A New Wave of Evidence-In Short

The evidence is consistent, positive, and convincing: families have a major influence on their children's achievement in school and through life. This fourth edition of Evidence confirms that the research continues to grow and build an ever-strengthening case. When schools, families, and community groups work together to support learning, children tend to do better in school, stay in school longer, and like school more.

How are the many ways that families are engaged in their children's education related to achievement? Many studies found that students with involved parents, no matter what their income or background, were more likely to

• earn higher grades and test scores, and enroll in higher-level programs.
• be promoted, pass their classes, and earn credits.
• attend school regularly.
• have better social skills, show improved behavior, and adapt well to school.
• graduate and go on to postsecondary education.

Several studies found that families of all income and education levels, and from all ethnic and cultural groups, are engaged in supporting their children's learning at home. White, middle-class families, however, tend to be more involved at school. Supporting more involvement at school from all parents may be an important strategy for addressing the achievement gap.

Do programs and special efforts to engage families make a difference?

Yes, several studies found that they do. For example, teacher outreach to parents was related to strong and consistent gains in student performance in both reading and math. The effective outreach practices included meeting face to face, sending materials home, and keeping in touch about progress. Workshops for parents on helping their children at home were linked to higher reading and math scores. Schools with highly rated partnership programs made greater gains on state tests than schools with lower rated programs.

How do higher performing schools engage families and community?

Schools that succeed in engaging families from very diverse backgrounds share three key practices. They

• focus on building trusting collaborative relationships among teachers, families, and community members.
• recognize, respect, and address families' needs, as well as class and cultural difference.
• embrace a philosophy of partnership where power and responsibility are shared.

What is the impact of parent and community organizing on improving schools?

This type of engagement is based outside schools and led by parents and community members, and it is growing nationwide. These efforts are aimed at schools that are low performing. Strategies of community organizing are different from traditional parent involvement and are openly focused on building low-income families' power and political skills to hold schools accountable for results.

A new group of studies found that community organizing contributed to these changes in schools:

• Upgraded school facilities.
• Improved school leadership and staffing.
• Higher-quality learning programs for students.
• New resources and programs to improve teaching and curriculum.
• New funding for after-school programs and family supports.

Summing up

When parents talk to their children about school, expect them to do well, help them plan for college, and make sure that out-of-school activities are constructive, their children do better in school. When schools engage families in ways that are linked to improving learning, students make greater gains. When schools build partnerships with families that respond to their concerns and honor their contributions, they are successful in sustaining connections that are aimed at improving student achievement. And when families and communities organize to hold poorly performing schools accountable, studies suggest that school districts make positive changes in policy, practice, and resources.

How can we put these findings into action?

• Recognize that all parents — regardless of income, education, or cultural background — are involved in their children's learning and want their children to do well.
• Design programs that will support families to guide their children's learning, from preschool through high school.
• Develop the capacity of school staff to work with families.
• Link efforts to engage families, whether based at school or in the community, to student learning.
• Build families' social and political connections.
• Focus efforts to engage families and community members on developing trusting and respectful relationships.
• Embrace a philosophy of partnership and be willing to share power with families.
  Make sure that parents, school staff, and community members understand that the responsibility for children's educational development is a collaborative enterprise.
• Build strong connections between schools and community organizations.
• Include families in all strategies to reduce the achievement gap among white, middle-class students and low-income students and students of color.

Introduction

The Office of the Superintendent of Public Instruction (OSPI) of the State of Washington recently published a literature review of 20 studies that examined the common characteristics of high-performing schools. These studies include several U.S. Department of Education studies, including Hope for Urban Education: A Study of Nine High Performing, High Poverty Urban Elementary Schools (Mayer, D. P., Mullens, J. E., & Moore, M. T., 2000), and Monitoring School Quality: An Indicators Report (Charles A. Dana Center, 1999). Their research found that high-performing schools tend to have a combination of many characteristics, which were narrowed into these nine areas:

1. A clear and shared focus.
2. High standards and expectations for all students.
3. Effective school leadership.
4. High levels of collaboration and communication.
5. Curriculum, instruction, and assessments aligned with state standards.
6. Frequent monitoring of teaching and learning.
7. Focused professional development.
8. A supportive learning environment.
9. High levels of parent and community involvement.

The purpose of this publication, A New Wave of Evidence: The Impact of School, Family, and Community Connections on Student Achievement, is to examine one of these identified characteristics of high-performing schools: parent and community involvement and its role in impacting on student achievement. This publication is the fourth in the series of Evidence publications authored or co-authored by Anne Henderson. It is also the second in the series of publications by SEDL’s National Center for Family and Community Connections with Schools (hereinafter referred to as the Center).

The Center’s first publication, Emerging Issues in School, Family, and Community Connections, is a research synthesis created to identify “key issues that must be addressed if research is to assist schools, families, and communities in working together to nurture high standards and academic success for all children” (p. 1). For Emerging Issues, SEDL staff reviewed a broad body of literature on the process and impact of school, family, and community connections. This body of literature reviewed is captured in full in an online, searchable annotated bibliography database, The Connection Collection: School-Family-Community Publications Database (2002) available at www.sedl.org/connections/resources/. (See “About the Studies” section, p. 13, for information about how we selected the studies.)

Although we tried to cover a vast range of topics, certain areas were deemed too important to cover briefly in this publication. These topics will be treated separately in future annual research syntheses published by the Center. The 2003 synthesis will focus on connecting families from diverse backgrounds with schools. In future years, the Center will take a closer look at the involvement of community organizations in the process of transforming schools into high-performing learning communities and connecting families and schools to support successful transitions through the education system.

We tried to write in reader-friendly language free of educational jargon. We also tried to explain and demystify some of the more complex statistical methods and results used in the research studies.

Some Definitions

Throughout this report, we frequently use the words “family” or “families” in place of “parent” or “parents.” We want to recognize that all family members—siblings, grandparents, aunts, uncles, and “fictive kin” who may be friends or neighbors—often contribute in significant ways to children’s education and development. If a study uses the terms “parent” or “parents,” we stick to the terminology used by the researchers.

For the purposes of this report, we use the terms “connection” and “involvement” interchangeably.

By “community” we mean:

• the neighborhood or the places around the school.
• local residents, who live in the area and may or may not have children in theschool, but have an interest in the school.
• local groups that are based in the neighborhood.

How This Report Is Organized

We have tried to organize this report in a way that will be easy to navigate. Here is a brief guide to what is in the report.

About the Studies

The section describes the methods used for selecting the studies, describes what the studies cover, and provides a table showing the studies by topic area, by age and grade level, and by design type (Tables 1–3, pp. 15–17). Limitations of the studies are indicated.

Synthesis of Research Studies

The first section of the synthesis sums up the findings briefly and provides some definitions. Following that, the studies are divided into three categories:

• Impact of Parent and Community Involvement on Student Achievement;
• Effective Strategies to Connect Schools, Families, and Community; and
• Parent and Community Organizing Efforts to Improve Schools.

The next section lists a series of recommendations designed to help people put these findings into use in a practical way, followed by the conclusion.

The Research Studies

This section provides summaries of the 51 studies described in this report.

Appendix: Looking Back—A Brief History and Key Studies, 1974–95

Because the studies in this report are all recent (1993–2002), we also include a short history of the research in this field for the past 30 years. Summaries of key studies mentioned in the brief history are also included in this section. A review of these studies with longer summaries is available in the previous edition, A New Generation of Evidence: The Family Is Critical to Student Achievement (1994), by Anne T. Henderson and Nancy Berla. It can be obtained from the Center for Law and Education in Washington, D.C., at www.cleweb.org.

About the Authors and Publisher

More information about the writers of this report and the Southwest Educational Development Laboratory is available in this section.

References

This section lists works cited.

Index

This report has a full index to help the reader find studies and topics of interest. We hope that this report will be a useful tool for educators, researchers, policymakers, funders, community leaders, and others interested in the impact of school, family, and community connections on children’s learning.

About the Studies

How We Selected the Studies

This review examines the growing evidence that family and community connections with schools make a difference in student success. It is the second in a series from the National Center for Family and Community Connections with Schools at the Southwest Educational Development Laboratory (SEDL). The first publication in the series was Emerging Issues in School, Family, and Community Connections. For Emerging Issues, SEDL staff reviewed a broad body of recent literature (published since 1995, with the exception of a few seminal studies) on the process and impact of school, family, and community connections. An annotated bibliography of more than 200 research studies, conceptual or theoretical pieces, practice and policy-oriented works, and literature reviews is available as an online, searchable database titled The Connection Collection (2002) on the SEDL Web site at www.sedl.org/connections/resources/.

For A New Wave of Evidence, SEDL staff identified about 80 research studies and literature reviews out of the documents they had reviewed. SEDL staff also did a further search in such major databases as the Educational Resources Information Center (ERIC) and Education Abstracts. This subgroup of studies focuses on the influence of family and community involvement on student academic achievement and other outcomes. (See the "Synthesis of Research Studies" section on p. 21 of this report for more detail on how these studies defined student achievement and family involvement.) In addition, we asked colleagues in the field to recommend other studies and send us copies of their research.

All studies were reviewed to make sure they met these standards:

1. Sound methodology: experimental, quasi-experimental, or correlational design with statistical controls. For qualitative studies, such as case studies, we looked for sound theory, objective observation, and thorough design.
2. Study findings that matched the data collected and conclusions that were consistent with the findings.

Our choices were, of course, limited to what was available, and published in the past eight or nine years. In choosing the 51 studies that were ultimately included, we looked for a range of studies that covered:

• early childhood through high school;
• all regions of the country;
• diverse populations (income, race/ethnicity, educational level, and occupation);
• community as well as parent and family involvement;
• a variety of methods, both quantitative and qualitative; and
• different sources of data (survey research, evaluations, case studies, experimental and quasi-experimental studies, and research reviews).

Finally, we included studies that attempted to break new ground, either in defining student outcomes, ways that families and community members were engaged, or theories of change. In the interest of focus and scale, we did not include studies on special education, educational policy, parent choice, or business partnerships. These topics will be covered in later reports.

What the Studies Cover

Tables 1-3 (pp. 15-17) group the studies by design type, general topic, and age and grade level. This grouping will help the reader find studies more easily and will display the many topics, methods, and grade levels covered. In classifying the studies by methods, we used the typology of empirical studies presented in Amy Baker and Laura Soden's review (1997).

• Pre-experimental studies: no comparison group, or the comparison group not randomly assigned and assessed at pretest.
• Quasi-experimental studies: no pretest comparability between treatment and comparison families (for example, comparing treatment students with students from the year before or in a different class).
• Ex post facto and correlational studies: level of involvement is naturally occurring, not randomly assigned. Parent involvement is a continuous variable that is related to a continuous dependent variable, without an intervention.
• Experimental studies: families are assigned to a treatment and control group at random, compared at pretest, received an intervention or not, then tested after the intervention.

Following the tables, a section on the limitations of this research provides more detailed standards for experimental studies.

*Where a study appears under more than one topic, the second mention is indicated by an asterisk.

*This table does not include the literature reviews.

Note: This table uses the typology from Baker and Soden’s literature review (1997).

Limitations of the Research

We feel confident in making a strong statement about the benefits of school, family,and community connections. We also feel confident that the studies we have selected were carefully done and thoughtfully interpreted. However, the research in this field shares many of the limitations found in other areas of educational research. Certainly, there are not enough experimental or quasi-experimental studies. We included the few we found.

There is also not enough long-term research because of the limits of funding for such ambitious work. Many studies have small samples, while others depend on self-reports rather than independent verification. Many conclusions have to be carefully hedged because little can be said about cause and effect. Some studies have mixed, ambiguous, or incomplete findings and conclusions.

