Engaging Parents and Their Fifth- and Sixth-Grade Latina Daughters in a Family Science Program

Abstract

This research study was conducted to pilot an out-of-school family science program for fifth- and sixth-grade Latina girls and their parents. Program goals included encouraging parents in supporting their Latina daughters in science, increasing the girls’ interest in science and increasing the families’ participation in science experiences together. The 41 families participated in a 7-week Saturday program on either rocketry or gardening. Each week, the parent–daughter dyads engaged in hands-on Family Problem-Based Learning activities together and then the parents and daughters met separately in Conversation Groups. To measure the impact of the program, surveys were administered to the parents and daughters separately at four points: pre-, mid-, post- and delayed-post (three months after the program). Parents reported increases over time for several aspects of their support for their daughters in science and also increases in frequency of science experiences with their daughters. The daughters reported increases over time in their science identity and their discussions with their parents about jobs in science. In addition, the examination of video-recordings of a subset of the parent–daughter interactions during the activities revealed that parental and daughter behaviors evolved over the course of the program. Implications for engaging parents in science education are discussed.

1. Introduction

This paper describes an empirical research study to pilot an out-of-school family science program for fifth- and sixth-grade Latina girls and their parents. Our project seeks to broaden the participation in science of women and Latiné communities, given their underrepresentation in numerous science, technology, engineering, and mathematics (STEM) majors and occupations (National Science Foundation, National Center for Science and Engineering Statistics, 2021). Previous research has shown that students, particularly girls and students from ethnic/racial minorities, often begin to lose interest in science by the middle-school years (Baram-Tsabari & Yarden, 2011; Fouad & Santana, 2017; Patrick et al., 2009). It is important to examine ways to counter this trend in order to encourage them to participate more in science courses and science-related careers as they get older.

Although efforts to expand STEM engagement and diversify the workforce have grown in recent years, minority groups remain underrepresented in many STEM fields, emphasizing the need for intentional initiatives to foster inclusivity (Dormer, 2023). Many traditional strategies to address these disparities focus primarily on students, often neglecting the influential role of families, particularly parents, in shaping academic pathways. This gap is especially relevant in Latiné communities, where educational aspirations and success are closely linked to family involvement and cultural values (Moll et al., 1992; Starr et al., 2022). Additionally, many STEM programs fail to incorporate the cultural knowledge, language, and lived experiences that Latiné families contribute to learning environments (Habig et al., 2021; Rincón & Rodriguez, 2021; Yosso, 2005). Rincón and Rodriguez (2021) emphasize that engaging families in STEM learning and drawing on their cultural and linguistic strengths can bolster Latina students’ science identities. Goals of our family science program include encouraging parents in supporting their Latina daughters in science, increasing Latina girls’ interest in science, and increasing the families’ participation in out-of-school science experiences together.

The fifth- and sixth-grade levels were chosen in our study because it is a developmental period when enrichment programs may be especially impactful. This grade range bridges the middle-school transition in most U.S. school districts. At the beginning of middle school, children are often starting to make more choices for themselves regarding the selection of elective coursework at school and extracurricular activities, which may influence their future majors and careers. Another reason for the focus on fifth and sixth grade is that it is well documented that a shift in science interest by gender often occurs in later childhood and adolescence (Baram-Tsabari & Yarden, 2011; Britner & Pajares, 2006; Tenenbaum, 2009). The findings from work with fourth graders indicate no gender differences in science assessment scores at this grade level (Short-Meyerson & Sandrin, 2016). When the shift does occur, it is particular to certain science disciplines. Girls especially are more likely to maintain interest in life science than physical science. This also provides rationale, in part, for our decision to offer separate programs with a physical science topic (rocketry) and a life science topic (gardening).

Children of this age, unlike adolescents in high school, often still spend a significant amount of their out-of-school time with their parents. Furthermore, parents may have more influence on their daughters at the early stages of change. At this age, many children may feel they have “outgrown” traditional out-of-school science activities, such as some summer day camps and children’s museums. Parents often search for things to do with their tweens and the proposed activities were designed to engage parents and keep their tweens interested in STEM subjects. Furthermore, parental support is likely to be enhanced by fostering parents’ (and other family members’) engagement in their children’s learning. For example, Pushor and Amendt (2018) state:

When parents engage their children in cultural teachings and ceremony, when they talk with them about such things as current events or what they are reading, when they take them places and engage in activities with them, when they establish hopes, dreams, and expectations with and for their children, they are uniquely, contextually, culturally engaged in their children’s education (p. 207).

With this familycentric approach, the family is at the center and their engagement with their child’s schooling is authentic and meaningful. The extensive pilot work described in this paper was conducted over an 18-month period. It consisted of a series of six iterations of the program in two locations in the western U.S. (two rocketry and one gardening program in California; and two gardening and one rocketry program in Arizona). The initial programs were co-created with families and the Community and Family Engagement (CAFE) specialists at the district level. The main purpose of the pilot was to implement the programs and incorporate the voices of the participating families to refine the programs. The first iterations included only a small number of families, partly to maintain health/safety measures such as social distancing during COVID-19, with the number of families increasing with each successive iteration. Throughout the pilot, parents were involved in the design of the programs, which reflected participatory design (Ishimaru, 2020). We viewed them as co-designers and we intentionally provided opportunities for them to assist with decision making. For example, when we (i.e., the researchers and authors of this paper) began the pilot, we planned to have two conditions: one condition in which the families engaged in only one component of the program (Family Problem-Based Learning, described below) and a second condition in which the families engaged in two components (Family Problem-Based Learning and Conversation groups, described below). However, after the first iterations, which involved both components, the families expressed that they did not see how one component could exist without the other. Their reasoning was sound, and they helped us see that the two components could not be teased apart. Therefore, we decided to keep both components for all of the iterations.

Furthermore, surveys that included items seeking feedback on the program and suggestions for ways to improve it were administered to parents and daughters mid-way through each program and again at the end of each program. In addition, at the end of each program, focus groups facilitated by a researcher were conducted with groups of parents and groups of daughters, separately. This gave the parents the opportunity to describe their experiences, discuss their ideas together, and share valuable insights with each other and with the researcher. Many of the families also provided unsolicited feedback and shared their knowledge throughout the programs, which helped shape the programs. We had intentionally planned six iterations so that following each iteration, our research team was able to reflect on what we had learned from the families. Thus, the families’ knowledge and ideas were used to co-create and shape each current program as well as the future programs.

Our study builds upon previous research on parents’ participation in science with their children in various out-of-school settings, including museums and at home (Fender & Crowley, 2007; Short-Meyerson et al., 2016; Tenenbaum & Leaper, 2003; Vandermaas-Peeler et al., 2016; Zimmerman et al., 2009). A limited number of research studies in informal science settings have investigated parent–child interactions of Latiné families (Navarro et al., 2007; Shirefley et al., 2020; Short-Meyerson et al., 2019; Siegel et al., 2007). In our study we have created, enacted, and evaluated the impact of two complimentary enrichment strategies: Family Problem-Based Learning activities (FPBL), which involve daughters and parents collaborating on hands-on science activities, and Conversation Groups (CG), where girls and parents meet separately. During each 7-week Saturday morning program, families delved into FPBL on either the physical science topic of rocketry or the life science topic of gardening. In this paper, we provide an overview of the program and share the results obtained from six iterations of the program conducted in two different locations, California and Arizona.

The impact of the family science program has been examined in two ways. First, surveys were administered to the parents and daughters (separately). Parent surveys focused on their knowledge and confidence regarding supporting their daughters in science and their participation in science experiences with their daughters. Daughter surveys focused on their science interest, attitudes, and participation in science experiences with their parents. They all completed the surveys at four different times: before (pre-), during (mid-), and at two points after the program (post- and delayed post-, three months later). Second, a subset of the parent–daughter interactions during the hands-on rocketry and gardening activities (FPBLs) were video- and audio-recorded. Both the girls’ behaviors and the behaviors of the parents to support their daughters in science were analyzed. Examining data from these multiple sources gives a comprehensive picture of the impact of the program.

1.1. Theoretical Framework

Bronfenbrenner’s (1979) ecological model emphasizes the role of environmental context, especially sociocultural influences, on development. Bronfenbrenner proposed that there are five environmental systems that are pertinent to development: microsystem, mesosystem, ecosystem, macrosystem, and chronosystem. Although many of these systems are relevant to our present research study, we focus on the microsystem and the macrosystem. Our study highlights the importance of the microsystem, which most directly impacts the child’s development and learning and includes the family and school. The role of parents and families is detailed in the following literature review and throughout the paper.

An indirect influence is the macrosystem, which has been described as “the larger cultural and social context within which the other systems are embedded” (Siegler et al., 2020, p. 338). This includes attitudes, beliefs, and traditions of groups. While Bronfenbrenner was central to moving our field as educators towards a greater recognition of the role of culture, Yosso (2005) further extended that work by framing the context of culture in a child’s life as cultural wealth. In her asset-based model, she delineates six specific forms of cultural wealth: familial, aspirational, linguistic, social, navigational, and resistance, which all have relevance to our project. For instance, Yosso refers to familial capital as “those cultural knowledges nurtured among familia (kin) that carry a sense of community history, memory, and cultural intuition” (2005, p. 79), which influenced the instrument and curriculum development used in this project.

This study is also grounded in the sociocultural work of Rogoff (1994), including the concept of communities of learners. This acknowledges that “learning occurs as people participate in shared endeavors with others” (p. 209). Furthermore, the work of Radziszewska and Rogoff (1988, 1991) on guided participation emphasizes the significant role that parents have as partners in tasks with their children. Their work shows that parents can guide their children in a number of ways beyond hands-on activities and verbal exchanges. However, the child may also benefit from participating as an observer, as Silva et al. (2010) found in their study of the interactions of 5- to 11-year-old U.S. Mexican heritage children with their parents. As documented in their study, the children were attentive and learned through observation. Moreover, participation and learning may look different, depending on the task at hand (Rogoff et al., 2003).

1.2. Literature Review

The present study expands upon previous research on the influence of parents on their elementary-age children’s science interests, attitudes, and problem solving. In a small-scale study, Short-Meyerson et al. (2016) investigated the influence of child gender and parent gender on parent–child science problem-solving behaviors. They examined interactions of parents (mothers and fathers) with their second- and fourth-grade children while solving a variety of hands-on science problems together. They found that both mothers and fathers provided more encouragement to their sons than to their daughters. Similarly, sons were praised more often than daughters for being smart. In addition, daughters were asked more questions than sons and mothers used questioning more than fathers.

The present study also expands upon a study by Tenenbaum and Leaper (2003) that included sixth- and eighth-grade children, as well as the mother and father of each child. Most of the parents identified themselves as European American and from middle-income households. Each parent completed a questionnaire in which they were asked to rate their child’s interests and abilities in science and each child completed a questionnaire about their self-efficacy in various academic domains, including science, and their academic interests and aspirations. In addition, the researchers observed and video-recorded the families in their homes as they performed science (biology, physics, and technology) tasks. The results from the parent questionnaires indicated that parents believed that science was less interesting and more difficult for daughters than for sons. Furthermore, mothers’ beliefs that science was difficult for their children were negatively correlated with their children’s self-efficacy and interest. Fathers’ beliefs that science was difficult for their children, however, were negatively correlated with their children’s self-efficacy only. From the behavioral data, during the physics task only (not the other tasks) fathers used more cognitively demanding speech (involving conceptual questions, causal explanations, and scientific vocabulary) than mothers in their interactions with their children.

Similar research with families of young children has been conducted in the context of museums. For example, families of 3- to 8-year-old children were observed by Fender and Crowley (2007) at a science museum exhibit. They found that there were several advantages of parental participation, such as the children remaining engaged for a longer duration and learning more about how the exhibit worked. In addition, Vandermaas-Peeler et al. (2016) found that parents of young children (4- to 6-year-olds) who receive guided instruction on how to discuss the topic at an interactive science museum exhibit asked their children more “how” and “why” questions than parents who did not receive the instruction. In families with older children (5- and 12-year-olds) who regularly attend science museums, the children as well as the adults may take the lead on sharing content in science discussions (Zimmerman et al., 2009).

A limited number of research studies in informal settings have examined the support that parents provide to their Latiné children in science. Short-Meyerson et al. (2019) qualitatively analyzed parents’ interactions with their Latiné and non-Latiné fourth-grade children during informal science activities. There were a variety of hands-on activities such as a sink–float activity, completing an electric circuit, and predicting how one’s heart rate would be affected by exercise. Through content analysis of the video-recorded interactions, the researchers explored and characterized parental behaviors that may influence their children’s participation in science activities. The themes that emerged included seeking deeper understanding of the activity (such as providing or asking for explanation of concepts) and parental questioning (which was usually to prompt the child to respond or act), as well as parental intrusive support by taking over the materials or telling the correct answer immediately rather than letting the child work through the problem.

The repertoire of support that parents provide to their Latiné and non-Latiné 8- to 12-year-old children during hands-on science activities was examined by Short-Meyerson et al. (2024). They found that parents often provided help, and it appeared in many forms, including stating an observation about the science problem, providing explanations, and directing or telling their child what to do. Parents also asked a variety of different types of questions, such as those aimed at determining or deepening their child’s understanding and prompting their child to perform an action. However, parents provided little encouragement to their children during the science activities (e.g., “Good job” or “Keep going, almost there”). Interestingly, Latino sons received less encouragement than non-Latino sons, although there was no difference in the amount of encouragement received by Latina and non-Latina daughters. For parental helping behaviors and questioning there were no differences between Latiné and non-Latiné daughters or Latiné and non-Latiné sons.

Navarro et al. (2007) investigated eighth-grade Mexican American children’s math/science self-efficacy. Specifically, they addressed whether their perceived support from their parents and others (such as teachers, classmates, and close friends) was related to the children’s math/science self-efficacy. They found that perceived support from parents, but not from others, predicted the students’ math/science self-efficacy. Similarly, Bhanot and Jovanovic (2009) found that mothers’ encouragement of their daughters’ participation in science was associated with their daughter having a higher self-assessment of their science ability.

In an investigation of collaborative behaviors of Mexican descent families of 3- to 9-year-olds, Siegel et al. (2007) explored two contexts: a children’s museum which involved open-ended interactions and a sink–float science activity at home. They found an interaction between child gender and context. During museum visits, parents of daughters displayed more collaborative behaviors than parents of sons. During at-home activities, however, parents of sons displayed more collaborative behaviors than parents of daughters.

Unfortunately, many parents may not feel confident in their ability to support their children in science. Moreover, Silander et al. (2018) found that Latiné parents are less likely than White parents to feel confident in helping their children in science. One goal of our family program is to encourage parents, regardless of their own level of expertise in science, to support their daughters in science in a variety of ways.

Based on self-efficacy theory (Bandura, 1977) and work on the role of self-efficacy in women’s career choices (Betz & Hackett, 2006; Hackett & Betz, 1989), we must consider the complex nature of factors that may contribute to underrepresentation of women and ethnic minorities in some STEM fields (e.g., see review by Fouad & Santana, 2017). In addition to self-efficacy, outcome expectations, and interests, underrepresentation also involves contextual factors such as socio-economic status (SES), mentoring (or lack thereof), racism, sexism, and more. In their review, Fouad and Santana (2017) conclude that parental support is related to math and science self-efficacy of middle school and high school students. In addition, the studies they reviewed supported “the relationship of math and science self-efficacy and outcome expectations to math/science interests and intentions to pursue STEM goals” (pp. 27–28). Their findings, consistent with Lent et al.’s (1994) work, suggest a need to build parental support to promote science career interventions for female students and those from ethnic minority groups. Furthermore, they suggest that future research expands upon earlier work by examining the intersection of ethnicity and gender, particularly research that examines the efficacy of programs involving parental support.

Although the existing literature has provided many insights about the influence of parents on their elementary-age children’s engagement in science, gaps in the literature remain. One of the areas that has often been overlooked by previous researchers is the behaviors of parents and children as they work on science activities together. In most of the previous research, parent–child dyads participated in a single session (e.g., Short-Meyerson et al., 2016, 2024; Tenenbaum & Leaper, 2003). Our current study, in which the families participated across seven weeks, was designed to examine changes in their interactions over time. Furthermore, there are a very limited number of studies on parent–child interactions in an informal science context with elementary- and middle school-age children. The current study addresses that need and, moreover, compares two grade levels (5th and 6th) to examine the developmental point at which informal science programs may be most impactful. Finally, our program also features two topics (reflecting two different science disciplines) in two distinct locations. This provides insights into whether parent and child behaviors during their science activities together vary by the specific program.

1.3. Research Questions

In the present study, there were two sets of research questions.

• The first set of research questions was based on parents’ survey data and daughters’ survey data.
  • What effect did the family science program have on the parents’ support for their daughters in science (including their understanding of their daughters’ science interests and abilities, and their confidence in providing support for their daughters in science) and their participation in science experiences with their daughters outside of the program?
  • What effect did the family science program have on the daughters’ science identity, science interests and attitudes, and their participation in science experiences with their parents outside of the program?
• The second set of research questions was based on the parent–daughter interactions during the hands-on science activities that they participated in together (i.e., the FPBLs).
  • Did the parent and/or daughter behaviors vary as a function of week in the program (from the beginning to the end of the seven weeks)?
  • Did the parent and/or daughter behaviors vary as a function of daughters’ age?
  • Did the parent and/or daughter behaviors vary as a function of the program implementation (Rocketry-CA, Rocketry-AZ, Gardening-CA, Gardening-AZ)?

2. Materials and Methods

2.1. Participants

Forty-one families participated in the 7-week Saturday morning program. Each week, a parent participated with their fifth- or sixth-grade Latina daughter. The term “parent” is used for the primary adult participating with the child, whether another adult family member or guardian, such as a grandparent or older sibling. The primary adult participant was the person who completed the parental surveys. Similarly, the child is designated as a “daughter”, whether they are the adult’s child or had a different relationship, such as granddaughter or sister.

The girls ranged in age from 10 to 13 years (mean = 10.92 years) and the parents ranged in age from 29 to 71 years (mean = 42.63 years). Additional information attained from a demographic form completed by the parent included the daughter’s and parent’s first language and language spoken at home (shown in

Table 1. Daughters’ and parents’ first language and langue spoken at home.

Language Measure	English	Spanish	English and Spanish
Daughters’ first language	18	21	1
Daughters’ language spoken at home	16	10	14
Parents’ first language	16	18	3
Parents’ language spoken at home	16	18	3

), parent level of education, annual family income, and parent’s occupation.

Socioeconomic status (SES) was indicated by parent level of education and annual family income, as reported on the demographic form. Parent education included six levels: 1 = no high school, 2 = some high school, 3 = high school diploma or GED, 4 = some college, 5 = bachelor’s degree, and 6 = graduate degree. Parent level of education means for the four implementations (Rocketry-CA, Rocketry-AZ, Gardening-CA, Gardening-AZ) ranged from 2.83 to 4.14 and an ANOVA revealed no significant differences in parent education among the four implementations, p = 0.163. Annual family income included eight levels: 1 = less than $15,000, 2 = $15,000–$29,999, 3 = $30,000–$44,999, 4 = $45,000–$59,999, 5 = $60,000–$74,999, 6 = $75,000–$89,999, 7 = $90,000–$104,999, and 8 = $105,000 or greater. Annual family income means ranged from 2.67 to 5.00 for the four implementations and Tukey multiple comparisons of means revealed no significant differences in family income among the four implementations, p = 0.057.

Twelve of the parents reported being stay-at-home parents. The parents who were employed outside of the home reported a variety of occupations. Some (one to three) parents reported occupations in each of the following areas: agriculture, business/finance/insurance, caregiver, contractor/supervisory, cosmetology services, food/beverage services, healthcare services, housekeeping/landscaping, law enforcement, retail, and social services. Three parents reported being unemployed.

The families were recruited from two school districts (one in Arizona and one in California) with a high proportion of Latiné students and students of low socio-economic status. In the Arizona and California school districts, respectively, 66.1% and 69.7% of students are Latiné and 64.5% and 48.8% of students are low-SES (i.e., eligible to participate in the federal free and reduced-price meal program). To offset the costs of participation, each family received compensation to offset costs of transportation and childcare for younger siblings at home. All participants gave their informed consent prior to their inclusion in the study.

2.2. Procedure

Throughout the study, the rocketry and gardening programs occurred at both the California and Arizona sites. From Spring 2021 through Spring 2022, six implementations of the program took place (two Rocketry-CA, one Rocketry-AZ, one Gardening-CA, and two Gardening-AZ). Three to nine families participated in each implementation. In the early implementations, the number of participants was small to enable social distancing during COVID-19, with the number growing as restrictions were lifted as the study progressed. A total of 24 families (15 in California and 9 in Arizona) participated in rocketry and 17 families (11 in Arizona and 6 in California) participated in gardening.

The family science program was held at local public schools for seven weeks, for 2.5 h each Saturday morning. It was run by bilingual Project Leaders, university student research assistants, and volunteers, under the direction of university faculty members (the second and third authors).

2.2.1. Family Problem-Based Learning

Each weekly session began with Family Problem-Based Learning (FPBL) activities. (The only exception was the first week, in which the order of the activities was changed to accommodate the pre-surveys described below.) FPBL evolved from work on a previous project (by two of the authors) and is based on the rationale that when the language in which academic content is delivered to students is not accessible, their academic success is jeopardized (Wright, 2015). There are methods that can enhance emergent multilingual students’ acquisition of language and mastery of rigorous curriculum (Von Esch & Kavanagh, 2018), which include focusing on language scaffolds, linking background knowledge and culture to learning, and stimulating higher level thinking (Levine et al., 2013). Combining language-based theories, specific methods for working with multilingual students, and PBL, Jimenez-Silva et al. (2016) and Rillero et al. (2017) led a team that developed an approach called Problem-Based Enhanced-Language Learning (PBELL). While the approach attends to emergent multilingual students’ needs, the language additions to PBL make it a valuable experience for all learners. Working from PBELL as a base, the family science program uses longer-term problems and has the family as the unit for learning. We are using the term Family Problem-Based Learning (FPBL) to describe the approach.

The FPBLs involved hands-on science activities that the parent–daughter dyads worked on together, alongside the other parent–daughter dyads. There were two distinct program topics (rocketry and gardening), both aiming to foster interactive experiences and cultivate interest and curiosity in science. Throughout the seven weeks, problem-based learning opportunities were woven into the experience. The rocketry program encompassed a series of launching projectiles and parachute activities that culminated during the last week with the launch of model rockets using information and skills developed in previous weeks. The gardening program included planning, planting seeds in raised beds, and caring for, measuring, and harvesting vegetables for the culminating tostada party. Both the rocketry and gardening curriculum were designed to be interactive and promote science interest and inquiry. Each family was assigned a Science Notebook to write plans and record observations. Each FPBL was approximately 90 min in duration and used both English and Spanish. For each session, one or two family’s FBPL interactions were video- and audio-recorded for later analysis (described below).

2.2.2. Conversation Groups

Following the FPBLs and a brief snack break, parents and girls participated separately in Conversation Groups (CG). The CGs consisted of one of the researchers leading a discussion of a relevant science topic. Topics for the parent groups included discussing ways in which to continue to support their daughters’ interests in science at home and in the community, engaging in conversations about gender-typing, and developing their confidence in supporting their daughters in science. Topics for the girls’ conversation groups included the exploration of STEM careers, learning about Latina scientists, and advantages of being bilingual. The parent groups were run using both English and Spanish. The girl groups were held predominantly in English with some Spanish. Each CG was approximately 30 min in duration.

2.3. First Data Set: Surveys

Separate parent and daughter surveys were used to address our first set of research questions. Four surveys were administered to each parent and daughter: pre-survey (week 1), mid-survey (week 4), post-survey (week 7), and delayed post-survey (3 months after the program). The surveys were developed by the authors, based partly on instruments used by previous researchers, such as the Child and Adolescent Social Support Scale (CASS; Malecki et al., 2000), the Trends in International Mathematics and Science Study survey (Institute of Education Sciences, 2011), and the National Research Center on Latino Children and Families report (Silander et al., 2018). Completion time for each survey was approximately 20 min.

The parents also completed a demographic form at the beginning of the program. The demographic form was developed by the authors and it included items about the child’s age, the parent’s level of education, the family’s annual income, and more. All instruments were available in Spanish or English and paper/pencil or online format.

2.3.1. Parent Surveys

The parent survey items focused on parent support of their daughters’ science interest and parent participation in science experiences with their daughters (See Appendix A). There were 23 items, with 5-point Likert scale ratings. For example, “I feel more confident in my ability to support my daughter in science compared to other subjects”, rated from Strongly Disagree (1) to Strongly Agree (5) and “How often do you talk with your daughter about high school courses in science”, rated from Never (1) to Always (5).

