The effects of interactive (TIPS) homework on family involvement and science achievement of middle grade students


Material Information

The effects of interactive (TIPS) homework on family involvement and science achievement of middle grade students
Physical Description:
x, 135 leaves : ill. ; 29 cm.
Van Voorhis, Frances Landis
Publication Date:


Subjects / Keywords:
Psychology thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 2000.
Includes bibliographical references (leaves 127-134).
Statement of Responsibility:
by Frances Landis Van Voorhis.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 025875055
oclc - 47103540
System ID:

This item is only available as the following downloads:

Full Text






For Kerry, Carolyn, & Ken


A dissertation is never the result of one person's efforts. I have benefited from the

support of many intelligent and patient people throughout this two-year process. I thank

each of them for helping me to complete a study of which I am very proud.

First I thank my family. My husband Kerry endured many late evening

conversations about my homework study, my graduate work, and my career. I thank him

for his patience, constructive criticism, encouragement, and reading of my work. His

gentle manner in communicating is one of the traits I most love and respect.

One can do amazing things with the support of parents and siblings. My mother and

father always communicate to me their faith in my ability to succeed. When I was

growing up, they never laughed when I proposed a big plan for my life (I have probably

proposed about 500 or so). Their belief in me inspires me to give that same

encouragement to others! I thank both of them for their involvement in my life and my

schooling. Whether they know it or not, I am stronger and more confident because of the

time they took to show me I was important. I hope to give my own children the same

feeling of love and support that my parents gave me.

My sisters, Johanna and Gwyneth Landis, are always quick to say a kind word to me

and to let me know how proud they are of my efforts. They each have so many talents

and gifts that I know they are putting to great use! I love them both and thank them for

reminding me that there is more to my life than my studies.

I am so proud and thankful to have had Dr. Scott Miller as my advisor and friend

through the dissertation phase of my graduate work. Whether he knows it or not, his

genuine eagerness to work with me two years ago prevented me from quitting graduate

school. His role in this process has kept me moving forward, and I appreciate the weekly

encouragement he gave me during our phone conversations, the quickness with which he

read drafts and gave me feedback (usually within a day or two), and the flexibility he

displayed in working with me from afar on a topic that was not his specialty. I respect

his ability to balance his many roles as teacher, researcher, mentor, colleague, father, and

friend. His care for me as a person and not just as a student has made all the difference in

my graduate studies.

My complete admiration and thanks also go to Dr. Joyce Epstein, the guru of school,

family, and community partnerships. It is such a pleasure to work for someone who

applies her research to practice and who treats teachers and administrators with the

respect they deserve. Words cannot express how much I appreciate the time she spent

teaching me about research, practice, patience, and hard work. She constantly taught me

things when she probably did not even realize it! I thank her for the many opportunities

she gave me to further my research and career, and I look forward to many future fruitful

endeavors together.

This study would not have been possible without the support of the science teachers

and principal at Pikesville Middle School. I thank Barbara Walker, the principal, for her

willingness to support this study. I thank her for listening to Kerry explain my research

when we first moved to Baltimore, and for extending her hand to me to help me earn this


I can never, at least in this lifetime alone, thank Chris McChesney, Natalie

Dewberry-Moore, Leah Blind, Chris McGuiness, and Julie Homung enough for their

dedication to learning how to involve families in homework. Each of them taught me

about the sacrifices they often make to do the most important job in the world, teaching.

I respect their commitment to this research and to their students.

Nobody knows the trouble I've seen--unless it another graduate student. Many

fellow graduate students or recent graduates at the University of Florida and Johns

Hopkins University have shared their struggles and celebrated their successes with me.

Beth Simon and I worked through this process together, shared knowledge, and

encouraged each other along the way. She provides constant support and inspiration, and

I look forward to continued learning from and with her as a colleague.

I thank Dr. Elizabeth Kemper for meeting with me each week for lunch to discuss

my progress. This weekly meeting inspired me to accomplishment so that I would not be

embarrassed in front of someone I deeply respect as a researcher and phenomenal

woman. I also thank Laurel Clark, Dr. Steven Sheldon, Joann Benigno, Cynthia Koenig

and Nicole Alea for all of the suggestions, encouragement, and fun times shared.

Finally, I thank Dr. Jeff Farrar, Dr. Pat Ashton, Dr. Pat Miller, and Dr. Carolyn

Tucker for their feedback on my qualifying exam, on the proposal phase of my work, and

on my final dissertation. I respect their abilities to balance the demands of research and

teaching at such a premier institution as the University of Florida.



ACKNOWLEDGMENTS............................................ ...... ......................... iii

ABSTRACT ........................................................................................................... ix


1 IN T R O D U C T IO N .................................................................................. ..................... 1

T heoretical B background ...................................................................... ...................... 3
Focus of the Literature Review ...................................................... ..................... 6
Parental Involvem ent...................................................................... .........................
Developmental differences in parental involvement levels ..............................8...
Factors affecting the involvement decision.................................................... 9
Factors affecting parents' choice of involvement.............................................. 11
H om ew ork ...................................................................................... .................... 12
Developmental differences.............................................................................. 14
Parental involvement in homework.............................................................. 16
N negative findings ................................................................... ................... 17
Positive findings........................................... 19
Teachers Involve Parents in Schoolwork (TIPS) interactive homework ...........21
T IP S w writing .. .................... ........................................ .................... 22
T IP S m ath................................................... ............................................. 23
T he Present Study .............................................................................. ...................... 25

2 M E T H O D S................................................... ............................................................. 29

Sam ple .............................................................................................. ............. ........... 29
D design ............................................................................................................. ............ 29
M materials ................................................................................... ............................... 3 1
Interactive (TIPS) Assignments ............. .................................................. 32
Non-Interactive (ATIPS) Assignments...................................................................... 33
TIPS/ATIPS Test Questions .......................... .................................................... 33
Student/Family Surveys .................................................................................. 33
P rocedure........................................................................................................... ............ 34
Letter to Fam ilies ............................................................... ................................. 34
Science Assignments.......................... ..................................... 34
Science Assessments ..................................................................................... 35


S u rveys ...................................................................................................................... 36
M easures............................. ....................................................... ...................... 36
Background Measures ..................................................................................... 36
Homework Completion Measures....................................................................... 36
Survey M measures .............................................................................. ................... 37
Fam ily involved ent....................................................................... ..................... 38
Family background....................................................................................... 38
T im e on hom ew ork ....................................................................... ..................... 38
Opinions of the TIPS/ATIPS homework assignments......................................... 39
G general interest in science.......................................................... ..................... 39
Science Achievement Measures........................ .................................................41
Research Questions, Hypotheses, and Analyses...................................................... 41
Family Involvement ..................... ... ........................41
Homework Assignments ................................................................................. 42
Science A chievem ent.................................................................. ......................... 43
Concordance in Parent and Child Surveys................................ 43

3 R E SU LT S .................................................................................................................... 45

D escriptive Statistics ........................................................................... ....................... 45
Family Involvement Research Questions................... ....................................... 45
Homework Completion and Accuracy.................................................................... 52
Science A chievem ent ........................................................................... ...................... 60
Concordance and Discrepancy in Parent and Student Reports..................................... 64

4 D ISC U SS IO N ........................................................ .................................................. 77

M major Findings ...................................... ....................... ..................... 77
TIPS Increased Family Involvement in Science Homework................................. 77
Family Involvement Relates to Homework Completion........................................ 79
TIPS Students Earned Higher Report Card Grades than ATIPS Students. ..............80
Results Often Differed Developmentally and by Gender. ......................................82
Parent and Student Categories Provide New Insights into Achievement Results.....83
L im itatio n s .................................................................................................................... 8 3
L ength of the Study ......................................................................... ...................... 84
Need for Standardized Achievement Scores............................... 84
Lack of Teacher Implementation Measures. ........................... .........................85
Lack of Measures for the Quality of Homework Interaction ................................. 86
Lack of Homework Adaptation by Class Ability Grouping.................................. 87
T he Future of H om ew ork ....................................................................... .................... 88


A HOMEWORK ASSIGNMENT ORDER BY TEACHER..................................... 91

B SAMPLE TIPS ACTIVITY .................. ... ......................... 92

C SAMPLE ATIPS ACTIVITY .............. ..................................................... 94

D TIPS/ATIPS TEST QUESTIONS.................... ................................................. 96

E STUDENT AND PARENT SURVEYS............................................................ 107

F LETTERS TO PARENTS ................ ...................................................... 1 16

G DESCRIPTIVE DATA FOR ALL STUDY VARIABLES................................. 118

R E FE R E N C E S ........................................................ ............................. ..................... 127

BIOGRAPHICAL SKETCH ..................................................................................... 135


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Frances Landis Van Voorhis

December, 2000

Chairman: Scott A. Miller
Major Department: Psychology

The purpose of the study was to investigate the effects of interactive and non-

interactive science homework assignments on family involvement in homework,

homework completion and accuracy, student science achievement, and student and parent

attitudes about science. Most previous research on homework has examined what

parental involvement results naturally, without prompts or instruction from teachers. In

contrast, this study experimentally examined the effects of teacher prompts to parents for

involvement in their children's homework. Two hundred and fifty-three students from

10 classes of sixth and eighth grade students participated in the study that lasted 18 weeks

of the school year. Six classes of students completed the TIPS (Teachers Involve Parents

In Schoolwork) interactive homework assignments, and four classes completed the non-

interactive assignments that contained the same content and questions as the TIPS

assignments. The TIPS students received instructions to involve a parent or other family

partner in certain sections of the homework assignment. The ATIPS students received

the same assignment with no instruction to involve another person. Results indicated that

TIPS students more often involved parents in their science homework assignments than

did ATIPS students. However, TIPS science students reported no more parental

involvement in homework than ATIPS students in subjects where teachers did not assign

interactive homework. Therefore, the TIPS instructions elicited more parental

involvement in homework than the ATIPS assignments. The TIPS students did not differ

from ATIPS students in accuracy or the percent of homework returned. Students who

rated the homework more positively and who regularly involved their families returned

more homework assignments than students who did not do so. The TIPS students did

earn significantly higher science report card grades than ATIPS students after controlling

for background variables, teacher effects, and percent of homework returned.

Exploratory analyses of matched student and parent data suggest the importance of both

student and parent data in research of interactive homework. The results of the study

indicate the benefits of well-designed interactive homework for levels of family

involvement, science attitudes, and science achievement.


Education reform in the elementary and secondary grades constantly weighs on the

minds of families, educators, business people, community members, and politicians in the

United States. Researchers, educators, and politicians worldwide have proposed

numerous strategies to improve current educational practices: reduction of class size,

better preservice teacher training, end of social promotion, standards-based reform,

voucher programs, and increased and informed family involvement in children's


For solutions to current educational shortcomings, one must consider the goal of the

efforts or changes. Though the specifics of the goals differ, most people would probably

agree that yesterday's, today's, and the future's main concern is to better educate our

children to survive in this increasingly "informational" age. Children remain at the

center of the issue and are influenced not only by their school environments, but also by

their family and community environments (Epstein, 1992).

Therefore, any solution affecting change only in the school environment (like

standards-based reform or reduction of class size) may result in improvements but will

probably fail to completely address the current multidimensional educational issues

(Coleman et al., 1966). Problems in education are complicated. Schools vary in teaching

philosophies and standards; represent children of various developmental levels,

backgrounds, and cultures living in varied communities; and identify similar and different

challenges. Part of the solution to today's educational concerns must recognize these

intricacies and include inter-environmental strategies like family involvement or school-

family-community partnerships.

Many research findings indicate the positive impact that parental involvement may

have on student development and academic achievement (Booth & Dunn, 1996; Chen &

Stevenson, 1989; Dornbusch, Ritter, Roberts, & Fraleigh, 1987; Eccles & Harold, 1996;

Epstein, 1988; Epstein, Simon, & Salinas, 1997; Fehrmann, Keith, & Reimers, 1987; Ho

Sui-Chu & Willms, 1996; Keith et al., 1993). Most parents recognize the importance of

involvement in their children's education. Students desire positive interactions with their

parents, guardians, or family members regarding their educational endeavors, and

teachers want students to communicate with their families about school subject matter.

Despite interest from students, teachers, and parents themselves in parental involvement

in a student's academic progress, parents still continually seek direction by asking, "How

may I help my child with her education at home?" (Epstein, 1998).

Though all involved parties seem to desire connections between home and school, it

takes a concerted and coordinated effort on the part of teachers, families, school

administrators, researchers, and policymakers to make these partnerships a permanent

foundation of our educational institutions. Since the publication of A Nation At Risk

(National Commission on Excellence, 1984), researchers and policymakers have

presented parental involvement as one possible and promising strategy to improve

education. Such policy efforts have continued with the passing of Goals 2000: Educate

America Act that includes a call for "partnerships that will increase parental involvement

and participation in promoting the social, emotional, and academic growth of children."

(National Education Goals Report, 1994, p. 11). The media also have directed attention

to parental involvement through articles like "How to make your kid a better student:

Yes, parents matter. Here's how." (Willis, 1998).

The present study focused on a specific form of parental involvement in student

education, namely parental or family involvement in student homework. Though

homework is an accepted and expected part of school life, it has long been a controversial

part of the education process, and debate has been renewed of late as evidenced by a

flurry of recent media attention. Writers frame the concerns of many educators and

families in titles including "The homework ate my family. Kids are dazed. Parents are

stressed." and "Homework bound: Schools are assigning more and more to younger and

younger children. First grader? No excuse! A hundred years later, the homework wars

rage on." (Ratnesar, 1999; Winnerip, 1999). Can teachers, students, and parents do more

to add value to the homework process and quell some of the quarreling? Research in the

areas of homework and parental involvement can help to inform this debate.

Theoretical Background

Both homework and parental (family) involvement in education link two

environments that influence the development of children: the family and the school.

Epstein proposed a theoretical model to describe family-school connections (1987) and

described four possible relationships between families and schools that reflect paradigm

shifts in the way that people have perceived the needs of families, schools, and children:

separate, sequenced, embedded, and shared (1992).

Separate responsibilities refer to unshared goals of schools and families and their

separate contributions to society (Waller, 1932; Weber; 1947). This perspective stresses

the independent functions of families and schools and therefore represents rare

communications between the two unless students experience problems in school or at


Sequenced responsibilities build on the theme of separate responsibilities but

recognize a progression of influence. This perspective emphasizes a sequence of critical

stages beginning with parents first contributing to their children's education in

preparation for school. Once children are school age, educators take over primary

responsibility for educating children. Both separate and sequenced responsibilities of

families and schools have been questioned as true representations of today's school-

family relationships as many mothers enter the workforce, use child care early in the lives

of their children, and have more education to know how to promote the positive

development of their children.

Embedded relationships refer to nested connections between individuals and larger

groups and organizations. This represents an ecological model in which individuals are

embedded within systems or layers (concentric circles), with those closest to the child

having the most direct impact and those farther away having a less direct effect on

development (Bronfenbrenner, 1986). Embedded relationships emphasize the

importance of social factors and contexts and how these affect children's cognitive

development (Markus & Kitayama, 1991; Miller, 1993; Wetsch & Tulviste, 1992).

