abuse and Alcoholism or the National
Institutes of Health.
References
1. Lambert SF, Brown TL, Phillips CM, et al. The relationship between perceptions of neighborhood
characteristics and substance use among urban African American adolescents. American Journal of
Community Psychology. 2004; 34(3-4):205–218. [PubMed: 15663207]
2. Leventhal T, Brooks-Gunn J. The neighborhoods they live in: the effects of neighborhood residence
on child and adolescent outcomes. Psychological Bulletin. 2000; 126(2):309–337. [PubMed:
10748645]
3. Byrnes HF, Chen MJ, Miller BA, et al. The relative importance of mothers' and youths'
neighborhood perceptions for youth alcohol use and delinquency. Journal of Youth and
Adolescence. 2007; 36:649–659.
4. Frank R, Cerda M, Rendon M. Barrios and burbs: residential context and health-risk behaviors
among Angeleno adolescents. Journal of Health and Social Behavior. 2007; 48(3):283–300.
[PubMed: 17982869]
5. Shaw, CR.; McKay, HD. Juvenile delinquency and urban areas. Chicago: The University of Chicago
Press; 1942.
Byrnes et al. Page 11
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
6. McCulloch A. An examination of social capital and social disorganisation in neighbourhoods in the
British household panel study. Social Science and Medicine. 2003; 56(7):1425–1438. [PubMed:
12614694]
7. Sampson, RJ.; Lauritsen, JL. Violent victimization and offending: Individual-, situational-, and
community-level risk factors. In: Reiss, AJ.; Roth, JA., editors. Understanding and preventing
violence. Vol 3: Social influences. Washington, DC: National Academy Press; 1994. p. 1-114.
8. Sampson RJ, Morenoff JD, Earls F. Beyond social capital: Spatial dynamics of collective efficacy
for children. American Sociological Review. 1999; 64(5):633–660.
9. Sampson RJ, Raudenbush SW, Earls F. Neighborhoods and violent crime: A multilevel study of
collective efficacy. Science. 1997; 277:918–924. [PubMed: 9252316]
10. Elliott D, Wilson W, Huizinga D, et al. The effects of neighborhood disadvantage on adolescent
development. Journal of Research in Crime and Delinquency. 1996; 33(4):389–426.
11. Andersen RM. Revisiting the behavioral model and access to medical care: Does it matter? Journal
of Health and Social Behavior. 1995; 36(1):1–10. [PubMed: 7738325]
12. Ellen IG, Mijanovich T, Dillman KT. Neighborhood effects on health: Exploring the links and
assessing the evidence. Journal of Urban Affairs. 2001; 23(3-4):391–408.
13. Auchincloss AH, Van Nostrand JF, Ronsaville D. Access to health care for older persons in the
United States: Personal, structural, and neighborhood characteristics. Journal of Aging and Health.
2001; 13(3):329–354. [PubMed: 11813730]
14. Andersen RM, Yu H, Wyn R, et al. Access to medical care for low-income persons: How do
communities make a difference? Medical Care Research and Review. 2002; 59(4):384–411.
[PubMed: 12508702]
15. Fossett JW, Perloff JD, Peterson JA, et al. Medicaid in the inner city: The case of maternity care in
Chicago. Milbank Quarterly. 1990; 68(1):111–141. [PubMed: 2215426]
16. Perloff JD. Health care resources for children and pregnant women. Future of Children. 1992;
2:78–94.
17. Pappas G, Hadden WC, Kozak LJ, et al. Potentially avoidable hospitalizations: inequalities in rates
between US socioeconomic groups. American Journal of Public Health. 1997; 87(5):811–816.