Nevertheless, we feel that the findings from the research reviewed here can be useful to our country's efforts to improve the policies and practices of schools. Some more detailed comments about the limitations follow.

1. Studies of programs. In their critique of research on early childhood programs, Karl White and his colleagues (1992) point out that few evaluation studies are based on rigorous standards for validity. (See White et al. in Appendix, p. 216.) These standards for experimental studies include:

• Children to be studied are assigned at random either to a treatment or a control group.
• The two groups are comparable in terms of family background. This is verified by interviews with families.
• The two groups stay together from the beginning to the end of the study.
• The interventions are described in detail, and are fully carried out.
• Trained testers assess the students in a neutral place.

Only a few studies in this review, all of programs or interventions, met these standards. These were conducted on Early Head Start, Head Start Family Math, and the HIPPY program (Mathematica et al., Starkey and Klein, and Baker et al.). The studies on Project EASE, Interactive Math Homework, and TIPS Science were quasi-experimental because the control groups were not chosen by random selection (Jordan et al., Balli et al., Van Voorhis). Most other studies about the effects of parent involvement on student achievement used a correlational method, with statistical controls (Clark, 2002; Clark, 1993; Epstein, Simon, and Salinas; Invernizzi et al.; Marcon; Moore; Shaver and Walls; and Westat/Policy Studies Associates). The report by Dryfoos reviewed findings from studies done by others, but did not critique their methods.

The correlational studies compared children of more highly involved parents with children of less involved parents, rather than with a control group. Neither group of children was chosen by random assignment. The researchers used statistical methods to analyze the relationship between level of involvement and improved student outcomes. Then they introduced controls for family income, occupation, and education levels About the Studies Southwest Educational Development Laboratory 19 to see if the effects could be explained by other factors. This method is considered reliable, but it may miss or fail to measure some factors that could account for the findings.

2. Studies using survey data. Many studies, and all on middle and high school students, use large databases such as the National Educational Longitudinal Study (NELS). These studies use correlational methods. In interpreting the findings, we need to be aware of some limitations in this type of research.

• NELS:88 and other survey data about parent involvement are based almost entirely on what parents, students, and educators report in structured interviews or questionnaires. There was no way in the data collection design to confirm that reported behavior matched real behavior. NELS:88, for example, covers 25,000 students. A few studies use a data source that includes information from open-ended interviews with a small subsample of parents or students (Miedel and Reynolds, Gutman and Midgley, Sanders and Herting, Shumow and Miller). This offers more information but it still is self-reported. Three studies cross-checked parent responses with student and teacher reports (Keith and Keith, Miedel and Reynolds, Trusty). Jerry Trusty found that students' reports about their parents' involvement had the strongest effects. In other words, the more students perceive their family's involvement and support, the better they tend to do in school.

• Studies using survey data are looking after the fact. They ask how much involvement has happened or is happening. That approach can make what is called "directionality" hard to determine. This is a problem with all studies that collect data at one point in time. In these studies, we can see that more parent involvement and higher achievement are related in some areas. But which came first? Perhaps higher achieving children attract more parent involvement, rather than the other way around. Some studies attempted to address this issue by controlling for prior achievement.

  Miedel and Reynolds checked to see if children's kindergarten readiness scores were more powerful than parent involvement in predicting later achievement. They found that the number of activities parents took part in during the early years of schooling had an effect on eighth-grade achievement that was independent of readiness. Controlling for achievement, Catsambis found that students with both low and high grades seemed to benefit from discussions about school and planning for college with their families. In other words, parent involvement is related to achievement gains for both high- and low-achieving students.

• Survey data tend to cover many topics, but without probing them deeply. They don't tell us why parents, students, or teachers responded the way they did or what they might like us to know. The relationships among parents, teachers, and students are complex and influenced by many factors. From survey research, we can only conjecture what is going on. As Baker and Soden put it in their review (1997):

  Closed-ended self-report surveys cannot fully capture the dynamic transactional nature of parents' involvement in their children's education. Many of these A processes could better be explored through open-ended and observational techniques that would produce rich data, shed light on complex processes, and generate new hypotheses. (p. 15)

How to Get Copies of the Studies

Many of the studies covered in this overview are available through the Educational Resource Information Center (ERIC) system. If the study summary includes a listing of ED or EJ, followed by a six-digit number, the publication is indexed in ERIC. The ERIC numbers are the unique identifiers assigned to each ERIC entry. For more information about ERIC, visit the Web site at www.eric.ed.gov.

ED Numbers. ERIC numbers that begin with "ED" (e.g., ED 435484) refer to documents indexed in ERIC. You can use the ED number to find the resource in the ERIC database online at www.eric.ed.gov. Although you can't read the whole document online, you can read an abstract or summary. Most documents can be ordered from the ERIC document reproduction service at www.edrs.com. Choose from the following formats: Downloadable PDF file, a print copy, or microfiche.

The ERIC Document Reproduction Service also has a phone number, 1-800-443-ERIC. Another option is to visit one of the ERIC Resource Collections. They provide access to full-text ERIC documents via microfiche or electronic indexes. Find the one closest to you through the Directory of ERIC Resource Collections: http://oeri4.ed.gov/BASISDB/EROD/eric/SF/.

EJ Numbers. ERIC numbers that begin with "EJ" (e.g., EJ 674533) refer to journal articles. The least-expensive way to obtain a full copy is to consult a library. If your local library does not subscribe to a particular journal, or have what is called "full-text electronic access" to that journal, you can get copies through interlibrary loan (ILL). For a fee, there are article reproduction services that will provide a copy. Ask your library to suggest one.

How the Studies Define Family Involvement and Student Achievement

In this review, we look at 51 recent studies, all but two published from 1995 to 2002. Every one sheds some light on the relationship between parent involvement and, in some cases, community involvement and improved student achievement. Together they cover children and youth of all ages, from birth through high school and into the postsecondary years. These studies also cover a wide range of perspectives and approaches.

Some studies evaluate programs that are designed to engage families in improving achievement, while others look at high-performing schools or students to study how parent involvement may have contributed. Several studies analyze long-term databases drawn from large-scale surveys of families, students, and educators, while others focus closely on how families and educators interact in a single setting.

In general, the studies fall into three broad categories:

1. Studies on the impact of family and community involvement on student achievement.
2. Studies on effective strategies to connect schools, families, and community.
3. Studies on parent and community organizing efforts to improve schools. These studies comprise a new, still developing arena of research that forecasts some interesting trends in both research and practice.

The next sections will cover the findings for each category, make some recommendations for putting these findings into action, and draw a brief conclusion. Before we describe this research in more detail, let’s look at how the studies define parent involvement and student achievement.

How do the studies define parent involvement? Joyce Epstein and her colleagues at the Center on Family, School, and Community Partnerships at Johns Hopkins University, have developed a useful framework of six types of parent involvement. Table 4, adapted from the Sophia Catsambis study (1998), is based on this work. It shows how parent involvement is frequently broken down and defined. Many researchers used some variation of this framework.

Definitions of parent involvement in elementary school were similar to those given in the table for grade 8. Some researchers condensed this list into parent involvement at home and at school, using definitions like these:

• Engaging in learning activities at home, including helping with reading skills and checking homework.
• Supervising children and monitoring how they spend their time out of school.
• Talking about school and what children are learning.
• Attending school events, going to parent-teacher conferences, meeting with teachers, and volunteering in the classroom or school.

In their analysis of middle grades achievement, for example, Esther Ho Sui-Chu and Douglas Willms (1996) identified four basic types of involvement. Two are based at home, two at school:

• Discussing school activities.
• Monitoring out-of-school activities.
• Contacts with school staff.
• Volunteering and attending parent-teacher conferences and other school events.

Rebecca Marcon (1999) put an interesting spin on this distinction in her study. She grouped involvement by whether parents were active and “in charge,” or passive and “reacting to the school.” Deborah Bugg Williams (1998) used an educational productivity model based on Herbert Walberg’s research (1984):

• Parent effort: contacts with school, expectations of student, and discussions with student.
• Instructional support: how much time student spends learning outside school.
• Environmental support: learning at home, quality of school (parent rating), knowing student’s friends, and out-of-school activities.

The studies that consider community-based initiatives to improve schools take into account the social and political context that leads to poor performance. These movements for better schools take a political approach, demanding more resources, higher teacher quality, smaller schools, and new programs to improve student achievement. Underway mainly in urban areas, these parent and community-driven efforts are focused on holding the school system accountable for low student performance.

How do the studies define student achievement? The studies were fairly uniform in how they defined and measured student academic achievement. Measures of student achievement and other outcomes most commonly used were:

• For young children: teacher ratings (using instruments like the Vineland Adaptive Behavior Scales) of school adjustment, vocabulary, reading and language skills, social and motor skills.
• For school-age children: report card grades, grade point averages, enrollment in advanced classes, and standardized test scores.
• Attendance, staying in school, and being promoted to the next grade.
• Improved behavior and healthy development (for example, less substance abuse and disruptive behavior).

Studies on the Impact of Parent and Community Involvement on Student Achievement

From these studies, one overarching conclusion has emerged:

Taken as a whole, these studies found a positive and convincing relationship between family involvement and benefits for students, including improved academic achievement. This relationship holds across families of all economic, racial/ethnic, and educational backgrounds and for students at all ages. Although there is less research on the effects of community involvement, it also suggests benefits for schools, families, and students, including improved achievement and behavior.

Among the studies reviewed here, the benefits for students include

• higher grade point averages and scores on standardized tests or rating scales,
• enrollment in more challenging academic programs,
• more classes passed and credits earned,
• better attendance,
• improved behavior at home and at school, and
• better social skills and adaptation to school.

Contributing to this conclusion are several key findings that clarify and deepen our understanding. Before presenting them, however, we would like to sound a caution. As we point out in the introduction, it takes more than engaged parents to produce high student achievement. Many studies of high-performing schools identify several key characteristics associated with improvement. These include high standards and expectations for all students and curriculum, as well as instruction and assessments aligned with those standards. They also include effective leadership, frequent monitoring of teaching and learning, focused professional development, and high levels of parent and community involvement.

As expected, while the effect sizes in many of these studies are statistically significant, they are small to moderate. A number of studies found that some forms of parent involvement with the school (communications with school, volunteering, attending school events, parent-parent connections) appeared to have little effect on student achievement, especially in high school. A few found that parent involvement with homework and parent-initiated contacts with school were negatively related to grades and test scores (Catsambis, 1998; Fan and Chen, 1999; Izzo et al., 1999; Shumow and Miller, 2001).

What does this mean? Does helping children with school work, monitoring their behavior, or contacting the school impel them to get lower grades and scores? Are there other factors that influence both achievement and parent intervention? Do parents offer more guidance to children who are struggling? In her study of 13,500 families, Sophia Catsambis found that certain forms of involvement (contacting the school, encouraging teens to graduate from high school rather than go to college, and supervising behavior) were associated with lower student achievement. When she controlled for problem behavior (coming to school late or unprepared, cutting classes), the negative effects disappeared.

Lee Shumow and Joe Miller (2001) found that at-home involvement is related to students’ positive attitudes toward school but negatively related to grades and test scores. They interpreted their findings to mean that parents of struggling students provide more help at home than parents of successful students. In their analysis of data from 25 studies, Xitao Fan and Michael Chen (1999) found a similar pattern. They also suggest that parents impose more controls when children are not doing well. These studies suggest that parents whose children have academic or behavior problems tend to supervise them more and seek help from the school.