Unlike the separate and sequenced perspectives, the embedded perspective does

recognize the effects of multiple environments like the family, school, neighborhood,

media, and community on the child. However, the embedded perspective fails to address

questions about developmental change or longitudinal effects of multiple environments

(Epstein, 1992).

Overlapping responsibilities refer to dynamic relationships between schools and

families that can be pushed together or pulled apart by practices and interpersonal forces.

Pictorially, the school and family each may be represented by spheres that overlap in

specific areas, not embedded within each other. This model of overlapping spheres of

influence does not presuppose that all responsibilities of families and schools overlap, but

it does assume that there are some shared goals for students between the two institutions

during most of a child's development. In addition to linking family and school, the

model is developmental in that it affords changes in degree of overlap with student age

and grade level (Epstein, 1987, 1995).

Two levels of interactions may occur within this model: institutional (external) and

interpersonal (internal). The external structure of the model includes the two spheres

representing the school and family that are affected by three forces: time or development

of the child, experience or philosophy of the family, and experience or philosophy of the


The internal structure of the model emphasizes the interpersonal relationships,

within or between organizations. There are also two levels of interaction: 1) standard,

organizational communications such as information to all parents about school policies,

and 2) specific, individual communications that occur between parents and teachers of an

individual child (Epstein, 1987). Therefore, the model accounts for the histories of the

major institutions and the dynamic nature of the changing skills of individual (Epstein,


The child is at the center of the areas of overlap between schools and families (and

communities, Epstein, 1995). "Children interact with, influence, and are influenced by

their schools, especially their teachers; and by their families, especially their parents,"

and by changes in school, teacher, and family practices (Epstein, 1987, p. 130). The

theory permits various research questions about the institutional and individual

connections that ameliorate student learning.

Because the present study was grounded in psychology and the individual level of

analysis, its focus featured the internal structure of the overlapping spheres model.

Therefore, information about the historical context of family-school relationships or

about the impact of political, economic, and social events on these relationships was not

analyzed in the study. The purpose of this study was to investigate how interactive and

non-interactive science homework affect family involvement in homework and science

attitudes and achievement of middle grade students as well as the attitudes and behaviors

of their family members or other homework partners.

Focus of the Literature Review

The literature review that follows first describes the different relationships between

parental involvement and student academic achievement. Next is a presentation of

findings related to homework and student academic achievement and attitudes. The final

section demonstrates how parental or family involvement and homework findings may be

integrated to create interactive homework that is designed to increase communication

between the home and the school (Epstein, in press; Epstein, Salinas, & Jackson, 1995).

Though these topics cross many disciplines, this review focuses on research from the

areas of psychology, sociology, and education.

Parental Involvement

Parental involvement is a broad term or construct like "aggression" or "altruism"

that means different things to different people. Therefore, researchers have constructed

varying definitions of parental involvement depending on the measures and outcomes

they use (Epstein, 1992; Georgiou, 1997; Greenwood & Hickman, 1991; Grolnick,

Benjet, Kurowski, & Apostoleris, 1997; Grolnick & Slowiaczek, 1994; Hoover-Dempsey

& Sandler, 1997; Ho Sui-Chu, & Willms, 1996; Keith et al., 1993; Mannan & Blackwell,

1992; Meighan, 1989, Stevenson & Baker, 1987).

Early studies of parental involvement often utilized one measure to explain the

construct such as attendance at school events (e.g., Stevenson & Baker, 1987). More

recent work addressing parental involvement involves multidimensional conceptions

including both home- and school-based activities related to children's learning (Epstein,

1995; Grolnick and colleagues, 1997; Keith and colleagues, 1993). For example, Keith

and colleagues (1993) include parental educational aspirations for the child, parent-child

communication, amount of home structure, and parent participation in school activities as

measures of parental involvement. Hoover-Dempsey and colleagues (1992) collected the

following data to gauge parental involvement: parental help with homework, parental

participation in other educational activities, parent volunteer work at school, telephone

calls with teachers, and parent-teacher conferences.

Most researchers separate parental involvement from parental beliefs. Involvement

is defined as home- and school-based observable activities related to children's learning

at school (Hoover-Dempsey & Sandier, 1997). Parental beliefs, which certainly affect

decisions to become involved, refer to the cognitive, non-observable processes behind the

actions of parental involvement (Miller, 1988).

Researchers often study parental involvement within a particular domain and in so

doing use the term parental involvement in conjunction with a modifier, a specific task or

area. For example, the proposed study focused on parental or family involvement in

middle grade students' interactive science homework assignments. This represents one

of six types of involvement among schools, families, and the community outlined by

Epstein (1992, 1995). The six types include the following: parenting (basic

responsibilities of parents), communicating (two-way communication between school

and home and home to school), volunteering, learning at home, decision-making (parent

representation on school decision-making bodies), and collaborating with the community.

Parental involvement in homework of course relates to Type 4: Learning at Home.

Developmental differences in parental involvement levels

Despite differences in definitions of parental involvement, many researchers have

found that parental involvement in schooling (however it is defined) tends to decline as

children grow up (Cooper, 1989, 1999; Eccles & Harold, 1996; Epstein, 1988, 1998;

Stevenson & Baker, 1987). Stevenson and Baker (1987) found a negative relationship

between age of child and level of parental involvement in schooling. Parents tended to be

more involved when their children were young, but involvement declined as time passed.

Those mothers with higher educational status tended to be more involved in their

children's school activities. Most important, children of involved parents performed

better in school than children of parents who were less involved in their schooling

activities. While the reasons for decline in parental involvement have not been proven,

many hypotheses have surfaced.

Factors affecting the involvement decision

Hoover-Dempsey and Sandler (1997) proposed a model that frames many of the

explanations of decreased involvement levels with age, emphasizing three variables that

particularly affect parental involvement decisions: parental role construction, parental

efficacy, and level of invitations and demands from children and the school.

Parental role construction refers to parental beliefs or expectations of their role in

their children's educational progress. These expectations are absorbed by parents from

their own culture, including family, church, educational background, and other forces

providing guidance in the rearing of children. Therefore, many researchers believe that

the way parents construct their roles directly relates to their actions or involvement

behaviors. For example, Sheldon (1999) examined the relationship between parental

(mothers only) role construction and parental involvement behaviors using Likert scale

surveys with questions like the following: "It is a parent's responsibility to help her child

understand his or her homework." Results revealed that parental role construction

significantly correlated with and predicted mother's involvement behaviors both in and

out of school.

Parental efficacy, another type of parental belief, signifies the degree to which a

parent thinks he may effectively help his child with the skills he has available (Hoover-

Dempsey & Sandier, 1995, 1997). Studies have indicated a positive relationship between

parental education (SES) and parental involvement in student education, namely the fact

that more educated parents tend to be more involved in their children's education (Ames,

De Stefano, Watkins, & Sheldon, 1995; Astone & McLanahan, 1991; Lareau, 1989;

Sheldon, 1999; White, 1982). Lack of education also may relate to a parent's sense of

efficacy such that less educated parents often feel less efficacious than more educated

parents in their efforts to help their children with academic tasks, especially as children

grow up (Hoover-Dempsey et al., 1992). Sheldon (1999) also assessed the relationship

between parental efficacy and parental involvement at school and at home. Like parental

role construction, self-efficacy significantly correlated with parental involvement at

home. However, the relationship was not as strong as the relationship between parental

role construction and parental involvement. Self-efficacy correlated (though not

significantly) with parental involvement at school. Therefore, both parental self-efficacy

and parental role construction are types of beliefs that represent important variables to

consider in studies of antecedents to parental involvement at home.

General invitations and demands relate to requests from the school and the child for

the parent to take part in educational activities with the child. As highlighted by Epstein,

most studies fail to address the school's role in encouraging parental involvement in

student education (Dauber & Epstein, 1993; Epstein, 1991; Epstein & Dauber, 1991).

Epstein and Dauber (1991) investigated ways in which 171 teachers in inner-city

elementary and middle schools attempted to involve parents. Their findings pointed to

another reason besides parental role construction and parental efficacy as to why parental

involvement decreases between the elementary and middle grades. They found that

teachers in the middle grades provide fewer opportunities for parental involvement than

do their elementary grade counterparts. Extending this study, the researchers then

investigated the reports of parents on their perceptions of invitations/ requests for parental

involvement. Regardless of parent education levels, they found that parents are more

likely to be involved if they perceive the school to be actively working on involving them

in their children's education.

A study by Ames, Khoju, and Watkins (1993) supports these results and included an

investigation of the way that teacher communication practices affect parental

involvement. Their findings indicated that when teacher communications to the home

include information that supports parental efficacy, parent comfort with the school

increases. This in turn influences parental perceptions of the child as learner. When

these effects result, parent involvement may increase.

In summary, early studies of parental involvement pointed to fixed characteristics of

parents (such as parent education and socioeconomic status) as most directly related to

parental involvement levels and student achievement outcomes. However, investigators

now highlight modifiable variables (variables that may change as a result of information,

guidance, and/or invitations) like parental role construction, parental efficacy, and

especially general invitations from the school as more helpful ways to increase parental

involvement between the elementary and middle grades.

Factors affecting parents' choice of involvement

The review above included a summary of three main factors influencing parents'

decisions to be involved in their children's education. Hoover-Dempsey and Sandier

(1995) delineate a second level of three additional factors influencing the specific forms

of involvement parents choose: specific domains of skill and knowledge, the mix of

employment and other family demands, and specific invitations and demands from the

child and school.

Parents choose involvement forms that use their specific skills. For example,

parents with strengths in math are likely to help with math studies. Those who feel

comfortable leading decision-making may volunteer to head the PTO but may choose not

to help out in the school office with copying or typing. The form of involvement also

relates to the demands on a parent's time--such as employment, childcare, or other family

responsibilities (Lareau, 1989). Finally, invitations from the child and the school will

affect how the parent chooses among involvement alternatives (Hoover-Dempsey &

Sander, 1995).

In sum, these three factors (specific domains of skill and knowledge, the mix of

employment and other family demands, and specific invitations and demands from the

child and school) further elucidate the specific forms of involvement parents choose.

They do not, however, negate parents' initial decisions about whether to be involved or

not. If parents place a high premium on involvement based on their views of parental

role construction, self-efficacy, and general invitations, they will find a way to be

involved regardless of their specific skills, demands on time, and general invitations

(Hoover-Dempsey & Sandier, 1995).


Homework has been defined as "tasks assigned to students by school teachers that

are meant to be carried out during non-school hours" (Cooper, 1989, p. 7). Because it is

most effective in promoting student achievement when it is completed during non-school

hours (Keith, 1998), homework (as suggested by its name) is often worked on in the

home and may serve as an important link and form of communication between the school

and the home (Corno, 2000; Epstein, 1995).

Both the school and settings outside the school may contribute positive and negative

characteristics to the student homework experience (Corno, 2000). For example, teachers

who design homework that balances challenge and skill may encourage classes of

students to value the importance of the homework experience, promote student

achievement, and encourage positive attitudes toward particular school subject areas. In

contrast, if homework is too challenging, too easy, non-varied, or too frequent, classes of

students complain and fail to see the benefits of the homework experience. Similarly,

family members, peers, and community members may add value to homework by

offering judicious assistance and new perspectives to homework questions. On the more

negative side, these influences from outside the school may distract the student from

completing homework.

A brief look at history demonstrates shifting public opinion over the homework

topic (Cooper, 1989, 1998; Winnerip, 1999). During the 1800s, teachers emphasized the

importance of memorization as a key homework strategy. Educators saw memorization

as an excellent mental exercise vital to new knowledge acquisition, and suitable for work

at home. By the early 1890s, people began voicing concerns over constant recitation, and

the first anti-homework movement began with charges that homework impedes children's

physical and mental health. By the 1920s, some school districts even abolished

homework arguing that constant drill activities often discourage student interest in

learning and interfere with home life. Attitudes again shifted in favor of homework

during the 1950s with "progressive" educators calling for more creative homework

assignments. The success of Sputnik in 1957 launched an era of global competition

wherein homework was used as a means to complete with the Soviets in math and

science. In the 1980s, the United States began to view the disciplined and industrious

people of Japan as its new world competitor, sparking a call for intensified homework as

a means to compete. This brings us to the end of the century with an awareness that

homework is here to stay as an everyday part of school life.

Historical attitudes have portrayed the topic of homework as an "all-or-none"

debate, with some proposing a need for more homework and others arguing for none.

Though researchers now focus on more complex issues, the first studies of homework

focused on whether or not students benefit scholastically from completing homework.

Early studies compared students completing homework to those not completing

homework. Meta-analyses indicate that completion of homework does have a positive

effect on student achievement as measured by class grades or achievement test scores

(Cooper, 1989; Paschal, Weinstein, & Walberg, 1984).

Developmental differences

Though completion of homework positively relates to student achievement, the age of

the student moderates the relationship. Achievement scores of high school students more

positively relate to homework completion than do those of elementary students. In a

meta-analysis of studies conducted before 1989, Cooper found that high school students

completing homework outperformed 75% of students in a no-homework class. Junior

high students outperform about 40% of students not assigned homework; and homework

assignments in the elementary years had small effects on student achievement gains

(1989, 1998). Though completion of homework in the elementary grades is not strongly

related to student achievement, the practice of homework, when it helps young children

develop good study habits, has a long-term positive effect on secondary school

achievement (Cooper, 1998) and student understanding of responsibility, and an

immediate effect on the mastery of skills.

With the discovery that completion of homework often correlates with increased

achievement, researchers then began to investigate how much time investment was

needed to achieve benefit (Cooper, 1989; Keith, 1982; Keith & Cool, 1992; Paschal,

Weinstein, & Walberg, 1984; Walberg, Fraser, & Welch, 1986). Again, age of student

moderated these correlational relationships. In the elementary grades, many studies show

that increasing time spent on homework correlates with lower school grades and

achievement, a negative relationship. In contrast, by the middle grades and high school,

this relationship becomes positive such that more time spent on homework relates to

higher grades and achievement. In fact, Keith's (1982) results indicate that a low-ability

high school student completing one to three hours of homework weekly could

theoretically achieve grades commensurate with an average student who fails to complete

homework (p. 251).

However, these correlational relationships fail to tell us what causes the relationship

between time on homework and achievement to be negative in the elementary grades and

to be positive in the high school grades. Epstein (1988, 1998) offered several plausible

explanations. First, elementary students often are assigned the same type of homework

activity night after night. These activities often take the brighter students less time to

finish; therefore, the brighter students earning higher grades spend less time on

homework, a negative correlation. In addition, Epstein has found that there is also a

strong negative relationship between minutes of parent help and achievement in the

elementary grades. Therefore, it seems that slower students and their parents work

together in the elementary years on the mastery of skills. We do not know what happens

during the middle and high school years, but Epstein (1998) suggested the following

possibilities: teachers may stop giving low-achieving students homework; parents may

stop monitoring their children's homework; or low-achieving students may stop

completing assigned homework. Muhlenbruck and colleagues (2000) conducted a study

to address this developmental relationship and found weak evidence that young students

who were struggling took more time to complete homework.