[PubMed: 9184511]
18. Bach PB, Pham HH, Schrag D, et al. Primary care physicians who treat blacks and whites. New
England Journal of Medicine. 2004; 351(6):575–584. [PubMed: 15295050]
19. Fiscella K, Williams DR. Health disparities based on socioeconomic inequities: Implications for
urban health care. Academic Medicine. 2004; 79(12):1139–1147. [PubMed: 15563647]
20. Freeman HP, Chu KC. Determinants of cancer disparities: Barriers to cancer screening, diagnosis,
and treatment. Surgical Oncology Clinics of North America. 2005; 14:655–669. [PubMed:
16226685]
21. Shah-Canning D, Alpert JJ, Bauchner H. Care-seeking patterns of inner-city families using an
emergency room. A three-decade comparison. Medical Care. 1996; 34(12):1171–1179. [PubMed:
8962583]
22. Eslami MH, Zayaruzny M, Fitzgerald GA. The adverse effects of race, insurance status, and low
income on the rate of amputation in patients presenting with lower extremity ischemia. Journal of
Vascular Surgery. 2007; 45(1):55–59. [PubMed: 17210382]
23. Niederhauser VP, Stark M. Narrowing the gap in childhood immunization disparities. Pediatric
Nursing. 2005; 31(5):380–386. [PubMed: 16295153]
24. Salmond C, Crampton P, Sutton F. NZDep91: A New Zealand index of deprivation. Australian and
New Zealand Journal of Public Health. 1998; 22(7):835–837. [PubMed: 9889455]
25. Perloff JD, Jaffee KD. Late entry into prenatal care: the neighborhood context. Social Work. 1999;
44(2):116–128. [PubMed: 10718077]
26. Wells BL, Horm JW. Targeting the underserved for breast and cervical cancer screening: the utility
of ecological analysis using the National Health Interview Survey. American Journal of Public
Health. 1998; 88(10):1484–1489. [PubMed: 9772849]
27. Gelberg L, Gallagher TC, Andersen RM, et al. Competing priorities as a barrier to medical care
among homeless adults in Los Angeles. American Journal of Public Health. 1997; 87(2):217–220.
[PubMed: 9103100]
Byrnes et al. Page 12
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
28. Kumpfer KL, Alvarado R, Whiteside HO. Family-based interventions for substance use and
misuse prevention. Substance Use and Misuse. 2003; 38(11-13):1759–1787. [PubMed: 14582577]
29. Kumpfer, KL. Special populations: Etiology and prevention of vulnerability to chemical
dependency in children of substance abusers. In: Brown, BS.; Mills Arnold, R., editors. Youth at
High Risk for Substance Abuse. Rockville, MD: NIDA Monograph; 1987. p. 1-71.
30. Tobler, N.; Kumpfer, KL. Meta-analysis of effectiveness of family-focused substance abuse
prevention programs. Rockville, MD: Center for Substance Abuse Prevention; 2000.
31. Tobler NS, Stratton HH. Effectiveness of school-based drug prevention programs: a meta-analysis
of the research. Journal of Primary Prevention. 1997; 18(1):71–128.
32. Heinrichs N, Bertram H, Kuschel A, et al. Parent recruitment and retention in a universal
prevention program for child behavior and emotional problems: Barriers to research and program
participation. Prevention Science. 2005; 6(4):275–285. [PubMed: 16075192]
33. Prinz RJ, Smith EP, Dumas JE, et al. Recruitment and retention of participants in prevention trials
involving family-based interventions. American Journal of Preventive Medicine. 2001; 20(1
Suppl):31–37. [PubMed: 11146258]
34. Spoth R, Redmond C. Research on family engagement in preventive interventions: Toward
improved use of scientific findings in primary prevention practice. The Journal of Primary
Prevention. 2000; 21(2):267–284.
35. Ansari Z, Carson NJ, Ackland MJ, et al. A public health model of the social determinants of health.
Sozial- und Praventivmedizin. 2003; 48(4):242–251. [PubMed: 12971112]
36. Cohen DA, Linton KL. Parent participation in an adolescent drug abuse prevention program.
Journal of Drug Education. 1995; 25(2):159–169. [PubMed: 7658296]
37. Cohen DA, Rice JC. A parent-targeted intervention for adolescent substance use prevention:
Lessons learned. Evaluation Review. 1995; 19(2):159–180. [PubMed: 20871746]
38. Toomey TL, Williams CL, Perry CL, Murray DM, Dudoritz B, Veblen-Mortenson S. An alcohol
primary prevention program for parents of 7th graders: The amazing alternatives! home program.
Journal of Child and Adolescent Substance Abuse. 1996; 5:35–53.