The Gutman and Midgley (2000) study of fifth- and sixth-grade students from 62 families in a Michigan district found that parent involvement as a single variable did not appear to be related to students’ grades. It is interesting, however, that their definition of parent involvement (talking to students about school, checking homework, attending events and volunteering) contains factors that Catsambis and others found were not significantly related to achievement.

Having expressed these cautions, let’s take a closer look at the findings on the relationship between parent and family engagement and improvements in student outcomes.

Almost all these programs are aimed at families with younger children, from birth through kindergarten, then in elementary school. Two studies, developed by Joyce Epstein and her colleagues, examine Teachers Involving Parents in Schoolwork (TIPS) in middle school. TIPS was originally designed for elementary schools, but now includes designs for seventh and eighth grades. The studies on parent involvement for middle and high school students look at family processes and family-school interactions and are drawn mostly from survey research. These findings are presented under the next set of findings.

Birth through preschool. Early Head Start is a federal program that serves low-income families with infants and toddlers. Either through home visits or classes at a central location, mothers learn ways to stimulate their children’s mental, physical, and emotional development at home. The program includes early education, parenting education, health services, and family support services. In an experimental study, Mathematica Policy Research and the Center for Children and Families at Columbia University (2001) examined initial results from this new program.

The research team looked at programs in 17 sites, studying about 3,000 children and their families. At each site, children were randomly assigned to the program or to a control group.

When they were two years old, the Early Head Start children scored higher on cognitive development scales, used more words, and spoke in more complex sentences than control-group children. The program families were also more likely to support their children’s development and literacy skills than families of control-group children.

Project EASE (Early Access to Success in Education) is a literacy program in Minnesota that offers home and school activities for kindergartners and their families. In sessions at school, trained parent educators coach mothers in developing literacy skills. Then teachers send home book-related activities for parents to do with their children. Gail Jordan, Catherine Snow, and Michelle Porche (2000) looked at the project in four lower-income schools in a mostly middle-class district. Over one year, the students in Project EASE made significantly greater gains on language scores than children in a control group. The more activities a family completed, the higher their students’ gains. Children who started out with the lowest skills gained the most.

HIPPY, the Home Instruction Program for Preschool Youngsters, is a program with similar goals but delivered entirely through home visits. It is aimed at poor and immigrant families with four- and five-year-old children. During biweekly home visits, a trained paraprofessional models the lesson through role-play. Mothers read books to their children, then engage them in learning activities. The home visitors, recruited from backgrounds similar to their assigned families, are trained and supervised by professional HIPPY coordinators.

Amy Baker, Chaya Piotrkowski, and Jeanne Brooks-Gunn (1998) examined outcomes for 182 HIPPY program and control-group children in two cohorts in a New York school district. The study extended through the two-year program and followed up one year later, at the end of first grade. The results were mixed. For Cohort I, the researchers found positive gains in the HIPPY children’s school performance, both during the program and in first grade, compared with the control group. For Cohort II, there were no significant differences between the HIPPY and control children.

Baker and her colleagues concluded that these results are promising but tentative. Programs that develop young children’s learning skills are important because children who start out as high performers tend to remain that way, while children who have a poor start tend to remain poor students. From their analysis of the data, the researchers concluded that “we may be seeing naturally occurring variations on the effects of programs within communities. . . . Our findings also alert us to the importance of replication studies and caution us about generalizing positive or negative results from singlesample, single-site evaluations” (p. 584). They call for further research on HIPPY.

For 10 years, a team of researchers in Turkey studied a program based on HIPPY (Kagitcibasi et al., 2001). In an experimental study, children were randomly assigned to the HIPPY program or to three other settings. The four programs or settings studied were

• home care provided by mothers with training, home visits, and discussion groups (HIPPY);
• home care provided by mothers with no support;
• childcare without education; and
• educational nursery schools.

In the short term, children in both HIPPY and nursery school settings made greater progress than children in the other two groups. Seven years after completing the programs, however, the HIPPY children showed greater gains than children in the other groups, including those who had gone to nursery school. They earned higher scores in reading and math and in social development. They also were more likely to stay in school. As with the Baker study, these results should be treated with caution, for there may be unidentified cultural differences that influenced the outcomes.

Studies of a program to develop math skills in Head Start children found more consistent results. Prentice Starkey and Alice Klein (2000) did two experimental studies of a four-month intervention. At two sites in the San Francisco area, one serving African- American families and the other Latino families, staff gave classes for mothers and children and loaned math activity kits to use at home. About 30 families were involved at each site. Half were randomly assigned to the program, and half to a control group. In both sites, the researchers found that parents were willing and able to work with their children on math when given training and materials. The children in both programs developed greater math knowledge and skills than the control-group children.

Elementary and middle school. A study of standards-based reform practices done by Westat and Policy Studies Associates (2001) for the U.S. Department of Education looked at their impact on student achievement in 71 Title I elementary schools. (Title I is the largest federal program for elementary and secondary education. It is aimed at improving the academic skills of low-income students.) The study used an advanced statistical method to analyze the relationship between student test scores and these practices:

• Visibility of standards and assessments.
• Basic or advanced teaching techniques.
• Teacher preparation and teachers’ skills in math instruction.
• High or low ratings (by teachers) of professional development.
• Focus on assessment and accountability.
• District standards policies.
• Outreach to parents.

Outreach to parents measured how much teachers communicated with parents of low-achieving students through

• meeting face to face.
• sending materials on ways to help their child at home.
• telephoning both routinely and when their child was having problems.

The researchers found that teacher outreach to parents of low-performing students was related to improved student achievement in both reading and math. Of the eight other practices studied, only professional development that was highly rated by teachers was as consistently linked to student gains in both subjects. In schools where teachers reported high levels of outreach to parents, test scores grew at a rate 40 percent higher than in schools where teachers reported low levels of outreach.