2.3.2. Daughter Surveys

The daughter survey items focused on their science interests and attitudes and their participation in science experiences with their parents (See Appendix B). There were 29 of these items with 5-point Likert scale ratings, similar to the items on the parent survey. For example, “I’m interested in science”, rated from Strongly Disagree (1) to Strongly Agree (5) and “How often do your parent/s participate with you in science activities at home, such as going on a nature hike, observing the night sky, or helping with the family garden or animals”, rated from Never (1) to Always (5). There was also an item on science identity, rated on a 7-point scale.

2.4. Second Data Set: Recorded Parent-Daughter Interactions During FPBLs

In order to address our second set of research questions (regarding behavioral data from the parent–child interactions), a subset of FPBLs were recorded and analyzed. For each session, one or two families who consented were video- and audio-recorded. This resulted in 49 codable video sessions (8 rocketry-CA, 9 rocketry-AZ, 13 gardening-CA, and 19 gardening-AZ) from 13 families. The ages of the daughters in this subset ranged from 10 to 12 years old, with a mean age of 10.85 years.

2.4.1. Transcription and Coding of Parent-Daughter Interactions

The video- and audio-recordings from each FPBL session were transcribed by undergraduate and graduate students trained on Systematic Analysis of Language Transcript conventions (SALT; Miller & Iglesias, 2008). Portions where there was direction, instruction, or conversations with the researchers were not transcribed or analyzed. Spanish segments were translated into English for coding. The mean duration of the transcripts was 48.76 min (range = 8.68–79.12 min), depending on the activity of the day. In addition to transcribing dialog, the transcribers made notes about non-verbal behaviors to aid in interpreting the verbalizations. A graduate student was trained by the first author to code the parent and daughter turns for each behavior (described below) while watching the video-recorded interactions and following along on the transcripts. A turn consisted of one or more utterances by a speaker, which ended either when they paused for at least 2 sec or when the other person spoke. Inter-rater reliability of coding was calculated as a percent agreement for each dependent measure. The mean percent agreement was 97% (range = 84% to 100%).

Some of the dependent measures were based on previous studies (Short-Meyerson & Sandrin, 2016; Short-Meyerson et al., 2016, 2019) and some emerged from a preliminary review of the video-recordings from the current study. They included parent behaviors, daughter behaviors, and interactivity of the parent–daughter dyad, as described below.

2.4.2. Parents

Each conversational turn by the parents was coded for the following behaviors:

• Helping: Providing direction, providing explanation, asking a guiding question, and nonverbal help, such as providing a demonstration.
• Cautioning: Cautioning the child, such as “Watch out! This is a little loose”.
• Slowing: Trying to get the child to slow down or wait, such as “Hold on” or “Wait for me”.
• Encouragement and praise: Examples include “Keep going! You’ve almost got it”., “Do the best you can”, “Good job!”, and “You’re so smart”.
• “Aha” moment: Being amazed by something new or learning a new way to think about something, e.g., “That’s cool” when they are amazed by how something works.
• Moment of frustration: Being frustrated or discouraged by something related to the activity. For example, “Darn! This isn’t working”.
• Rate of parent utterances: This was the number of parent utterances per minute in the session, which was used as a measure of the amount of parents’ talk.

2.4.3. Daughters

Each conversational turn by the daughters was coded for the following behaviors:

• Help-seeking: Asking for assistance with the content, the materials, problem-solving process/procedure, instructions, or for an explanation about the science activity.
• Encouragement and praise: See description above.
• “Aha” moment: See description above.
• Moment of frustration: See description above.
• Rate of daughter utterances: This was the number of daughter utterances per minute in the session, which was used as a measure of the amount of daughters’ talk.

2.4.4. Parent–Daughter Collaboration

We examined collaboration, or interactivity, between each parent and their daughter. Each 3 min interval was coded for whether the parent and daughter were collaborative (i.e., working together equally on the activity).

3. Results

3.1. Parent Surveys and Daughter Surveys

The parent and daughter surveys were examined to assess the impact of the family science program. Repeated measures ANOVAs were used to analyze survey responses over the four time intervals (pre-, mid-, post- and delayed post). Covariate demographic variables were used to block and explain variation. These were family income, parent education level, child age, program topic (rocket, gardening) and location (CA, AZ). The reported Means and Confidence Intervals (CIs) indicate the magnitude of the changes over time. Only statistically significant results (p < 0.05) are reported.

Regarding our first research question (1a), parents reported that several aspects of their support for their daughters in science significantly increased over time. Specifically, there were increases in knowing their daughter’s science abilities (p = 0.009), knowing how to be supportive when their daughter seeks help in science (p = 0.024), knowing when to let their daughter work through a science problem on her own (p = 0.014), and feeling confident in their ability to support their daughter in science (p = 0.011).

Table 2. Parents’ science support significant survey results.

Item †	T1 Mean (CI) ‡	T2 Mean (CI)	T3 Mean (CI)	T4 Mean (CI)	p-Value
I know my daughter’s science abilities.	3.38 (±0.30)	3.89 (±0.31)	3.75 (±0.30)	4.11 (±0.35)	0.009
I know how to be supportive if my daughter asks for help on a science problem.	3.37 (±0.34)	3.97 (±0.34)	3.96 (±0.34)	3.89 (±0.37)	0.024
I know when to let my daughter work through a science problem on her own.	3.43 (±0.28)	3.84 (±0.29)	3.93 (±0.28)	4.02 (±0.32)	0.014
I feel more confident in my ability to support my daughter in science compared to other subjects.	3.36 (±0.38)	3.80 (±0.39)	4.14 (±0.38)	3.75 (±0.43)	0.011

^† Scale: 1 = strongly disagree, 2 = disagree, 3 = undecided, 4 = agree, 5 = strongly agree. ^‡ Abbreviation: CI, Confidence Interval.

shows the Mean and Confidence Interval (CI) for each time interval (T1, T2, T3, and T4) for each item.

Similarly, for many of the science experiences, parents reported significant increases in participation over time. Specifically, there were increases in telling their daughter that she is good at science (p = 0.001), asking their daughter about how things are going in her science classes (p = 0.003), talking about high school courses in science (p = 0.001), talking about careers in science (p = 0.005), and telling their daughter how important performing well in science will be in her future (p = 0.002).

Table 3. Parents’ science participation significant survey results.

How Often Do You Do Each of the Following with Your Daughter? †	T1 Mean (CI) ‡	T2 Mean (CI)	T3 Mean (CI)	T4 Mean (CI)	p-Value
Tell your daughter that she is good at science.	3.75 (±0.35)	4.33 (±0.35)	4.39 (±0.34)	4.65 (±0.40)	0.001
Ask your daughter about how things are going in her science classes.	3.97 (±0.83)	4.37 (±0.85)	4.52 (±0.83)	4.51 (±0.83)	0.003
Talk about high school courses in science.	2.74 (±0.40)	2.99 (±0.40)	3.74 (±0.40)	3.78 (±0.47)	0.001
Talk about careers in science.	3.06 (±0.39)	3.36 (±0.39)	3.72 (±0.39)	4.08 (±0.45)	0.005
Tell your daughter how important performing well in science will be for her future.	3.48 (±0.41)	3.94 (±0.41)	4.25 (±0.42)	4.54 (±0.48)	0.002

^† Scale: 1 = never, 2 = a little, 3 = sometimes, 4 = a lot, 5 = always. ^‡ Abbreviation: CI, Confidence Interval.

shows the Mean and Confidence Interval (CI) for each time interval (T1, T2, T3, and T4) for each experience.

Regarding research question 1b, there were fewer significant effects of the program for the daughters than for the parents. However, the girls’ science identity (p = 0.019) and their discussions with their parents about jobs in science (p = 0.016) increased significantly over time.

Table 4. Daughters’ significant survey results.

Item †	T1 Mean (CI) ‡	T2 Mean (CI)	T3 Mean (CI)	T4 Mean (CI)	p-Value
Science identity (on 7-point scale)	3.63 (±0.60)	4.05 (±0.61)	4.33 (±0.61)	4.95 (±0.68)	0.019
Talk about jobs in science.	2.36 (±0.43)	2.96 (±0.43)	3.17 (±0.43)	3.34 (±0.47)	0.016

^† Scale: 1 = never, 2 = a little, 3 = sometimes, 4 = a lot, 5 = always. ^‡ Abbreviation: CI, Confidence Interval.

shows the Mean and Confidence Interval (CI) for each time interval (T1, T2, T3, and T4) for each item.

3.2. Parent-Daughter Interactions During FPBLs

In order to explore additional impacts of the family science program, the interactions of the parent–daughter dyads during their hands-on science activities together were also examined. All of the findings reported for the parent–child interactions during FPBLs have p < 0.05.

Logistic regression was used to identify significant variables for binary response variables, specifically responses categorized as “yes” or “no”. Covariate demographic variables were used to block and explain variation. These were family income, parent education level, child age, program topic (rocket, gardening) and location (CA, AZ). In the logistic regression results, odds ratios are reported in parentheses. An odds ratio of 1 indicates no change, whereas odds ratios greater than 1 indicate an increase, and odds ratios less than 1 indicate a decrease. For the variables “Week” and “Daughter’s Age”, when we increase by 1 unit (week or year, respectively), the odds ratio indicates how much the odds have increased or decreased. For instance, if the odds ratio was 0.89 for the “Week” variable, then for each passing week, the odds of the observed behavior diminish by 11%.

Conventional linear regression was used for the numerical response variables. For linear regression results, the slope coefficients are reported in context.

3.2.1. Change Across Time (Research Question 2a)

Three behaviors of the daughters changed across time. They sought less help (odds ratio: 0.89) and were less talkative at the end of the program than at the beginning (child utterances per minute dropped by 0.36 each week), whereas they had more frequent “aha” moments at the end of the program than at the beginning (odds ratio: 1.21). Interestingly, for the parents, the same three behaviors changed across time. Parents provided less help (odds ratio: 0.93) and were less talkative at the end of the program than at the beginning (parent utterances per minute dropped by 0.53 each week), and they had more frequent “aha” moments at the end of the program than at the beginning (odds ratio: 1.15).

3.2.2. Daughters’ Age (Research Question 2b)

Regarding daughters’ age, older daughters had fewer “aha” moments (odds ratio: 0.65) and produced fewer instances of encouragement and praise (odds ratio: 0.63) than younger daughters. Parents provided more help (odds ratio: 1.18) and had many more moments of frustration (odds ratio: 2.79) with older daughters than with younger daughters. In addition, parents had fewer “aha” moments (odds ratio: 0.43), provided less encouragement and praise (odds ratio: 0.60), and had fewer instances of slowing their daughters (odds ratio: 0.68) with older daughters than with younger daughters.

3.2.3. Program Implementation (Research Question 2c)

As shown in Table 4, daughters’ and parents’ behaviors varied as a function of program implementation (i.e., the combination of topic and location). Topic and location had a complex interaction for many models, and so we are unable to characterize topic alone or location alone; instead, all four implementations are presented for comparison. The entries in

Table 5. Modeled predictions of daughter and parent behaviors.

	Program Implementation
	Rocketry		Gardening
Behavior †	CA	AZ	CA	AZ
Daughter Behaviors
Help-Seeking	0.181	0.277	0.277	0.277
“Aha” Moment	0.004	0.002	0.007	0.015
Moment of Frustration	0.021	0.037	0.021	0.021
Encouragement/Praise	0.031	0.018	0.007	0.015
Rate of Utterances (per minute)	8.22	5.09	5.09	5.09
Parent Behaviors
Help	0.369	0.482	0.543	0.478
Cautioning	0.003	0.003	0.015	0.003
Encouragement/Praise	0.060	0.060	0.094	0.060
“Aha” Moment	0.006	0.003	0.029	0.009
Moment of Frustration	0.000	0.020	0.008	0.003
Slowing	0.037	0.020	0.037	0.020
Rate of Utterances (per minute)	7.13	4.84	7.13	4.84
Parent–Daughter Interactivity
Collaboration	0.874	0.632	0.618	0.73

^† The probabilities in Table 4 are predicted probabilities from the statistical models fitted. They assume a 10-year-old daughter participated in Week 1 in a 60 min session; from there we can observe the relative effect of the four program implementations.

report probabilities, except the last row which reports rates (utterances per minute).

Daughters were more talkative in Rocketry-CA than in the other implementations. Rocketry-CA is also the implementation in which they sought the least help and provided the most encouragement/praise to their parents. Daughters experienced the most “aha” moments in the Gardening-AZ implementation, whereas they experienced the most moments of frustration in the Rocket-AZ implementation.

Parents were more talkative (rate of utterances) and used more behaviors to slow down their daughters in the Rocketry-CA and Gardening-CA implementations than the other two implementations. Parents provided the most help, most cautioning, and most encouragement/praise and experienced the most “aha” moments in Gardening-CA. Parents experienced the most moments of frustration in Rocket-AZ. The parent–daughter dyads were most collaborative in Rockets-CA.

4. Discussion

This study examined science learning among fifth- and sixth-grade Latinas with their parents, using a familycentric approach (Chen & Pushor, 2023) and creating a community of learners (Rogoff, 1994), at a younger developmental level than has often been investigated. The results of this study provide insights into Latinas’ science learning in an informal setting with parents, an area of research under-investigated in the research literature. For example, the findings from the parents’ surveys demonstrate that the family science program had multiple effects on the parents’ support of their daughters in science, including knowing how to be supportive if they ask for help and knowing when to let them work through a science problem on their own. They also reported an increase in their feeling of confidence in supporting their daughters in science, compared to other subjects. Furthermore, their knowledge of their daughters’ science abilities increased and remained high even three months after the program. These findings point to the impact of the program on expanding Latiné parents’ social and navigational cultural capital (Yosso, 2005; Yosso & Burciaga, 2016) in ways that will impact their daughters immediately and in their future experiences with science.

Parents’ reports of their participation in science experiences with their daughters outside of the program revealed an interesting pattern. Over the course of the program (from the first week to three months after the program), parents reported increases in telling their daughter that she is good at science, as well as discussing her current and planned high school science courses, the importance of performing well in science for her future, and possible careers in science. However, there was no change reported for their engagement with their daughters in specific out-of-home (e.g., going to a zoo, observing the night sky) and in-home science experiences (e.g., watching science television shows together, talking about current events related to science). This suggests that the program’s greatest impact on participation in science experiences (outside of the program) was on parents engaging in meaningful discussions with their daughters about science in their lives. It is important to not assume that the lack of participation is due to a lack of desire or ability to engage in the experiences. Rather, as Bronfenbrenner (1979) argues, the macrosystem of families may include many different factors, such as financial or time constraints, which are crucial to acknowledge. Informal science programs similar to ours may help mitigate access to opportunities to engage in science experiences.

The findings from the daughters’ surveys demonstrate that the family science program had two effects, which may be related to the findings from the parental surveys. First, daughters reported that their conversations with their parents about jobs in science increased throughout the program and extended to the delayed post-survey three months later, which is consistent with parents’ survey responses. In addition, as the program progressed, daughters described themselves with a stronger science identity (indicated by their image of themselves as a science professional). This trend also continued beyond the end of the program, to the delayed post-survey three months later. The increase in daughters’ science identities may be a direct impact of the program and/or it may be due, at least in part, to the discussions that their parents were having with them about science in their lives. These findings are consistent with our previous research suggesting that involving parents in garden-based learning had significant positive effects on their Latina daughters’ science identity and science career aspirations (Rillero et al., 2025). Similarly, Simpkins et al. (2006) focused on older Latiné youth. Their work emphasizes the importance of middle school students’ participation in informal, out-of-school STEM learning environments on their later enrollment in high school and college STEM courses. Our study captured the various ways in which parents support students’ learning, which may have implications for increasing Latinas’ participation in STEM careers.

The second set of research questions in this study involved the examination of parent and daughter behaviors as they worked on the rocketry and gardening activities together. This revealed that their behaviors evolved across the seven weeks of the program. Both the parents and daughters were more talkative at the beginning of the program than towards the end. This may be attributed to there being more that they needed to talk about in the initial weeks of the program, when they were new to activities, the materials, the people involved, and so on. As the weeks passed, they may have developed a routine or script for what they needed to do and how to work together, sometimes without even needing to discuss it as much. Additionally, parents provided less help and daughters sought less help as the program progressed. The finding that parents provided less help over time is positive because it may indicate that parents recognized that their daughters needed much support initially, but less as the program progressed, and so they adjusted their support accordingly. Similarly, the girls sought less help over time, indicating that they became more capable of working on the activities independently as the program progressed. It is important to note, though, that parents and daughters remained very engaged in the activities throughout the seven weeks of the program, demonstrated by no change in our measure of collaboration over time (in addition to the fact that we did not need to prompt them to remain engaged or on-task). This fits with Rogoff’s (1995) concept of the bidirectional social influence that occurs in interactions over time. As Rogoff states, “through engagement in an activity, individuals change and handle a later situation in ways prepared by their own participation in the previous situation” (p. 142). The familycentric routines and consistent structure of our program across the seven weeks helped the parent–child dyads coordinate their participation in the shared endeavor of problem-based learning. For example, the planned activities provided a space and a voice for parents to share their own knowledge and experiences related to the science concepts being presented. In the structure of our program, the conversation groups also provided a time and place for the parents to share with each other, similar to the adult conversation circles implemented by Chen and Pushor (2023). Finally, it is not surprising that both the parents and daughters had more “aha” moments at the end of the program compared to the beginning. The curriculum of both the rocketry program and gardening program were designed to build up to the culminating activity that concluded during the final week and the families reacted with more “aha” moments.

We found that several behaviors differed as a function of daughters’ age, even though the range of daughters’ ages was small (10 to 12 years old). Parents provided more help to older daughters than to younger daughters, which was somewhat surprising because typically young children need more help. For example, in a small-scale study, Short-Meyerson et al. (2016) examined the interactions of 2nd- and 4th-grade children with their parents while solving hands-on science problems together, finding that the 2nd graders received more parental help than the 4th graders. The nature and duration of the activities in their study were different, however, than the 7-week programs in our study. Another age difference found in our study was that parents had more instances of frustration with older daughters than with younger daughters. However, they provided more encouragement/praise and more slowing behaviors with younger daughters, indicating that parents may have been sensitive to younger daughters’ needing more of those types of support. Regarding daughter behaviors, younger daughters had more “aha” moments compared to the older daughters, perhaps indicating that some of the specific rocketry and gardening activities were more novel to them. By sixth grade, some of the girls may have already experienced some of the activities, which were new to the fifth graders. Younger daughters also provided more encouragement/praise than the older daughters, perhaps as a way of showing their enthusiasm and support to their parents. In general, younger girls may have been more willing to show enthusiasm than older girls.

A limitation of this study is its relatively small sample size. However, the fact that a number of statistically significant and interesting findings were revealed suggests that further investigation of this topic is warranted. Indeed, we have subsequently begun a larger follow up study that includes more families.

A strength of this study is that we had purposefully chosen to pilot two program topics (rocketry and gardening) in two locations (California and Arizona). We found that both topics were successful, as were both locations (e.g., the facilities, the partnerships with the school districts, the research teams, etc.). The discovery that there were some differences in findings among the programs was not surprising and it highlights the complexities of family science programs. For instance, the topics of the programs reflected two different science disciplines: physical science (rocketry) and life science (gardening), and previous research has shown that girls tend to be more interested in life science than physical science (Short-Meyerson et al., 2016). Furthermore, some things are idiosyncratic and unpredictable, especially when the program runs over multiple weeks (e.g., needing to make adjustments during the gardening program due to a temporarily broken sprinkler system). Interestingly, the statistical analyses revealed that for some of the models there was an interaction between topic and location, rather than effects of topic or location separately. For example, several of the parental behaviors that we investigated were more prominent in the gardening program in California than in any of the other three implementations. This highlights the fact that working with families and embracing a family-centric approach in informal learning environments creates distinct contexts in which not all experiences (or all participating families) are the same. This corresponds with Hong’s (2019, p. 161) suggestion that “schools celebrate the diversity of communities and strive to be unique, not uniform”. Our finding that there were differences between the programs underscores this point. Although many aspects of our programs were consistent across iterations, each of our programs was unique, meeting the needs of each community of learners. Rather than a one-size-fits all approach, it is crucial to be responsive to each particular group of families.

Our study indicates the need to engage parents in teaching and learning, which is a focus of the familycentric approach (Chen & Pushor, 2023). This may be fostered by providing families opportunities to engage in authentic and meaningful activities and conversations. In our program, this includes the hands-on family problem-based learning activities as well as the discussions that the parents had with their daughters about their current science classes, future science courses, and careers. Furthermore, throughout our program we considered and supported the needs of families. For example, we adapted the start time of the weekly program to be more convenient for families and we moved the location from the library to a multi-purpose room with larger tables to accommodate more interactions among the families. We also anchored what we did in the curriculum to bring in culture. This included involving the local community when possible, such as bringing in examples of local plants during the gardening program and discussing the impact of the local water shortage.

Ishimaru (2020) describes her research developing and examining a partnership between a public school district, which served many Latinx students learning English, and parents and community leaders. The goal of the partnership was to foster equitable collaborations. Two key principles that came from Ishimaru’s research guided our work: “begin with families and communities” (Ishimaru, 2020, p. 51) and “build reciprocity and agency” (Ishimaru, 2020, p. 53). As described above, we valued the expertise of the parents in our programs as they co-created the programs with us. Furthermore, their expertise was also highlighted during the programs. For example, we chose the location where the program was held with the parents and the CAFE specialists. This was based on parents’ input that there was a lack of out-of-school programs in that community. Two-way on-going communication between the families and the researchers was important to ensure that the parents had a voice. This looked different for every group of families. As one example, we had initially planned to share information about what middle school and high school classes were required for students who planned to apply to universities. Parents provided feedback on the plan, adding that they had older children in high school already and that it would be appreciated if we could also provide information for those older students. We adjusted our program to meet the families’ requests and consequently, not only did we meet the needs of older children, but we also had one parent who we mentored as she herself applied to a local four-year college.

5. Conclusions

As stated previously, the overall aim of our project is to broaden the participation in science of women and Latiné communities. Specific goals of the program were met. These included encouraging parents in supporting their Latina daughters in science, increasing Latina girls’ interest in science, and increasing the families’ participation in some out-of-school science experiences together. This was indicated through self-report by parents and daughters on the surveys and also by behavioral data gathered from the video-recorded FPBL sessions.

Our findings from the parent–daughter interactions during the FPBLs expand upon previous work. In particular, the design of our study enables the opportunity to examine behaviors over time, whereas the designs of most previous research studies have included only one session for each parent–child dyad (e.g., Short-Meyerson et al., 2016, 2019; Siegel et al., 2007; Tenenbaum & Leaper, 2003). The findings of our study have implications for designing, implementing, and researching in- and out-of-school science programs, especially those where families are involved. For instance, the increase in “aha” moments by daughters and parents over time indicates the importance of conducting programs for sufficiently long durations. Programs that are shorter than ours, such as a single session or a few weeks, may not be of long enough duration to have some of those experiences. “Aha” moments are important because they spark interest and enthusiasm, creating experiences that the girls (and their parents) are most likely to remember. They may even continue to discuss them at home after the program has ended.

Overall, the findings from this study add to the literature on the importance of exposure to a variety of science disciplines at the elementary level. We have produced a theory-supported model for procedures supporting parents in promoting positive attitudes and self-efficacy in science for their daughters to broaden participation in the field. In particular, parent support focuses on three dimensions. First, parents are interested in their daughters’ science endeavors, which will motivate parents to provide support. Second, parents believe that they can help shape their daughters’ science interests and abilities. That is, they have a sense of self-efficacy, regardless of their own level of science expertise. Finally, they know how they can support their daughters’ science interests and participation, such as through everyday conversations and experiences related to STEM (e.g., observing the night sky or examining a plant’s growth) and also through becoming aware of resources in the community (e.g., opportunities to attend a school STEM night or to volunteer to help with animals at the humane society or zoo). This model may enable practitioners and others in the field to better understand what Latiné families value in science education and what activities and prompts promote the most familial interactions and build on existing capital (i.e., Community Cultural Wealth).

The findings have implications for multiple learning contexts, in- and out-of-school. Given what we have learned and the finding that there were not changes in participation in out-of-home and in-home activities, we recommend providing take-home kits that replicate or expand upon the activities implemented in the family programs. For example, many of the families expressed a desire to build and launch water bottle rockets with their other children at home. We are confident that if we had provided those materials, they would have been used at home.

Consistent with the familycentric approach, we see tremendous value in working to create education programs with families and not only for families. From our experience, co-constructing the programs was key in providing opportunities that were meaningful for both the families and the education researchers. As we did in the current study, we recommend that in future studies and programs, parents who have experience participating in the program co-construct all aspects of the next iteration of the program (e.g., location, design, curriculum). We encourage those who design both in-school and out-of-school experiences to be intentional in their work with families and create comfortable spaces in which they can share experiences, support each other, and cultivate their children’s learning.