39. Williams CL, Perry CL, Dudovitz B, et al. A home-based prevention program for sixth-grade
alcohol use: Results from Project Northland. The Journal of Primary Prevention. 1995; 16(2):125–
147.
40. Spoth RL, Redmond C, Kahn JH, et al. A prospective validation study of inclination, belief, and
context predictors of family-focused prevention involvement. Family Process. 1997; 36(4):403–
429. [PubMed: 9543661]
41. Wilson, WJ. The Truly Disadvantaged: The Inner City, the Underclass and Public Policy. Chicago:
University of Chicago Press; 1987.
42. SAMHSA. Substance Abuse and Mental Health Services Administration. Results from the 2008
National Survey on Drug Use and Health: National Findings (Office of Applied Studies, NSDUH
Series H-36, HHS Publication No. SMA 09-4434). Rockville, MD: 2009.
43. Spoth R, Redmond C, Lepper H. Alcohol initiation outcomes of universal family-focused
preventive interventions: one- and two-year follow-ups of a controlled study. J Stud Alcohol
Suppl. 1999; 13:103–111. [PubMed: 10225494]
44. Bauman KE, Foshee VA, Ennett ST, et al. Family Matters: A family-directed program designed to
prevent adolescent tobacco and alcohol use. Health Promotion Practice. 2001; 2(1):81–96.
45. Spoth R, Redmond C, Shin C. Randomized trial of brief family interventions for general
populations: Adolescent substance use outcomes 4 years following baseline. Journal of Consulting
and Clinical Psychology. 2001; 69(4):627–642. [PubMed: 11550729]
46. Spoth R, Reyes ML, Redmond C, et al. Assessing a public health approach to delay onset and
progression of adolescent substance use: Latent transition and log-linear analyses of longitudinal
family preventive intervention outcomes. Journal of Consulting and Clinical Psychology. 1999;
67(5):619–630. [PubMed: 10535229]
47. Bauman KE, Ennett ST, Foshee VA, et al. Influence of a family program on adolescent smoking
and drinking prevalence. Prevention Science. 2002; 3(1):35–42. [PubMed: 12002557]
Byrnes et al. Page 13
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
48. Bauman KE, Foshee VA, Ennett ST, et al. The influence of a family program on adolescent
tobacco and alcohol use. American Journal of Public Health. 2001; 91(4):604–610. [PubMed:
11291373]
49. DeMarsh J, Kumpfer KL. Family-Oriented Interventions for the Prevention of Chemical
Dependency in Children and Adolescents. Journal of Children in Contemporary Society. 1986;
17:117–151.
50. Kumpfer, KL.; Molgaard, V.; Spoth, R. Preventing childhood disorders, substance abuse, and
delinquency. Thousand Oaks, CA, USA: Sage Publications, Inc; 1996. The Strengthening Families
Program for the prevention of delinquency and drug use; p. 241-267.
51. Molgaard, V.; Kumpfer, KL.; Fleming, CB. The Strengthening Families Program: For Parents and
Youth 10-14. Ames, Iowa: University Extension; 1997.
52. Spoth R, Redmond C, Shin C. Randomized trial of brief family interventiosn for general
populations: Adolescent substance use outcomes 4 years following baseline. Journal of Consulting
and Clinical Psychology. 2001; 69(4):627–642. [PubMed: 11550729]
53. Spoth RL, Redmond C. Project family prevention trials based in community-university
partnerships: Toward scaled-up preventive interventions. Prevention Science. 2002; 3(3):203–221.
[PubMed: 12387555]
54. Glanz, K.; Lewis, FM.; Rimer, BK. The scope of health promotion and health education. In: Glanz,
K.; Lewis, FM.; Rimer, BK., editors. Health behavior and health education. San Francisco: Jossey-
Bass; 1997. p. 3-18.