Does offering workshops at school enhance parents’ skills to help their children? Ann Shaver and Richard Walls (1998) looked at the impact of school-based parent workshops on the achievement of 335 Title I students in nine schools in a West Virginia district. In addition to attending sessions designed to their interests, parents received learning packets in reading and math, as well as training in how to use them. The students’ gains were compared with pretest scores, then measured against average national gains, on the Comprehensive Test of Basic Skills. The researchers found that

• students with more highly involved parents were more likely to gain in both reading and math than children with less involved parents. This finding held across all income and education levels.
• younger students (grades 2–4) made greater gains than older students (grades 5–8).
• parents were more likely to be involved when their children were in elementary school (grades 2–4) than in middle or junior high school.
• students from lower-income families made fewer gains than students from higherincome families, no matter how involved their families. However, low-income students with more involved parents made greater gains than low-income students with less involved parents.
• a family’s income level did not affect its level of involvement. Low-income familieswere as likely to attend regularly as higher-income families.

Do school efforts to engage families make a difference in student achievement? A study by Joyce Epstein and her colleagues of a family-school partnership program adopted by 80 Baltimore elementary schools found positive results. These schools are members of a network that receives technical assistance to develop six types of partnership, from working with children at home to being engaged in school decisions (see Table 9, p. 91). In schools with more highly rated partnership programs, students made small but significant gains on writing and math tests, compared with schools with lower-rated programs. Attendance also improved at the more highly rated schools (Epstein, Clark, Salinas, and Sanders, 1997).

Do programs that engage families in children’s learning at home have effects on older children? Joyce Epstein and her colleagues at Johns Hopkins University have developed an interactive homework program called TIPS (Teachers Involving Parents in Schoolwork). TIPS can be adopted by both elementary and middle schools. In a study of TIPS for Writing in two Baltimore middle schools, Epstein and her colleagues found that parent involvement in TIPS boosted sixth- and eighth-grade students’ writing scores. Almost 700 sixth- and eighth-grade students and their families took part in the study. The more TIPS homework students completed, the better their grades in language arts (Epstein, Simon, and Salinas, 1997).

In a quasi-experimental study of TIPS for Science in a suburban middle school, Frances Van Voorhis (2001) found similar results. Three classes from two sixth-grade teachers and two classes from two eighth-grade teachers, a total of 253 students, took part in the study. The students were a cross-section of those in the school (about half white and half a mix of African American, Asian, Hispanic, and Russian). In sixth grade, they were in low, average, and honors classes; in eighth grade, they were in average and honors classes. The teachers assigned TIPS homework to six classes and noninteractive homework to four classes. The study covered two marking periods. After controlling for prior grades, family background, and amount of homework turned in, TIPS students earned significantly higher grades in science than the control group.

A study with a similar design on interactive math homework is discussed in the next section (Balli, Demo, and Wedman). Although the authors found no significant differences in posttest math achievement, they did discover that families of students assigned interactive homework were significantly more involved in math homework than families who did not. The researchers noted that the small sample size may have affected the results.

Summing up. Early childhood, preschool, and kindergarten programs that train parents to work with their children at home tend to have significant, positive effects (Baker et al., Kagitcibasi et al., Mathematica, Starkey and Klein). Children’s grades and ratings from teachers tend to improve the longer they are in the program, and they make greater gains than children not in the program (Jordan et al., Shaver and Walls). The studies that compared levels of involvement found that achievement increased directly with the extent to which parents were engaged in the program (Jordan et al., Epstein et al., Shaver and Walls, Westat/Policy Studies Associates). Children from all family backgrounds and income levels made gains. In some cases the children having the most difficulty in school made the greatest gains (Jordan et al., Westat/Policy Studies Associates).

Older children benefit as well. Such simple programs as weekly homework assignments in which students engage their parents are linked to improved grades for elementary and middle grade students (Epstein, Simon, and Salinas; Van Voorhis). One study shows that if schools fully adopt well-designed practices to engage families, their students’ test scores tend to rise and attendance improves (Epstein et al.). Standardsbased reform practices are more likely to have a positive effect on students’ test scores when teachers communicate regularly with parents (Westat/Policy Studies Associates).

One study, on the HIPPY program, reported inconsistent results (Baker et al.). The first cohort of students made significant gains compared with the control group, but the second did not. All the treatment and control groups attended a high-quality preschool program. Baker and her colleagues attributed the uneven result to natural variations in program impact. The HIPPY study done in Turkey found comparable results between the HIPPY group and the nursery school control group at the end of the program intervention (Kagitcibasi et al.). Several years later, however, the HIPPY graduates were doing significantly better in school than the control-group students. It may be that the influence of the program on the home environment helped to sustain longerterm effects.

Another study mentioned briefly in this section, the Balli, Demo, and Wedman study of interactive math homework, found no significant test score gains in the treatment group. The researchers speculate that the small sample size and short term (18 weeks) of the study may have influenced the results.

Taken together, we feel that these studies make a solid case that programs to engage families can have positive effects on student academic achievement and other outcomes.

A three-year study of 1,200 urban students in a New England district by Charles Izzo and his colleagues (1999) found that parent involvement, both at home and at school, was related positively to student achievement. They followed randomly selected students in 27 schools from kindergarten to third grade. Each year, teachers rated the parents’ involvement using these measures:

• the frequency of parent-teacher contacts each year
• constructive working relationships with parents (agree/disagree)
• parent participation in activities at school (yes/no)
• parents’ educational activities at home (yes/no)

Parents’ home activities were related to the widest range of gains on math and reading tests, compared with the other forms of parent involvement. The researchers also found that involvement at home remained steady, while involvement at school declined over time. This consistency, they concluded, may explain why the home activities had a stronger influence (Izzo et al., 1999).

Another study compared students whose parents are more highly involved with students whose parents are less involved. Rebecca Marcon (1999) looked at 700 African- American preschoolers in Washington, D.C. Using teacher reports of parent involvement, she compared students’ grades and skill ratings. Parents with high involvement ratings, compared with those with low or median ratings, tended to have children with higher grades and scores. This finding held across all family income levels and backgrounds.

In their retrospective study, Wendy Miedel and Arthur Reynolds (1999) analyzed interviews from 700 parents of eighth graders in Chicago. In addition to their background and expectations for their children, parents reported on their involvement when their children were in preschool and kindergarten. Seventy percent had been engaged in Chicago Parent Centers, which offered workshops and information about children’s learning, as well as activities to help parents be involved at school. To confirm parents’ reports, teachers rated parents’ participation in school activities. These teacher ratings closely matched what the parents said.

Miedel and Reynolds compared results for students based on how much their parents had been involved. Between first and eighth grades, students whose parents took part in a greater number of activities did consistently better in school. They tended to earn higher scores on reading tests, spend less time in special education, and pass from one grade to the next. These findings held across all family backgrounds.

How do families’ practices at home relate to middle and high school students’ achievement? Several studies use data from a long-term study of eighth graders, called the National Educational Longitudinal Study (NELS:88). NELS:88 provides an easily available source of information for researchers. Starting in 1988, the National Center for Educational Statistics (NCES) followed 25,000 eighth graders from 1,000 schools. It surveyed them at two-year intervals through 1994, and again in 2000. NCES also surveyed their parents, teachers, and school principals and collected data from high school transcripts. In 1988, the base year, and in follow-up years, students also took tests in math, reading, science, and social studies. The other studies of middle and high school students also use large national databases, such as the Survey of Parents and Children, done by the National Commission on Children, and the Longitudinal Study of American Youth.

These studies asked questions like these:

• What form does parent involvement take as children move from elementary school to middle school and from middle school to high school?
• How are parents involved at school versus at home?
• Do some forms of involvement have different effects than others? Under what conditions?

Exploring these layers beneath the surface has led to a richer, deeper definition of parent involvement. It has also allowed us to see that different types of involvement have different effects, at different ages, and in families of different backgrounds. Results from these studies will be covered under this and the next two key findings.

Support from home and school. What helps students through the transition to middle school? Looking at low-income African-American students from 62 families during the transition between fifth and sixth grades, Leslie Gutman and Carol Midgley (2000) asked what helped them through the change. They found that the combined effect of parent and school support had a significant impact on middle school grades. Students reported on three key influences:

1. Parent involvement: talking to students about school, checking homework, attending events, and volunteering at school.
2. Support from teachers: taking time to help students and being supportive rather than critical.
3. Belonging at school: feeling accepted, respected, and included at school.

Relating these factors to grade point averages, the researchers found that no single one appeared to have an effect. When the researchers combined parent involvement with the other two factors, however, another picture emerged. Students reporting high parent involvement and a high sense of belonging, or high parent involvement and high teacher support, had higher average grades than students who reported low support at home and school.

This study suggests that if children don’t feel connected to school, parent involvement alone will not make a significant contribution to student achievement. Students must also feel that they belong at school and that their teachers support them. Sanders and Herting (2000) looked at similar influences for 800 urban African-American middle-grade students. The question they addressed was why African-American girls tend to do better in school than African-American boys. On a 1–5 scale, students rated these factors:

• Support from teachers (like feeling comfortable asking the teacher for help).
• Support from parents (like giving praise for doing well in school).
• Involvement in church (like belonging to a church group).
• Attitudes toward school (like feeling it’s important to work hard in school).
• Academic self-confidence (like feeling they do good work in school).
• Behavior in school (like being sent out of class for misbehaving).

Support from teachers and parents and involvement in church were positively related to attitudes, self-confidence, and behavior. These, in turn, influence achievement. The girls in this study felt more support from their family and teachers, and are more active in church, than the boys. The researchers were not surprised that African-American girls also reported more positive attitudes about their ability, less disruptive school behavior, and higher achievement than the boys did.

As mentioned earlier, some studies found that parent supervision, involvement with homework, and contacts with school seemed to be negatively related to grades and test scores (Catsambis, Fan and Chen, Izzo et al., Shumow and Miller). The consensus among the researchers was that parents are more involved in these ways when their students are struggling. In their analysis of data from 25 studies, Xitao Fan and Michael Chen (1999) suggest that parents impose more controls when children are not doing well. It is not clear if this strategy helps or hinders their children, because survey data are not designed to identify cause and effect. Catsambis and colleagues conclude, “the most effective types of parent involvement are not those geared towards behavioral supervision, but rather, those geared towards advising or guiding teens’ academic decisions” (p. 24).

Efficacy. Parents’ sense of their efficacy influences how they are involved during middle and high school. Efficacy means the power to produce an effect. In a study using a national sample of 900 families with children aged 10–17, Shumow and Lomax (2001) examined parents’ feelings of success in guiding their children. Parents have a high sense of efficacy when they believe that they can

• help their children do well in school, be happy, and be safe.
• overcome negative influences and keep their children away from troublemakers, illegal drugs, or alcohol.
• have a positive impact such as improving quality of the school and making the neighborhood a better place.

The higher parents’ sense of efficacy, the more closely they monitored their children and the more they were involved with school. The researchers then related efficacy to student outcomes. They used parent reports of their children’s grades, academic level (high, medium, or low), and behavior at school, and students’ reports of well-being (optimism, worries). They found that the higher parents’ feelings of efficacy, the more their children reported doing better in school and feeling happy, safe, and stable. Families who live in safe, higher-income areas with good programs for young people had more sense of efficacy than families living in lower-quality areas.

A protective effect. When students report feeling support from both home and school, they tend to do better in school. They say that they have more self-confidence and feel school is more important. Data indicate that they also are less disruptive, earn higher grades, and are more likely to go to college (Gutman and Midgley, Sanders and Herting, Shumow and Lomax, Trusty). Although several studies compare home versus school settings for parent involvement to see which have stronger effects, it is clear that both are important:

• At school, parents learn how to engage their children in learning at home, get help if their children are struggling, and form a constructive relationship with teachers (Izzo et al., Miedel and Reynolds).
• At home, parents guide their children toward postsecondary education, make sure they read and do their homework, and stress the value of education. They also steer children away from risky behavior, help them maintain positive attitudes, and support them through problems at school (Catsambis, Fan, and Chen; Shumow and Lomax; Trusty).

For many children, home and school are two very different worlds. Comparing forms of parent involvement in the primary grades (K–3), Izzo and his colleagues found that taking part in activities at school was positively related to students’ school engagement. The quality of parent-teacher interactions (as rated by teachers) was positively related to students’ social and emotional adjustment. When their parents are involved at school, in other words, children are more likely to adjust to school, take part in class, and feel they belong and fit in.

In his analysis of NELS:88 data, Jerry Trusty found a similar protective influence. Parent involvement in eighth grade, as reported by students, influenced students’ expectations to finish college six years later. Students who felt that their parents communicated with them and supported their learning were more likely to continue studies past high school. In other words, the more students felt their parents’ involvement and support, the longer they planned to stay in school. For students, families are a continuing presence, while schools are shorter-term resources.

In their study of African-American eighth graders, Sanders and Herting focused on why girls do better in school than boys. African-American girls felt they received greater support from parents and teachers and were more likely to belong to church groups. They also reported higher grades, better behavior, and more self-confidence in school. For both boys and girls, family, church, and teachers were positive influences that were connected to higher achievement.

Further, higher levels of parent involvement appear to have more impact than lower levels. This does not just mean that more is better than less. It also means that active types of involvement may have a stronger effect than more passive types. In her study, Marcon classified the four types of involvement in her study as active (volunteering and visiting the classroom) or passive (getting information from the teacher at conferences or home visits). She found that active involvement in preschool was related to higher marks both on teacher ratings and report cards.

In a study of how families manage their children’s time, Reginald Clark (1993) surveyed families of 1,171 third graders of all backgrounds in Los Angeles. After dividing the students into high and low achievers based on standardized test scores, he correlated the ways they spend their out-of-school time with grades, family background, and other factors. He found that the way children spent their time at home, not the family’s income or education level, predicted their success in school. Most parents reported that they talk to their children about homework, read to their children, and make sure they do their assignments. However, families with high achievers reported more time engaged in home learning activities than families with low achievers. For example, high-achieving children spend more time on homework, reading, and using materials like the dictionary.

Clark identified four variables that comprise what he calls “parents’ press for academic success.” Together, these factors explained 47 percent of the variation between lowand high-achieving students in the study:

• Parent knowledge about homework assignments.
• Parent perception of child’s engagement in homework.
• Child knowledge of how to use a dictionary.
• Parent expectations for child’s education.

Low achievers tended to come from homes where the parents were younger, did not work outside the home, had not been to college, and were low-income. Even though higher-achieving students often had parents who were not home to monitor their late afternoon activities, having parents in the workforce was related to higher test scores. Despite the relationship between achievement and family resources, Clark found that high achievers came from a wide variety of family backgrounds. “Let us recall that 51.3 percent of the mothers of high achievers possessed no more than a high school education. Almost 40 percent . . . lived in single parent households. Almost 43 percent of the high achievers were Hispanic and 21.8 percent were Black” (p. 103).

In their study of NELS:88 data, Esther Ho Sui-Chu and Douglas Willms (1996) found that involvement at home had the greatest effect on student achievement. Compared with volunteering and attending school activities, parents’ talking about school with their children and helping them plan their education programs were more highly related to higher grades and test scores.

Although children from higher-income families tend to do better in school, students of all backgrounds gain when their parents are involved. Ho Sui Chu and Willms found that higher-income families were slightly more involved in some ways, but the effect was small. On the whole, the researchers concluded, higher-income and two-parent families were not more involved with their children’s education than lower-income and single-parent families. The types of involvement vary somewhat, however, by race and ethnicity.

An interesting twist is that children from all backgrounds tend to score higher in both math and reading if they attend a school where the average family income is higher. Ho Sui-Chu and Willms surmise that this is because schools in higher income areas appear to have a culture of greater parent involvement.

Using NELS:88 data, Sophia Catsambis (1998) studied 13,500 families whose children stayed in school through 12th grade. She measured the connection of six types of involvement (see Table 4, p. 22) with high school student achievement. Enhancing learning at home, she found, had the strongest effect.

• Forms of involvement with less effect: Parenting practices, communications with school, attending school events, and contacts with other parents.
• Forms of involvement with more effect: Expressing high expectations, discussing going to college, and helping students prepare for college.

When their families guided them to classes that would lead to higher education, students were more likely to enroll in a higher-level program, earn credits, and score higher on tests. The connection was somewhat greater for math and science than for English, and for earning credits than scoring well on tests. Looking back from parent involvement in grades 8 through 12, Catsambis found that parents’ expectations for their students to do well and attend college had the strongest effect on grade 12 test scores in all subjects. These findings held across all family backgrounds.

In their meta-analysis of 25 studies, Xitao Fan and Michael Chen (1999) also found that parents’ aspirations for their children were associated with higher grades, test scores, and passing rates. By aspirations, they mean expectations for their children to do well in school, graduate, and go on to higher education. In describing the connection, they explain,

The overall relationship between parent involvement and students’ academic achievement is close to .30. Although an average correlation of .30 may appear low to many people . . . this represents a medium effect size in social sciences . . . certainly a meaningful effect. (p. 18) In practical terms, this means that students from families with above-median parent involvement showed success rates that were 30 percent higher than those from families with below-median parent involvement. “This is not trivial by any standard,” Fan and Chen conclude (p. 18).

Gender influences. Is gender a factor? In their study of 60 middle school students’ families, Lee Shumow and Joe Miller found that involvement at home contributed to positive attitudes toward school, while involvement at school contributed to higher grades. Relating gender to different types of involvement, the researchers found: