A mixed methods study of parents’ social connectedness in a group-based parenting program in low-income communities

Abstract

Introduction:

Group-based parent training (PT) is an evidence-based approach for strengthening parenting skills and reducing child behavior problems. However, there has been little research on the social connections (SC) formed among PT participants, particularly in low-income communities where parents may be more socially isolated. This study describes SC formed among parents in a group-based PT program implemented in their children’s school and its association with changes in child behavior.

Methods:

Using a convergent mixed-methods design, data collection occurred between 2020-2022. Parents (n=97) completed measures of their SC to other parents in their PT group and their child’s behavior. Qualitative interviews with a representative subsample of parents (n=17) were also conducted to understand parents’ perceptions and experiences of SC within their PT group.

Results:

Parents reported high levels of SC (M=4.45 [range=3.04-5 on scale of 1-5]; SD=0.4). From baseline to post-intervention, the number of children with child behavior problems significantly decreased (32.12%, 37.5% behavior intensity and problems respectively). The magnitude of decline in child behavior problems was significantly related to parents’ SC (b=−11.52, p=0.02, SE=4.99). Qualitative data confirmed high levels of SC, which parents linked to improvements in their parenting and children’s behavior. Themes focused on the building of connections, committing to a safe space with parents who share similar goals, supporting one another, and gaining connections within the school environment and during the COVID-19 pandemic.

Conclusions:

Results highlight the potential synergistic effects of SC formed in the context of group-based PT with implications for strengthening parenting skills and children’s well-being.

Introduction

Social connectedness (SC) is the sense of belonging and trust one has to other people within their community (Haslam et al., 2015; Lee & Robbins, 1995). SC can have important health benefits as it may reduce social isolation, a phenomenon that has been linked with several health risks such as depression and premature death (Bess & Doykos, 2014; Holt-Lunstad et al., 2015; Keating-Lefler et al., 2004; Rank et al., 1998). Research surrounding SC has generally focused on exploring its relationship to loneliness and isolation within older adult populations and in those with mental health disorders (Hare Duke et al., 2019; Haslam et al., 2015; O’Rourke & Sidani, 2017). However, there is growing interest in the role of SC among other populations at risk of social isolation, such as parents of young children in low-resourced settings (Bess & Doykos, 2014; Keating-Lefler et al., 2004; Plesko et al., 2021; Rank et al., 1998)

In 2022, The Global Alliance for Behavioral Health and Social Justice published a policy brief highlighting the importance of research surrounding social connectedness within the built environment of individuals and communities (Hassen et al., 2022). Parents who are raising children in low-resourced settings are more likely to face barriers to establishing social connectedness within their communities. For example, if parents feel their communities are unsafe, they may avoid socializing outside their home in order to protect their children from harm (Parsons et al., 2019). Another barrier parents face in low-resource settings include difficulty accessing resources that help to facilitate social connections, such as reliable modes of transportation (Plesko et al., 2021). Therefore, creating opportunities for parents raising young children in low-resourced settings to be able to meet and engage with other parents in their communities may be important for helping reduce social isolation in parents as well as allow them to gain access to important community resources through gained social connections (Rangel et al., 2020).

Social connectedness as it relates to parents’ health is an important area of study as children’s health and well-being are integrally linked to that of their parents’ (National Academies of Sciences, Engineering, and Medicine, 2016). Evidence suggests SC is protective of the health and well-being of parents and their children living in under-resourced communities. For example, increased SC has been linked to lower parenting stress and is a potential mediator for decreased rates of parental aggression and child abuse (McLeigh et al., 2018; Plesko et al., 2021; Prendergast & MacPhee, 2020). Despite these findings, there are gaps in the literature surrounding the impact of SC on parents and their children in low-resourced settings. In particular, existing literature is lacking parents’ perspectives. There is a need to further identify how parents’ SC relates to their parenting behaviors, and in turn, shapes their children’s outcomes.

Social Connectedness in Group-based Parenting Interventions

Group-based parent-training (PT) programs are evidence-based interventions that strengthen participants’ parenting skills. There is strong research evidence supporting group-based PT for the management of disruptive child behavior problems (Epstein et al., 2015; Gross et al., 2018). Typically, these PT programs are brief interventions, lasting several weeks or months, led by trained group facilitators who support parents’ learning with information about a range of behavior management skills through lecture, discussion, handouts, and practice assignments. Often, these programs are embedded within community programs or are delivered within school settings. This environment provides parents the opportunity to meet and engage with other parents within their own community, potentially providing opportunities to gain SC and its related benefits.

The Chicago Parent Program

The Chicago Parent Program (CPP) is a 12-week, group-based PT program specifically designed for supporting parents raising young children in low-resource settings. CPP is effective in reducing children’s disruptive behavior, with sustained effects up to one year later (Gross et al., 2009; Bettencourt et al., 2019). CPP was developed in collaboration with an advisory board of Black/African American and Hispanic parents from low-income communities and has been shown to improve parenting skills and confidence, and reduce behavior problems in young children 2-8 years old (Gross et al., 2009). The intervention is led by trained PT facilitators, employed by the school, who teach parenting skills to the groups of parents through video vignettes, group discussions, role play, and weekly practice assignments Bettencourt et al., 2019). The program is structured in a way that supports parents in a non-stigmatizing and non-judgmental forum, facilitates problem solving, and encourages parents to share their ideas and experiences with each other. As part of the first session, group members are oriented to a list of ground rules for participating in the group such as “respect everyone’s perspective”, “have an open mind”, and “one person speaks at a time”. These group rules are reviewed at the beginning of each CPP session and parents are also encouraged to add additional rules if they feel they are important to the group.

In a recent study, CPP was offered universally to parents of pre-kindergarten children in 12 Title 1 Baltimore City Public Schools (i.e., schools serving a large proportion of students from low-income communities: Bettencourt et al., 2019). Results showed that CPP was associated with a 37% decrease in the number of children with clinically significant behavioral problems after the intervention. After the study ended, participating schools continued offering CPP in partnership with a community organization and with financial support from the school district (Fund for Educational Excellence, 2022). One of the benefits reported by school staff and CPP facilitators was the effect the program seemed to have on parents’ relationships with one another formed within the CPP group. Despite the program’s documented effectiveness in supporting parenting and lowering child disruptive behavioral problems, the potential SC formed within this group-based PT program has yet to be explored.

Current Study

This mixed-methods study was nested within an on-going parent study clinical trial (clinical trial number withheld for peer review) evaluating CPP effects among parents who had experienced childhood traumas and adversity. The purpose of this nested study was to examine the extent to which group-based PT, delivered in schools serving students from predominantly low-income urban communities, generates SC among parents and if that SC is associated with decreases in disruptive child behavior problems, an important outcome of PT interventions. For the quantitative phase, we examined the extent to which SC was formed among parents within the context of the PT group and whether SC was associated with improvements in their children’s behavior. We hypothesized that greater SC would be associated with greater reductions in their children’s behavior problems. For the qualitative phase, we sought to understand how parents with varying degrees of SC described their connectedness to other parents from their PT group, including factors that facilitated or hindered its emergence, and the impact they believed SC had on the parenting skills they learned in PT.