55. Medicine Io. The future of public health. Washington, DC: National Academy of Sciences; 1989.
56. Kish, L. Survey Sampling. New York: John Wiley and Sons, Inc.; 1965.
57. Crichton N. Information point: Odds ratio. Journal of Clinical Nursing. 2001; 10(2):268–269.
58. Beyers JM, Bates JE, Pettit GS, et al. Neighborhood structure, parenting processes, and the
development of youths' externalizing behaviors: a multilevel analysis. American Journal Of
Community Psychology. 2003; 31(1-2):35–53. [PubMed: 12741688]
59. Rosenfeld, E. Social support and health status: A literature review. Adelaide: South Australian
Community Health Research Institute; 1997.
60. Harachi TW, Catalano RF, Hawkins JD. Effective recrutiment for parenting programs within
ethnic minority communities. Child and Adolescent Social Work Journal. 1997; 14(1):23–39.
61. Yancey AK, Ortega AN, Kumanyika SK. Effective recruitment and retention of minority research
participants. Annual Review of Public Health. 2006; 27:1–28.
62. Duncan SC, Strycker LA, Duncan TE, et al. Telephone recruitment of a random stratified youth
sample for a physical activity study. Journal of Sport and Exercise Psychology. 2002; 24:347–358.
63. Verboncoeur CJ, Stewart AL, King AC, et al. The use of refusal postcards in recruiting older
adults. Annals of Behavioral Medicine. 2000; 22(4):330–333. [PubMed: 11253445]
64. Aquilino WS. Telephone versus face-to-face interviewing for household drug use surveys.
International Journal of the Addictions. 1992; 27(1):71–91. [PubMed: 1537641]
65. Burton, LM.; Price-Spratlen, T. Through the eyes of children: An ethnographic perspective on
neighborhoods and child development. In: Masten, AS., editor. Cultural Processes in Child
Development. The Minnesota Symposia on Child Psychology. Vol. 29. Mahwah, NJ: Lawrence
Erlbaum; 1999. p. 77-96.
66. Duncan SC, Duncan TE, Strycker LA. A multilevel analysis of neighborhood context and youth
alcohol and drug problems. Prevention Science. 2002; 3(2):125–133. [PubMed: 12088137]
67. Bronstein P, Duncan P, Clauson J, et al. Preventing middle school adjustment problems for
children from lower income families: A program for aware parenting. Journal of Applied
Developmental Psychology. 1998; 19:129–152.
68. Hahn EJ, Simpson MR, Kidd P. Cues to parent involvement in drug prevention and school
activities. Journal of School Health. 1996; 66(5):165–170. [PubMed: 8735580]
Byrnes et al. Page 14
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Byrnes et al. Page 15
Table I
Parent and Child Characteristics by Agreement to Participate in the Study
Variables All families
(N = 744) Agreed to participate
(n = 351) Did not agree to participate
(n = 393) Statistics
Parent
Average Age
M
(
SD
) 41.93 (7.05) 41.40 (7.38) 42.40 (6.73) t=1.789
% Graduated from College 45.2 44.4 47.8 χ2=0.414
Ethnicity (%)
White 37.8 39.5 33.3 χ2=1.502
African American 32.3 36.9 20.2 χ2=11.950
**
Asian 16.8 13.8 24.8 χ2=8.178
**
Hispanic 17.1 16.8 17.8 χ2=0.075
Pacific Islander 1.0 0.9 1.6 χ2=0.430
American/Alaskan Native 3.1 4.0 0.8 χ2=3.259
Child
% Female 47.4 49.6 45.5 χ2=1.205
** p
< .01
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Byrnes et al. Page 16
Table II
Parent and Child Characteristics by Enrollment
Variables All families
(N = 351) Enrolled
(n = 214) Did not enroll
(n = 137) Statistics
Parent
Average Age
M
(
SD
) 41.40 (7.38) 42.39 (7.16) 39.89 (7.48) t=-2.901
**
% Graduated from College 44.4 53.7 29.3 χ2=19.808
***
Ethnicity (%)
White 39.5 48.6 25.4 χ2=18.335
***
African American 36.9 30.7 46.3 χ2=8.404
**
Asian 13.8 16.8 9.0 χ2=4.289
*
Hispanic 16.8 14.1 20.9 χ2=2.704
Pacific Islander 0.9 0.9 0.7 χ2=0.034
American/Alaskan Native 4.0 4.2 3.7 χ2=0.048
Child
% Female 49.6 50.9 47.4 χ2=0.407
*p
< .05,
** p
< .01,
*** p
< .001
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Byrnes et al. Page 17
Table III
Neighborhood Characteristics: Mean Percentages across Census Block Groups
Variables All families
(N = 744) Range
Low SES M (SD)
Unemployment 6.44 (5.40) 0.00 - 43.54
Below poverty line 10.45 (10.00) 0.00 - 61.64
On Public assistance
a
4.73 (5.99) 0.00 - 40.10
High school dropout 17.55 (14.22) 0.00 - 72.77
Female headed households
a
11.35 (9.12) 0.00 - 55.80
Residential Instability M
(
SD
)
Moved past 5 years 17.22 (8.62) 0.00 – 69.80
a
Variable refers to percentage of
households
in neighborhood, whereas other variables refer to percentage of
persons
.
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Byrnes et al. Page 18
Table IV
Associations between Neighborhood Disorganization and Agreement to Participate
βSE βWald's χ2peβ (odds ratio
Sample 1
Individual
Child Gender -0.20 0.16 1.50 0.221 .82
Parent age -0.02 0.01 2.90 0.089 0.98
Parent white -- -- -- -- --
College graduate -- -- -- -- --
Neighborhood
% Unemployment -0.31 0.12 6.75 0.009 0.73
**
% below poverty line 0.17 0.16 1.17 0.279 1.18
% on Public assistance -0.23 0.14 2.58 0.108 0.80
% High school dropout 0.20 0.14 2.03 0.154 1.22
% Female headed 0.19 0.12 2.56 0.109 1.21
% moved past 5 years -0.03 0.09 0.12 0.724 0.97
Sample 2
Individual
Child Gender -0.07 0.24 0.09 0.758 0.93
Parent age -0.02 0.02 1.77 0.183 0.98
Parent white 0.31 0.27 1.34 0.248 1.36
College graduate -0.03 0.26 0.01 0.904 0.97
Neighborhood
% Unemployment -0.28 0.16 2.81 0.094 0.76
% below poverty line 0.11 0.24 0.22 0.641 1.12
% on Public assistance -0.27 0.22 1.52 0.218 0.76
% High school dropout 0.33 0.23 2.11 0.147 1.39
% Female headed 0.36 0.18 4.20 0.040 1.43
*
% moved past 5 years -0.11 0.12 0.80 0.370 0.90
*p
< .05,
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Byrnes et al. Page 19
** p
< .01
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Byrnes et al. Page 20
Table V
Associations between Neighborhood Disorganization and Enrollment
βSE βWald's χ2peβ (odds ratio
Sample 3
Individual
Child Gender -0.26 0.25 1.08 0.298 .772
Parent age 0.04 0.02 4.35 0.037 1.04
*
Parent white -- -- -- -- --
College graduate -- -- -- -- --
Neighborhood
% Unemployment 0.12 0.17 0.49 0.485 1.13
% below poverty line 0.08 0.25 0.11 0.743 1.09
% on Public assistance 0.31 0.22 1.92 0.165 1.36
% High school dropout -0.58 0.21 7.57 0.006 0.56
**
% Female headed -0.16 0.16 1.00 0.317 0.85
% moved past 5 years 0.08 0.14 0.34 0.561 1.09
Sample 4
Individual
Child Gender -0.20 0.26 0.61 0.436 0.82
Parent age 0.01 0.02 0.57 0.448 1.01
Parent white 0.73 0.30 6.07 0.014 2.08
*
College graduate 0.74 0.28 6.88 0.009 2.09
**
Neighborhood
% Unemployment 0.12 0.18 0.47 0.493 1.13
% below poverty line -0.09 0.26 0.11 0.736 0.92
% on Public assistance 0.41 0.23 3.25 0.071 1.51
% High school dropout -0.38 0.22 3.09 0.079 0.68
% Female headed -0.10 0.17 0.34 0.562 0.91
% moved past 5 years 0.13 0.15 0.84 0.361 1.14
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Byrnes et al. Page 21
*
p < .05,
** p
< .01
J Behav Health Serv Res
. Author manuscript; available in PMC 2013 April 01.
