Nate Shannon IV
SEE THE DATA PREPROCESSING AND FACTOR ANALYSES HERE :)
Parental involvement in education is as unique and multifaceted as the families who navigate our school systems. Picture parents staying up late to review homework, juggling schedules to attend school events, or quietly nurturing their child’s confidence with heartfelt conversations about dreams and aspirations. These acts of engagement reflect more than just academic ambition—they are deeply rooted in cultural traditions, economic realities, and personal values.
Yet, the ways families participate in their child’s education are far from uniform, shaped by systemic barriers and cultural expectations. For some, school-based activities like PTA meetings or volunteering may fit naturally into their routines. For others—particularly families of color or those from lower socioeconomic backgrounds—these activities can be fraught with challenges, from inflexible work schedules to feelings of exclusion in school spaces. But these differences don’t reflect a lack of care or commitment; they reflect context.
The question is: how can we ensure that all forms of parental involvement are valued? This data brief explores the diverse ways families engage in their children’s education, challenging narrow definitions of involvement and celebrating the strengths different communities bring to the table. By doing so, we aim to reframe how we view engagement—not as a deficit to be corrected but as a rich spectrum of practices that deserve recognition and support.
Let’s start:
library(dplyr)
library(psych)
library(readr)
library(ggplot2)
library(waffle)
library(ggthemes)
library(directlabels)
library(stringr)
library(tidyr)
library(sjPlot)# Variables to convert to factors
variables_to_factor <- c(
"PAR1_RACE", "PAR2_RACE", "PAR1_Relation",
"PAR2_Relation", "ELL_STATUS", "HOME_LANG")
involvement_data[variables_to_factor] <- lapply(involvement_data[variables_to_factor], as.factor)Data Overview This study draws from the High School Longitudinal Study conducted by the National Center for Education Statistics (NCES), a nationally representative dataset capturing the educational journeys of approximately 25,000 students from over 240 schools. These students, who began high school in 2009, participated in surveys that shed light on family demographics, educational engagement, and the nuanced ways families support their children.
Central to this analysis are two dimensions of parental involvement: School-Based Involvement (SBI) and Home-Based Involvement (HBI).
SBI encompasses participation in school-centric activities like parent-teacher conferences, meeting with counselors, or attending school events—eight items in total.
HBI focuses on activities that foster a love of learning outside the classroom, such as museum visits, library trips, or STEM-related discussions—seven items in total. Together, these measures reveal the diverse and often overlooked ways families contribute to their high school student’s education.
Let’s take a closer look at the racial makeup of the survey population:
# Scale the problem, overview of the population
race_demo <- table(involvement_data$PAR1_RACE)/sum(table(involvement_data$PAR1_RACE))*100
# Convert race_demo table into a data frame for ggplot
race_demo_df <- as.data.frame(race_demo)
colnames(race_demo_df) <- c("Race", "Percentage")
# Integer values to sum exactly 100, that took too long to figure out! >:(
race_demo_df$Tiles <- floor(race_demo_df$Percentage)
remaining_tiles <- 100 - sum(race_demo_df$Tiles)
# Leftover tiles got to the category with the largest percentage
race_demo_df$Tiles[which.max(race_demo_df$Percentage)] <-
race_demo_df$Tiles[which.max(race_demo_df$Percentage)] + remaining_tiles
# Waffle chart
ggplot(race_demo_df, aes(fill = factor(Race), values = Tiles)) +
waffle::geom_waffle(
n_rows = 10,
size = 0.5,
colour = "black",
flip = TRUE
) +
scale_fill_manual(
values = c("#f94144", "#f9c74f", "#43aa8b", "#277da1", "#f3722c"),
name = "Race Categories:",
labels = c("Asian", "Black", "Hispanic", "Mixed", "White")
) +
theme_minimal() +
theme(
panel.background = element_rect(fill = "#e8d8c3", color = NA),
plot.background = element_rect(fill = "#e8d8c3", color = NA),
panel.grid = element_blank(),
axis.text = element_blank(),
axis.title = element_blank(),
legend.position = "bottom",
legend.title = element_text(size = 10, face = "bold"),
legend.text = element_text(size = 9)
) +
ggtitle("Proportion of Families by Parental Race") +
theme(plot.title = element_text(size = 14, face = "bold", hjust = 0.5))
The demographic makeup of the study’s population provides essential context for understanding the data. White families account for 68% of the total, followed by Hispanic families at 13%, Black families at 9%, Asian families at 7%, and mixed-race families at 3%.
This mosaic of diversity lays the foundation for exploring how families engage in their child’s education. It reminds us that every family brings a unique story, shaped by culture, resources, and life experiences (see the literature on cultural capital if you get a chance). Understanding these stories is key to rethinking parental involvement and creating more inclusive educational systems.
Now, let’s look what general involvement looks like across parents of different races.
#Min-Max Scaling
min_max_scale <- function(column) {
(column - min(column, na.rm = TRUE)) / (max(column, na.rm = TRUE) - min(column, na.rm = TRUE))
}
#Total_Inv and scaled variables
involvement_data <- involvement_data %>%
mutate(
Total_Inv = HBI_Tot + SBI_Tot,
Total_Inv_scaled = Total_Inv
)
#Combine data and compute mean
race_summary_total <- involvement_data %>%
filter(!is.na(PAR1_RACE)) %>% # Remove rows with NA in PAR1_RACE
group_by(PAR1_RACE) %>%
summarise(
Total_Inv_scaled = mean(Total_Inv_scaled, na.rm = TRUE)
)
# Race labels for better readability
race_summary_total <- race_summary_total %>%
mutate(
PAR1_RACE = recode(
PAR1_RACE,
`2` = "Asian",
`3` = "Black",
`4` = "Hispanic",
`6` = "Mixed",
`8` = "White"
)
)
# Bar chart for Total Involvement Score
ggplot(race_summary_total, aes(x = PAR1_RACE, y = Total_Inv_scaled, fill = PAR1_RACE)) +
geom_bar(stat = "identity", position = "dodge", colour = "black") +
scale_fill_manual(
values = c("#f94144", "#f9c74f", "#43aa8b", "#277da1", "#f3722c"),
name = NULL # Remove legend title
) +
scale_y_continuous(
breaks = seq(0, max(race_summary_total$Total_Inv_scaled, na.rm = TRUE), by = 1) #Axis scaling
) +
labs(
title = "Total Parental Involvement by Race",
x = "Parental Race/Ethnicity",
y = "Total Involvement Score"
) +
theme_minimal() +
theme(
legend.position = "none", # Remove legend if redundant
panel.background = element_rect(fill = "#e8d8c3", color = NA),
plot.background = element_rect(fill = "#e8d8c3", color = NA),
plot.title = element_text(size = 14, face = "bold", hjust = 0.5),
axis.text.x = element_text(size = 12, angle = 45, hjust = 1)
)
When we look at overall levels of involvement, by combining home- and school-based activities, we see clear patterns. Mixed-race parents report the highest levels of engagement, followed closely by Black and White parents. Hispanic and Asian families, meanwhile, report lower levels of involvement.
But what do these numbers really mean and where is the involvment happening? They are not a reflection of how much families care about their student’s education but rather the diverse ways families participate—ways that are often shaped by cultural norms, systemic barriers, and personal circumstances. To truly understand these differences, we must separate school-based from home-based involvement.
Let’s take a closer look:
#Min-Max Scaling
min_max_scale <- function(column) {
(column - min(column, na.rm = TRUE)) / (max(column, na.rm = TRUE) - min(column, na.rm = TRUE))
}
#Add scaled variables to the data
involvement_data <- involvement_data %>%
mutate(
SBI_Tot_scaled = min_max_scale(SBI_Tot),
HBI_Tot_scaled = min_max_scale(HBI_Tot)
)
#Compute mean scores, and include scaled variables
race_summary <- involvement_data %>%
filter(!is.na(PAR1_RACE)) %>% # Remove rows with NA in PAR1_RACE
group_by(PAR1_RACE) %>%
summarise(
SBI_Total_scaled = mean(SBI_Tot_scaled, na.rm = TRUE),
HBI_Total_scaled = mean(HBI_Tot_scaled, na.rm = TRUE)
) %>%
pivot_longer(
cols = c(SBI_Total_scaled, HBI_Total_scaled),
names_to = "Involvement_Type",
values_to = "Mean_Score"
) %>%
mutate(
Involvement_Type = case_when(
Involvement_Type == "SBI_Total_scaled" ~ "SBI",
Involvement_Type == "HBI_Total_scaled" ~ "HBI",
TRUE ~ Involvement_Type
)
)
#Grouped bar chart
ggplot(race_summary, aes(x = Mean_Score, y = Involvement_Type, fill = factor(PAR1_RACE))) +
geom_bar(stat = "identity", position = "dodge", colour = "black") +
scale_fill_manual(
values = c("#f94144", "#f9c74f", "#43aa8b", "#277da1", "#f3722c"),
name = NULL,
labels = c("Asian", "Black", "Hispanic", "Mixed", "White") # Adjust labels as needed
) +
scale_x_continuous(
breaks = seq(0, max(race_summary$Mean_Score, na.rm = TRUE), by = 0.1) # Axis Scaling
) +
labs(
title = "Comparison of School- and Home-Based Involvement by Race",
x = "Involvement Score (Scaled)",
y = "Involvement Type"
) +
theme_minimal() +
theme(
legend.position = "bottom",
panel.background = element_rect(fill = "#e8d8c3", color = NA),
plot.background = element_rect(fill = "#e8d8c3", color = NA),
plot.title = element_text(size = 14, face = "bold", hjust = 0.5),
axis.text.y = element_text(size = 12, angle = 0, hjust = 1)
)
(note: variables scaled to be on the same level, from 0-1)
Involvement in education isn’t one-size-fits-all—it’s a reflection of cultural, economic, and personal contexts that shape how families support their children.
The bar graph above demonstrates that home-based involvement (HBI) is more prevalent overall among parents than school-based involvement (SBI). This aligns with both our own experiences as teens and the scientific literature, which suggests that as students mature, parents often shift their focus toward fostering autonomy, emphasizing engagement at home over direct participation in school activities.
Yet, how do these patterns of involvement translate into academic outcomes? Understanding the link between different forms of parental engagement and student achievement provides critical insights into how families can best support their child’s success. To explore this, we analyzed the relationship between home- and school-based involvement, as well as communication with school staff, and their impact on students’ STEM GPA.
Parents play a vital role in shaping their student’s academic success, but how much does their involvement actually affect achievement outcomes? To answer this, we analyzed the relationship between different forms of parental involvement—like school-centered activities, home enrichment, and communication with school staff—and students’ STEM GPA.
Our analysis also accounted for important factors like socioeconomic status (SES) and the parent’s role in the child’s life (mother, father, etc.) to isolate the effect of involvement itself. Even with these adjustments, we found significant differences in predicted STEM GPA across racial groups. Students with Black, Hispanic, and Mixed-Race parents tend to have lower STEM GPAs compared to their peers with White or Asian parents. In fact, the gap is as large as 0.85 GPA points for students with Black parents.
Below, we’ve generated some graphs to describes the trends in our analysis:
# Regression model predicting STEM GPA from 3 types of involvement and with main effects and interactions
#As a part of this data brief, we also performed an exploratory factor analysis (EFA) to determine if there are some more granular components to involvement. We identified the following sub-components of SBI and HBI. (link at top of document).
####SBI "school activities" (fundraiser, volunteering, school events) and "school-parent communication" (going to a PT conference, PTA meeting, counselor, or general school meeting).
####HBI "home enrichment" (taking child to a museum, show, library, and fixing a computer.
involvement_data$PAR1_RACE <- relevel(involvement_data$PAR1_RACE, ref = "2")
interaction_formula <- GPA_STEM ~ PAR1_RACE + home_enrichment + School_Activities + Parent_Sch_Comm + PAR1_Relation + home_enrichment:PAR1_RACE + School_Activities:PAR1_RACE + Parent_Sch_Comm:PAR1_RACE + SES_Q5
interaction_model <- lm(interaction_formula, data = involvement_data)
summary(interaction_model)##
## Call:
## lm(formula = interaction_formula, data = involvement_data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.9942 -0.5498 0.0363 0.5845 2.3145
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.3590720 0.0564352 41.801 < 2e-16 ***
## PAR1_RACE3 -0.8359104 0.0788910 -10.596 < 2e-16 ***
## PAR1_RACE4 -0.7101393 0.0701619 -10.121 < 2e-16 ***
## PAR1_RACE6 -1.0043038 0.1091942 -9.197 < 2e-16 ***
## PAR1_RACE8 -0.6216779 0.0592678 -10.489 < 2e-16 ***
## home_enrichment 0.0002133 0.0204743 0.010 0.991688
## School_Activities 0.0436458 0.0251499 1.735 0.082687 .
## Parent_Sch_Comm -0.0225205 0.0215994 -1.043 0.297130
## PAR1_Relation2 0.0456230 0.0159943 2.852 0.004345 **
## SES_Q5 0.1781583 0.0051276 34.745 < 2e-16 ***
## PAR1_RACE3:home_enrichment -0.0036777 0.0283618 -0.130 0.896828
## PAR1_RACE4:home_enrichment 0.0471324 0.0258876 1.821 0.068680 .
## PAR1_RACE6:home_enrichment 0.0696804 0.0372765 1.869 0.061603 .
## PAR1_RACE8:home_enrichment 0.0496743 0.0218172 2.277 0.022810 *
## PAR1_RACE3:School_Activities 0.1198470 0.0340631 3.518 0.000436 ***
## PAR1_RACE4:School_Activities 0.0768318 0.0318306 2.414 0.015801 *
## PAR1_RACE6:School_Activities 0.1171689 0.0430451 2.722 0.006497 **
## PAR1_RACE8:School_Activities 0.1148402 0.0265345 4.328 1.52e-05 ***
## PAR1_RACE3:Parent_Sch_Comm -0.0386690 0.0287876 -1.343 0.179212
## PAR1_RACE4:Parent_Sch_Comm -0.0404324 0.0269201 -1.502 0.133134
## PAR1_RACE6:Parent_Sch_Comm 0.0173940 0.0365047 0.476 0.633735
## PAR1_RACE8:Parent_Sch_Comm -0.0297871 0.0229473 -1.298 0.194286
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.8031 on 14451 degrees of freedom
## (9030 observations deleted due to missingness)
## Multiple R-squared: 0.2198, Adjusted R-squared: 0.2187
## F-statistic: 193.8 on 21 and 14451 DF, p-value: < 2.2e-16# Interaction plot for Parent-School Communication and PAR1_RACE
plot_model(interaction_model, type = "pred", terms = c("Parent_Sch_Comm", "PAR1_RACE")) +
scale_color_manual(
values = c("#7d0204", "#f9c74f", "#43aa8b", "#277da1", "#ff975e"),
name = "Race Categories",
labels = c("Asian", "Black", "Hispanic", "Mixed", "White")
) +
theme_minimal() +
theme(
panel.background = element_rect(fill = "#e8d8c3", color = NA),
plot.background = element_rect(fill = "#e8d8c3", color = NA),
panel.grid = element_blank(),
legend.position = "bottom",
legend.title = element_text(size = 10, face = "bold"),
legend.text = element_text(size = 9),
axis.text = element_text(size = 10),
axis.title = element_text(size = 12),
plot.title = element_text(size = 14, face = "bold", hjust = 0.5)
) +
labs(
title = "Effect of Parent-School Communication and Race on STEM GPA",
x = "Parent-School Communication",
y = "Predicted GPA_STEM"
)## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.
Direct parent-teacher communication, such as meetings with counselors or teachers, shows no significant impact on STEM grades overall. In fact, for some racial groups, frequent communication is associated with slightly lower predicted STEM GPAs—potentially reflecting cases where communication is reactive rather than proactive, tied to other academic struggles.
# Interaction plot for home_enrichment and PAR1_RACE
plot_model(interaction_model, type = "pred", terms = c("home_enrichment", "PAR1_RACE")) +
scale_color_manual(
values = c("#7d0204", "#f9c74f", "#43aa8b", "#277da1", "#ff975e"),
name = "Race Categories",
labels = c("Asian", "Black", "Hispanic", "Mixed", "White")
) +
theme_minimal() +
theme(
panel.background = element_rect(fill = "#e8d8c3", color = NA),
plot.background = element_rect(fill = "#e8d8c3", color = NA),
panel.grid = element_blank(),
legend.position = "bottom",
legend.title = element_text(size = 10, face = "bold"),
legend.text = element_text(size = 9),
axis.text = element_text(size = 10),
axis.title = element_text(size = 12),
plot.title = element_text(size = 14, face = "bold", hjust = 0.5)
) +
labs(
title = "Effect of Home Enrichment and Race on STEM GPA",
x = "Home Enrichment",
y = "Predicted GPA_STEM"
)## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.
As home enrichment activities increase (moving from left to right), there is a slight upward trend in STEM GPA for most racial groups. This suggests that more home enrichment is generally associated with improved STEM performance.When it comes to the impact of home enrichment activities on STEM performance, the story is anything but uniform across racial groups.
Asian students, who begin with the highest predicted STEM GPAs, show only minimal gains as home enrichment activities—like STEM-related discussions, library visits, or museum trips—increase. Their already high performance suggests that other factors may play a larger role in their academic success.
White, Hispanic, and Mixed-Race students, on the other hand, see steady improvements as home enrichment levels rise. These groups experiences a stronger boost from these activities than any other, highlighting how additional academic support at home can reinforce learning and contribute to better grades in STEM.
For Black students, the picture is less encouraging. Starting with lower predicted GPA boost from involeement, they see only slight improvements from home enrichment. This limited benefit points to broader systemic challenges, such as disparities in access to high-quality educational resources, which may overshadow the potential gains from these activities.
These patterns underscore a broader reality: while home enrichment can play a role in shaping academic outcomes, its impact is deeply intertwined with structural and cultural contexts that vary across racial groups.
# Interaction plot for School Activities and PAR1_RACE
plot_model(interaction_model, type = "pred", terms = c("School_Activities", "PAR1_RACE")) +
scale_color_manual(
values = c("#7d0204", "#f9c74f", "#43aa8b", "#277da1", "#ff975e"),
name = "Race Categories",
labels = c("Asian", "Black", "Hispanic", "Mixed", "White")
) +
theme_minimal() +
theme(
panel.background = element_rect(fill = "#e8d8c3", color = NA),
plot.background = element_rect(fill = "#e8d8c3", color = NA),
panel.grid = element_blank(),
legend.position = "bottom",
legend.title = element_text(size = 10, face = "bold"),
legend.text = element_text(size = 9),
axis.text = element_text(size = 10),
axis.title = element_text(size = 12),
plot.title = element_text(size = 14, face = "bold", hjust = 0.5)
) +
labs(
title = "Effect of School Activ. and Race on STem GPA",
x = "School Activities",
y = "Predicted GPA_STEM"
)## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.
Participation in school-centered activities—like fundraisers, school events, or parent-teacher meetings—has a more noticeable impact on STEM grades for all high school students. This effect is particularly pronounced for Black, Hispanic, Mixed-Race, and White families, where every additional level of involvement adds as much as 0.12 GPA points. These activities seem to serve as a bridge, helping historically marginalized groups better connect with school systems.
This research sheds light on the complex relationship between parental involvement and STEM achievement, emphasizing the need to move beyond one-size-fits-all definitions of involvement. By analyzing various forms of engagement—home enrichment, school-centered activities, and parent-school communication—we uncovered significant racial disparities in their effects. While school activities provide a measurable boost for many students, home enrichment shows uneven benefits, particularly for Black families, where systemic barriers may limit its impact.
At the same time, our findings illustrate the meaningful, culturally specific ways families support their children.
Hispanic parents often foster educational values through family discussions (Jeynes, 2010; Dotterer, 2022), while Asian parents emphasize academic discipline and resilience, leveraging cultural expectations and peer networks to guide success (Chua, 2011; Sue & Okazaki, 2022). Mixed-race parents, who demonstrate the highest levels of involvement, appear to benefit from broader social capital—such as cross-cultural fluency and diverse networks—that enable a uniquely adaptive approach to supporting their children. Notably, despite reporting lower levels of traditional involvement, Asian students achieve the highest GPAs, suggesting that culturally specific strategies can be as impactful as conventional measures of engagement.
These findings underscore the importance of recognizing and respecting the diverse ways families contribute to their children’s education. Schools and policymakers must prioritize inclusive approaches that value all forms of involvement—whether they occur in the classroom, at home, or through cultural practices. When every family’s efforts are acknowledged and supported, we can ensure that all students have the opportunity to succeed and thrive. The time to act is now!
Thanks for reading!
Chua, E. (2011). Battle hymn of the tiger mom. New York: Penguin Press.
Dotterer, A. M. (2022). Diversity and complexity in the theoretical and empirical study of parental involvement during adolescence and emerging adulthood. Educational Psychologist, 57(4), 295-308.
Jeynes, W. (2010). The salience of the subtle aspects of parental involvement and encouraging that involvement: implications for school-based programs. Teachers College Record, 112, 74.
Sue, S., & Okazaki, S. (2022). Asian-American educational achievements: A phenomenon in search of an explanation. In The New Immigrants and American Schools (pp. 297-304). Routledge.