Methods

This study was approved by Johns Hopkins University’s and the school district’s Institutional Review Boards. Participants provided informed consent prior to participating in each phase of the study. A prospective convergent mixed-methods study design was chosen in order to provide an in-depth understanding of parents’ SC related to their experience in the PT program. Consistent with a convergent approach, quantitative and qualitative data were collected and analyzed during the same phase of the study and then merged to compare findings for interpretation. The quantitative strand used a single group pre-test/post-test design. Quantitative data collection occurred from October 2020 through July 2022 across four academic semesters in which CPP was offered in Baltimore City Public Schools. Qualitative interviews occurred between November 2021 and April 2022. During the study period, 20 CPP groups were implemented across 10 Baltimore City Public Schools. More information about the study protocol has been previously described in publications (Plesko et al., 2022).

In March of 2020, Baltimore City Public Schools shut down in-person programing due to the COVID-19 pandemic. With this change, CPP groups transitioned from in-person to synchronous virtual group sessions via online video platforms, such as Zoom. Although the implementation format changed, the CPP content and curriculum remained the same as for in-person. In September of 2021, Baltimore City Public Schools reopened for student classroom instruction. However, CPP groups continued exclusively in a virtual format due to ongoing restrictions within the schools surrounding extracurricular programming. Therefore, all of the PT groups described in this study were conducted virtually.

Quantitative Methods

Participants

Participants were eligible for this study if they were (a) over 18 years old, (b) a parent/caregiver of a child aged 2-8 years old who was enrolled in school (e.g. enrolled in the school’s Head Start, Prekindergarten, or elementary school program) (c) the parent enrolled in the CPP parent group at their child’s school, and (d) spoke either English or Spanish. Parents were excluded from this study if they never attended any CPP sessions, as they would be unable to provide a valid assessment of their SC to other members of the CPP group. Parents were not excluded if they had attended a previous CPP program at their child’s school.

During the two years of data collection, 204 parents participated in the CPP program. Of these parents, 135 parents (66.18%) enrolled in this study; 65 parents (31.87%) did not respond after multiple invitations to participate and 4 (3.0%) declined participation. For the 135 parents who were successfully recruited, 19 (14.1%) were ineligible to participate and 19 (14.1%) were lost to follow up at Time 2 and leaving a final sample size of 97 participants for the quantitative phase. A CONSORT flowchart (Schulz et al., 2010) of the recruitment of participants is included as Supplement 1. Of the 97 participating parents, 23.71% had previously participated in another CPP group.

Chi-square tests were run to assess for differences between the 97 participants included in the final sample and the 38 parents who were excluded from the study sample. There were no statistically significant differences between those who were included and those who were excluded from the final sample based on the demographic covariates included in the final statistical models (parents’ job status, race, education level, and age or whether they participated in an English- or Spanish-language CPP group). There were also no differences between those included and excluded from analyses based on parents’ relationships with other parents in the group at baseline or baseline child behavior problems.

Measures

Social Connectedness.

Participants’ social connectedness (SC) was measured at the conclusion of the CPP intervention (T-2) using a 25-item measure, the Social Connectedness in Group Environments Scale, an adapted version of the Intervention Group Environment Scale (Wilson et al., 2008). The measure was selected for use because it assesses key components of the construct of SC and was designed for use with group-based interventions. The measure was modified by the study team to assess SC within the CPP intervention prior to being translated and back translated from English to Spanish. Modifications included changing language from “Group leaders” to “Group facilitators” to match language used within the CPP program. The Likert-type scale measures social connectedness formed in the context of group-based interventions, and includes three subscales measuring cohesiveness (i.e., sense of belongingness to other group members, n=8 items, Cronbach’s α = 0.88); group leader preparedness (i.e., participant’s perceptions of their engagement in group activities and the group facilitator’s contributions, n=11 items, Cronbach’s α = 0.93 for this study sample); and negative or counterproductive interactions among group members (i.e., extent to which participants perceived negative group interactions, n=6 items, Cronbach’s α = 0.86 in this study sample). Item responses range from 1 (strongly disagree) to 5 (strongly agree) and total scale scores are averaged and can range from 1 (lowest level of SC) to 5 (highest level of SC; total scale Cronbach’s α = 0.97 for this study sample).

Child Behavior Problems.

Parent-reported child externalizing behavior problems (e.g. aggression, inattention, destructive behavior) were assessed before and after the CPP intervention using the Eyberg Child Behavior Inventory (Eyberg & Pincus, 2018). The ECBI is a widely used and well-established measure that evaluates child externalizing behavior on two scales: the Intensity Scale, which asks parents to report how frequently child behavior problems are occurring (n=36 items, response options range from “Never” = 1 to “Always” = 7; total scores are the sum of responses to 36 items which can range from 36-352) and the Problem Scale, which asks parents to report whether the child’s behavior is problematic for them (n=36 items, response options are “Yes” = 1 or “No” = 0; total scores are sum of responses to 36 items which can range from 0-36; (Eyberg & Pincus, 2018). The ECBI is normed and has clinical points to indicate the presence of clinically significant behavior problems on the Intensity Scale (scores ≥ 131) and the Problem Scale (scores ≥ 15). The measure has high internal consistency (Cronbach’s α = 0.98 and 0.96 for Intensity and Problem Scales respectively) and has been a sensitive indicator of change in child behavior problems in prior studies using CPP (d=0.31 intensity and d=0.33 problem scale changes). The ECBI is available in English and Spanish.

Baseline Parent Relationships.

To help control for how well parents knew each other prior to starting the CPP intervention, data were collected at baseline asking parents to indicate how well they knew each of the other parents enrolled in the program before attending the CPP group. Participants were given a list of names of the other parents signed up for their CPP group and asked to rate how well they knew each parent prior to beginning CPP using the following response options: 1- Don’t know them, 2-Know of them, but we’ve never talked, 3-Know them, but I don’t know them well, 4- Know them fairly well, 5-Know them very well. Average scores were calculated for each parent as their “Baseline Parent Relationship” score.

Quality of Parent PT Group Engagement.

Parent’s participation quality was measured using the Parent Engagement Rating Scale to describe the extent to which each participant actively engaged in the CPP group discussions. Quality of parents’ engagement was assessed by CPP group facilitators for each parent in the group at the conclusion of the CPP program. The scale consists of seven questions ranked on a Likert-scale from 1- Not at all to 4 (Most of the time) and includes questions regarding parents’ engagement in discussions with other group members, completing practice assignments, and overall quality of participation (example: “In your opinion, how much did this parent actively participate in the discussions during the parent groups they attended”). Scores range from 7-28, with higher scores indicating greater engagement. Previous studies have demonstrated positive associations between parents’ PT group engagement and changes in their child behavior problems after the CPP intervention (Garvey et al., 2006). Cronbach’s alpha for this scale was 0.91.