Age also plays an important role in students' understanding of the responsibility to

complete homework. For example, Warton (1997) investigated second, fourth, and sixth

grade Australian children's understanding of responsibility for homework. Though all

children provided a reason for completing homework, there were developmental

differences in terms of children's understanding of responsibility. Sixth grade children

provided responses like the following: "It's your work, not your Mum or Dad's"; while

second grade children responded "It would be a favor to Mum if I remembered (to do my

homework)." Whereas older children understand and can articulate the fact that

homework is their responsibility to initiate and complete, younger children often have not

yet integrated this understanding that it is their job to remember their homework

assignments, not their parents'.

Parental involvement in homework

Parents contribute in various ways to their children's education. Scott-Jones (1995)

proposed a four-level framework to organize the different types of parent-child

interactions that affect school achievement: valuing, monitoring, helping, and doing.

Valuing refers to the fact that parents communicate the importance of education to their

children. Like valuing, monitoring affects student motivation and relates to parental

supervision of children's academic performance. Helping interactions involve working

together on basic skills. Doing interactions exemplifies an overly-involved parent who

disregards a student's need for autonomy and does homework for his child. Therefore,

this framework highlights both positive and negative forms of parental involvement that

emerge in home-based parental help.

Negative findings. Several researchers have investigated the role of parental

involvement in the homework process (Cooper 1998, in press; Epstein, 1988, 1991;

Epstein & Lee, 1995; Epstein et al., 1997; Hoover-Dempsey et al., 1992; Hoover-

Dempsey, Bassler, & Burrow, 1995; Leone & Richards, 1989; Levin, 1997; MacBeath,

1998; McDermott, Goldman, & Varenne, 1984; Moller 1994; Xu & Como, 1998). For

example, Cooper (in press) conducted an investigation of 709 parents (from grades 2, 4,

6, 7, 8, 10, 11, and 12) regarding involvement in their children's homework. He found

that two-thirds of the parents reported providing a form of help that was negative or

inappropriate such as helping the child in order to finish more quickly, helping the

student despite knowing that the work needed to be completed independently, or helping

the child and knowing that the help made the work harder for the student. Although we

often assume benefit for students through parental involvement, it can be

counterproductive if parents are not given explicit instruction from teachers on how best

to be involved in the homework process.

Several studies highlight this need for improved communication between school and

home regarding parental help with homework (Balli, 1998; Dauber & Epstein, 1993;

Epstein, 1991: Epstein & Dauber 1991; Epstein & Lee, 1995; Hoover-Dempsey et al.,

1995; Hoover-Dempsey & Sandier, 1997; Lehrer, 1997; Scott-Jones, 1995; Xu & Como,

1998). According to Hoover-Dempsey and colleagues (1995), most parents of

elementary students felt it was their duty to help their children with homework, but many

felt unprepared to help with certain subjects. Some parents openly discussed

embarrassment regarding their lack of skills to help or discussed frustration with

spending most of their time trying to communicate ideas using vocabulary that, according

to the children, differs from the teacher's.

Xu and Como (1998) conducted a case study investigation of a small group of well-

educated families involved in their children's third-grade homework. Even this small and

educated sample of families spent much of their time trying to improve and make their

children's homework assignments more interesting. Though this involvement may be

positive for these families, one must consider what happens to the children of parents

who do not have the skills or time to improve existing homework and to inspire the

interest of their children.

Negative forms of family involvement are not limited to the elementary grades.

Epstein and Lee (1995) studied homework practices of middle grades students as part of a

comprehensive investigation of school, family, and community partnerships. Using the

National Educational Longitudinal Study of 1988 (NELS: 88), the researchers used data

from 1011 middle schools, including reports from teachers and principals related to over

20,000 eighth grade students and their parents. Though 91% of parents in the study

reported that they believe homework is valuable, over 50% never, seldom, or infrequently

helped their children with homework. The authors suggested that this finding relates to

the common practice of teachers in the United States to assign homework requiring

independent work on the part of the student and no interaction or discussion with family.

Reports of principals supported this explanation with over 75% of them stating that less

than half of parents at their respective schools receive regular information from teachers

as to how they may help their children with homework. Given this lack of information, it

is not surprising that 85% of principals in the study report that less than 50% of parents at

their school help their children at home or assist with homework.

Results from these studies emphasize the need for parents to have explicit

information on how they may help their children with learning activities like homework.

Though nearly all parents and families understand the importance of school practices

such as homework, many do not know what forms of involvement may be most helpful to

their children at different points in development and often do nothing at all to support the

homework process.

Positive findings. Because most schools currently do not provide practices or

information to facilitate parental involvement with homework, many findings in the

literature are negative or null. However, the following are some examples of positive

findings related to family involvement in current school practices. Epstein (1988) found

that children in the elementary grades who enjoyed talking about homework and school

with their parents tended to be better students. In fact, 80% of parents of middle grade

students reported talking with their children about middle school (Epstein & Lee, 1995).

Most young adolescents reported that they would interact with their families if homework

was designed to encourage family interaction (Connors & Epstein, 1994; Epstein,

Jackson, & Salinas, 1994), and studies have shown that adolescents in the middle grades

who complete homework with their parents tend to be better students (Leone & Richards,


Research also shows the benefits of parental-child discussion about school-related

topics on homework completion and academic achievement in the middle and high

school years (Fehrmann et al., 1987; Ho Sui-Chu & Willms, 1996; Keith et al., 1993;

Simon, 2000). Keith and colleagues examined the effects of parental involvement

(educational aspiration, parent-child communication, amount of home structure, and

participation in school activities) on time spent on homework and achievement of eighth

grade students. Parental involvement affected achievement more strongly than

socioeconomic status; therefore, parental involvement had a strong effect on achievement

independent of family background measures. Similar findings emerged for Fehrmann

and colleagues' study of high school seniors in the High School and Beyond (HSB) data

set (1987).

Research conducted by Ho Sui-Chu and Willms (1996) also indicated that certain

forms of parental involvement are more beneficial to student achievement than others.

The researchers examined the effects of four types of parental involvement (home

discussion, school communication, home supervision, and school participation) on the

math and reading achievement of eighth grade students from the NELS data set.

Discussion of school-related activities at home had the strongest effect of all the parent

involvement variables on math and reading achievement. In fact, parental involvement

reduced the effect of SES on achievement by seven percent.

In summary, homework activities often are tedious and fail to generate creativity;

parents do not feel prepared to discuss some homework concepts with their children; and

parents and teachers need to know what strategies are appropriate for parental help at

home. Parental involvement in general exerts the most positive effects when a child

perceives the involvement to be developmentally appropriate and a nice fit with the

school's expectations (Eccles et al., 1993; Hoover-Dempsey & Sandier, 1995). For

example, if parents develop accurate knowledge of the their children's abilities (Miller,

1995; Stevenson, Lee, Chen, Stigler, Hsu, & Kitamura, 1990), their strategies for

involvement will be perceived more positively by students. Similarly, given that both the

school and family have standards for parental involvement, the better the agreement

between parental involvement activities and the school's expectations for parental

involvement, the more appropriate the child will perceive the activities to be (Hoover-

Dempsey & Sandier, 1995).

Therefore, though current homework practices often discourage familial interaction

on homework assignments, those schools that do encourage, instruct, and expect

developmentally-appropriate interaction may provide students, parents, and teachers a

promising strategy for increasing beneficial forms of family involvement. Such efforts

may also promote interest in particular subjects and more positive student achievement.

Teachers Involve Parents in Schoolwork (TIPS) interactive homework

Based on findings from studies of homework and parental involvement, Epstein and

colleagues developed a framework for interactive homework assignments by subject as

well as sample activities for students in the elementary and middle grades (Epstein et al.,

1995). These assignments include clear objectives for learning, instructions for

completion, and explicit instructions to the student for involving family members in

certain portions of the assignment. TIPS Interactive homework assignments differ from

traditional homework in that they are assigned once a week or twice a month; students

are given several days to complete the activity (to permit time to involve family); certain

sections of the activity prompt students to involve family members in their completion;

and parents provide feedback as to how effective the activity was for them and their

children. As with conventional homework, TIPS activities are graded, integrated with the

curriculum, and designed to extend student learning beyond the classroom.

The TIPS Interactive homework framework helps schools with the many

shortcomings in current homework practices. Use of the research-driven framework

helps teachers identify topics in the curriculum that require interaction for better

comprehension. Through the assignment, teachers provide instructions to students as to

how they may involve family members in certain sections of the assignment. To protect

family members from embarrassment for not knowing specific concepts, teachers design

interactive questions that parents can answer without detailed knowledge of the subject at

hand. This promotes discussion of what is being learned in school without embarrassing

family members because they may not know about specific concepts. Also, the

assignment includes a section asking the parent to communicate with the teacher about

the effectiveness of the assignment. If the student finds the assignment to be too

challenging or too easy, the parent and student may provide feedback to the teacher to

alter the assignment for future use.

TIPS writing. Researchers have assessed the impact of TIPS Interactive Homework

in language arts and math. Epstein and colleagues (1997) conducted a developmental

investigation of TIPS- Language Arts Interactive Homework with 683 sixth- and eighth-

grade students from inner-city schools. They investigated relationships between TIPS

completion, students' writing skills (as measured by scores on three writing samples

collected in the fall, winter, and spring), report card grades, and parental attitudes about


Results indicated that the number of TIPS assignments completed positively related

to language arts report card grades; therefore, the more assignments completed the higher

the language arts grade. In addition, parent participation on the TIPS assignments

positively influenced overall quality of student's writing skills and also related to parents'

overall liking of TIPS. Both teachers and parents reported liking TIPS, and parents noted

that TIPS helped inform them of what their children were learning in class. Students

varied in their liking of school and TIPS, but they did report understanding how to use

TIPS with their family members.

Use of TIPS writing homework assignments promoted positive reactions from both

teachers and parents about the process and resulted in positive scholastic outcomes for

middle grades students.

TIPS math. Balli (1995, 1998) conducted an experimental investigation of suburban

middle school students and their use of TIPS math in three ways. Seventy-four sixth-

grade students taught by the same teacher were placed in one of three groups: (1) TIPS

activities designed with both family and student prompts for family involvement; (2)

TIPS activities without family prompts for family involvement; and (3) conventional

math homework with no prompts to the student or family for family involvement.

Family involvement was higher for both TIPS groups than for the conventional math

group, and families in these groups reported knowing more about what their children

were learning in math class. However, though family involvement for the math TIPS

group was higher in math than the other two groups, the TIPS families were not more

involved in homework in other subject areas. Findings of this experiment allow one to

conclude that the design of TIPS caused more family involvement than the conventional

homework design. Despite the fact that all groups experienced achievement gains as

measured by a posttest of questions related to information from the homework

assignment, no significant differences emerged by group.

Both of these TIPS studies alert researchers to the importance of addressing outcome

variables besides achievement, as measured by grades or test scores. Other outcome

variables worthy of investigation relate to attitudes and motivation for the subject in

which TIPS activities are utilized. Though no achievement differences emerged in the

Balli (1995) investigation, continued use of the TIPS activities requiring family

involvement, further into the school year, might have resulted in more student interest in

and motivation for math than students not involving families. Research does suggest that

student motivation and learning as well as classroom perceptions relate to classroom

performance and degree of cognitive engagement (Ames & Archer, 1988; Pintrich &

DeGroot, 1990). Achievement gains and increased interest or motivation often come

hand-in-hand in intervention studies (Epstein, 1997; Paris, Yambor, & Packard, 1998).

Therefore, any attempt to measure the results of an intervention should include attitude or

motivation measures as well as achievement measures because achievement effects often

take more time or more comprehensive interventions to see positive results.

Therefore, TIPS Interactive homework assignments may provide teachers with an

important strategy for improving current homework practices that often discourage

interaction, promote confusion, and emphasize incessant drilling. Most important, TIPS

assignments provide a way for teachers to communicate with and invite families to

interact with their children on homework. TIPS interactive assignments relate directly to

school objectives for learning, encourage creativity, and provide students the opportunity

to initiate and discuss issues with their families that do not require that family members

be an expert on the topic of the assignment. Few schools currently utilize the TIPS

framework as part of curriculum, and TIPS are not an accepted part of school practices

today. However, studies of TIPS use can demonstrate what may result in the future if

more schools work to improve their current homework practices in meaningful and

research-driven ways.

The Present Study

The current developmental and experimental investigation of TIPS Science

homework use in grades 6 and 8 extends current knowledge of TIPS use and

effectiveness in six important ways. First, the study involved the examination of a

subject area not yet investigated, science. Goal 5 of the National Education Goals

focuses on math and science achievement: "By the year 2000, United States students will

be first in the world in mathematics and science achievement." Despite this goal, many

researchers have alerted us to the fact that current science education often discourages

girls from pursuing studies in the physical sciences (Burkam, Lee, & Smerdon, 1997).

Students, especially girls, reported that science often feels detached from real-life and

desire discussion and experimentation to make it a more meaningful experience. Studies

of the NELS data indicated that students learned more in science classes that included

active learning experiences like regular laboratory activities where students write up lab

reports, perform experiments, and discuss concepts. Activities such as these help

students learn more than when they simply watch the teacher perform specific

experiments (Burkam et al., 1997). Because the TIPS framework for science includes a

"Lab Report or Data Chart" section as well as opportunities for performing experiments

and discussing conclusions, these assignments should prove beneficial to student

learning. Therefore, integrating the TIPS process into this often homework-neglected

subject serves as one possible strategy for furthering current science achievement and

motivation levels in the middle grades.

In addition to extending knowledge related to science curriculum, the experimental

nature of this design permitted the investigation of the role of interaction on homework

completion, science interest, and science achievement. This study included two groups: a

TIPS interactive science homework group and a non-interactive group. Half of the

students in each grade completed TIPS interactive science assignments as they are

designed to be conducted, while the other half completed non-interactive assignments

(the same content of the TIPS assignment without prompts to the student to involve

family members). Very few investigations of homework have included experimental

designs (Balli, 1995; Elawar & Como, 1985; Miller, Duffy & Zane, 1993), and the

findings from this study allow comparisons between the performance of students in the

noninteractive homework group and students in the TIPS interactive homework group.

Thirdly, the proposed study informs our current knowledge of developmental and

gender differences in student homework practices. Students in this study from grade 6

just completed elementary school, and their experiences are very different from students

in grade 8 who are finishing their last year in middle school. Whether or not the age and

gender experience differences of students in these two grades relate to differences in the

way that TIPS are used or received was an important question that is addressed in the


Another important contribution of this study was its examination of results with

different ability levels of students and different teachers. Balli's experimental

investigation involved students of the same ability level and teacher. Though these

controls allowed interpretations of effects for similar high ability students with the same

teacher, it remains important to understand how the intervention generalizes to different

teaching styles and students because teachers often vary in their teaching practices and

homework feedback (Elawar & Como, 1985).

The fifth contribution of the study relates to the various science achievement, family

involvement, homework completion, and science attitude measures from parents and

students. Having different measures such as these allowed interpretations of not only

achievement or intellectual effects, but also of social and attitudinal effects that are also

important goals of any educational intervention.

Finally, having both parent and child opinions and reports of various questions

permitted analysis of the correspondence of their reports and how this relates to various

outcome measures for children. Much of the research on intergeneration agreement

makes a point of noting the importance to parents of correspondence between parent and

child, but there may be various cases where divergence in parent and child opinions may

not cause conflict (Goodnow, 1992). For example, Smetana (1988) studied the opinions

of parents and adolescents, finding that differences in opinions often did not result in

major problems. She distinguished between "disagreement with conflict" and

"disagreement without conflict." Preliminary analyses of these patterns of positive and


negative agreement as well as patterns of disagreement were explored to begin to

understand how these reports relate to student science achievement and student attitudes.

The six contributions of the present investigation underscore both its research and

practical implications. The study examined the effect of inviting and instructing parents

to be involved in homework on student science achievement and interest. It also provides

information on the development, implementation, and evaluation of interactive

homework assignments that may be useful to middle school administrators and teachers

in their curriculum decisions related to science homework.



All participants in the study came from one public middle school in Baltimore

County in the state of Maryland. The school serves a diverse body of students in grades

six, seven, and eight. Fifty-three percent of students in the study were white; 36% were

African-American, 5% were multi-racial; and the remaining 6% of students considered

themselves Asian-American, Hispanic, or other. Two teachers from sixth grade and two

teachers from eighth grade utilized the homework intervention over the course of the first

two marking periods (18 weeks) of the 1999-2000 school year. The study included a

subsample of these teachers' classes, three classes from each sixth grade teacher and two

classes from each eighth grade teacher. Therefore, a total of 253 male and female

students participated in the investigation: 143 sixth grade students (mean age = 11.39

years) and 110 eighth grade students (mean age = 13.28 years).


The design of the experimental study was a 2 x 2 between-subjects design with two

conditions (TIPS and ATIPS) and two grades (sixth and eighth grades). None of the

teachers had two classes of the same ability level students. The purpose of this

investigation was to compare the effects of completing TIPS homework assignments to

ATIPS homework assignments on the average student, not those students at the extremes

in either "gifted and talented" or "inclusion" classes.

At the outset of the study, the author assigned classes to either the TIPS or ATIPS

condition keeping class ability level constant across condition but not across teacher.

Because each participating teacher contributed both an "average" and "above average"

class to the experiment, it was not desired to confound condition with ability level.

As planned, ability level was held constant across condition but not across teacher

in the eighth grade sample. Due to experimental error in homework assignment

implementation, ability level was not held constant across condition; it was held constant

across teachers in the sixth grade. Once the author discovered the error in week 2 of the

intervention, she collected data from an additional two classes of low ability sixth grade

students in the TIPS condition. At this point in the study, parents had already received

letters explaining their roles. The TIPS family letter encouraged families to assist

children with their homework assignments when children asked. ATIPS family letters

included no information about parents assisting children with homework. Therefore, it

proved impossible to change a class from the TIPS condition to the ATIPS condition. If

such a change happened, it would have confused both the students and parents in these

classes. The only way to compensate in some way for the error was to include two

additional sixth grade TIPS classes of low ability students in the study.

One hundred and forty-three students came from sixth grade science classes, and

one hundred and ten students came from eighth grade science classes. The classes varied

in the ability levels of students, including low-ability, average-ability, and honors classes

in sixth grade, and average-ability and honors classes in eighth grade. Six classes

completed the Teachers Involve Parents In Schoolwork (TIPS) process of science

handouts, and four classes completed ATIPS assignments. Each sixth grade teacher

(Teachers A and B) taught one honors and one low ability class using TIPS, and one

average class using ATIPS. One eighth grade teacher (Teacher C) taught a TIPS average

ability class and an ATIPS honors class, while the other teacher (D) taught a TIPS honors

class and an ATIPS average class. See Table 2-1 below for more details on the number

of students and the study groups.

Table 2-1. Number of students by condition, ability level, and grade (N = 253)


Student ability level Grade 6 Grade 8 Grade 6 Grade 8

Low 36

Average 23 53 22

Honors 54 33 32

Total 90 56 53 54


During the summer of 1999, teachers from grades six and eight worked with the

author to develop TIPS science homework assignments linking to the school curriculum.

Teachers from each grade agreed to develop at least 18 activities, allowing the study

period to span two marking periods with one assignment for each week (August 30,

1999-January 25, 2000).

A teacher from sixth grade created a total of 18 TIPS activities so that one TIPS

assignment could be assigned each week over the course of the first two 9-week marking

periods of the 1999-2000 school year. Two teachers from eighth grade created a total of

40 TIPS activities so that one TIPS assignment could be assigned each week over the

course of the entire 1999-2000 school year; though this study involves the analysis of

only the first 18 assignments. Appendix A lists the assignment order for each of the sixth

and eighth grade teachers. Both sixth grade teachers teach their units of work in the same

order: skills and processes, oceanography, and machines. However, due to lack of

textbooks, the eighth grade teachers cannot teach their 4 curriculum units of the academic

year (ecology, heredity, geology, chemistry) in the same order. Teacher C taught the

geology unit during the first 9 weeks followed by the ecology unit during the second 9

weeks, and Teacher D taught the ecology unit followed by the heredity unit. Both eighth

grade teachers did, however, teach the ecology unit during one of the marking periods of

the study.

Interactive (TIPS) Assignments

All the developed TIPS activities include eight important components: 1) Letter to

Parent or Guardian (including a brief description of the activity as well as the due date of

the assignment), 2) Objective (s), 3) Materials (common, inexpensive items in most

homes that are easily obtainable), 4) Procedure (steps for experimentation or scientific

thought about a topic), 5) Lab Report or Data Chart (show work or write complete

sentences on the topic under investigation), 6) Conclusions and/or Discussion (conclude/

summarize information and/or guide a discussion to extend information into the "real

world"), 7) Home-to-School Communication (allows parents to share comments and

observations with teachers), and 8) Parent or Guardian Signature (Epstein, Salinas, &

Jackson, 1995). In addition, each TIPS activity linked directly to the Baltimore County

curriculum indicators in a meaningful and appropriate way, was doable, interactive, the

student's responsibility, easy to read and understand, attractive, and 2 pages (a front and

back side of a piece of paper). See Appendix B for a sample eighth grade TIPS activity

for the geology unit.

Non-Interactive (ATIPS) Assignments

After the teachers developed the TIPS activities, the author produced a second set of

"altered" (ATIPS), non-interactive, activities for the experimental investigation. These

activities included the same content in terms of information and questions, yet they

included no prompts to the student or family regarding family involvement (with the

exception of when safety is a concern). Therefore, these non-interactive assignments do

not include any prompts for interaction with the fair, l., like (1) the letter to the parent or

guardian, (2) the home-to-school communication section, or (3) any questions

encouraging students to ask their family members questions. Appendix C displays a

sample ATIPS activity. This corresponds to the sample TIPS activity presented in

Appendix B.

TIPS/ATIPS Test Questions

In addition to developing the science TIPS homework assignments during the

summer, teachers wrote 36 total test questions (2 questions for each of the 18

assignments for the first 2 marking periods). Because the content/information of both the

TIPS and ATIPS assignments was identical, teachers used the same test questions for

both conditions of the study. See Appendix D for a complete listing of the TIPS/ATIPS

test questions utilized in the study.

Student/Family Surveys

Most of the measures needed to assess the effectiveness of TIPS and ATIPS

assignments came from the homework assignments themselves. However, surveys of

students and families provided additional information to learn more about levels of

parental involvement in science and other subject areas for students in both conditions,

time spent on homework, and student interest in science and the ATIPS/TIPS science

homework assignments. See Appendix E to examine items on the student and family

sixth and eighth grade surveys.


Letter to Families

The teachers and I wrote a letter appropriate for each study condition to be sent to

the home of the student with the first TIPS/ATIPS assignment. Each letter described

these new homework assignments to students and parents. The ATIPS group letter

described the fact that students would receive weekly science assignments on green paper

linked to the science curriculum. There was no mention of family involvement in this

letter as these assignments were not designed for such interaction.

Conversely, families of students in the TIPS group received a letter explaining

weekly assignments on green paper; however, parents were notified of the importance of

family involvement in some sections of the activity. See Appendix F for copies of the

letters sent home to student families. As noted in the manual for TIPS use, it was made

clear that it is the student's responsibility to involve parents, not the parent's

responsibility (Epstein et al., 1995).

Science Assignments

Each teacher assigned one TIPS or ATIPS activity per week over the course of two

marking periods for a total of 18 assignments. Teachers assigned the first activity the

first week of school, and they continued to assign an activity weekly through the study

period ending January 25, 2000. Teachers in both grades chose to assign the TIPS

activities on Tuesdays and Wednesdays (with some exceptions for holiday), and they

were due the following Tuesday. All teachers graded the assignments on a 10 point

scale and tested the content of the assignment as they do any other homework. Prior to

distribution of any TIPS homework assignments, teachers made it clear to the students

that material from the TIPS/ATIPS assignments would be included on classroom


Science Assessments

Teachers in the study used the science test questions linked to each TIPS/ATIPS

science activity to assess the science achievement of students. Teachers were given

flexibility in terms of when these test questions were given, whether on a test or quiz, but

all teachers included the relevant test questions on their classroom examinations.

Teacher A for sixth grade gave four exams over the course of the 18 week study period

with questions from homework assignments 1-5 on the first test, 6-12 on the second, 13-

16 on the third, and 17-18 on the last exam. Teacher B designed 5 exams and included

TIPS/ATIPS questions from assignments 1-4 on the first exam, 5-7 on the second, 8-11

on the third, 12-13 on the fourth, and 14-18 on the final exam. Teacher C from eighth

grade gave 6 exams including the questions over the study period (1-5; 6-9; 10-11; 12-13;

14-15; 16-18), and Teacher D from eighth grade gave 9 exams over the course of the 18-

week study period that included TIPS/ATIPS test questions (1; 2-4; 5; 6; 8-9; 10- 11; 12-

13; 14-17; 18).


After students completed all relevant TIPS and ATIPS assignments and related

science exams through the first two marking periods, they completed a confidential in-

class survey. Children also delivered their parents a survey and informed consent form to

complete and return to school.


This study included a variety of independent and dependent measures. Because

participants in this study varied in terms of ability level, prior science achievement, and

family background, the following independent measures were collected to control for

such differences: prior achievement, mother education level, student ability level,

student race, gender, and grade. Condition (TIPS or ATIPS) served as the experimental

variable in the study. Homework completion, family involvement, time on homework,

science achievement, and science attitudes served as the dependent constructs in the

study. See Table 2-2 for a more complete description of the measures used to define the

dependent constructs.

Background Measures

Most of the background measures including overall science grade from the year prior,

previous standardized reading and math scores, gender, grade, and class ability grouping

came from the student record or class roster. Age of student and other measures not used

in the analyses came from the student surveys.

Homework Completion Measures

The author recorded homework data each week by student number for various

measures of homework completion. Specific homework data collected from both the

TIPS and ATIPS assignments included the following: return of homework assignment

(yes or no); lateness of the assignment (turned in late or not turned in late); if late, the

number of days late (0 or more days late); completion of the homework assignment (yes-

all sections answered, only some sections answered, no sections answered); marking of

teacher-assigned homework grade (1-10 points); and total points received for the

interactive family sections (0-6 points) on the TIPS assignments or the related questions

from the ATIPS assignments.

Additional data from the TIPS assignments only included the name or relationship

of the family partner to the student, gender of the family partner (male, female, could not

be determined), the presence of a parent signature (yes or no). the answers to the

questions in the Home-to-School Communication Section, and any parent or other family

member comments made on the assignment itself. Each Home-to-School

Communication Section of every TIPS science homework assignment included the

following three questions: "My child understood the homework and was able to discuss

it; My child and I enjoyed the activity; This assignment helped me know what my child is

learning in science." For each of the TIPS assignments, the author recorded a response of

"yes," "no," or missing data.

Survey Measures

The purpose of the survey was to gather necessary information about family

involvement from both the ATIPS and TIPS groups, additional family background

information, time on homework, opinions of the science assignments, and student interest

in science. Where possible, questions were written so that the parent and child responses

could be compared to investigate how the role of discrepancy and concordance in parent-

child responses relates to some of the dependent variables of interest. Surveys used in

previous studies of parental involvement in homework were collected to create and

design the specific components of the parent and child surveys in the present study (Balli,

1995; Epstein & McPartland, 1978; Epstein, Simon, & Salinas, 1997; Grolnick, Ryan, &

Deci, 1991; Levin et al., 1997).

Family involvement

Three questions from the student survey (la, lb, Ic) were used to gauge levels of

family involvement in the TIPS and ATIPS groups for math homework, language arts

homework, and the science homework assignments. The same questions appeared on

the parent survey to observe differences in student and parent reports (la, lb, Ic). Two

additional questions on both the parent and student surveys required the respondent to

gauge how often the student worked with a parent on the science green sheets or someone

else at home other than a parent (2a, 2b). Finally, all students recorded all the people

who worked with them on any of the TIPS/ATIPS homework assignments as well as the

person who helped the most (2c, 2d).

Family background

Students answered several survey questions about themselves and their families.

Questions included their ages, the number of adults and children living at home, race,

their mother's education level, and employment status of their mothers and fathers (12b,

c, d, e, f, g, & h).

Time on homework

Students and parents verified how much time is typically spent on various

homework tasks. Both groups checked how much time the student spent on homework

for all subjects in an average week, for science homework in an average week, and on

one typical TIPS/ATIPS science homework assignment. The final question asked how

much time a parent or family partner spent with the student on a typical green sheet

(student- 3, 4, 5, 6; parent- 4, 5, 6, 7).

Opinions of the TIPS/ATIPS homework assignments

Students and parents answered 11 of the same questions regarding their opinions of

the green science homework sheets (student- 7a, b, c, d, e, f, g, i,j, I, & m; parent 3a, b, c,

d, e, f, g, h, i, j, & k). Students only answered one additional question about their liking

of the green sheets as compared to regular homework (7h). Epstein and colleagues

(1997) used similar questions in their investigations of TIPS Language Arts interactive

homework in the middle grades.

General interest in science

Three questions from the Commitment to Classwork subscale of the Quality of

School Life Scale (Epstein & McPartland, 1978) were included on the student survey (8,

9, 10). The purpose of the Commitment to Classwork subscale is to measure student

interest in classwork in general or by subject. For purposes of the present study,

commitment to and interest in science classwork was of most interest.

In addition, during the first few weeks of the intervention, all students completed a

nine-question survey of their interest in science, math, English, life sciences, physical

sciences, and specific areas of science covered by that grade's curriculum. The same

nine questions were included on the final survey given after all 18 assignments were

completed to analyze change in science interest over the course of the 18 weeks (11).