Quantity of Participation in CPP group.

Parents’ participation was also measured by number of CPP sessions attended. Attendance ranged from 1 to 12 sessions.

Demographic Background.

The following sociodemographic variables were collected from the parents at baseline and assessed as potential covariates: parent’s caregiver role (e.g., mother, father, grandparent, foster parent); parent age, race, ethnicity, education, job status; CPP intervention language (English or Spanish); annual household income; child age; and whether parent had previously attended the CPP program.

Procedures

To recruit participants for this study, research assistants attended “Meet and greet” CPP information sessions or the first CPP session to present the study to parents. Group facilitators also advertised the study using electronic flyers that were distributed to parents enrolled in CPP. Parents who were successfully recruited then completed survey measures at baseline (Time 1) and post-intervention (Time 2), with approximately 3 months between the two survey times. All recruitment and study materials were available in English and Spanish and both English and Spanish-speaking research team members were available to support recruitment and data collection depending on participants’ language preferences. All data were collected virtually by phone or videoconferencing. Quantitative data were entered directly into and managed within the study’s REDCap account, a secure web application for collecting and storing data (Harris et al., 2009).

Quantitative analyses

Data were analyzed using STATA statistical software (StataCorp, 2019). Data were first checked for completeness and explored to ensure assumptions of linear regression were met. Two multiple linear regression models were created to analyze the associations between SC and child behavior problems, one using the change in the ECBI Problem Scale scores and a second using the change in the ECBI Intensity Scale score. Multi-level analyses were used to control for participants being nested within CPP groups at 10 schools.

Covariates were considered for inclusion in the final model based on their relationship with SC and the child behavior scores as well as the variance inflation factor to assess for multicollinearity. A p-value of 0.1 was used as the cut off for the relationship between the variables and the covariates to include within the model. Using a larger p-value than 0.05 when selecting covariates is a common statistical practice to help reduce Type II errors when selecting these variables (Dales & Ury, 1978; Mickey & Greenland, 1989). Both adjusted and unadjusted models were explored. There were no missing data for the independent and dependent variables of the models. However, there were missing data for PT group engagement, education level, baseline parent relationships, child age, and household income. Missing data ranged from 1% (baseline parent relationships) to 12.4% (quality of participant engagement in CPP groups) of responses for these variables. Multiple imputation was used to handle these missing data.

Qualitative Methods

Participants

Purposive and theoretical sampling (Strauss & Corbin, 1998) were used to recruit participants in the qualitative arm. Participants were recruited from eight CPP groups conducted across five schools to maximize variability in SC scores and school locations. Participants were recruited based on their SC scores to represent variability and explore the complexity of SC in participants’ social-psychological world. A mean split of SC scores was used to identify “low” and “high” SC. Of the 97 participants included in the quantitative phase, 27 were contacted for interview; three declined to participate in the additional interview and seven did not respond to recruitment efforts. The three parents who declined and five of the seven parents who did not respond to recruitment efforts had SC scores below the sample mean.

Procedures

An interpretative phenomenology approach (IPA; Smith & Osborn, 2003) was used for the qualitative phase of this study. IPA is a qualitative methodology with a central focus on meaning and a goal to understand the content and complexity of those meanings in respondents’ social-psychological world (Smith & Osborn, 2003). IPA was well-suited for this phase because of our interest in the meaning and complexity of SC from the perspective of parents in their social-psychological world (e.g., being a parent of a young children in low-resourced communities). Consistent with IPA, the meanings of SC are not transparently available, we obtained them through a sustained engagement with the interview transcript and a process of interpretation (Smith & Osborn, 2003).

English qualitative interviews were completed by the first author, while Spanish qualitative interviews were conducted by the fourth author, who is fluent in both English and Spanish. Qualitative participants were interviewed once in a virtual face-to-face interview using a video-conferencing platform, however only the audio portion was recorded. Participants joined the virtual interview from a private place of their choice. The average length of interviews was 30:18 minutes (Range=17:58 - 50 minutes). Interviews were transcribed verbatim by a professional transcription service. Spanish interviews were first transcribed in Spanish then translated into English by professional translators. Final translated transcripts were then reviewed by the fourth author to ensure accuracy.

Interview guide

A semi-structured interview guide was developed by the study team and pilot tested with the first three qualitative participants. Additional interview questions were added in response to emerging themes and driven by the theoretical sampling (Strauss & Corbin, 1998). See Supplement 2 for the final interview guide.

Qualitative analyses

Qualitative data collection and analysis were conducted concurrently and iteratively. Weekly meetings were held to discuss qualitative analyses and data between the study’s PI and qualitative coders. F4 Analyze, a qualitative analysis software, was used to manage qualitative data analysis, document memos, and generate a directory of participant’s phrases that support related themes (audiotranskription, 2022).

Consistent with the interpretive phenomenology analysis, the first stage of qualitative analysis involved looking in detail at the first three interviews before moving on to examine further interviews. Transcripts were read multiple times prior to coding in order to become as familiar as possible with the content. Similarities and differences were compared across the three interviews to develop depth in the analysis. At this stage, the analysis resembled free textual analysis, and codes were developed that represent events, actions, or interactions relevant to SC.

The codes from the first three interviews were used to comprise the initial code book which guided our subsequent analysis. The code book constantly evolved throughout the coding process to respect the convergences and divergences in the data. New codes were added into the code book if the existing codes failed to capture events, actions, or interactions relevant to SC and clustering codes were collapsed into a theme. The next stage of analysis involved looking for connections between codes (Smith & Osborn, 2003). In this stage, we engaged in an iterative and interactive process of (1) discovering emerging themes, or codes that cluster together, (2) making sense of the connections between the emerging themes, and (3) checking in the transcript to make sure our interpretation reflect the actual words of the participant. A final codebook of themes and sub-themes were generated after all transcripts had been analyzed using this interpretative process. After piloting the interview guide, participants were recruited to enrich the emerging conceptual framework and recruitment continued until data saturation was achieved, that is, no new themes and subthemes were found in the new interview analysis (Morse, 2000).

Qualitative Rigor

Distinguishing what the participant said and the analyst’s interpretation is critical for IPA (Smith & Osborn, 2003). Data analysis was conducted collaboratively by and discussed frequently among the research team, which prevented the imposition of individual interpretations over those of participants (Sandelowski, 2016). To ensure trustworthiness, written memos were used as an audit trail throughout the analysis. Later in the analysis, member checking was used to seek confirmation on the study findings from the participants. Reflexivity was maintained through using direct quotes from participants in this manuscript. To increase transferability of the study’s findings, our interview guide and in-depth description of the data analysis process are provided in this manuscript.

Mixed-Methods Analysis

Following separate analyses, the quantitative and qualitative results were integrated and interpreted. The mixed data analysis consisted of evaluating for convergent and divergent themes across the data (Creswell & Plano Clark, 2018). Figure 1 represents the joint display of participants’ experiences in the group by their SC score.

Figure 1.

Joint Display of Qualitative Responses Based on Participant Social Connectedness Score

Results

Quantitative Results

Participant demographic data are described in Table 1. CPP groups ranged from 5-18 parents per group. The majority of study participants were female (94.85%) with most self-identifying as mothers (82.47%). Additionally, 68.04% of the study participants reported annual household incomes that meet low-income criteria of $40,000 or less. Descriptive participant data related to the CPP intervention are reported in Table 2. To be eligible for the study, parents had to attend at least one CPP session. On average, participants attended 9.88 (SD=2.58) or 82.3% of the 12 CPP session. On average, group facilitator ratings of parents’ PT group engagement quality were high (M=25.18, SD 3.65).

Table 1.

Demographic Characteristics of Participating Parents

Variable	Quantitative Sample N=97	Qualitative Sub Sample n=17
Parent Gender, n (%)
Female	92 (94.85)	15 (88.24)
Male	5 (5.15)	2 (11.76)
Caregiver role, n (%)
Mother	80 (82.47)	12 (70.59)
Father	5 (5.15)	2 (11.76)
Stepmother	1 (1.03)	0
Foster parent	1 (1.03)	0
Grandmother	7 (7.22)	3 (17.65)
Other: “Legal Guardian”	3 (3.09)	0
Parent age, mean (SD)	25.36 (8.9) Range 20-64	37.94 (10.55) Range 23-64
Child age, mean (SD)	4.55 (1.52)* Range 2-8	4.33 (1.50)** Range 2-7
Race, n (%)
Black	59 (60.82)	9 (52.94)
White	7 (7.22)	2 (11.76)
Multiracial	7 (7.22)	3 (17.65)
Other	24 (24.74)	3 (17.65)
Ethnicity, n (%)
Hispanic or Latino	30 (30.93)	5 (29.41)
Non-Hispanic or Latino	67 (69.07)	12 (70.59)
Highest Education Level, n (%)
High School Education or Less	45 (46.39)	7 (41.2)
Post High School Education	51 (52.58)	10 (58.8)
Decline to Answer	1 (1.03)	0
Household Income, n (%)
<$40,000 per year	66 (68.0)	15(88.2)
≥$40,000 per year	27 (27.84)	2 (11.76)
Declined to answer	4 (4.12)	0
Current job status, n (%)
Working full-time	26 (26.8)	4 (23.53)
Working part-time	17 (15.53)	4 (23.53)
Going to school	3 (3.09)	0
Not working	49 (50.51)	9 (52.94)
Other	2 (2.06)	0
Marital Status, n (%)
Married or Living with Partner	40 (41.24)	6 (35.29)
Single	57 (58.76)	11 (64.71)
CPP Language, n (%)
English	73 (75.26)	14 (82.35)
Spanish	24 (24.74)	3 (17.65)
Previously Attended CPP, n (%)
Yes	23 (23.71)	8 (47.06)
No	74 (76.29)	9 (52.94)

^*n=81;

^**n=12

Table 2.

Parent Relationships and Social Connectedness Before and After CPP groups

	Quantitative Arm N=97		Qualitative Arm n=17
Variable	Mean (SD)	Range	Mean (SD)	Range
Baseline Relationships	1.55(0.72)	1-4.25	1.71(0.66)	1-3.1
Number of Sessions Attended	9.88(2.58)	1-12	10.18(2.1)	7-12
Social Connectedness Score	4.45 (0.40)	3.04-5	4.48 (0.51)	3.04-5
Cohesiveness Subscale	4.32 (0.52)	2-5	4.36 (0.74)	2-5
Preparedness Subscale	4.5 (0.43)	3.55-5	4.48 (.50)	3.55-5
Counterproductive subscale	4.6 (0.49)	3.5-5	4.65 (0.51)	3.5-5
Quality of Parent PT Group Engagement	25.19(3.65)	10-28	28.82(3.23)	15-28

Note: There were data missing from one parent for baseline relationships (n=96) and for twelve parents for quality of parent PT group engagement (n=85).

Baseline Parent Relationship and SC scores are reported in Table 2. Baseline Parent Relationship scores assessing how well participants knew each other prior to starting the CPP group were low, with an average score of 1.55 (SD = 0.72); 36% of parents indicated that they did not know any of the other parents in their CPP group when they started and 75% of the parents had an average score of less than 2 (“Know of them, but we’ve never talked”).

Following CPP, parents’ ratings of their SC to other members of their CPP group were high. As shown in Table 2, the average SC score was 4.3 out of a maximum possible score of 5. SC scores were not significantly correlated with either parents’ attendance scores (r=0.11, p=0.30) or group facilitators’ ratings of parents’ quality of PT group engagement (r=0.11, p=0.30).

Child behavior scores are displayed in Table 3. Changes in child behavior intensity and problems were not correlated with parents’ attendance (Intensity r=−0.07, p=0.49; Problem r=−0.14, p=0.17) or their quality of PT group engagement (Intensity r=0.03, p=0.78; Problem r=0.04, p=0.74). A clinically meaningful way of assessing behavior change on the ECBI is to evaluate if behavior has improved, worsened, or stayed the same related to the clinically significant cut off score. Table 4 displays results from a chi-squared test showing changes in children’s clinical behavior status before and after the intervention. For child behavior intensity, 65 children were within the normal behavior range at both Time 1 and Time 2, 13 children who had clinically significant behavior problems at Time 1 improved to the normal range by Time 2, 15 children were in the clinically significant behavior range at both Time 1 and Time 2, and 4 children who had normal behavior at Time 1 had clinically significant behavior at Time 2. For the child behavior problem scale, 60 children were within the normal range at both Time 1 and Time 2, 17 children who had clinically significant behavior problems at Time 1 improved to the normal range by Time 2, 15 children were in the clinically significant behavior range at both Time 1 and Time 2, and 5 children who had normal behavior at Time 1 had clinically significant behavior at Time 2.

Table 3.

Mean (SD) Child Behavior Intensity and Problem Scores and Percent with Scores in the Clinical Range at Baseline and Post-Intervention

	Quantitative Phase N= 97				Qualitative Phase n=17
Variable	Mean (SD)	Median	Range	Clinically significant (n, %)	Mean (SD)	Median	Range	Clinically significant (n, %)
ECBI Intensity Scale
Time 1	104.1 (42.59)	99	36-230	28 (28.87)	100.65 (41.7)	99	45-183	5 (29.41)
Time 2	100.3 (38.58)	91	43-214	19 (19.59)	96.94 (34.91)	86	43-198	1 (5.88)
ECBI Problem Scale
Time 1	10.14 (8.76)	9	0-35	32 (32.99)	12.18 (8.83)	14	0-31	7 (41.18)
Time 2	9.27 (8.15)	8	0-31	20 (20.62)	9.35 (9.64)	7	0-29	5 (29.41)

Table 4.

Chi-squared child behavior clinical range change between time 1 and time 2

	Child Behavior Intensity post CPP
Child Behavior Intensity Before CPP	Normal Range	Clinically Significant	Total
Normal Range	65 94.20%	4 5.80%	69 100%
Clinically Significant	13 46.43%	15 53.57%	28 100%
Total	78 80.41%	19 19.59%	97 100%
Pearson chi2 28.86 p=0.00
	Child Behavior Problem post CPP
Child Behavior Problem Before CPP	Normal Range	Clinically Significant	Total
Normal Range	60 92.31%	5 7.69%	65 100%
Clinically Significant	17 53.12%	15 46.88%	32 100%
Total	77 79.38%	20 20.62%	97 100%
Pearson chi2 20.11 p=0.00

A paired t-test was conducted and determined there were significant decreases in behavior between Time 1 and Time 2 scores. For behavior intensity scores, there was a significant decrease between Time 1 and Time 2 scores for the 28 children who had the clinically significant behavior intensity at the beginning of the intervention (p=0.00; mean score Time 1: 158.36, Time 2=139.93). However, there was not a significant difference when comparing the entire sample of 97 participants’ Time 1 and Time 2 intensity scores (p=0.15, mean score Time 1: 104.10, Time 2: 100.36). Similarly, for the behavior problem scale there was a significant decrease between Time 1 and Time 2 problem scores for the 32 children who were in the clinically significant range at Time 1 (p=0.00; mean score Time 1=20.47, Time 2=15.75), but no significant decrease when comparing the scores for the entire sample (p=0.2, mean score Time 1=10.14, Time 2=9.27).

As shown in Table 5, SC and child behavior problems were significantly associated with several parent demographic background variables. Based on their significant associations with SC and child behavior problems, the following covariates were included in the final models: parent age, parent race, CPP intervention language, highest educational level achieved by parent, current job status, and baseline child behavior. Table 6 reports the adjusted and unadjusted models for the associations between SC and changes in ECBI intensity and behavior scores. A significant relationship was found between parents’ SC and the change in their child’s behavior based on the ECBI intensity scale score in both the unadjusted (b= −13.72, p=0.00) and adjusted models (b= −11.52, p=0.00). This indicates that the more parents were socially connected within the CPP group, the greater their reduction in child behavior intensity on the ECBI was following the intervention. In the model assessing parents’ SC and change in their child’s behavior based on the ECBI problem scale, there was a significant relationship in the unadjusted model (b= −1.77, p=0.05), but no significant relationship when controlling for the covariates (b= −0.73 p=0.062).

Table 5.

Correlation Table of Potential Covariates and Study Variables

	Social Connectedness	ECBI Intensity Change	ECBI Problem Change
Baseline Relationships	−0.03	0.04	0.01
Parent Race	0.22**	−0.01	−0.04
Parent Ethnicity	0.26**	0.07	0.08
CPP Intervention Language	−0.31**	−0.05	−0.01
Parent Gender	0.09	0.09	0.00
Previously Attended	0.08	0.13	0.14
Marriage Status	−0.11	−0.04	−0.11
Income	0.16	0.01	−0.07
Highest Education Level	0.18*	−0.06	−0.15
Parent Age	0.31**	−0.03	−0.06
Attendance	0.11	−0.07	−0.14
Parent PT Group Engagement	0.11	0.03	0.04

^*p ≤ 0.1,

^**p ≤0.05

Note: Ethnicity was significant in the correlation table but was dropped from the final model due to multicollinearity with the race and language variables. Job status was not included in the correlation table because it is a categorical table. Instead, regression was used for job status and each variable which showed a significant relationship at the p=0.1 level with ECBI intensity change (p= 0.0) and ECBI problem change (p = 0.06), therefore it was also included as a covariate in the model.

Table 6.

Unadjusted and Adjusted Multi-level Linear Regression Model for Associations between Parents’ Social Connectedness and Changes in Child Behavior Problems from Baseline to Post-Intervention

	Coefficient	SE	p-value	95% CI Low	95% CI High	ICC Group	ICC School
ECBI Intensity Scale
Unadjusted
Social Connectedness	−13.72	3.88	0.00	−21.31	−6.12	0.10	0.00
Adjusted
Social Connectedness	−11.52	4.99	0.021	−21.31	−1.74
Race (Other)
Black	−6.66	5.24	0.20	−16.93	3.61
Language (English)
Spanish	−16.01	5.84	0.01	−27.47	−4.57
Job Status (Not Working)
Working Full-Time	−2.20	5.10	0.67	−12.20	7.81
Working Part-Time	−7.47	2.60	0.00	−12.57	−2.38
Student	37.59	7.43	0.00	23.04	52.16
Education (Less than High School)
Post High School Education	−2.36	4.02	0.56	−10.25	5.52
Parent Age	0.25	0.16	0.13	−0.07	0.56
Baseline behavior intensity	−0.29	0.07	0.00	−0.42	−0.15
ECBI Problem Scale
Unadjusted
Social Connectedness	−1.77	0.91	0.05	−3.55	0.00	0.00	0.00
Adjusted
Social Connectedness	−0.73	1.48	0.62	−3.63	2.17
Race (Other)
Black	−0.64	1.73	0.71	−4.03	2.75
Language (English)
Spanish	−3.58	1.61	0.03	−6.73	−0.43
Job Status (Not Working or Student)
Working Full-Time	−2.60	1.88	0.17	−6.28	8.90
Working Part-Time	−0.31	0.53	0.56	−1.33	0.72
Student	2.74	3.14	0.38	−3.43	8.90
Education (Less than High School)
Post High School Education	−1.20	1.45	0.41	−4.03	1.64
Parent Age	0.03	0.06	0.54	−0.08	0.15
Baseline behavior problem	−0.39	0.08	0.26	−5.54	−0.23

Qualitative Results

Analysis of the qualitative interview transcripts identified five major themes related to parents’ experience of SC within the CPP group: 1. A spectrum of connectedness, from lacking connection to building connection, 2. Committing to a safe space with shared goals, 3. Teaching, learning, and supporting each other, 4. Connections within the School Environment and 5. Connecting during the COVID-19 pandemic. Within each of these categories, several sub-themes related to parents’ experiences emerged and are described below.

A Spectrum of Connectedness: From Lacking Connection to Building Connections

Many parents interviewed described a feeling of strong connection to the other parents in their group. Several parents remarked how they made friends within the group and some even described the relationships they developed with the other parents as closer than friends, and more like family:

“I think that we are a close-knit group, we are more like a family group now. So it really brought everybody together.” (P.6, 12 sessions, SC 5, high)

One participant, who was in a group that consisted of all women, also commented that the group became close over time:

“Everybody got along extremely well. It kind of became like a sisterhood” (P1, 8 sessions, SC 5, high)

One parent, who had previously participated in the CPP program, remarked that her experience feeling like she belonged with the other parents in the group was different each time she participated:

“The first time, I was like, I was the youngest one in it. So I kind of felt out of place. But now it was like I’m older and I’ve experienced parenting longer. So I kind of felt like I fit in more this time than I did the first time.” (P2, 12 sessions, SC 4.2, low)

The parent reporting the lowest SC score of the sample, described a lack of connectedness to the other parents in the group:

“[For] me personally…I don’t feel as though there was a connection. Unfortunately…I previously knew [some] parents, and it wasn’t like we had a contact before, and the group could either brought us together or set us further apart. And it didn’t really do either.” (P7, 7 sessions, SC 3.04, low)

Committing to a Safe Space with Shared Goals

Several themes emerged surrounding parents’ perspective on factors that either aided or hindered their ability to connect with the other parents. A sense of mutuality, either in shared experiences or shared goals for their children, was frequently discussed as an aid to forming these connections:

“I think that’s what made the dynamic of the group the best is everyone was from, you know, we all were raising the children, so we made a child, whether it was theirs or adopted or grandparent or foster, whatever it was, we all could still meet at the factor of we’re here because we want to be better to our, to be better with raising our children.” (P15, 12 sessions, SC 5, high)

One participant, a grandmother, remarked how having a mixture of parents and grandparents in the group helped her to feel like she belonged:

“Now, if I was the only grandparent in there. No, I wouldn’t have really liked that. But it was a different age group. It was a mixture. Yeah, it was kids, I keep saying kids- there was young adults in there- my kids’ age, you know. And then it was people in there my age. So like I said, I enjoyed it.” (P4, 7 sessions, SC 4.28, low).

Another facilitator for establishing connections within the group was the parents’ perception of the group being a safe space and a place free of judgment, where they could trust the other parents and share openly about their experiences:

“So we were just there, like no matter what your situation was, even if you, we thought it was foolish, we were not there to judge. That wasn’t the reason. But we were there to help you find a solution so that you can get through that situation.” (P6, 12 sessions, SC 5, high)

Others remarked the group was a place where people felt respected:

“If one person was talking. Everybody kind of listened, nobody was rude. So I really enjoyed it. It was wonderful for me.” (P1, 8 sessions, SC 5, high)

Parents also discussed how the group facilitators played a role in helping to contribute to an open environment between the parents:

“[The group facilitators] always brought us in, kept us connected, and always helped us relate to each other.” (P10, 12 sessions, SC 5, high)

Even parents who were not well connected with the other group members felt the support of the group facilitators:

“I didn’t feel a sense of connection to the other parents. But the group [facilitators] did try to make it so we could all see things from each other’s perspectives. So if I saw a situation one way and another parent saw it a different way, then she would mediate and say, well, what is it that makes you feel as though you see it this way. And then you know, we would explain it or define it. And then the other parent would give their explanation or definition so that we could all try to be on the same page.” (P7, 7 sessions, SC 3.04, low)

However, not all parents felt a sense of mutuality with other group members or that the group was a safe space to share. One parent with Low SC discussed her frustration when sharing her experience in the group while other parents did not agree with her as well as difficulties relating to other parents’ life experiences in the group:

“I’m the kind of mom that’s like, look, if you don’t act right, you will get your hand smacked… and there’s just certain moms to be like… ‘I don’t think that’s correct’. Or ‘I think you could have handled the situation in a different manner’… So it was a lot of like feedback where certain moms would say one thing and another mom would be like, ‘oh, but I do this, and I do that’, and it’s like, ‘Sweetie, this is reality. No one’s kid is perfect’… I just still feel as though that in some of the groups that they were just taking it to an extreme or exaggerating and not keeping it realistic.” (P7, 7 sessions, SC 3.04, low)

Another barrier to parents’ gaining connection within the group was how often they attended the group sessions and their commitment to the group.

“I mean we experience that feeling of belonging more with the parents who have attended consistently. Because since we already know each other a little bit more, then we could connect over that. But with the parents who are a bit sporadic, who go once and then don’t go and don’t come back. We keep them in mind, yes, but I don’t know, I don’t know, like we didn’t establish that solidarity or that camaraderie that we have with the parents we always see.” (P10, 12 sessions, SC 5, high)

Teaching, Learning, and Supporting Each Other

Parents frequently discussed themes of teaching, learning, and supporting the other parents they connected with, beyond what was taught in the group sessions. Parents frequently shared their own experiences with other parents and gained support from the group:

“Believe it or not, a lot of the parents were actually going through the same thing I am with temper tantrums and things like that. So we sort of give each other advice and help each other out. Like one’s sort of struggling with something. We just say it’s ‘okay, sweetheart, you got this. It’s fine’. We kind of, like, help each other out and build each other’s self-esteem up. Make it feel like they’re worth something. Because a lot of times when they’re trying to do stuff for a child and they’re getting overwhelmed, and they feel like they’re not worth anything. So we try to help each other out. Like if we feel like we’re struggling on one thing, we always go and say: ‘it’s okay. You got this. You got this. Just take a deep breath. Count to ten. Relax. Give them a crayon, give them a coloring book, talk to them like an adult. Give them an option. And let them understand the difference between right and wrong. Don’t stress yourself out’. And a lot of times when we’re teaching each other how to do these things, it actually makes it easier for everybody else because we give each other feedback and figure out a way to make it more easier to do stuff.” (P8, 12 sessions, SC 5, high)

Parents also related their connections with others helped with their uptake of the program and the PT skills taught:

“… when you um, hear the stories of how the other people [try the strategies taught in CPP], it makes you want to do it.” (P9, 11 sessions, SC 4.72, high)

The group discussion in addition to the parent training videos was an important part of the program for many parents:

“So that was one of my favorite parts, was watching the videos and hearing everybody else’s experience or their opinion and what they would do and what they wouldn’t do.” (P5, 11 sessions, SC 4.28, low)

Several parents remarked that they only tried the parent training strategies within the program because there were other parents in the group that had tried them and shared that the strategy had worked for them:

“It allows you to hear perspectives of other people, hear situations, connect to them in that, and it also allows you to give advice and take advice. Without that, I wouldn’t be able to get those advice. Um, of course, the videos were a big part of the group as part of the presentation, but at the same time they’re paid actors¹ from which, you know, I got to actually talk to real people.” (P15, 12 sessions, SC 5, high)

Connecting with other parents in the group also created a sense that parents were not alone in their parenting struggles:

“As a parent, sometimes you feel like you’re the only one that’s going through a lot of stuff. But when we hear other parents talk about their kids and the stuff they’re going through, you know, it don’t make you feel alone. So it was really, it really did help me” (P5, 11 sessions, SC 4.28, low)

These connections also helped to normalized parents’ experiences:

“Realizing that the things that you go through were normal, the frustrations that you go through as a parent, the happy times, the ups, the downs again, all of those things is normal. The behavior of your child, everything was just normal behavior.” (P15, 12 sessions, SC 5, high)

Connections within the School Environment

While a few parents shared that their participation in the CPP program encouraged them to become more engaged in their child’s school, many of the parents interviewed stated that they were already very involved in the school prior to the program:

“So I’m definitely involved with the education. Um, and yeah being involved in the, in the program just served as a motivation to keep doing it um and keeps driving at it. Sometimes it can be hard, especially when I come home from work and tired. But I just got to keep working at it.” (P9, 11 sessions, SC 4.72, high)

Other parents commented that the program helped them continue to stay involved within the school:

“It just showed me that uh, just to step out and get involved in different things, to show me how much more um, how, how much the teachers actually care and how many, you know, how much help I can get from just reaching out and using tools and the things that they have at the school…. I’ve been more involved. Um, I actually communicate more uh, sending (the teacher) emails and having her to contact me if she sees something that needs to be addressed.” (P1, 8 sessions, SC 5, high)

Additionally, parents shared that the school offering the CPP program made them feel valued as parents by the school and made them feel positively about the school and their child’s teachers:

“It makes me more confident and more positive that the teachers aren’t just there for their paycheck and that they’re really trying to give these young children education and help the parents understand the education so that we can progress you know, throughout our future because our children are our future.” (P7, 7 sessions, SC 3.04, low)

Connecting During the COVID-19 Pandemic

Themes emerged describing factors that influenced parents’ connections within the group. Most notably, the parents who participated in this study did so virtually during the COVID-19 pandemic. Some parents discussed how the virtual programming helped them connect with the parents:

“Well, if it was in person, I probably would have been more held back because I’m a little shy, but it seemed like since it was virtual, it was, I didn’t have nothing to worry about because I physically. I’m not physically in the same room with them. You know, we can be just on the video… [if in person], once I would have got to know them, maybe after two sessions, then I would have got, you know, would have opened up a little bit more. But it seemed like because it was virtual I was just open from day one.” (P3, 10 sessions, SC 4.52, high)

Parents even comment on the timing of the program during the pandemic helped them connect during an otherwise isolating time:

“[Virtual CPP was helpful because] I had somebody talk to. I had somebody to express how I felt. Because without virtual, I’m like, I feel like I’m losing it. I can’t talk to nobody. I can’t physically see somebody. I can’t tell somebody how I feel when I’m going through” (P8, 12 sessions, SC 5, high)

Other parents found virtual CPP inhibited their connections with parents. One parent, who had previously attended the CPP program in person, compared her experience between the in-person and virtual program saying:

“I think it was more so like being face to face with the people just made it easier because you’re like looking into their eyes while you’re telling your story, then they are telling their story, it’s like you connect with them better.” (P2, 12 sessions, SC 4.2, low)

One parent with low SC felt that the virtual setting made parents act differently than they would have if they had been together in person:

“But it was also with it being virtual, because when you’re virtual like, you’ll say stuff online that you wouldn’t say if you’re sitting close to somebody, like even if you have a six-foot distance. I guarantee more than 75% of those people wouldn’t have made the same remarks or same expressions that they would have made if you were in the same room.” (P7, 7 sessions, SC 3.04, low)

Discussion

This convergent mixed-methods study examined the SC generated among parents participating in a group-based PT program, the Chicago Parent Program (CPP) implemented virtually through their child’s school during the COVID-19 pandemic and how SC was associated with improvements in their children’s behavior following PT. The results of this study indicate that although parents did not know each other before enrolling in CPP, they formed strong social connections with one another by the end of the program. Moreover, SC was significantly associated with reductions in the intensity of their children’s behavior problems.

Qualitative findings also confirmed high SC among the parent participants. Parents discussed how forming connections with other parents was aided by a group environment that felt safe and was comprised of peers that could relate to their experiences. Parents also discussed how their connections within the group encouraged them to try the parenting strategies at home that were taught in the group, which may help explain the significant relationship between parents’ SC scores and decreases in child behavior intensity scores. This significant relationship may also be explained by parents’ SC and discussions with the other parents normalizing child behavior, potentially leading to parents rating their child’s behavior less severe after the intervention. To our knowledge, this is the first study that has explored the relationship between parents’ SC and PT outcomes. These results suggest that SC is an important therapeutic component of group-based PT that needs to be routinely measured and assessed for its effects on improvements in parenting skills and children’s behavior.

The participants in this study demonstrated high SC to other group members. Originally, a SC score of 3 or less was determined to represent “low” SC. However, the lowest SC score amongst all participants was a 3.04 and only 11.34% of participants had SC scores less than 4. When merged, qualitative results mirrored these quantitative findings, with little variability among parents’ perceptions of SC within the group. We found most parents reported positive experience with the other parents in the group, even those with “low” SC scores relative to the rest of the sample. Nonetheless, the parent reporting the lowest SC did describe feeling unwelcome in the group and judged by the other parents. The agreeance between these qualitative and quantitative findings furthers our confidence that both the SC quantitative measure and the qualitative interview guide captured parents’ experiences well.

The overall high SC scores speak highly to the safety and comfort parents felt within the group from both the group facilitators and among the other parent participants. CPP group leaders are trained to support parents as the experts of their children and maintain a non-judgmental environment. This allows for parents to share and learn from one another and strengthen their parenting skills and confidence. Additionally, the CPP program is offered universally to all parents of PreK through 2^nd grade students at each school, reducing the chance for parents to feel stigmatized when joining the program compared to a program that specifically targets parents of students with behavior problems. Despite offering the program universally to parents, almost one third of the children of parents who participated had clinically significant behavior problems at program entry. Having groups comprised of parents with a range of skills, strengths, and challenges may increase their therapeutic value. More research is needed to understand the relationships among PT group facilitation strategies, parent characteristics, and SC, as well as how SC affects a wider range of PT outcomes. The results of this study highlight the importance of building safe, non-judgmental, supportive group environments to foster SC among parents and optimize PT outcomes.

Parent attendance rates and group facilitators’ ratings of parents’ engagement in the CPP groups were high; average attendance was 9.88 or 82.33% of group sessions and average PT engagement scores was 25.19 (maximum possible score of 28). These data demonstrate that parents consistently attended and were highly engaged in the CPP groups. It is important to note that to ensure valid estimates of SC based on PT group participation, only parents who attended at least one CPP session were included in these analyses. However, their exclusion may have contributed to the limited variability in parent attendance and engagement data, and the non-significant associations with SC.

Child behavior problems also decreased significantly between the two data collection time points. This is consistent with previous studies evaluating the CPP intervention, which have also found sustained reductions in child behavior up to one year later (Gross et al., 2009). While there were significant decreases in both child behavior intensity and problem scores, SC scores were significantly related to the changes in children’s behavioral intensity and not to changes in the problem scale scores when the model was adjusted for multiple co-variates. It remains unclear why changes in parents’ perceptions of whether their child’s behaviors were problematic were unrelated to SC in the adjusted model though the result may be a function of the sample size and limited power to detect smaller effects. It would be important to replicate this finding using larger samples.

The results of this study are consistent with the theories that underlie the CPP structure. With theoretical underpinnings in Attachment Theory, CPP practices are focused on parents learning skills that help them to be more responsive, consistent, and nurturing with their children and reinforces the centrality of parents in children’s lives. CPP development was also guided by Social Learning Theory, which supports parents learning through watching others in common parenting situations and practicing parenting skills through role playing exercises and home practice assignments (Breitenstein et al., 2020). The qualitative themes that emerged from interviews within this study highlight these theoretical bases of CPP. The theme of Teaching, Learning, and Supporting Each Other demonstrates how parents benefited from being in a group with their peers, adding credibility to the skills being taught in the video vignettes by sharing their own personal experiences while supporting each other in their roles in their children’s lives. These findings highlight the power of group-based interventions, as parents benefit from learning from one another in addition to the content being taught by the facilitators.

This study coincided with the COVID-19 pandemic and therefore groups were facilitated online rather than in-person. This was the first time that the program had been implemented in a virtual setting for parent participants. However, by the beginning of data collection for this study, Baltimore City Public Schools had been conducting all school programming virtually for six months due to the pandemic. This helped the transition to virtual CPP, as parents already had vast experience with online programing and the district had supplied technology for virtual homeschooling to families in need. The existing format of CPP with structured video vignettes and facilitated group discussion also lent itself well to the transition to virtual meetings. Group facilitators were still able to play videos over the virtual meeting and lead meaningful discussions with participants. Measures were taken by facilitators to help groups run smoothly and promote connection of parents, such as requiring cameras of participants to be on (if possible) and muting themselves when they were not talking.

Despite the virtual format of the program, SC scores were high among participants as were attendance and group engagement scores. This is promising as virtual programming continues in many spaces, even as the pandemic resolves. However, what is not understood is how well parents are able to maintain their connections with one another after the program ends, and if SC varies based on the parents participating in a PT program online vs. in-person. It is possible that high levels of SC in the virtual format during the pandemic was a novelty effect, as many of the participants may have felt socially isolated and hungry for the safety of the social interactions made possible through virtual meetings. Future research is needed to determine if SC in virtual groups remains high in a time when virtual options are less essential for social interactions. A study by van Leuven and colleagues (van Leuven et al., 2022) comparing virtual and in-person PT group delivery during the pandemic found no differences in PT outcomes by format. However, group facilitators reported many challenges implementing the program and felt parents were often distracted and less able to concentrate in the virtual setting (van Leuven et al., 2022). While virtual formats allow some parents who work, are disabled, have limited access to transportation, or multiple competing responsibilities greater flexibility to participate, they also allow for more distractions during group sessions which may impact their ability to connect with the other parents. Some parents also remarked how they were more comfortable participating and sharing in an online meeting rather than in person, while others felt they could connect better had it been in person. Future research comparing virtual and in-person group-based PT formats is recommended to better understand the effects on SC, parents’ willingness to share and participate within the group based on the program format, and other PT outcomes post-pandemic.

The parents in this study enrolled in CPP with a shared goal of strengthening their parenting skills and their relationships with their children. About 24% of the parents had previously participated in CPP in their child’s school. As reported in a previous study evaluating CPP in schools, parents’ main reason for attending the program was their goal of wanting to strengthen their parenting skills and the parent-child relationship (Gross & Bettencourt, 2019). That 24% of parents had participated in CPP multiple times through their child’ school highlights the acceptability of universal PT programs in public schools and the need to create safe spaces for parents to connect with one another, particularly for schools located in under-resourced communities.

There are several limitations within this study. First, conducting this study during the COVID-19 pandemic presented many challenges with participant recruitment. For example, the study team was unable to enter school buildings to recruit parents directly, requiring research assistants to “cold call” parents by phone. This made it more difficult to contact parents to describe the study, results in lower enrollment. It is possible that with a larger sample size and greater variability in scores, more effects may have been detected. The lack of a control group for the intervention also limits the ability to make causal inferences about the associations between parents’ SC and changes in their children’s behavior. It is possible that the higher SC scores obtained at post-intervention may have been a function of parents becoming more engaged with their child’s school through other events, a confound that could have been controlled using a randomized, experimental design. Another limitation is the small number of parents interviewed reporting low SC scores. Using the SC sample mean to purposefully recruit parents for the qualitative phase of this study, nine parents with high SC scores and eight parents with “low” SC score were interviewed. However, 10 other parents purposefully selected to participate in the qualitative interviews based on their SC scores did not participate (three declined; seven were lost to follow up). Of these 10, eight had “low” SC scores and two had “high” SC scores. It is possible that the qualitative results reflect more positive experiences than would have been described if more parents with low SC scores had shared their perspectives. It is also possible that the underlying qualities of parents who did not feel comfortable sharing and engaging within the PT group setting may also be the same qualities that prevented them from wanting to be interviewed, such as external stressors in their lives or discomfort with sharing personal information with others in their school community.

Nonetheless, this study has a number of strengths. First, it is among the first to examine SC and its potential to enhance the benefits of group-based PT in school settings, helping to fill a major gap in the literature. Although PT groups in schools may be challenging to initiate and sustain, the results of this study suggest that there may be multiple benefits of these group-based programs for parents, children, and schools. Second, using both quantitative and qualitative methods helped to generate a more comprehensive understanding of parents’ experience with SC within PT groups. Finally, by controlling for the possibility that parents’ SC may have been linked with SC formed before the group began, we learned that most parents did not have prior relationships with the other parents enrolled in their group, strengthening the interpretation that SC scores obtained at post-intervention were achieved through their participation in the PT group.

Conclusion

This mixed methods study describes how a universal group-based PT program (the Chicago Parent Program) implemented in elementary schools located in low-income, under-resourced urban communities generated social connections among parents that were associated with significant improvements in their children’s behavior. Parents attributed feeling that the group was a safe space to share and being surrounded by parents who shared similar experiences helped to foster their connections to other parents. Whereas perceptions of judgment from the other parents hindered parents’ abilities to connect. Parents also discussed how their connections with other parents helped to facilitate their uptake of the PT program skills. These findings suggest that SC is a promising construct for facilitating outcomes related to group-based PT. Furthermore, imbedding group-based PT programming in schools can be an important avenue for parents to build connections with other parents and their school, in turn strengthening their children’s behavioral well-being.

Public Policy Relevance Statements:

Social connectedness is an important construct that has been linked to improved mental and physical health outcomes. Findings from this study highlight the importance of investing in school-based preventive parenting interventions that improve young children’s behavioral health and build socially connected parent communities. These investments are particularly needed in low-income, under-resourced communities where parents may experience more psychosocial adversity and isolation.