Table 2-2. Independent and dependent constructs and variables



Background variables

Independent variables
Variables of interest

Dependent variables
Homework completion

Science achievement

Science attitudes

Time on homework

Family involvement
in homework

Previous science, English, and math achievement
Mother education level
Employment status of mother and father
Age of student
Class ability grouping of student
Race of student


Percent of TIPS/ATIPS homework assignments returned
Percent of assignments completed
Percent of assignments returned late
Average points earned per assignment
Percent of family points earned per assignment

Percent of TIPS/ATIPS test questions correct
Average test score of tests including TIPS/ATIPS
Science report card grades

Interest in science
Interest in science class
Opinions of the TIPS/ATIPS homework assignments*

Time on all homework each week*
Time on all science homework each week*
Time on a typical science TIPS/ATIPS assignment*
Family time on a typical TIPS/ATIPS assignment*

Family involvement in science homework*
Parent involvement in science homework*
Non-Parent family involvement in science homework*
Family involvement in language arts homework*
Family involvement in math homework*

* Both parent and student reports were collected.

Science Achievement Measures

Three main types of achievement measures were collected: report card grades in

science, percent of TIPS/ATIPS assignments answered correctly, and average science

test/quiz score for those exams including TIPS/ATIPS science questions. Each teacher

assigned different point values to the TIPS/ATIPS test questions; however, each student's

score for the percent correct was computed based on the total points earned over the total

points possible for each teacher. Average science test/quiz score was adjusted similarly

so that average score was based on points assigned to each test rather than the average of

the exam score grades which would not fairly account for the individual teacher's grading


Research Questions. Hypotheses, and Analyses

Family Involvement

The first research question concerned the relationship between homework condition

(interactive and non-interactive) and reported levels of family involvement with science

homework. Because the TIPS assignments were specifically designed to elicit family

involvement, it was hypothesized that higher mean levels of family involvement would

be associated with this homework condition.

Several questions on the survey asked students and parents to rate family

involvement levels in science, English, and math homework. The second family

involvement research question asked whether parental involvement levels differed across

subjects by science homework condition. Based on results from Balli (1995), it was

predicted that parental involvement levels would be higher for the TIPS group in science;

however, involvement levels in other subject areas (math and language arts) would not

statistically differ by science condition (interactive and non-interactive). If this proved

true, we would have additional evidence that the design of TIPS science assignments

elicited more involvement than traditional homework assignments in other subject areas.

Ordinary least squares regression analyses helped to answer these two family

involvement research questions.

Homework Assignments

Students in the study completed 18 assignments over the course of the study period.

The next research question addressed how the number of homework assignments turned

in as well as the average number of points per assignment relates to homework group and

grade of student. The hypothesis predicted that students in the TIPS condition would turn

in more assignments and earn more points per assignment than students in the ATIPS

condition because they completed the assignments with a family partner. Regression

analyses examined the effects of the various background measures and the experimental

variable of condition on homework completion measures.

Extending the research question further, did those students who more regularly

involved family members (regardless of condition) turn in more accurate assignments

than those not involving family members? The hypothesis predicted that TIPS students

who involved family members regularly would benefit more academically than ATIPS

students because the TIPS condition included instructions for such involvement.

Bivariate correlations of family involvement and homework completion measures

assessed the relationships for the entire sample of students as well as by condition and by


Science Achievement

Every educator wants to know the effectiveness of a homework intervention on

student achievement. Therefore, I investigated two questions related to science

achievement. I first examined whether or not there were significant differences in

posttest achievement of students by condition and by grade. Again, it was expected that

controlling for various background variables including mother education level and

previous grade in science, students in the TIPS group would outperform those students in

the ATIPS group because of the interactive nature of assignments that is likely to

promote student achievement.

Additionally, I predicted that regardless of condition, the more assignments

completed would be related to higher science test scores and or grades. Previous research

supports the hypothesis in that the amount of homework middle grade students completed

positively related to achievement (Cooper et al., 1998). OLS regression analyses again

examined these relationships.

Concordance in Parent and Child Surveys

The nature of the survey data permitted exploratory analyses of correspondence and

divergence in parental and child reports of family involvement and opinions of the

TIPS/ATIPS science homework. Looking at the parent and student surveys, one may

note the similarity in wording to allow direct comparisons of parent and student views on

the same question.

Researchers are interested in how agreement or discrepancy in parent and child

views may relate to various outcomes for students (Goodnow, 1992, 1995). Therefore,

correlations determined overall agreement /discrepancy in parent and student reports. It

was predicted that those students having student and parent reports that were more


positive and similar would earn the highest science grades, turn in more homework

assignments, and enjoy science class the most. OLS regression analyses and mean

comparisons by group examined the effects of the agreement in reports on selected

student achievement measures.


This study examined the effects of interactive (TIPS) and non-interactive (ATIPS)

science homework assignments on student science achievement, student and family

attitudes about science homework, and family involvement in homework. This chapter

includes descriptive and inferential analyses of data to answer the research questions

related to family involvement, homework completion, student achievement, and

concordance in parent and student reports of family involvement in homework, time on

homework, and opinions of science homework.

Descriptive Statistics

Appendix G includes descriptive data for all study variables. Means or frequency

and sample size data are provided for specific background, homework completion,

science achievement, interest in school, family involvement in homework, time on

homework, and opinions of the science homework variables. The sections that follow

highlight data from the appendix and selected correlations related to the regression

models. This chapter includes many references to "science homework" or "science

assignments," and these terms refer only to the TIPS and ATIPS weekly assignments.

Family Involvement Research Questions

The first research question examined the relationship between homework condition

(TIPS and ATIPS) and family involvement in science homework at home. Figure 3-1

displays levels of family involvement in science homework by condition.





never rarely sometimes frequently always
Level of Involvement

Figure 3-1. Student reported levels of family involvement in science homework by
condition (N = 226)

The white bars in Figure 3-1 represent ATIPS student reports (n_= 98), while the

black bars represent students in the TIPS condition (n = 128). Over 80% of students in

the ATIPS condition said their families were never, rarely, or sometimes involved in the

science homework assignments over the 18-week study period. In contrast, 80% of TIPS

students said their families were sometimes, frequently or always involved in science

homework assignments.

Of interest is whom students reported working with on science homework. Seventy-

four percent (74%) of students reported working with their mothers on at least one of the

science assignments (167/226). Forty-seven percent (47%) of students who completed

the survey said their fathers were involved in at least one homework assignment

(106/226). Eighteen percent of students reported help from a sister, and sixteen percent

reported help from a brother at least once. Twenty-seven percent of students said they

received homework help from a relative other than immediate family or a friend at least

once. These relatives and friends included cousins, grandparents, neighbors, aunts,

uncles, and others. Students also described who helped with science homework most

frequently. Fifty-eight percent (58%) of students named mothers helping with homework

most often, 17% of students named fathers, 11% named a sister or brother, and 5%

named a friend, and the remaining 9% named neighbors, aunts, uncles, grandparents, and


Parent reports of involvement followed the same pattern as the student reports with

most TIPS parents marking sometimes, frequently, or always involved, while most

ATIPS parents marked never, rarely, or sometimes involved in science homework. A

positive and significant correlation (r = .669, p < .001, n = 177) exists between parent and

student reports of family involvement in science homework. Because the student survey

return rate (89%) exceeded the parent return rate (71%) and the patterns of student and

parent were similar, student reports of family involvement served as the dependent

measure for family involvement in science homework.

Ordinary least squares (OLS) regression analyses calculated the effects of various

background measures and homework condition on family involvement in science

homework. Table 3-1 displays the results of the regression analysis.

To control for variations in the TIPS and ATIPS groups, previous science

achievement, mother education level, class ability grouping of student, race of student,

gender, and grade served as the background control variables. The first model includes

those variables and accounts for only 6% of the variance in family involvement levels in

science homework. Honors students had significantly higher family involvement levels

in model 1 than did average students (p = .26; p < .01), and sixth grade students reported

significantly higher levels of involvement than did eighth grade students (P3 = .25;

p < .01). Sixth grade students reported a mean level of involvement of 2.37, while eighth

grade students reported a mean level of 1.73 on a scale of 0 to 4.

Table 3-1. Predictors of student reported levels of family involvement in science
homework (N = 226)

Variable Model 1 Model 2
3 t 3 t

Prior science achievement .01 0.08 .02 0.29
Mother education level
(reference- college graduate)
High school .09 1.14 .07 0.95
Some college .03 0.32 .02 -0.21
Graduate school .01 0.16 .00 0.03
Class ability grouping
(reference- Average)
Low .09 1.12 -.15 -1.68
Honors .26 2.90** .01 0.05
(reference- White)
Black .00 -0.02 -.03 -0.35
Other .09 1.10 .05 0.68
(Male = 1, Female =0) .01 0.08 .03 0.36
(Grade 6 = 1, Grade 8 = 0) .25 2.94** .21 2.68**

(TIPS = 1, ATIPS = 0) .45 5.40**

R2/Adjusted R2 .11/.06 .24/.19

Note. AR =. 129 for Model 2 (p <.01). *p <.05. **p<.01.

Model 2 adds the effect of homework condition to the previous model and accounts

for 19% of the variance in student reported levels of family involvement in science

homework. Condition is the most predictive variable in the model such that TIPS

students reported significantly higher levels of family involvement in science homework

than did those in ATIPS classes (p = .451; p < .001). Again, sixth grade students reported

more involvement of families in science homework than eighth grade students, but

classroom ability grouping failed to explain a significant portion of the variance when

condition was added to the model. Also worthy of mention is the fact that previous

science achievement did not predict a significant portion of the variance in family

involvement for these TIPS/ATIPS science homework assignments. Therefore, students

of all class ability grouping levels and varying previous grades in science did not differ in

their reported levels of family involvement, but students in the TIPS condition did report

higher levels of involvement than ATIPS students.

Another survey question asked students to report specifically on how often their

parents (not just any family member) participated in the science homework. When this

variable is used in the previous model as the dependent measure of family involvement in

science, the same pattern emerged with condition (TIPS or ATIPS) as the strongest

predictor of family involvement in science. Subsequent analyses include the family

involvement measure, not the parent involvement measure.

One might wonder whether students in the TIPS condition had families who were

already more involved in student homework than the families of ATIPS students, or if

involvement in science homework generalized to other subject areas. The second

research question examined levels of family involvement of the TIPS/ATIPS science

grouping across subjects where TIPS/ATIPS assignments were not used.

Figure 3-2 displays the family involvement means by school subject for the TIPS

and ATIPS science homework conditions. According to the graph, the mean for TIPS

family involvement levels in science (M = 2.55) exceeds the ATIPS mean (M = 1.43).

The math and language arts family involvement levels do not dramatically differ by

science condition. Importantly, as with science, parent reports of involvement followed

the student reports, and parent and student reports are positively correlated for math and

language arts (r =.384, <.001, N= 176; r= .380, <.001, N= 176).

3 2.55

I2 1 1.43 1.56 1.4
S1.19 1

Science Language Arts Math

Figure 3-2. Student reported levels of family involvement in science, math, and language
arts (N = 226) by science homework condition (TIPS and ATIPS)

Table 3-2 displays results of the regression analysis of variables predicting family

involvement in language arts homework. None of the variables in models 1 or 2 predict a

significant amount of the variation in student reports of family involvement in language

arts homework. Unlike the previous analysis for science homework, TIPS science

students did not differ from ATIPS science students in their reports of family

involvement for language arts, a subject where interactive homework was not utilized.

No additional variance is explained by science condition (TIPS/ATIPS). The patterns for

math are similar to language arts in that none of the variables predict a significant portion

of the variance in math homework family involvement levels.

Table 3-2. Predictors of student reported levels of family involvement in language
arts homework (N=226)

Variable Model 1 Model 2
3 t 3 t

Previous English achievement .09 0.85 .11 -1.01
Mother education level
(reference- college graduate)
High school .12 1.34 .11 1.24
Some college .02 0.24 .01 0.12
Graduate school .09 1.01 -.09 1.03
Class ability grouping
(reference- Average)
Low .15 1.50 .09 0.76
Honors -.11 1.18 -.16 1.47
(reference- White)
Black -.02 -0.21 .03 -0.33
Other .05 0.59 .04 0.47
(Male = 1, Female = 0) .08 0.97 .08 0.99
(Grade 6 = 1, Grade 8 = 0) .15 1.77 .15 1.73

(TIPS = 1, ATIPS = 0) .09 0.95

R2/Adjusted R2 .18/.13 .19/.13

Note. AR = .005 for Model 2 (p= .901, NS). *2<.05. **p<.01

In summary, results of the family involvement analyses indicate that TIPS students

reported higher levels of family involvement in science homework than did ATIPS

students. Because these students and families received explicit information for family

involvement on science homework through a letter and instructions on each TIPS

assignment, it is not surprising that the results differ by condition. Math and language

arts family involvement levels did not differ by science condition. Therefore, one can

conclude that the TIPS condition and instructions for family involvement in science

elicited more family involvement in science homework, but not math or language arts

homework. This alerts researchers and educators to the importance of subject-specific

instructions for family involvement. The next analyses examined the effects of family

involvement in science homework on homework completion and accuracy.

Homework Completion and Accuracy

The second main research question investigated differences in the return rates and

accuracy of students' science homework by condition, teacher, grade, and gender. Table

3-3 presents the means and standard deviations of percent of assignments returned and

the average points earned, by homework condition (TIPS or ATIPS), gender, grade (6 or

8), and teacher (Teacher A, B, C, or D).

It was hypothesized that students in the TIPS condition would complete more and

turn in more accurate assignments than students in the ATIPS condition because of the

involvement of their families. In contrast to the hypothesis, the means by condition do

not reveal an advantage for the TIPS students in terms of turning in more complete and

accurate homework assignments. Table 3-3 shows that students in the ATIPS condition

(75.9%) turned in slightly more of the weekly assignments than did TIPS students

(71.8%) (with about the same average points earned per activity).

Developmental differences emerged with eighth grade students turning in slightly

more assignments than students in sixth grade. Results also differed by gender with

female students turning in more and earning more points for their homework assignments

than male students.

There were strong differences by teacher, with students taught by Teachers A and D

turning in more assignments than Teachers B and C. Teachers A and D made a point to

remind students more than once during the week of the importance of completing the

science homework. As a result of this instruction, more students in those classes

consistently returned and earned more points on their science homework assignments.

Additional investigation of the effects of the various background variables on

homework completion and average points per assignment required the use of OLS

regression analyses. Because the homework groups differed in terms of classroom ability

grouping and previous science achievement, it was necessary to control for these

differences to assess the true effects of the TIPS intervention on homework completion

and accuracy.

Table 3-4 displays four regression models. Model 1 includes background variables

and teacher effects and accounts for 32% of the variation in homework returned. Teacher

effects and previous science achievement accounted for a significant portion of the

variation in the percent of homework returned. Higher previous science achievement as

measured by the previous year's report card grades related to more assignments returned.

In addition, Teacher A's students returned significantly more assignments than any other

teacher's students. In fact, when the teacher dummy variables are entered separately,

they add significant predictive power to the model.

Table 3-3. Means, standard deviations, and number of students for homework
completion and accuracy by condition, grade, gender, and teacher over 18

Subsample Percent of assignment returned Average homework points


TIPS 132 71.8 26.3 6.2 2.6

ATIPS 101 75.9 23.4 6.3 2.2

Grade 6 128 72.6 27.1 6.1 2.6

Grade 8 105 74.8 22.6 6.4 2.2

Male 123 70.7 26.1 5.9 2.5

Female 110 76.9 23.6 6.6 2.4

Teacher A (6) 61 87.9 19.7 7.1 2.4

Teacher B (6) 67 58.7 25.3 5.2 2.5

Teacher C (8) 50 69.1 23.1 5.8 2.1

Teacher D (8) 55 80.0 20.9 6.9 2.2

Total 233 73.6 25.1 6.2 2.5

Note: Maximum number of homework points was 10.

Model 2 adds the effects of condition to the background variables and teacher

effects. Contrary to the hypothesis, TIPS and ATIPS students did not differ significantly

in the percent of homework assignments they returned, and condition failed to predict a

significant amount of variation in the model. Model 3 adds the student reports of family

involvement in science homework, which does account for a significant portion of the

variation. Students who more frequently involved family members in homework

returned more science assignments.

The most complete model, Model 4, includes all previous variables and adds the

effect of the student opinion of science homework. This measure is the average of a scale

of nine items on the student survey relating to student reports of how much they and their

parents liked the assignments, whether the assignments should be used next year, and

whether or not the student liked sharing science work with family partners at home

(a = .75). The more positive a student's opinion of the science homework, the more

assignments the student returned (P = .20, p < .01). Worthy of note is the fact that when

the student opinion rating is added to the equation, the predictive power of family

involvement is reduced (P = .139, p = .053).

Another homework variable of interest relates to average points earned per

assignment over the 18-week study period. Table 3-5 below includes the same predictors

used in Table 3-4.

Similar to homework assignments returned, condition failed to predict a significant

portion of the variation in average homework points earned per assignment. Unlike the

homework return model, working with a family partner on science homework (P3 = .156,

p < .03) remained significant in the model including student opinion of the science

homework (p = .212, p < .003). Worthy of note is the fact that teacher effects predicted

more variance in the previous two models than did students' grade in school.

Students in the low ability classes earned fewer points per assignment than did average

classes of students. Classroom ability level and previous grade in science were

significantly correlated (r = .365, p < .001, N = 239). However, being in an honors class

is even more strongly correlated with previous grade in science (r = .405, P < .001, N =

239) than being in a low class (r = -. 162, p < .05, N = 239). This strong relationship

between the honors class grouping and previous grade in science may explain the lack of

significance for the honors grouping (Honors) alone, as much of the variance in this

variable is explained by previous science grade.

The model also indicates that black students in the sample earned fewer points than

did white students. This was unexpected as previous science grade and class ability

grouping are controlled in the analysis. Therefore, one might expect there to be no

difference in accuracy of homework by race of student.

Tables 3-4 and 3-5 indicate a significant and positive relationship between family

involvement in science homework and homework completion and accuracy. Table 3-6

addresses the second homework completion research question about the relationship

between family involvement and homework completion by condition.

Looking at the correlations separately by condition, TIPS students who more often

involved family members turned in significantly more assignments (r = .29) and earned

more average points per assignment (r = .32) than did students not involving family

members as often.

Table 3-4. Predictors of Number of Homework Assignments Returned (N = 233)

Variable Model 1 Model 2 Model 3 Model 4
P t 13P t 13 t 3 t

Prior science achievement .29 3.64**
Mother education level
(reference- College graduate)
High school -.13 -1.80
Some college -.07 -1.02
Graduate school -.00 -0.03
Class ability grouping
(reference- Average)
Low -.14 -1.94
Honors -.13 -1.62

(reference- White)
(Male = 1, Female = 0)
(reference- Teacher A)
Teacher B (6)
Teacher D (8)
Teacher C (8)

(TIPS = 1, ATIPS = 0)

Student report of
family involvement

Student opinion of
science homework

R2/Adjusted R2

.29 3.61** .29 3.68** .29 3.13**

-.13 -1.79 -.14 -2.03* -.11 -1.63
-.08 -1.01 -.08 -1.01 -.07 -1.01
-.00 -0.03 -.00 -0.03 .01 0.18

-.14 -1.66 -.11 -1.37 -.10 -1.20
-.13 -1.37 -.13 -1.46 -.13 -1.47

-.11 -1.46 -.11 -1.45 -.10 -1.41 -.09 -1.31
-.07 -1.00 -.07 -0.99 -.08 -1.18 -.09 -1.32

-.08 -1.24 -.08 -1.24 -.08 -1.33 -.09 -1.47

-.54 -6.97** -.54 -6.94** -.54 -7.01** -.49 -6.41**
-.17 -2.21* -.17 -2.20* -.15 -1.88 -.06 -0.73
-.24 -2.78** -.24 -2.73** -.19 -2.18** -.16 -0.18

-.01 .99 -.08 -.98 -.05 -0.63

.19 2.68** .14 1.95

.20 3.01**



.39/.34 .43/.37

Note: AR = .000 for Model 2 (p = .991); AR2 = .027 for Model 3 (p = .008); AR2 = .032
for Model 4 (p = .003); *p < .05. **p < .01

Table 3-5. Predictors of accuracy of science homework assignments returned (N = 233)


Prior science achievement
Mother education level
(reference- College graduate)
High school
Some college
Graduate school
Class ability grouping
(reference- Average)
(reference- White)
(Male = 1, Female = 0)
(reference- Teacher A)
Teacher B (6)
Teacher C (8)
Teacher D (8)

(TIPS = 1, ATIPS = 0)

Model 1 Model 2 Model 3 Model 4
.32 4.03** .33 4.1t 4** t 424** 33 t4.34**

.32 4.03** .33 4.14** .33 4.24** .33 4.34**

-.13 -1.79
-.10 -1.36
.01 0.14


-.15 -2.12*
-.11 -1.47
-.01 0.12

-.12 -1.70
-.11 -1.49
.02 0.34

-.16 -2.27* -.20 -2.36* -.17 -2.05* -.15 -1.89
-.03 -0.43 -.07 -0.79 -.08-0.88 -.08 -0.88

-.18 -2.54* -.19 -2.59* -.19 -2.57* -.18 -2.50*
-.08 -1.15 -.08 -1.24 -.10 -1.46 -.10 -1.62

-.09 -1.36 -.08 -1.31 -.09 -1.42 -.10 -1.57

-.37 -4.74**
-.15 -1.74
-.09 -1.10

-4.76** -.36 -4.80**
-1.57 -.08 -0.97
-1.13 -.06 -0.76

-.31 -4.17**
-.05 -0.59
.03 0.41

.06 .81 -.03 -0.31 .01 0.07

Student report of
family involvement

Student opinion of
science homework

R2/Adjusted R2

.21 3.00** .16 2.24*

.21 3.19**

.38/.33 .38/.33 .41/.36


Note. AR2 = .003 for Model 2 (p = .652, NS); AR2 = .033 for Model 3 (p = .003);
AR2 = .035 for Model 4 (p = .002); *p <.05. **p <.01.

In the ATIPS group, no significant correlations exist between student reports of

involvement in science homework and homework completion or average points per

homework assignment. For both the ATIPS and TIPS conditions, higher opinions of the

science homework correlated significantly and positively with return of assignments and

average points per assignment. In the TIPS condition only, more positive homework

opinions correlated with higher levels of family involvement in science homework.

Table 3-6. Intercorrelations for homework completion, homework accuracy, family
involvement, and homework opinions by homework condition

Variable 1 2 3 4

TIPS (n = 132)

1. Homework completion -- .95** .29** .25**

2. Average homework points .95** -- .32** .24**

3. Family involvement .29** .32** -- .34**

4. Student homework opinions .25** .24** .34**

ATIPS (n= 101)

1. Homework completion -- .93** .01 35**

2. Average homework points .93** -- .00 34**

3. Family involvement .01 .00 -- .11

4. Student homework opinions .35** .34** .11 --

Note. *p<.05. **p<.01

Contrary to the hypothesis, TIPS students did not differ significantly from ATIPS

students in terms of the number of science homework assignments returned and the

average number of homework points earned. However, regression analyses indicated a

positive relationship between family involvement and student opinions of the

assignments on homework accuracy and completion. In fact, correlations of the TIPS

and ATIPS groups separately revealed an important distinction in the effects of these two

variables. TIPS students turned in more assignments when family involvement levels

were higher and their opinions of the science activities were more positive. Though

ATIPS students turned in more assignments when they rated them more positively,

family involvement showed no relation to homework completion or accuracy in this


This is interesting because both groups of students received the same science

questions on the assignments. ATIPS students who involved families without instruction

to do so did not receive the same benefit as TIPS students who involved families in

science homework in planned and productive interactions. This suggests the importance

of the letter to parents, instructions on the TIPS assignments for family involvement, and

the comments from teachers to students to involve family members in their homework


Science Achievement

The next research question addressed the impact of the TIPS intervention on student

science achievement. Three measures of student achievement were analyzed in the

study: 1) average science report card grades for two marking periods; 2) average percent

of homework test questions answered correctly; and 3) average test scores from tests

including the TIPS/ATIPS questions. Table 3-7 presents the means and standard

deviations of these three measures of science achievement by condition (TIPS or ATIPS).

For all three science achievement measures, TIPS students outperformed ATIPS

students. Further investigation of the relationship between family involvement and

science achievement remains vital, as these means do not take into account background

differences among students in the sample. Table 3-8 shows the regression analysis for

student science report card grades.

Table 3-7. Means and standard deviations for science achievement measures over 18
weeks by condition

n = 135 n = 105

1. Science report card grade 2.79 1.03 2.52 1.10

2. Homework test questions 78.07 14.68 72.50 14.82

3. Average science test score 78.65 13.00 75.78 11.19

Model I in Table 3-8 accounts for 50% of the variation in report card grades.

Higher previous science achievement related to higher report card grades in science.

Students with mothers having high school or some college education had lower report

card grades than students with mothers having a college degree. Race of student

predicted a significant portion of the variance with black students earning lower report

card grades than white students in the study. As with the homework completion models,

this finding is puzzling as previous science grade and class ability grouping are

controlled. Male students earned significantly higher report card grades than females.

Student reported levels of family involvement in science failed to predict a

significant portion of the variation in report card grades for the entire sample. When the

student report of family involvement is replaced with the student reported levels of parent

involvement, parent involvement does predict a significant portion of the variation in

Model 1 only (p = 132, i = .021). When percent of homework completed and condition

are added to the model, parent involvement fails to predict significant variation just as

family involvement does. Therefore, other variables in this model carry more of the

variance in predicting report card grades than either of the family involvement measures.

Another important science achievement research question asked how the percent of

assignments returned relates to science achievement. Model 2 demonstrates that students

who completed more of the science homework assignments earned higher report card

grades. This finding supports the fact that both the ATIPS and TIPS assignments had

clear purposes and objectives and both represented high quality homework assignments

that, when completed, should relate to higher report card grades. Studies of middle

school homework completion and achievement confirm this result and demonstrate the

positive relation between the two (Cooper, 1998; Keith et al., 1993).

Percent of homework assignments returned strongly related to report card grades

(p= .461, p< .001). After controlling for the background variables and percent of

homework assignment returned, students in TIPS still earned higher report card grades

than students in ATIPS. This finding is exciting because it suggests the importance of

not only completing homework, but also completing specifically the TIPS interactive


Table 3-8. Predictors of science report card grades (N = 240)

Variable Model I Model 2 Model 3
3 t P _t p

Prior science achievement
Mother education level
(reference- College graduate)
High school
Some college
Graduate school
Class ability grouping
(reference- Average)
(reference- white)
(reference- Teacher A)
Teacher B (6)
Teacher D (8)
Teacher C (8)
Student report of
family involvement

.39 5.72** .25 4.19** .26 4.39**

-.13 -2.18*
-.17 -2.62**
-.03 0.45

-.05 -0.73
-.00 -0.01

-.07 -1.27 -.07 -1.33
-.13 -2.36* -.15 -2.65**
-.03 -0.50 -.03 -0.57

.03 0.47 -.05 -0.80
.08 1.31 .00 0.06

-.22 -3.54** -.18 -3.22** -.18 -3.40**
-.11 -1.90 -.07 -1.43 -.08 -1.64

-.09 1.65

.12 1.85
-.36 -5.31**
-.12 -1.58
.07 1.15

Percent of homework

R2/Adjusted R2

.13 2.69** .14 2.89**

.37 5.65** .37 5.75**
-.29 -4.95** -.30 -5.18**
-.04 -0.56 -.02 -0.34
-.01 -0.17 -.06 -1.04

.45 7.75** .46 8.02**

.15 2.41*




Note. AR2 =.12 Model 2 (p= .00); AR2 = .01 Model 3 (p= .02); *p <.05. **p <.01.

Examination of the means by condition revealed the fact that TIPS students earned

higher report card grades, answered more TIPS/ATIPS test questions correctly, and

earned higher science test scores than ATIPS students. Even after controlling for

background measures, teacher effects, and percent of homework completed, TIPS

students still earned significantly higher report card grades than ATIPS students in

regression analyses. This model accounted for 65% of the variance in report card grades.

When these same variables were controlled for average test score in science, condition

failed to predict a significant portion of the variance above and beyond percent of

homework returned. Condition predicted a significant portion of the variance in percent

of TIPS/ATIPS test questions answered correctly, but the model only explained 7% of the

variance in the test questions.

Results from the first research question revealed the fact that TIPS students more

often involved family members in homework assignments than did ATIPS students (r =

.427, p < .001, N = 226). However, the regression analyses failed to demonstrate family

involvement as a significant predictor in this analysis. Therefore, it is unclear from this

study what it is about the TIPS condition that related to higher report card grades for

students. The analyses below explore the nature of the involvement relationship for

students and parents in both homework conditions to search for clues in the science

achievement patterns.

Concordance and Discrepancy in Parent and Student Reports

Family and student survey questions were worded similarly to allow comparison of

student and family partner answers to the same question. In most cases, the parent was

the one to complete the family survey (95%); therefore, the respondents to this survey

will be referred to as parents from this point on. The similarly worded survey questions

permitted exploration of how the degree of agreement and positive reports from parents

and students related to some of the dependent measures of interest, including measures of

homework completion.

Table 3-9 displays the means and standard deviations for the matched sample of

students and parents who responded to the survey questions. These variables included

questions about time spent on homework, family involvement in homework, and opinions

of the science homework.

Family involvement questions were coded on a scale of 0-4 with 0 indicating

"never" involved, I indicating "rarely" involved, 2 meaning "sometimes" involved, 3

meaning "frequently" involved, and 4 indicating "always" involved in homework. The

science homework opinion questions were coded on a scale of 1-4 with 4 indicating

strong agreement, 3 indicating agreement, 2 indicating disagreement, and I indicating

strong disagreement. Time questions were coded on different scales ranging from 0-5,

but higher numbers always indicate more time spent on homework. For example, time

spent on homework for all subjects was coded as 0 for 0 minutes, I for less than 30

minutes, 2 for 30 minutes to 1 hour, 3 for 1-2 hours, 4 for 2-3 hours, 5 for 3-4 hours, and

6 for more than 4 hours.

In 14 of the 19 cases, parents had more positive reports, on average, than did

students. Family members completing the surveys (74% mothers, 21% fathers, 5%

grandparents, neighbors, or other) tended to be more positive about science homework,

tended to estimate that students spent more time on homework, and rated their

involvement in homework for English, science, and math higher than students did. For

two of the 18 cases, the student and parent means did not differ. Their mean responses

for students liking to share their work with someone else at home other than a parent

were the same. Similarly, parents and students did not differ in their ratings of how often

parents took part in science homework. In three cases, parents' reports were lower than

students' reports. Parents reported spending slightly less time on science homework

than students reported that they did. A larger mean difference in parent and student

reports occurred when parents were asked if they needed more information from the

school about the green science homework sheets. Parents said they needed more

information than students reported that their parents needed. Finally, parents did not

report as high as students that students work as hard as they can in science.

A scale composed of nine of the survey items was constructed to examine the

relationships between student (x = .76) and parent (ax = .79) opinions concerning the

science homework assignments. These questions asked students and parents to rate their

liking of the TIPS/ATIPS science assignments, to record whether or not the homework

took too much student or family time, whether or not these should be used next year, and

several other questions.

The scale included the variables with a star (*) in the table above. A score of I

indicated strong disagreement, 2 indicated disagreement, 3 indicated agreement, and 4

indicated strong agreement. Students reported a mean of 3.11 (SD = .54) for these 9

items, while parents reported a mean of 3.36 (SD = .59). A significant and positive

correlation (r = .437, n = 164, p < .001) exists between the student and parent reports.

Table 3-9. Matched sample of student and parent means for survey questions

Construct/ Variable Description Student Parent
N = 175 N= 175

Time on homework for all subjects 3.30 1.54 4.39 1.61
Time on science homework only 1.68 0.76 2.09 0.92
Time on one science homework assignment 1.92 0.82 2.29 0.91
Family partner time on one science assignment 1.87 0.96 1.83 1.01

Family involvement in science homework 2.15 1.30 2.36 1.23
Parent involvement in science homework 2.28 1.63 2.27 1.61
Family involvement in English homework 1.14 1.14 1.49 1.16
Family involvement in math homework 1.55 1.25 1.64 1.14

These science assignments are a good idea.* 3.23 0.85 3.41 1.05
My family partner liked the science homework.* 2.85 0.94 3.16 1.01
Students should use these next year.* 3.20 0.98 3.28 1.18
The science homework did not take too much
student time.* 2.90 0.98 3.18 1.15
The science homework did not take too much of
family partner's time.* 3.01 0.97 3.29 1.05
These assignments help parent see what student
is learning in science class.* 3.46 0.83 3.54 0.90
Parent does not need more information about the
science homework.* 3.11 1.00 2.85 1.23
Parent likes to hear what student is learning in
science class.* 3.44 0.74 3.89 0.42
Student works as hard as s/he can in science class. 3.46 0.71 3.19 1.12
Student likes to talk about science with family.* 3.28 0.88 3.56 0.77
Student likes to share science work with someone 2.98 1.06 2.98 1.25
at home.

*Questions that comprised the scale of opinions of the science homework.

The student and parent scale means were used to construct categories for student and

parent reports of the science assignments. When the student rated homework on the low

end of the scale (below the mean for students) and his/her parent rated the homework on

the low end of the scale (below the mean for parents), a 0 was assigned to that student.

When a student rated the homework low and his/her parent rated it high, a 1 was

assigned. In the cases where student rated high and the parent rated low, a 2 was

recorded, and when both student and parent rated homework more positively, a 3 was


Table 3-10 displays the means for three science achievement variables including

report card grades, TIPS/ATIPS test questions, and average test score and for two

homework completion variables including percent of homework returned and average

points earned per assignment. The highest percentage of students (44%) are included in

the student high/parent high category, and the remaining three categories include between

15 and 20% of the students in this reduced sample.

The means for high student and parent opinions of the science assignments are

higher in all cases than the low student and low parent reports. In addition, the means for

student high and parent low are consistently higher than the student low and parent low

reports. Inconsistent relationships between the discrepancy patterns emerged with two

cases of the student high/parent low mean being slightly smaller than the student low/

parent high mean (report card grades and TIPS/ATIPS test questions). For both

homework completion variables, the student low/ parent high report (71.20, 5.99) fell

below the low student and low parent reports (79.03, 6.74).

Table 3-10. Means and standard deviations for student science achievement by parent
and student science homework categories

Homework Opinion Similarity Index Category

Student Low Student Low Student High Student High Total
Parent Low Parent High Parent Low Parent High
Variable n = 32 n = 25 n = 32 n = 69 N= 158

Science report 2.62 2.78 2.75 3.12 2.87
card grade (0.95) (1.14) (0.94) (0.94) (0.99)

Average science 77.78 77.97 78.07 80.68 79.13
test score (12.51) (11.85) (10.45) (11.85) (11.67)

Percent of TIPS/
ATIPS test 70.05 78.74 78.59 76.43 75.93
questions correct (13.16) (14.23) (12.47) (14.74) (14.11)

Percent of homework 79.03 71.20 82.25 88.03 82.47
returned (22.71) (22.31) (17.70) (17.78) (20.25)

Average points
earned per 6.74 5.99 6.77 7.71 7.06
homework (2.35) (2.19) (1.88) (1.82) (2.08)

Note: Maximum science report card grade = 4; Maximum average number of points
earned = 10.

The final research question examined the relationship between these student and

parent categories of enthusiasm for science homework and specific dependent variables

of interest. The following tables (Tables 3-11 and 3-12) include preliminary evidence

suggesting that these categories including both student and parent data explain more

variance in average points earned per assignment and percent of homework returned than

student reports alone. These tables are different from the previous table on homework

accuracy (Table 3-5) in that they include fewer background variables (due to the reduced

sample). Table 3-5 also did not include any parent reports as the following tables do.

Model 1 of Table 3-11 includes various background measures affecting the average

number of homework points students earned. As in many of the regression analyses, the

higher the student's previous science grade, the more science homework points he or she

earned. Students in the low classroom ability grouping did not earn fewer homework

points than did average students, but black students earned significantly fewer points than

white students. Male students earned fewer points than females. Model 2 adds the effect

of the student rating of science homework finding that students who rated the homework

more positively tended to earn more points on science homework.

Contrast Table 3-11 with Table 3-12. Model 1 of Table 3-12 includes the same

background variables used in Table 3-11. Model 2 of Table 3-12 includes the parent and

student categories of ratings of the homework instead of the student information alone.

The addition of the parent information predicts slightly more variance in average points

earned. The adjusted R2 for the model with the student opinion only accounts for 34% of

the variance in homework points while the model including the student parent categories

accounts for 36% of the variance.

Similar results emerged for the percent of homework returned. Student opinion only

accounted for 20% of the variance in percent of homework returned while the student and

parent categories accounted for 22% of the variance.

One may wonder what results emerge when the student opinion of the homework is

included in a model along with the parent opinion of the homework. This differs from

the previous model in that this does not take into account the "fit" between the parent and

student opinions for each subject. The variance explained with these two variables is

very close to the values computed for the parent and student categories (37% for average

points and 21% for percent of homework returned). Therefore, these analyses underscore

the importance of including both the family and student opinion in the models.

The unstandardized beta coefficients (B) provided the values to compute the

predicted average points earned by the same hypothetical student. For example, imagine

that we wanted to know the effect of changing the student or the student/parent category

for opinions of science homework on a hypothetical student. Imagine a sixth grade white

female student named Isabel in the average class ability group with a 3.0 grade point

average in science last year.

Y= 3.061 + .764 (3) .671 (0) (.00) (0) (1.14) (0) (.338) (0) 1.00 (0)

-.451(1) + .954 (3.5) = 8.241

According to Table 3-11, if Isabel reported high enthusiasm for the science homework,

she would earn 8.24/10 average points on homework assignments. Using the

unstandardized beta coefficients from Table 3-12, one computes 8.67 average points

earned when Isabel and her parent report high enthusiasm for science homework. In

contrast, when Isabel's report is high and her parent reports low enthusiasm for the

science homework, Isabel earns 7.65 average points on homework.

Figure 3-3 displays the predicted average points earned for Isabel when her

enthusiasm for science is high. There is a notable advantage to Isabel when her parent

also reports enthusiasm for the science homework. When her parent reports a more

negative view of the science homework, Isabel earns fewer points than if just her report

of the homework is included.

Table 3-11. Predictors of average points earned per assignment including only the
student rating of science homework (N = 153)

Variable Model 1 Model 2
B p t B p t

(Constant) 5.92 3.06
Prior grade in science 0.75 .31 3.74** 0.76 .32 3.96**
Class ability grouping
(reference- Average)
Low -0.93 -.14 -1.85 -0.67 -.10 -1.36
Honors -0.00 -.01 -0.09 -0.00 -.01 -0.14
(reference- White)
Black -1.10 -.25 -3.34** -1.14 -.26 -3.61**
Other -0.23 -.03 -0.45 -0.34 -.05 -0.68
(Male = 1, Female = 0) -0.99 -.24 -3.42** -1.00 -.24 -3.57**
(Grade 6 = 1, Grade 8 = 0) -0.13 -.03 -0.39 -0.45 -.11 -1.37

Student opinion of
science homework 0.95 .25 3.56**

R2/Adjusted R2 .32/.29 .37/.34

Note. AR2 =.055 for Model 2 (<.001). *p<.05. **p<.01.

Table 3-12. Predictors of average points earned per assignment including the
student/parent category ratings for science homework (N = 153)

Variable Model 1 Model 2
B t B t

(Constant) 5.92 6.95
Prior parade in science 0.75 .31 3.74** 0.78 .32 4.09**

Class ability grouping
(reference- Average)
(reference- White)
(Male = 1, Female = 0)

(Grade 6 = 1, Grade 8 = 0)
Parent/student category
(reference- Student High/
Parent High)
Student High/Parent Low
Student Low/ Parent High
Student Low/ Parent Low

R2/Adjusted R2

-0.93 -.14 -1.85 -0.75 -.11 -1.54
-0.00 -.01 -0.09 -0.21 -.05 -0.63

-1.10 -.25 -3.34**
-0.23 -.03 -0.45

-0.99 -.24 -3.42**

-0.13 -.03 -0.39

-1.19 -.27 -3.69**
-0.46 -.06 -0.92

-0.99 -.24 -3.55**

-0.61 -.15 -1.83

-1.03 -.20 -2.78**
-1.36 -.24 -3.34**
-1.45 -.28 -3.62**

.32/.29 .40/.36

Note. AR2 = .084 for Model 2 (p <.001). *p<.05. **p<.01.

The pattern when Isabel reports a low rating of science homework also deserves

attention. Figure 3-4 includes the predicted average points earned when she reports low

enthusiasm for science homework assignments. Also included in the graph are the points

earned for the two possible student/parent categories. The predicted points when Isabel

reports low and her parent reports high are slightly higher than when only Isabel's low


report is included in the model. As with high student reports, low reports by the parent

results in fewer predicted points than when just the student report is included.

9- 8.677





Student Only Student High- Student High-
High Rating Parent High Parent Low

Figure 3-3. Predicted average homework points earned for high student ratings of
science homework by the same hypothetical student

These regression analyses and graphs alert researchers to the importance of

exploring both parent and student responses when the two parties interact. Knowing the

student opinion is certainly most important, as homework is a student's responsibility to

complete. However, the nature of interactive homework requires parent or family

involvement and these opinions add variance to the regression model that may positively

or negatively influence student outcomes such as homework completion and accuracy.




7.5- .2 7.224


Student Only Student Low- Student Low-
Low Rating Parent High Parent Low

Figure 3-4. Predicted average homework points earned for low student ratings of
science homework by the same hypothetical student

Exploratory analyses of the reports of students and parents to various survey

questions revealed the tendency for parents to report higher or more positively than

students. Parent and student matched reports were significantly and positively correlated.

To summarize the results of the concordance results, regression analyses calculated

the additional variance parent and student categories explained beyond student opinions

of homework. Graphs of predicted homework points revealed an important pattern for

the same hypothetical student. Students may earn more points on homework if they rate

homework enthusiastically. A positive parent report in conjunction with a positive

student report relates to higher points earned than when just the student report is


measured. A negative parent report paired with a positive student report indicates a

decrease of average points earned.

When the student report of homework is low, the parent report of the homework

may relate to slightly higher points earned than when just the student report is included.

These findings are to be taken with caution as 1) they do not imply causation; 2) they do

not represent the entire sample of students in the study; 3) the model accounts for 36% of

the variance so much of the variance is still not explained by these models. Given these

limitations, future research should investigate these student and parent patterns to explore

the relationships between positive homework opinions from both the student and family

and student outcomes. Teachers, students, parents, and researchers would benefit from

deeper knowledge of how these opinions from both student and parent might predict or

"boost" completion or accuracy of homework.


Previous research on homework paid insufficient attention to the roles of homework

design and family interaction in homework. This study explored the effects of interactive

and non-interactive well-designed science homework assignments on family

involvement, student science achievement, and student science attitudes. The

combination of a longitudinal and developmental design involving multiple teachers and

varying ability levels of students speaks to real-life variations in middle schools

nationwide. This discussion highlights the major findings and limitations of this study,

describes "tips" for future research that would help to support and clarify the current

findings, and summarizes strategies for brightening the future of homework.

Major Findings

TIPS Increased Family Involvement in Science Homework

The TIPS intervention was successful in promoting higher levels of family

involvement in science homework than the ATIPS condition. One TIPS student wrote

the following comment on the survey: "I think these sheets (TIPS assignments) were a

very good idea because they help my family partner know what I'm doing. I think it is a

good practice for tests."

This result supports findings from Balli's (1995) experimental study of middle grade

students using TIPS math. The higher level of family involvement in science, however,

did not carry over to other subject areas such as language arts and math (subjects not

utilizing TIPS homework assignments). The teachers of these subjects did not use TIPS

homework assignments. Thus, this alerts researchers and educators to the importance of

implementing subject-specific involvement strategies to increase subject-specific results

(Epstein, 1991; Epstein, in press; Epstein & Van Voorhis, in press).

Most parents recognize the importance of reading to children at home as an

appropriate form of involvement to promote reading achievement, not math achievement.

Parents do not often know how to support their children in subjects other than reading

without explicit information on what to do. The TIPS design provides the student with

activities to involve family partners and instructions for guiding interactions so that

parents do not need to figure how they may be involved in their children's homework in

appropriate ways. Several parents commented on the survey about their appreciation of

the TIPS homework assignments as a means of working together: "They (the TIPS

assignments) are a great way for us to work together and keep informed of what is going

on at school." Another parent remarked: "I think it is a good idea for the kids to be able

to share and to demonstrate for their parents what they are learning in school."

Another important aspect of this finding is the fact that parental education as

measured by mother's education did not predict a significant amount of the variation in

involvement levels. This TIPS science homework design, therefore, provided all

families in the study, not just those in the middle or upper classes, with an opportunity to

be involved in their children's education. Similar findings emerged in a study of TIPS

writing in the middle grades (Epstein et al., 1997).

Family Involvement Relates to Homework Completion

Though it was hypothesized that TIPS students would turn in more assignments and

more accurate work than ATIPS students because of the family interaction, the results

failed to demonstrate significant differences by condition in homework completion or

accuracy. This may be because both the ATIPS and TIPS assignments contained

identical content and questions, except for guided family involvement. Teachers focused

their energies on developing high quality, purposeful homework assignments linked to

their science curriculum indicators. The difference between the two conditions related to

the instructions and prompts for family involvement in the TIPS condition only. All

students received a maximum of 10 points per assignment, and teachers tested the

information from homework on science exams. Therefore, all students had incentives for

completing the assignments, and condition (TIPS vs. ATIPS) did not explain whether

students did their homework.

Although there were no differences in return rates of TIPS and ATIPS students,

higher levels of family involvement and more positive student opinions of science

homework did relate to more average points earned per homework assignment. These

findings point to important interrelations among student science attitudes, homework

completion, and family involvement behaviors. Designing assignments that promote

more positive attitudes about science and that thoughtfully involve family partners is an

-4 important goal of interactive homework assignments (or any type of homework) that may

relate to positive student achievement outcomes. Twenty-three of the 98 students who

commented about the science assignments on the survey in both the ATIPS and TIPS

groups remarked about their enjoyment of the assignments with experiments. One

student commented "I think the green sheets are fun if there is an experiment. The

experiment makes learning fun!" Paying attention to these student comments and

providing more hands-on learning for students is important in the design of future

homework assignments for students at this age (Paris, Yambor, & Packard, 1998).

TIPS Students Earned Higher Report Card Grades than ATIPS Students

After controlling for student background characteristics including prior report card

grade in science, teacher effects, and percent of homework completed, TIPS students

earned higher science report card grades than ATIPS students. This finding suggests an

important link of the TIPS interactive homework with student success in school. This

regression analysis represents the strongest model of the study, accounting for 65% of the

variance in science report card grades.

Less clear are the specific aspects of TIPS that relate to higher achievement.

Because there were no significant differences in homework return rates or homework

accuracy of TIPS and ATIPS students, it is not possible to conclude that the science

achievement difference by condition is due to TIPS students completing more or more

accurate homework assignments. Also, the regression analyses do not suggest that the

TIPS advantage is due to the higher levels of involvement of the TIPS students. Deeper

investigations of the TIPS process at home and its effects on other variables like school

attendance, student perceptions of their teachers and science in general, and ratings of the

home interactions may help elucidate the processes of TIPS that contribute to higher

science grades.

The analyses of the predictors of science report card grades also emphasize the

importance of homework completion to science report card grades (3 = .46, p < .001).

Analysis of variance (ANOVA) omnibus tests indicated significant differences in the

homework return rates and average points earned by students in different class ability

levels ( F (2, 232) = 5.61, p < .005; F (2, 232) = 12.45, p < .001). Tukey HSD post-hoc

analyses determined that low students (60.1%) turned in significantly fewer assignments

than average (74.5%) or honors (76.7%) students, but average and honors did not differ

significantly in their return rates. In terms of homework accuracy, all groups differed

significantly from each other. Specifically, low students (4.5/10) earned the least average

points; average students (6.0/10) earned significantly more points than low students but

fewer points than honors; and honors (6.8/10) students earned significantly more points

than average and low students.

Basically, these results indicate the fact that some students in all classes are not

completing homework assignments. Contrast this study with Balli's (1995) investigation

of 74 sixth grade students of the same family background completing 100% of their

homework assignments.

The message here is clear; homework completion positively influences middle

school student achievement. Teachers, parents, and students must understand the

importance of students completing their homework assignments. When middle and high

schools students do not spend the time they need to complete homework, they are at an

academic disadvantage compared to students who do (Keith & Cool, 1992; Keith et al.,

*4 1993). Therefore, teachers and parents must encourage all students to complete their

homework assignments in an effort to narrow achievement gaps by class ability


Results Often Differed Developmentally and by Gender

Interesting developmental and gender differences emerged in this study. Sixth grade

students reported significantly higher family involvement levels and earned higher report

card grades in science than eighth grade students, but eighth grade students returned more

homework assignments. One might predict these developmental differences as sixth

grade students just finished elementary school in which parental involvement in

schoolwork is customary. Therefore, involving family members in homework

assignments was not new for these students whereas for the eighth grade students it was.

Similarly, as children grow, grading standards tend to become stricter which may account

for sixth grade students earning higher report card grades.

Though males and females did not differ significantly in levels of family

involvement in science or percent of homework assignments returned, male students

earned significantly higher report card grades than females. The gender differences

confirm results from previous studies alerting researchers to male advantages in math and

science by high school (Hyde, Fennema, & Lamon, 1990; National Science Foundation,

1993), and sometimes by the time students reach middle school (Lee & Burkam, 1995).

Continued longitudinal research on these interactive science activities, perhaps over the

entire school year, may or may not uncover clues to reducing the gender gap in science

achievement. Researchers should continue to investigate the factors that influence the

<4 gender difference to determine how changes in classroom teaching and practice as well as

interactive homework assignments may possibly reduce inequalities in achievement

across the genders.

Parent and Student Categories Provide New Insights into Achievement Results

This study explored the relationship between parent and student opinions and

student opinions alone. Homework may be designed for many different purposes, but in

the case of TIPS, homework is designed for interaction with a family partner. Given this

purpose, it is important to investigate both student and family partner opinions of the

interactive homework assignments.

Results indicate an interesting relationship. Students who rated the homework

positively completed more homework. The matched student and parent patterns revealed

an added bonus to those students whose parents also rated the homework enthusiastically.

Again, these findings underscore the importance of reexamining homework practice and

design. Teachers and parents must change their homework focus solely from time, to

homework quality and interest.


This study expanded current homework research in five important ways. This study

1) implemented a quasi-experimental design assigning half of the students to a non-

interactive (ATIPS) homework condition and the other half to an interactive (TIPS)

condition; 2) looked at the effects developmentally (sixth and eighth grade students); 3)

examined results with different ability levels of students and different teachers; 4)

collected various science achievement, family involvement, homework completion, and

science attitude measures, and 5) examined a new homework subject area, science.

Despite these strengths, future studies would benefit by examining the limitations of the

present study.

Length of the Study

The first limitation relates to the length of the study period. Three studies of TIPS

have been conducted: TIPS writing lasted one school year (Epstein et al., 1997); TIPS

math lasted two months (Balli, 1995); and the current TIPS science study fell between the

two with four months of observation. Some of the changes one might expect from a

homework intervention like TIPS might include changes in parent and student attitudes

about homework, changes in teachers' attitudes about homework and parental

involvement, and achievement gains as a result of completing the homework on a regular

basis. Two months and even four months is a short amount of time to expect a change in

attitude or achievement gains. Monitoring the homework over an entire school year

would permit analyses of results from last year to the current year. Obviously, the reason

this is rarely done is due to the cost in terms of time for researchers and educators to

coordinate the measures needed for such a comprehensive investigation.

Need for Standardized Achievement Scores

This study was also limited as standardized achievement tests in science were not

available at the end of the study period. These scores are important as they are not

influenced by individual teacher bias. Having these scores would permit researchers to

determine whether TIPS students also experienced achievement gains as measured by

standardized tests. If this proved true, one could conclude that the differences in TIPS

and ATIPS report card grades were not solely due to teacher bias in favor of TIPS

students. In this study, remember that the author assigned condition to entire classes of

students. Some honors classes completed TIPS while other honors classes completed

ATIPS. This assignment certainly invites less teacher bias than if condition were

assigned by student. In this case, teachers might teach a class of both ATIPS and TIPS

students in which bias would certainly come into play. Analysis of this question of

teacher bias using standardized test scores does indeed warrant investigation to be sure

this was not the sole cause of the TIPS achievement advantage.

Lack of Teacher Implementation Measures

Another study limitation relates to its lack of any specific measures of teacher

implementation, namely homework introduction and follow-up. Though the author

carefully instructed the teachers to keep introduction and follow-up of homework

consistent in terms of time across condition, teachers varied in how that time was utilized.

Introduction of the homework might include explaining the various questions and

materials needed for the assignment, and for TIPS students only, pointing out the sections

requiring family interaction. Follow-up includes grading (teachers graded each

assignment on a 10-point scale) and teacher comments, review of homework answers in

class, and talking with students and families about the importance of family involvement

if home-to-school communications were not completed.

Teachers did vary in their comments to students (and even to parents) about errors

and responses in the home-to-school communication section. For example, one of the

sixth grade teachers took time to acknowledge all parent comments and reports by

writing thoughtful responses to their messages on the homework assignment. For

example, one parent noted the following: "These experiments are more fun to do than

when I was in school, I may change careers." The teacher wrote the following statement

in response: "I am glad to hear that both you and your child enjoy these assignments. I

think they are fun myself!" One parent noted that the way a graph was presented seemed

a bit confusing. The teacher acknowledged the comment and wrote: "Thanks for paying

such close attention! I will change the assignment before I send it home next year."

Such teacher feedback to parents and students could positively influence both parties'

feelings about science homework and class.

Experimental studies of teacher feedback indicate that students of teachers who

provide constructive written feedback on homework had higher achievement and more

positive attitudes about the subject than students of teachers who did not provide such

feedback (Cardelle & Comrno, 1981; Elawar & Como, 1985). An interesting study of

feedback and interactive homework might examine the effects of variations in teacher

feedback on both parent and student homework attitudes and how this relates to student


Because the author conducted no classroom observations of teacher instruction,

there are no measures of possible teacher variation in implementation. Results revealed

strong differences by teacher with respect to student achievement. Therefore, future

studies must identify and investigate teacher differences in classroom homework

introduction and follow up to see what characteristics, if any, promote or hinder student

learning and likelihood of returning homework assignments.

Lack of Measures for the Quality of Homework Interaction

Another area of future study represents more detailed investigation of the use of

TIPS assignments in the home. The matched parent and student survey questions

permitted exploratory analyses of the student/parent interaction in both the TIPS and

ATIPS conditions. Though both parents and students reported on their individual liking

of the homework, no survey questions specifically addressed the content of or emotions

about working together on homework. One may think of numerous questions worthy of

exploration including the following: Was it enjoyable to work together? What was the

role of the student in the interactive homework? What was the role of the parent in the

assignment? Did the student feel the parent helped too much? Was there tension

associated with working together? Future studies of family interaction on homework

assignments should examine the findings from the few studies that have addressed the

circumstances in which adolescents do homework and how these relate to academic

performance (Leone & Richards, 1989; McDermott, Goldman, & Varenne, 1984; Patton,

Stinard, & Routh, 1983).

Lack of Homework Adaptation by Class Ability Grouping

An important note to both researchers and educators relates to adaptation of the

homework assignments to match the ability levels of students. In this study, all students

completed the same homework assignments for the purpose of keeping content of

assignments constant for all students in both conditions. Results indicated that students

in the low ability classes earned significantly fewer homework points than average

students. The sixth grade teachers who taught low ability students remarked that some of

the homework assignments proved a bit too challenging for many of these students.

Therefore, in practice, teachers should adapt homework assignments according to the

ability levels of the students by either deleting some of the more challenging questions,

providing more clues to an answer, or adding more challenging questions to promote

critical thinking. Homework that is too challenging or too easy can frustrate or bore

students rather than encourage learning and interest in a topic (Corno, 2000).

The Future of Homework

Numerous popular press books have been written to help parents survive the years

of homework their children bring home (Cholden, Friedman, & Tiersky, 1998; Radencich

& Shay Schumm, 1997). The titles indicate the frustration of both parents and students:

99 Ways to Get Your Kids to Do Their Homework (And Not Hate It)., How to Do

Homework Without Throwing Up, Ending the Homework Hassle : Understanding,

Preventing, and Solving School Performance Problems, and Homework Without Tears:

A Parent's Guide for Motivating Children to Do Homework and to Succeed in School.

Homework should never be assigned for the sake of homework. All homework

should be designed with a clear purpose in mind. Epstein identified 10 common reasons

for assigning homework: practice, preparation, participation, personal development, peer

interactions, parent-child relations, parent-teacher communication, public relations,

policy, and punishment (not a defensible purpose) (1998, 2000; Epstein & Van Voorhis,

in press).

Results of this study show that well-designed TIPS homework assignments in

science can help students practice skills, prepare for the next class, participate in learning

activities, develop personal jepo.riilit., t.i homework, promote parent-child relations,

develop parent-teacher communication, and fulfill policy directives from administrators.

Survey reports from students indicated that students liked both the ATIPS and TIPS

homework assignments, rated them better than standard homework, and suggested that

they be used next year in school. Hundreds of prototype TIPS activities (including the

science activities used in this study) in various subjects for middle and elementary grade

students are available to teachers to use and adapt to their classroom objectives (Center

on School, Family, and Community Partnerships, Baltimore, MD). In fact, researchers at

Johns Hopkins are currently collecting the "top" TIPS activities developed by educators

nationwide to share with schools in their efforts to improve homework practice.

Though the results of the TIPS program in science are positive, teachers should not

abandon other types of homework. Again, homework designs serve different purposes,

and independent homework still holds an important place on the educational agenda.

Utilizing a variety of homework designs with clearly defined and different purposes

would probably encourage more interest in the nightly ritual rather than promote

boredom and frustration.

Continued research in this area requires strong connections and communications

between researchers and educators. This particular study involved collaborations during

the summer to design and produce the science homework assignments, and through the

fall semester to monitor homework completion and collect various science achievement

and attitude data. University and school partnerships allow researchers to share the

findings of their work with educators who can use the results to improve teaching

materials and practice. Teachers must have professional development opportunities that

encourage them to reexamine their current homework practices to make improvements.

Teachers already understand the importance of quality classroom instruction. Homework

deserves the same level of attention as classwork, especially because homework enters a

student's home life.

Learning does not take place in a vacuum and research must investigate the effects

of educational interventions to understand who benefits and why. Three assumptions

should guide continued investigations in the area of parental involvement in homework

(Renninger, 1999, p. 212): instructional decisions must be based on research of how

students learn and develop; educators must be aware of developmental theory and

research; researchers must become more familiar with educational practice to consider

how their research may contribute more directly to educational practice.

Much of school homework today is monotonous, pointless, discouraging to students

and parents, and disruptive of family time. Time must be allocated for teachers to learn

more about the importance of homework to student achievement as well as strategies for

developing meaningful homework assignments to match the creativity found in their

classrooms. TIPS homework is one tool for teachers to use in their teaching repertoire to

develop students' skills and to inform parents of what is going on in the classroom. The

focus of homework must shift from an emphasis on homework time, to an emphasis on

homework quality and homework design. As noted by one parent, I think TIPS is great

3-way communication between teacher, parent, and student." Homework is here to stay,

but brightening the future of homework will require a concerted effort on the part of

teachers to reevaluate its varied purposes and design, emphasize its importance to

students, and communicate to parents the roles they play in the homework process.

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd