Causal explanations for mathematics performance given by low socioeconomic status African American mothers and their children

MISSING IMAGE

Material Information

Title:
Causal explanations for mathematics performance given by low socioeconomic status African American mothers and their children
Physical Description:
xiii, 123 leaves : ; 29 cm.
Language:
English
Creator:
Cyrus, Kenneth R., 1956-
Publication Date:

Subjects

Subjects / Keywords:
Foundations of Education thesis, Ph. D
Dissertations, Academic -- Foundations of Education -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1996.
Bibliography:
Includes bibliographical references (leaves 117-122).
Statement of Responsibility:
by Kenneth R. Cyrus.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 022852523
oclc - 34950637
System ID:
AA00014227:00001


This item is only available as the following downloads:


Full Text









CAUSAL EXPLANATIONS FOR MATHEMATICS PERFORMANCE
GIVEN BY LOW SOCIOECONOMIC STATUS AFRICAN AMERICAN MOTHERS
AND THEIR CHILDREN















By
KENNETH R. CYRUS













A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA








Copyright 1996

by

Kenneth R. Cyrus












To my grandparents,

Edward and Elizabeth Cyrus,
and
Anderson and Camilla Catchings,

to my parents,

Frank and Ethel Cyrus,

to my daughter,

Yvette Moran Cyrus,

and to the rest of the immediate and extended

Cyrus family













ACKNOWLEDGMENTS


I would first like to thank my doctoral committee for their support of my efforts to

complete this formidable task. My chairman, Dr. Barry Guinagh, with his sympathetic

heart and mind gave invaluable guidance for this project. I thank Dr. James Algina, for

his continued confidence in my ability and for his statistical and editorial expertise. Drs.

Walter Busby, Woodrow "Max" Parker, and Mary Howard-Hamilton deserve very warm

and special thanks for their valuable insight and encouragement. Two of the original

members of my committee, Dr. Donald Avila, and Dr. Robert Jester, did not live to see

this dissertation come to fruition. They remained an inspiration and a shining light when

the process became dark. My hat is off to these two men. I would also like to thank Dr.

Patricia Ashton who initially gave me confidence while pursuing my doctoral program at

the University of Florida. The faculty in Foundations, Dr. Arthur Newman, Dr. Robert

Sherman, Dr. Rodman Webb, Dr. John Newell, Dr. Hannalore Wass, Dr. A. O. White and

Dr. David Miller, were friends whose courses inspired me.

I thank the current Dean of the College of Education, Dr. Roderick McDavis,

who was responsible for my recruitment to the University of Florida, and who provided

guidance and inspiration through the many years I have known him.








Dr. Willa Wolcott, Director of the Reading and Writing Center, deserves special

thanks for her support, confidence, and the employment she provided for a number of

years. May God richly bless her. I would like to thank Diane Heaney, Dr. Diane

Stevenson, and Margaret Steptoe for their encouragement and support. They provided

enthusiasm when I did not have any. Ellen Burleson, a colleague and friend, gave

untiring support, confidence, loyalty, and belief in me. I believe that, without her

support, completing this task would have been unusually difficult.

To the Alachua County Housing Authority, I would like to say thank you for the

introduction to the participants in this study. Under the direction of Ms. Eula Williams, I

was able to have access to the Housing Projects in Gainesville. My deepest thanks go to

Mr. Terry Lee who provided masterful guidance in arranging interviews with low income

mothers and their children. Mr. Lee is a special friend, who assisted me in so many

ways. Sincerest thanks go to Mr. Larry Saunders, a dear friend and confidant who

provided enormous support for a number of years. Also, thanks go to Dr. James D.

Lockett.

Finally, I would like to thank my parents for their patience, confidence, and

undying devotion to my pursuit of the doctoral degree. I would also like to acknowledge

my siblings who would never let me quit. For my daughter, I pass on the truth that

enabled me to persevere: "Through the grace and providence of God, all things are

possible."

















TABLE OF CONTENTS



ACKNOWLEDGMENTS ................ ............... ........ iv


LIST OF TABLES ..................................... ............. viii


ABSTRACT .............................................. xii



CHAPTER 1 INTRODUCTION .......................................... 1

Statement of the Problem ........................................ 2
Purpose of the Study ............................................ 6
Significance of the Study ........................................ 7



CHAPTER 2 REVIEW OF THE LITERATURE ............................. 9

Weiner's Attribution Theory ... ................................... 10
Attribution Measurement Issues .................................. 14
Attribution Patterns of Mothers and Their Children ..................... 17
Mothers' Attributions and Gender Differences ........................ 23
Children's Predictions of Their Mothers' Attributions ................... 27
Summary .............. ... .................... .......... 28


CHAPTER 3 METHODOLOGY ........................................ 30

Hypotheses ............. .................. ........... 31








M ethods ........... .................. ..... ............ 32
Variables Under Investigation .................................... 40
Research Design ............................................... 40
Data Analysis ........................................ ...... 40



CHAPTER 4 DATA ANALYSIS .................. ...................... 41

Mothers' and Children's Success and Failure Attributions for Mathematics
performance: Rating Scale Method ............................ 42
Mothers' Predictions of Their Children's Mathematics Success and Failure
Attributions: Rating Scale Method ............................ 53
Children's Predictions of Their Mothers" Mathematics Success and Failure
Attributions: Rating Scale Method ............................ 59
Mothers' and children's cussess and Failure Attributions for mathematics
Performance: Chip Distribution Method ....................... 69
Mothers' Predictions of Their Children's Mathematics Success and Failure
Attributions: Chip Distribution Method ........................ 78
Children's Predictions of Their Mother's Mathematics Success and Failure
Attributions: Chip Distribution Method ........................ 83


CHAPTER 5 DISCUSSION ............................................ 94


APPENDICES ................. ............................ 114

APPENDIX A: Parental Consent Form ............................. 114
APPENDIX B: Student Assent Form ............................... 115

REFEREN CES ....................................... .......... 117

BIOGRAPHICAL SKETCH .................. ................ ..... 12














LIST OF TABLES


Table 3-1 Sample Size for Rating Scale Respondents ......................... 33

Table 3-2 Sample Size for Chip Distribution Respondents ..................... 34

Table 4-1 Mother's and Children's Mathematics Success Attributions by Grade
and Sex ....................................................... 45

Table 4-2 Summary of ANOVA for Success Attributions ...................... 46

Table 4-3 Success Mean Ratings as a Function of Respondent and Type of
Attribution ............. .................................. 47

Table 4-4 Success Mean Ratings as a Function of Sex and Type of Attribution ..... 47

Table 4-5 Success Mean Ratings as a Function of Respondent, Child's Sex, and
Type of Attribution ............................................. 48

Table 4-6 Success Attribution Means by Grade .............................. 48

Table 4-7 Pearson Correlation Coefficients Between Mother's and Children's
Success Attributions within the Family ............................... 49

Table 4-8 Mother's and Children's Mathematics Failure Attributions by Grade
and Sex .............. .............................. ............ 50

Table 4-9 Summary of ANOVA for Failure Attributions ....................... 51

Table 4-10 Failure Mean Ratings as a Function of Respondent and Type of
Attribution ..................... .. ...... .............. 52

Table 4-11 Pearson Correlation Coefficients Between Mothers's
and Children's Failure Attributions Within the Family ............ 52


viii








Table 4-12 Pearson Correlation Coefficients Between Children's Success
Attributions and Mothers's Predictions of Their Children's Success
Attributions W within the Family ..................................... 54

Table 4-13 Children's Mathematics Success Attributions and Mother's
Predictions of Their Children's Mathematics Success Attributions by
Grade and Sex ..................................... ............ 56

Table 4-14 Summary of ANOVA for Children's Mathematics Success
Attributions and Mother's Predictions of Their Children's
Mathematics Success Attributions ................................... 57

Table 4-15 Predicted Success Mean Rating as a Function of Respondent
and Type of Attribution ........................................... 58

Table 4-16 Pearson Correlation Coefficients Between Children's Failure Attributions
and Mothers Prediction of Children's Attributions Within the Family 58

Table 4-17 Children's Mathematics Failure Attributions and Mother's Predictions
of Their Children's Mathematics Failure Attributions by Grade and Sex ..... 60

Table 4-18 Summary of ANOVA for Children's Mathematics Failure Attributions
and Mother's Predictions of their Children's Mathematics Failure
Attributions ........................................ ......... 61

Table 4-19 Predicted Failure Mean Ratings as a Function of Respondent and
Type of Attribution ................. ............................. 62

Table 4-20 Pearson Correlation Coefficients Between Mother's Success
Attributions and Children's Predictions of Mothers's Success
Attributions Within the Family ........................... ........ 62

Table 4-21 Mother's Mathematics Success Attributions and Children's
Predictions of Their Mother's Mathematics Success Attributions by Grade
and Sex ................... ................... .. ........... .. 63

Table 4-22 Summary of ANOVA for Mother's Mathematics Success Attributions
and Children's Predictions of Their Mother's Mathematics Success
Attributions .............................................. 64

Table 4-23 Predicted Success Means by Respondent ......................... 65








Table 4-24 Pearson Correlation Coefficients Between Mother's Failure
Attributions and Children's Predictions of Mother's Failure
Attributions Within the Family ..................................... 66

Table 4-25 Mother's Mathematics Failure Attributions and Children's Predictions
of Their Mother's Mathematics Failure Attributions by Grade and Sex ..... 67

Table 4-26 Summary of ANOVA for Mother's Mathematics Failure Attributions
and Children's Predictions of Their Mother's Mathematics Failure
Attributions .................................................... 68

Table 4-27 Predicted Failure Means as a Function of Respondent x Type of
Attribution ............ ................................... 69

Table 4-28 Mother's and Children's Mathematics Success Attributions by Grade
and Sex........................................... 72

Table 4-29 Summary of ANOVA for Success Attributions ..................... 73

Table 4-30 Success Means as a Function of Respondent and Type of Attribution ... 73

Table 4-31 Pearson Correlation Coefficients Between Mothers' and Children's Success
Attributions Within the Family ............................... 74

Table 4-32 Mother's and Children's Mathematics Failure Attributions by Grade
and Sex ........................................................ 75

Table 4-33 Summary of ANOVA for Failure Attributions ...................... 76

Table 4-34 Failure Means as a Function of Respondent, Type of Attribution, and
Grade ................ ...................................... 77

Table 4-35 Pearson Correlation Coefficients Between Mothers' and Children's
Failure Attributions Within the Family ............................... 77

Table 4-36 Pearson Correlation Coefficients Between Children's Actual Success
and Mothers' Predicted Success Attributions Within the Family ........... 79

Table 4-37 Children's Mathematics Success Attributions and Mother's
Predictions of Their Children's Mathematics Success Attributions by
Gradeand Sex ................ ......... ....................... 81








Table 4-38 Summary of ANOVA for Children's Mathematics Success
Attributions and Mother's Predictions of Their Children's
Mathematics Success Attributions ............................. 82

Table 4-39 Predicted Success Means as a function of Respondent and Type of
Attribution ....................... ........... ............... 82

Table 4-40 Pearson Correlation Coefficients Between Children's Actual Failure
and Mothers' Predicted Failure Attributions Within the Family ............ 83

Table 4-41 Children's Mathematics Failure Attributions and Mother's Predictions
of Their Children's Mathematics Failure Attributions by Grade and Sex .... 84

Table 4-42 Summary of ANOVA for Children's Mathematics Failure Attributions
and Mother's Predictions of Their Children's Mathematics Failure
Attributions .............. ................................. 85

Table 4-43 Pearson Correlation Coefficients Between Mother's Actual Success
and Children's Predicted Success Attributions Within the Family .......... 86

Table 4-44 Mother's Mathematics Success Attributions and Children's
Predictions of Their Mother's Mathematics Success Attributions by Grade
and Sex ...................................................... 87

Table 4-45 Summary of ANOVA for Mother's Mathematics Success
Attributions and Children's Predictions of Their Mother's Mathematics
Success Attributions ................ ........................... 88

Table 4-46 Predicted Success Means as a Function of Respondent and Type of
Attribution ................ ..... .......................... 89

Table 4-47 Pearson Correlation Coefficients Between Mother's Actual Failure
and Children's Predicted Failure Attributions Within the Family ........... 89

Table 4-48 Mother's Mathematics Failure Attributions and Children's Predictions
of Their Mother's Mathematics Failure Attributions by Grade and Sex ..... 91

Table 4-49 Summary of ANOVA for Mother's Mathematics Failure Attributions
and Children's Predictions of Their Mother's Mathematics Failure
Attributions .................. .......................... 92

Table 4-50 Predicted Failure Means as a Function of Respondent, Type of
Attribution, Grade, and Sex ........................................ 93

xi













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

CAUSAL EXPLANATIONS FOR MATHEMATICS PERFORMANCE GIVEN BY
LOW SOCIOECONOMIC STATUS AFRICAN AMERICAN MOTHERS AND THEIR
CHILDREN

By

Kenneth R. Cyrus

May, 1996


Chairman: Barry Guinagh
Major Department: Foundations of Education


The present study examines the attributional beliefs of low SES African American

mothers and their 6th, 7th, and 8th grade children for the children's mathematics

performance. Using two attribution measures, the rating scale and chip distribution

methods, data from mothers and children were examined for (a) attributional agreement

about the causes of mathematics performance, (b) mothers' attributions and children's

gender, (c) attributional agreement within the family, (d) mothers' and children's

attributional predictions, and (e) whether results would vary depending on the method--

chip distribution or rating scale--used for data collection.

The results from the rating scale and chip distribution methods revealed similar

findings on all hypotheses except for mothers and gender-related attributions. Both

xii








methods showed that mothers and children generally explained mathematics success as a

result of effort, with some emphasis on school and home training. For mathematics

failure, mothers emphasized lack of effort, whereas children emphasized lack of home

training and lack of effort. Both methods showed that mothers and children within the

same household were not in agreement about the causes of mathematics performance.

The two methods differed on the issue of mothers and gender-related attributions. The

chip distribution method showed that mothers believed that boys and girls succeed and

fail in mathematics for the same reasons. The rating scale revealed mothers believed

boys have more ability in mathematics than girls and girls had to try harder than boys to

be successful in mathematics. For failure, mothers did not make distinctions related to

gender. Both methods also showed that mothers could not predict the attributions the

children would give for the children's mathematics performance and children could not

predict the attributions their mothers would give for the children's performance.

Therefore, generally, the rating scale and chip distribution methods provided comparable

results when examining the attributions of low SES African American mothers and

children, with the exception of gender-related attributions.

Attributional beliefs of the African American family and measurement of

academic attributions are discussed.


xiii










CHAPTER 1

INTRODUCTION


Attributional beliefs are causal explanations individuals hold for their successful

or unsuccessful performance at a task. According to attribution theory, these beliefs

influence motivation, which in turn influence achievement. Consequently, research on

attributions is valuable to the extent that it provides insights into student motivational

factors (Graham, 1991; 1994). Attribution theory was first postulated by Weiner

(Weiner, 1979; 1985; 1986), who believed that attributions were one of the most

important influences on behavior and motivation in achievement settings. The causal

attributions individuals make for their performance will influence the individuals'

expectations for success on future tasks and their emotional response to the event. Causal

attributions most often given by individuals are ability, effort, task difficulty, and luck.

Sixty additional attributions have since been documented by other researchers (Hau &

Salili, 1993). Attributions can be classified on three attribution dimensions: locus of

cause--whether the reason is internal or external to the individual; stability of cause--

whether the reason is temporary or permanent; and controllability of cause--whether the

reason is under the control of the individual or not.

Although individuals may use several attributions to explain success and failure,

causal dimensions help clarify the similarities and differences among the various causal

attributions (Weiner, 1985; 1986; Graham, 1991). For example, ability and effort are










similar attributions in that both attributions are internal to the individual. However,

ability and effort are different because ability is considered a stable attribution not subject

to change, whereas effort is an unstable attribution which may change from occasion to

occasion. Ability cannot be controlled by the individual but effort is controllable.

Further, task difficulty and luck are both classified in Weiner's (1985) attribution theory

as external attributions and, as such, are not under the control of the individual.

However, task difficulty and luck differ in that luck is an unstable attribution because it

varies from occasion to occasion, while task difficulty for a specific task is considered a

stable or permanent attribution.



Statement of the Problem

Causal attributions about school performance as they are held by a family unit and

the similarities and differences in attributional beliefs between parents and children have

been the focus of past empirical study (Cashmore & Goodnow, 1986; Hess, Chih-Mei, &

McDevitt, 1987; Holloway & Hess, 1982; Holloway, Kashiwagi, Hess, & Azuma, 1986;

Stevenson & Lee, 1990; Yamauchi, 1989). This research has revolved around five basic

issues: parents and children's agreement about attributions for school performance;

parent-child attributional agreement within individual families; comparisons of parent-

child agreement between cultures; the relation of gender of the child to causal attributions

of both parents and children; and children's accuracy in predicting their mother's

attributions about children's school performance.










Three basic problems have emerged from this research. First, there has been a

lack of consistency in the measurement of attributions across studies. Three methods

have been used: a ranking, a chip distribution, and a rating method. The ranking method

asks the subject to rank attributions (such as luck, effort, and ability) from most important

to least important. The chip method gives the subjects a limited number of chips (or

sometimes points -- e.g. Stevenson & Lee, 1990) that they use to designate the

importance of attributions. The rating scale method uses a scale that subjects use to

indicate the strength of each attribution. For example, Cashmore and Goodnow (1986)

employed a rank order measure when comparing Australian mothers' and children's

attributional beliefs. Holloway and Hess (1982) used a chip distribution measure to study

Caucasian American mothers and children while Parsons, Adler, and Kaczala (1982) used

a rating scale instrument with this population. A point distribution measure was used by

Stevenson and Lee (1990) to compare the attributional beliefs of Caucasian American,

Chinese, and Japanese mothers and children.

Studies of attributions have used one of these three methods, but not two or three

methods at the same time. Inconsistent findings in this area of attribution research may

have been caused by the way attributions were measured. Various measurement

techniques used in attribution studies have both advantages and disadvantages and

different results found in attribution research may vary according to the measurement

technique used. This raises questions of reliability and validity for attribution research

(Elig & Frieze 1979; Hau & Salili, 1993). The inconsistent use of methods and








4

subsequent results found in attribution research make predictability and generalization of

findings to other populations or situations difficult.

A second problem has been that results across various studies addressing parents'

and children's attributions for academic performance were inconclusive. For example, for

parents' and children's attributional agreement, Caucasian American mothers and

children were in agreement about the causes for academic success and failure (Parsons,

Adler, & Kaczala, 1982); however, Holloway and Hess (1982) reported Caucasian

American mothers and children were not in agreement about the causes of school

performance. In another example, relating to gender, Japanese, Anglo-Australian, and

Italian Australian mothers did not make the distinction in attributions between their sons

and daughters for their mathematics performance (Cashmore & Goodnow, 1986;

Holloway, Kashiwagi, Hess, & Azuma, 1986). In contrast, Parsons, Adler, and Kaczala

(1982), Holloway and Hess (1982), and Holloway, Kashiwagi, Hess, and Azuma (1986)

found that Caucasian American mothers did make the distinction between attributions for

their sons' and daughters' mathematics performance.

The third problem found in the attribution literature has been that research

involving African Americans parents and children was quite limited. Some research has

been done with African American children that compares them to Caucasian American

children (Graham, 1991, 1994). However, no research with African American mothers

and children has been done (Graham, 1991; 1994). In reviews of minority participation

and achievement in mathematics, Matthews (1984), Reyes and Stanic (1988) identified

mothers as an important variable for their children's successful mathematics achievement.








5

They reported mothers' race, education, attitudes, expectations, and aspirations influenced

their children's mathematics achievement. However, both reviews cited a lack of research

in this area and recommended more research with minority populations. Although there

are few empirical studies, some early research involving minority women scientists (Hall

1981; Kenschaft, 1981; Malcolm, Hall, & Brown, 1976) found that the mother's attitudes

about, and a positive belief in their children's ability to achieve success in mathematics,

positively influenced the children's participation and performance. More research on

motivation is also needed because the African American population has consistently

achieved at lower levels than other groups (Jenkins, 1989), particularly in mathematics

and especially for the low socioeconomic status (SES) African American population

(Reyes, 1984; Reyes & Stanic, 1985; 1988). Therefore, an understanding of the

attributional beliefs of low SES African American mothers and children in relation to the

children's mathematics performance would seem to be important to understanding their

mathematics achievement.

This study examined the causal attributions of low SES African American

mothers and children for success and failure in mathematics. The following issues were

examined: (a) mothers and children's attributional beliefs about mathematics

performance; (b) agreement within the family about that performance; (c) gender

differences; (d) the ability of children to predict attributions given by their mothers; and

(e) the ability of mothers to predict the attributions given by their children. Additionally,

because the research in this field has used both a rating system and a chip distribution










system, both systems will be used to assess these beliefs and to examine the

comparability of these methods.



Purpose of the Study

The primary purpose of this study was to compare the causal attributions of low

SES African American mothers and children for mathematics performance. The

secondary purpose was to compare two attribution instruments in their ability to measure

the attributions of low SES African American mothers and children.

The research questions addressed by this study were as follows:

1. Are there significant differences in causal attributions of low SES African

American mothers and their children for mathematics success and failure?

2. Do low SES African American mothers give different explanations for

mathematics success and failure for their sons and daughters?

3. Can low SES African American mothers predict their children's causal

attributions for mathematics success and failure?

4. Can low SES African American children predict their mother's causal

attributions for the children's mathematics success and failure?

5. Are there significant differences in causal attributions of low SES African

American mothers and their children for mathematics success and failure within the

family unit?

6. Will low SES African American mothers' and children's causal attributions

differ when measured by separate attribution instruments?










Significance of the Study

Examining the causal attributions of African American children and parents is

important for several reasons. First, achievement in school is important. However,

African American children have not done well in mathematics, even though their mothers

believe they are doing well and will continue to do well (Alexander & Enwistle, 1988;

Matthews, 1984; Reyes & Stanic, 1988). Next, achievement is related to motivation and

motivation is related to attributions (Weiner, 1986). This connection between attributions

and motivation can be seen in other cultures. For example, Japanese and Chinese

mothers' belief in hard work or high effort has shown to be highly motivating for their

children's mathematics achievement (Holloway, Kashiwagi, Hess, & Azuma, 1986;

Stevenson & Lee, 1990). Parents influence the attitudes of children. Interviews of

African American parents find they want their children to do well in school (Alexander &

Enwistle, 1988; Stevenson, Chin, & Uttal, 1990). Do the attributions held by African

American children about motivation correlate to those attributions held by their parents,

or are there differences in attributions between mothers and children?

Additionally, this research is important because it attempts to clarify some

measurement concerns found in the parents and children's attribution literature. The past

research has used numerous methods to measure causal attributions. This may be the

cause of some of the inconsistent findings. The present study was designed to determine

whether results would be consistent when different methods are used to collect the

attribution data. Therefore, this study will determine whether two widely used

measurement instruments in attribution research are comparable for assessing the causal










attributions of low SES African American mothers and children for the children's

mathematics performance.

This study also provides further testing of attribution theory. Understanding the

attributional beliefs of African Americans may assist in understanding the nature of the

beliefs that help create the academic conditions experienced by this population.

Attribution theory could provide a model to investigate attribution patterns that hinder

achievement motivation in mathematics (Graham, 1991; 1994).

Finally this study will assist in establishing future research directions for the study

of motivation in the African American family. Graham (1994) has argued for more

research involving African American parents and children. She states,


Future studies conducted within a motivational
framework should systematically examine the sources of African-
American parents' beliefs about their children's ability and
prospects, whether and how these beliefs get communicated to the
child, and how they get played out in parenting practices that have
the potential to either undermine or enhance achievement strivings.
(p. 107)












CHAPTER 2

REVIEW OF THE LITERATURE



The primary purpose of this study was to compare the causal attributions of low

SES African American mothers and their children for mathematics performance. A

second purpose was to examine whether the rating scale instrument and the chip

distribution instrument yield similar and consistent results when measuring mathematics

attributions of African American mothers and children.

While mothers' and children's academic attributional beliefs of numerous cultures

have been studied (e.g., Bar-Tal & Guttmann, 1981; Cashmore & Goodnow, 1986; Hess,

Chih-Mei, & McDevitt, 1987; Holloway, & Hess, 1982; Stevenson & Lee, 1990), African

American mothers and children have not been the focus of any studies in the attribution

literature. Related literature was therefore chosen for review. The review will be

discussed as follows: first, a general overview of Weiner's attribution theory; second,

attribution measurement issues; third, attribution patterns of mothers and their children

for school performance; fourth, mothers' attributions and gender differences; fifth,

children's accuracy in predicting parent's causal attributions for the children's

mathematics performance. The chapter concludes with a summary of the review.










Weiner's Attribution Theory

"Attribution theory is a theory of motivation and emotion, with achievement

strivings as the theoretical focus" (Weiner, 1985, p. 549). The attribution approach to

achievement motivation was originated by Heider (1958) and later elaborated by Weiner

(Weiner, 1979; 1986; Weiner & Kukla, 1970). "A central assumption of attribution

theory is that the search for understanding is the basic spring to action" (Weiner, 1979, p.

3). In a school setting the search for understanding of success and failure experiences

often leads to attributional questions such as "Why did I succeed?" or "Why did I fail?"

According to Weiner (1985), the individual's desire for mastery of his or her environment

and situations that are unexpected are contributors to searching for causal explanations.

Perceived Causality

In their initial statement regarding the perceived causes of success and failure

Weiner, Frieze, Kukla, Reed, Rest, and Rosenbaum (1971) argued that in achievement

related contexts the causes perceived as most responsible for success and failure are

ability, effort, task difficulty, and luck. They asked participants to rate the importance of

four causes of success and failure. The participants were to rate the relative contribution

of high or low ability, high or low effort, ease or difficulty of the task, and good or bad

luck to account for their success or failure on a task (see also, Weiner, 1986). However,

in later studies (Weiner, 1974; Weiner, Russell, & Lerman, 1978) suggested other factors

such as mood, fatigue, illness, and bias could serve as necessary and sufficient reasons for

achievement performance. Many investigations have been conducted that more

systematically examine causal perceptions, particularly the perceived causes of success










and failure (Weiner, 1985). Two research procedures were followed to determine if

subjects attributed success and failure to other factors. First, the free response method, in

which subjects were provided with only outcome information, namely a success or

failure, has taken place. Participants were asked to list all possibilities that come to mind

for the success or failure. In the second procedure, subjects were given large lists of

causes and were asked to rate the contribution of each cause to the hypothesized outcome

(Weiner, 1985, 1986). From these two procedures, a variety of causes have been

documented (Weiner, 1986). Among the numerous causes, seven attributions received

the most attention including ability, immediate and long-term effort, task characteristics,

intrinsic motivation, teacher's competence, mood, and luck. It should be noted that the

most dominant causes were ability and effort. In most cases success was attributed to

high ability and hard work and failure was thought to be caused by low ability and poor

effort (Weiner, 1979; 1985; 1986). This also holds true for a number of cultural groups

(Triandis, 1972).

Structure of Perceived Causality

Weiner's (1979; 1985) theory classifies the types of attributions made by

individuals by using a three-dimensional taxonomy of causes of success and failure. The

first dimension, labeled locus of causality, determines whether a cause is within the

person or within the environment. Rotter's (1966) construct of locus of control was the

first to classify people as internals or externals. Ability and effort causal attributions are

internal to an individual, whereas task difficulty or luck are external.








12

The second dimension of causality, referred to as stability, characterizes causes on

a continuum. Causal attributions can be stable, or invariant, and unstable, or variant.

Weiner (1979; 1985) proposes that ability and task difficulty are likely to be perceived as

fixed, whereas luck and effort are more variable or unstable.

The third dimension in Weiner's three dimensional taxonomy is labeled

controllability. Effort is classified as controllable, while luck, ability, and task difficulty

are classified as uncontrollable.

Numerous investigations have provided support for Weiner's taxonomy (Meyer,

1980; Meyer & Koelbl, 1982; Passer, 1977; Stem, 1983; Wimer & Kelly, 1982). Meyer

(1980) presented questionnaires with 16 different situations and asked subjects to indicate

in each situation how much each of nine attributions was responsible for the outcome.

Using factor analysis, the results showed a correspondence between the three dimensions

of locus, stability, and controllability. Causal attributions theoretically can be classified

within one of eight cells: 2 levels of locus x 2 levels of stability x 2 levels of control

(Weiner, 1979).

According to Weiner's theory, each causal dimension has psychological

implications for thought and action. The locus of causality dimension has implications

for self-esteem. Self-esteem is higher when success is caused by internal factors (effort,

and ability) rather than external factors (luck or task difficulty). Self-esteem is lowered

when failure is caused by internal factors rather than external factors.

The causal dimension of stability is linked to expectancy for future success.

Failure that is caused by low ability or difficultly of task decreases the expectation of










future success more than failure attributed to bad luck, bad mood, and lack of effort

(Weiner, 1979). Weiner, Nirenburg, and Goldstein (1976) showed that American college

students who were successful at completing block designs because of ability and ease of

task anticipated a successful performance in the future, compared to students who were

successful because of effort and luck.

The dimension of control is linked to personal responsibility. The dimension of

control focuses on inferences about others and how beliefs about another's responsibility

for success and failure influence an individual's reactions toward that person. When a

person fails for controllable reasons, that person is blamed by others. Therefore, when a

person fails for uncontrollable reasons, that person does not receive blame.

Attributions. Expectations. and Consequences

Within Weiner's three-dimensional model, the stability dimension is posited to be

related to goal expectancy. According to Weiner (1985, p. 556), "the attributional

position is that the stability of a cause rather than its locus, determines expectancy shifts."

Therefore, when attributions are classified on the stability dimension they have a link to

expectations for future success. For example, if a subject attributed success on an

important examination to high ability, expectancy of future success will be high because

ability is a stable attribution which in most cases does not change from task to task. But,

if the reason for success was attributed to high effort one will not maintain such high

expectations for future success because effort may fluctuate from one task to the next.

Conversely, for the failure situation, if the cause for failure is one's low ability, the

expectancy of success might be decreased because one cannot change ability; but if








14

failure was attributed to inadequate effort, the positive expectancy for success might stay

high because the subject believes that trying hard will bring future success. Thus, if

failure is perceived as being caused by lack of effort rather than low ability, expectations

of future success will continue to stay high. Although ability and effort are both internal

attributions on the locus dimension (Weiner, 1985), ability and effort are classified

differently on the stability dimension. This means ability is a stable cause for success and

failure, while effort is perceived as unstable in most situations involving failure. Effort is

perceived as relatively stable in most situations of success. The attributions of effort may

bring about positive expectations in the future with both success and failure, whereas

ability attributions will create high expectations of failure.

Attributional theory, therefore,

is a theory of motivation and emotion in which causal attributions play a
key role. The perceived causes of success and failure share three common
properties: locus, stability, and controllability. The perceived stability of
causes influences changes in expectancy of success. All three dimensions
of causality affect a variety of common emotional experiences including
anger, gratitude, guilt, hopelessness, pity, pride, and shame. Expectancy
and affect, in turn, are presumed to guide motivated behavior. (Weiner,
1985,p.548)



Attribution Measurement Issues

One of the problems in attribution research has been the lack of consistency in the

methods used to measure attributions (Elig & Frieze, 1979; Hau & Salili, 1993;

Maruyama, 1982; Whitley & Frieze, 1985). In a thorough review examining the

measurement of achievement attributions, Hau and Salili (1993) investigating methods,










question contents, and measurement formats. They concluded that no one method is

superior to another. They found that the appropriate use of any attribution measure

requires that the instrument specifically match the research question; yet they are not

clear on how this should be accomplished. Despite the authors' statement that the method

and the research question need to be matched, their review indicates that no one method

is superior to another. While the rating scale was the most popular structured method

used by attribution researchers, the authors found that all measurement methods are

equally satisfactory when used properly. However, they did not answer the question as to

whether all measurement instruments would be equally valid for the same research

question. Elig and Frieze (1979) also cautioned that various measurement instruments

can yield different results for the same research question.

Researchers who have been interested in study of parents' and children's causal

attributions for school performance have commonly used three structured measurement

techniques. The first type of measure is the chip distribution measure in which parents

and children have a limited number of plastic chips that they place on different choices

weighting the importance of causal attributions (Dunton, McDevitt & Hess, 1986; Hess,

Chih-Mei & McDevitt, 1987; Hess, Holloway & King, 1981; Hess & McDevitt, 1985;

Hess, McDevitt & Cheng, 1984; Holloway & Hess 1982; 1984; Holloway, Hess & King,

1981; King, Hess & Holloway, 1981). For example, a child would be asked to place 10

chips on 5 choices according to the importance of each attribution for that child. If 8

chips were placed on the ability attribution option only two chips would be left to place








16

on the remaining attribution options. A variation of this method uses a limited number of

points to indicate the importance of each attribution (Stevenson & Lee, 1990).

A second measurement instrument of parents' and children's attributions for

school success and failure is the structured rating scale format. Parents and children rate

on Likert-type scales (frequently 5-, 7-, or 9-point) the importance of each attribution

(Bar-Tal & Guttmann, 1981; Bird & Berman, 1985; Parsons, Adler & Kaczala, 1982;

Yamauchi, 1989; Yee, 1984; Yee & Eccles, 1988). The rating scale method allows

participants a chance to rate the importance of each attribution without one rating

effecting other ratings.

A third type of measure is the rank order format in which parents and children

rank the order of importance of each attribution in relation to other attributions

(Cashmore, 1980; 1982; Cashmore & Goodnow, 1986; Holloway, 1986). The rank order

method usually has a limited number of attributions and participants must decide which

attribution is first, second, or third (or so on) in its contribution for determining success or

failure.

It has been recommended (Hau & Salili, 1993) that there should be a match

between the research question and the measurement instrument. Attribution researchers

who study parents' and children's beliefs about school performance have continued to use

a variety of measures to investigate the same research question: "Do parents and children

in families hold similar beliefs about the children's success and failure in school?"

The results of this line of research have been inconsistent. Holloway and Hess

(1982), using a chip distribution method, found that mothers and their children did not










hold similar beliefs about the causes of children's academic success and failure.

However, Parsons, Adler, and Kaczala (1982), using a rating scale, found that parents and

children held similar beliefs about causes the children's academic performances. Further,

Cashmore and Goodnow (1986), using rank order method, did not find agreement

between parents and children. It is unknown as to whether these inconsistent findings

were due to the participants in these studies or to the variation in measurement

techniques. It is also unclear as to whether the researchers attempted to match the

measurement instrumentation with the research question. One weakness in this literature

is that no studies were found that simultaneously used two or more methods (chip, rating,

or ranking) in order to show the comparability of those measurement instruments.

Attributional Patterns of Mothers and Their Children

Because it is generally believed that parents transmit values, beliefs, or traits to

the next generation, one might expect to find agreement about these values, beliefs, or

traits between parents and their children (see Cashmore & Goodnow, 1986; Holloway &

Hess, 1982). Some attribution researchers examined the possibility that parents convey

attributional beliefs regarding academic success and failure directly to their children.

However, findings were inconsistent across studies.

Disagreement. In her study investigating parental belief systems about their

children's school performance, Cashmore (1980) found little agreement between parents'

and children's attributions of 100 Australian children and their parents. Parents and

children were asked to rank order three attributions: ability, effort, and teaching. Parents

viewed ability as the most important cause of success and effort the most important cause










of failure. Children believed that effort was the most important cause for success while

ability was given as the cause for failure. The quality of teaching at school did not seem

to be important for either parents or children.

In a follow-up study, Cashmore and Goodnow (1986) compared attributions of

mothers, fathers, and their children from an Anglo Saxon Australian sample and an

Italian Australian sample. They were asked to put in rank order the attributions of talent,

effort, and teaching for six skill areas. Results indicated parents and children's responses

varied across these skill areas. Children from both samples cited effort as the most

important cause for school performance, while parents in both samples cited talent

(ability).

Holloway and Hess (1982) investigated the causal attributions given by mothers

for their upper elementary school children to explain their reasons for high and low

academic performance. Participants were given chips to place on cards listing

attributions in order to weight the importance of ability, effort, personality, and training.

Mothers' cited children's ability as the main cause of success, while lack of effort was

viewed as the main reason for failure. Children held the opposite belief, that effort was

more important for success while lack of ability contributed to failure.

Bar-Tal and Guttmann (1981) compared teachers', students,' and

parents' attributions for the students' academic performance. Using a rating scale subjects

were asked to rate 10 causes: ability, interest, difficulty of material, effort, teacher's

explanation, home conditions, parent's help, luck, diligence, and test difficulty. Findings

indicated that parents attributed their children's success to home conditions, and teacher's








19

explanations. Failure was attributed to bad home conditions and children's low level of

interest and ability. Children attributed success to their own effort and teacher's

explanation and failure to lack of effort and low ability.

While it was thought by researchers that parents and children might be in

agreement, indicating the conveyance of values and beliefs, these studies show a lack of

agreement in the attributional beliefs of parents and children about academic

performance.

Agreement. In contrast to these findings, some researchers did find agreement

between parents and children regarding the attributional beliefs for academic success and

failure.

In a study done by Parsons, Adler, and Kaczala (1982), agreement between

parents and children in their causal attributions was documented. Using a rating scale

measuring attitudes and beliefs about mathematics achievement, children's attributions

matched those of their parents. The authors of this study concluded that children's

attributions were influenced more by their parents' beliefs about their abilities than by

their own previous mathematics achievement history.

Yamauchi (1989) conducted a study that compared the attributions of children

with mothers' attributions, and children's predictions of their mothers' attributions for

their child's mathematics performances in Japan. The results, from a rating scale

procedure, showed that more than half of the children and their mothers attributed school

performance to the effort attribution on both success and failure outcomes.








20

Numerous conclusions can be drawn from these studies. First, it is not clear from

the findings of studies in this literature if parents and children hold similar beliefs or

attributions within the same culture or vary across different cultures. Second, this

literature does not inform us on the best method to measure parents' and children's

attributions about school performance. No study explained why a method was chosen to

investigate the research question. Finally, this literature provides little guidance as to

why these findings have been so inconsistent.

Cultural Patterns.

When parents and children from different countries were compared to examine

their attributional beliefs about school performance, findings were different.

Stevenson (1983) conducted a study of parents and children in Japan, Taiwan, and

the United States. Participants were first asked to rank four attributions: effort, natural

ability, difficulty of school work, and luck regarding the children's achievement. Next,

they were asked to divide up 10 points across the 4 attributions. In all three cultures the

rank order of importance was the same: effort first, followed by ability, task difficulty,

and luck. However, the point distribution method (a variation of the chip distribution

method) indicated more emphasis on effort for the Asian populations (Taiwan and Japan)

than for the participants from the United States. This study provides some evidence that

attributional beliefs may have cultural variations.

In a comparison of Japanese mothers' and children's with American mother's and

children's attributions, Hess et al. (1986) found there was a contrast between the two

cultures. Using the chip distribution method (see Holloway and Hess, 1982), Japanese








21

mothers and children weighted lack of effort as the most important cause for mathematics

failure, while American mothers' and children's explanations were evenly divided among

ability, effort and training at school. Japanese mothers were less likely to blame training

at school as a cause of their children's low achievement.

In a further study of Japanese mothers and children who were compared to

American mothers and children for causal attributions about mathematics performance

cultural differences in attributions, transmission of beliefs from mothers to their children,

and the effect of gender of child on attributions were examined by Holloway, Kashiwagi,

Hess, and Azuma (1986). Using the chip distribution method, this study also found that

there were cultural differences between American and Japanese mothers and their

children. Japanese mothers and children tended to stress effort as a cause for success and

failure in mathematics, while American mothers and children tended to stress ability.

Similar results were reported by Stevenson, Lee, and Stigler (1986) when

comparing Chinese, Japanese and American mothers on beliefs about causes of their

children's mathematics success and failure. Using a point distribution system, findings

showed that while American mothers gave the largest number of points to ability for

explaining why their children succeed or fail in mathematics, the Asian mothers gave the

most points to effort.

In another study to investigate cultural variations in family beliefs, Hess, Chih-

Mei, and McDevitt (1987) examined beliefs about children's performance in mathematics

through interviews with mothers and their 6th grade children involving three cultures, the

Peoples' Republic of China, Chinese American, and American. The three groups showed








22

different attribution patterns with a chip distribution method for the attributions of ability,

effort, school training, home training, and luck. Mothers in the Peoples' Republic of

China viewed lack of effort as a major cause of low performance. The Chinese American

mothers also viewed lack of effort as important concerning failure, but assigned

responsibility to other sources. The American mothers distributed responsibility more

evenly across the attribution options. Children in the three cultural groups had patterns of

attributions similar to those of their parents.

Stevenson and Lee (1990) studied mothers from Japan, Taiwan, and the United

States. Participants were asked to rate the influences of the factors of studying hard,

intelligence, study habits, a good teacher, home environment, parents' assistance, the

curriculum, and luck on academic achievement. American mothers gave higher ratings to

study habits, good teacher, home environment, and curriculum than they did to effort. On

the other hand, Asian mothers rated effort highly and the only factor they rated higher

than effort was good teachers.

These studies provide support for differences in attributions made by Asian

parents and children and those made by American parents and children for the children's

academic performance. These studies show that Asian parents and children tend to

emphasize the effort attribution more strongly than American parents and children.

These studies may also lend support for the differences of meaning of effort and ability in

the two cultures. A limitation of these studies is that they do not address non-Asian

cultures or sub-cultures within the United States. Different attributions could be held by










South American parents and children or Hispanic or African American parents and

children.

Summary

In summary, several conclusions can be drawn from the parents' and children's

attribution literature regarding academic performance. First, several measurement

techniques have been used by researchers in studies that make-up this literature. The

methods used were primarily structured interviews. When rating causes, participants

were asked to rate on a 5- or 7-point scale. When weighting causes, subjects were asked

to place 10 chips or assign 10 or sometimes 100 points to 4 or 5 attributional cards. A

few studies asked subjects to rank order attributions. Both rating scale and chip

distribution methods were chosen for this study.

Second, researchers have examined and compared attributions of students,

teachers, parents, and numerous cultural groups. African American mothers and children

are considered to be a sub-culture in the United States. Additionally, they are a

population that had not been studied.

Third, a wide variety of attributions have been included in instruments. However,

all of the instrument included ability and effort. The instruments chosen therefore needed

to include these attributional elements.

Finally, it may be difficult to generalize from findings in this literature to other

groups because findings have not always been consistent. It was hoped that this study

would address this issue.










Mothers' Attributions and Gender Differences

Examining gender differences in mother attributions is important for two reasons.

Boys and girls have long been shown to have different attitudes and different

performance levels in mathematics (Fox, Tobin, & Brody, 1979; Holloway & Hess, 1984;

Reyes & Stanic, 1988).

Cashmore (1980) examined the causal attributions of 100 Australian mothers and

children across 6 achievement areas. Subjects were asked to rank order the attributions of

talent, effort, and teaching. Parents did not rate boys and girls differently. The findings

found no differences.

Bar-Tal and Guttmann (1981) examined the causal attributions for mathematics

performance of boys and girls by Israeli parents, teachers and students. Participants were

asked to rate 10 attributions for their contribution to mathematics success and failure.

Results indicated that fathers and mothers believed boys and girls succeed in mathematics

because of home conditions and teacher's explanations while they both fail because of

bad home conditions, low interest, and low ability. The author's concluded there were no

gender differences.

Hess, Holloway, and King (1981), Holloway, Hess and King (1981), and King,

Hess, and Holloway (1981) compared the causal attributions for 21 mothers and their 5th

and 6th grade children. Participants were given open-ended questionnaires to ascertain

the children's spontaneous attributions about doing well or poorly in school. Participants

were then asked to place 10 chips on 5 attributional cards. The attributions were ability,

effort, personality and good teaching. Results showed that mothers attributed both sons'










and daughters' academic successes to ability. Moreover, mothers thought that lack of

effort was a more important explanation for poor performance than lack of talent,

regardless of gender.

Holloway (1986) conducted a study of children's achievement in mathematics and

their motivation to achieve in mathematics. Holloway, found that mothers of girls made

more attributions to ability and task difficulty for success and were more satisfied with

past performance than were mothers of boys, but exerted less academic achievement

pressure. Mothers of boys were found to make more attributions to effort and luck for the

success of their sons than mothers of girls. However, mothers exerted more academic

pressure on their sons to achieve. These differences were, however, minor and the author

concluded that there were few differences overall.

In contrast to the studies showing no differences, there is research that indicates

that mothers do hold gender related beliefs about their children's school performance.

In a study by Holloway and Hess (1982), 21 mothers and children in 5th and 6th

grade were asked to place 10 chips on 5 attributional cards for success and failure in a

subject of high achievement and for failure in a subject of low achievement. Findings

indicated that mother of boys believed boys had more ability than girls while girls had to

exert more effort to be successful in school.

Parsons, Adler, and Kaczala (1982) compared the causal attributions for

mathematics performance of 5th through 1 Ith grade students and their parents.

Participants were asked to rate on 7-point Likert scales their attitudes and beliefs about

mathematics performance. Results showed that parents held sex-differentiated








26

perception's of their children's mathematics aptitude despite the similarity of performance

of boys and girls. Parents of girls thought their daughters had to work harder to do well

in mathematics than did parents of sons. Parents of boys believed advanced mathematics

was more important for their sons when compared to their daughters.

In a study of Holloway and Hess (1984) investigating gender differences in

teacher and in parental beliefs, no significant sex differences on an achievement

assessment for mathematics performance were found for boys' and girls' mathematics

performance in 6th grade.

In a study by Yee and Eccles (1988), 7th grade students and parents were asked to

complete questionnaires at home about beliefs, expectations, and causal attributions for

their children's mathematics achievement. The results indicated that parents' perceptions

and expectations were commensurate with their child's level of mathematics ability.

Mothers attributions for mathematics successes differed for boys and girls.

Dunton, McDevitt, and Hess (1988) conducted a longitudinal study investigating

the origins of mothers' attributions about their daughters' and sons' academic performance

at 5 and 6 years of age and again at age 12. The findings indicated that there are both

similarities and differences in the sources of attributions given by mothers for their

children's performance. Mothers of boys attributed the source of their attributions to their

son's prior performance.

Mothers of son's attributed relative success to having ability and the relative

failure to lack of effort. The source of the mother's attributions for their daughters was

the affective relationship they have with them. Mothers of girls attributed their daughters'










relative success or failure to ability or effort depending on the type of affective

relationship.

Several conclusions can be drawn from the attribution literature related to

mother's attributions and gender differences. Evidence supporting mothers' contributions

to gender related attributions is mixed. Some studies show that mothers do not hold

gender related differences for their children's school performance, particularly

mathematics. Additionally, researchers have used several methods and cultural groups

when examining mothers' attributions and gender differences. There were no studies that

examined African American mothers and their gender related attributions. Finally,

generalizing findings from this literature to other cultural groups is difficult because the

findings are inconsistent.

Children's Prediction of Their Mother's Causal Attributions

Parental accuracy in predicting their children's causal attributions for success and

failure at school has not received much research attention. Yamauchi (1989) asked

Japanese children to predict their mother's attributions for the child's mathematics

performance. Using a rating scale methodology, results showed a discrepancy between

children's predictions of mothers' attributions and mothers actual attributions which

varied by children's school performance. The researcher concluded that children's

predictions of their mothers' mathematics attributions were not related to their mother's

actual attributions. Further research was recommended. As mothers' predictions of their

children's attributions about mathematics performance were not investigated, they might

yield more insight into this area.










Summary

While mothers' and children's attributional beliefs regarding mathematics

performance in numerous cultures have been studied, the attributional beliefs of African

American mothers and their children for the children's mathematics performance has been

neglected in the attribution literature. The primary purpose of this study was to compare

the causal attributions of low SES African American mothers and their children for

mathematics performance; a second purpose was to examine whether the rating scale

instrument and the chip distribution instrument yield similar and consistent results when

measuring mathematics attributions of African American mothers and children.

The review examined Weiner's attribution theory, attribution measurement issues,

attribution patterns of mothers and their children for school performance, mothers'

attributions and gender differences, and children's accuracy in predicting parent's causal

attributions for the children's mathematics performance.

Weiner's attribution theory can be used to study achievement motivation.

Attributions, or the reasons given by an individual to explain success and failure, are

dimensionalized in relationship to the individual (e.g. internal/external;

controllable/uncontrollable; stable/unstable) and provide meaning that has expectations

and consequences for that person. Numerous achievement attributions have been

documented in the literature, but most prominent among them are ability, and effort,

usually in conjunction with additional attributions.

An examination of the literature indicated that a variety of attribution

measurement instruments have been employed by researchers. Common among them are










weighting instruments, rating scales, and rankings. The findings from studies have

frequently been compared to other studies without regard for the measurement instrument

used. This indicated a need for some studies that simultaneously used two or more

methods to see if the different instruments did, in fact, produce comparable results.

Studies of attribution patterns of mothers and their children for school

performance were examined to see if attributions were transmitted within the family.

While numerous studies using a variety of methods and looking at several cultures

examined this issue, the results were inconclusive. Ability and effort are often cited as

reasons for both success and failure, but mothers and children did not always give the

attributions the same importance.

Because gender differences have been found to exist in children's mathematics

performances, the literature regarding mothers' attributions and gender differences was

examined. Again, using a variety of methods and cultures the results were not consistent.

Children's accuracy in predicting parent's causal attributions for the children's

mathematics performance was examined. This was seen as an additional way it might be

determined that beliefs about academic achievement, especially for mathematics, were

transmitted. Further research was found to be needed in this area. In conclusion, the lack

of consistency found in the attribution literature supports the need for additional studies,

especially with a population that has received little attention, African American mothers

and their children, and to utilize two instruments to determine their comparability. It was

to this end that this study was undertaken.












CHAPTER 3

METHODOLOGY





In this study the causal attributions made by low SES African American mothers

and their children for the children's mathematics success and failure were compared. In

addition, two attribution instruments for measuring mothers' and children's causal

attributions were compared.

While a variety of methods and populations have been used in studies of parents

and children's attributions for the children's school performance, low SES African

American mothers and children have not been the focus of this research. Attribution

research with low SES African American mothers and children appears warranted,

especially in the area of mathematics, where low SES African American children achieve

poorly, and are under represented in mathematics courses in high school and in

mathematics related careers (Matthews, 1984; Reyes & Stanic, 1985, 1988).



Hypotheses

The causal attributions of African American mothers for their children's school

performance were compared to the causal attributions of their 6th, 7th, or 8th grade

African American children. Based on the parents' and children's attribution literature

30










involving school performance, the following hypotheses were expected to be supported

by this study:

1. Low SES African American mothers will not agree with their children about

causal attributions for mathematics success and failure.

2. Low SES African American mothers will not make different attributions for

boys and girls about their mathematics success and failure.

3. Low SES African American mothers will not be able to accurately predict

attributions their children give for the children's mathematics success and failure

performances.

4. Low SES African American children will not be able to accurately predict the

attributions their mother will give for the children's mathematics success and failure

performances.

5. Low SES African American mothers and children, within the same family, will

not be in agreement regarding the children's mathematics success and failure

performances.

6. Conclusions about the attributions will not depend on the method--chip

distribution instrument or rating scale instrument--used for data collection.



Methods

Participants

This study was conducted in government subsidized housing in Alachua County,

which is located in north central Florida. The participants were one hundred low SES










students from the 6th, 7th, and 8th grades, and their mothers. Approximate ages of the

students were 12, 13, and 14. This age group was chosen because there is evidence from

the developmental literature that children do not differentiate consistently between ability

and effort attributions until 12-13 years of age (Nicholls, 1978; Karabenick, & Heller,

1976; Kloosterman, 1983; Eccles, 1985). Although 6th, 7th and 8th grade students are

distinguished from each other in the design so that the middle school years could be

examined, there are no research questions related to the specific grade or age level.

Criteria for Selection

From a list of 325 mothers with children in the 6th, 7th, and 8th grades obtained

from the Alachua County Housing Authority, 100 volunteers were chosen. Fifty mothers

and their children were assigned to for the rating scale procedure and fifty mothers and

their children were assigned to for the chip distribution procedure. An attempt was made

to select an equal number of boys and girls from the volunteering families.

Pre-Selection Process

The researcher interviewed all potential mothers who volunteered for this study.

The interview consisted of a series of questions:

"Is your child in middle school--6th, 7th, or 8th grade?"

"Is your child in special educations classes: this includes learning disabled,

mentally handicapped or emotionally handicapped classes?"

"Do you receive a copy of your child's report card, and if you are selected for the

study, can you locate, and bring with you, your child's most recent report card?"










Participating families lived in government subsidized housing. All children

selected for the study met the following criteria:

1. Students were in the 6th, 7th, or 8th grade regular education classes. Although

a large number of low SES African American children are in remedial or special

education classes, students who spent a majority of their day in regular classrooms were

classified as being in regular classes for this study.

2. Participating families had a copy of the student's most recent report card.

Student grades, however, were not a selection factor.

3. Only one child per family was selected.

Mothers and children not meeting selection criteria were thanked for their interest,

but were not selected.

The distribution of mothers and children by grade, gender, and method is shown

in Tables 3-1 and 3-2.

Table 3-1

Sample Size for Rating Scale Respondents


Grade Child's Sex Mothers Children
6 M 9 9
F 8 8
7 M 9 9
F 7 7
8 M 10 10
F 7 7

TOTALS 50 50
















Table 3-2

Sample Size for Chip Distribution Respondents


Grade Child's Sex Mothers Children
6 M 6 6
F 4 4
7 M 12 12
F 10 10
8 M 14 14
F 4 4

TOTALS 50 50


Instrumentation

In the present study, two methods, the chip distribution method and the rating

scale method, were selected from methods used in previous research (see Holloway &

Hess, 1982; 1984; Parsons, Adler, & Kaczala, 1982; and Yamauchi, 1989). For both

samples, the instruments consisted of questions about attributions and predictions for the

children and for the mothers.

Instrumentation for the children was identical to that used by mothers except for

word changes that make the instrument appropriate for children. The questionnaires were








35

designed to elicit 22 responses from the child and 22 responses from the mother in each

procedure (See Appendix C and D).

The Chip Distribution Method. The chip distribution procedure was used in the

research of Hess, Chih-Mei, and McDevitt (1987), Holloway and Hess (1982; 1984), and

Holloway, Kashiwagi, Hess, and Azuma (1986). Ten chips are given to the mother or

child, who is then asked to place the ten chips in 5 categories. The five categories

represent the five attributions in the study (ability, effort, training at school, training at

home, and luck). The ten chips can be distributed in any pattern. The chip distribution

instrument consisted of 5 choices for success, 5 choices for failure, 5 choices for success

prediction, and 5 choices for failure prediction. Each card had an attribution phrase on it.

For the success condition these phrases were used for the mothers:

My child has natural ability for math.

My child tries hard in math.

My child has had good training at school in math.

Ny child has had good training at home in math.

My child has been lucky in math.

Wording changes were made to adapt these phrases to the other conditions.

The Rating Scale Method. The rating scale procedure came from the research of

Bar-Tal and Guttmann (1981), Parsons, Adler, and Kaczala (1982), Yee and Eccles

(1988), and Yamauchi (1989).

For the rating scale instrument, the Holloway and Hess (1982; 1984) distribution

of chips was replaced by 9-point rating scales like those used by the Graham and Long








36

(1986) study. Every portion of the instrument was identical except instead of using chips

to weight the attributions, ratings scales were used to rate the attributions.

The prediction portion of the chip distribution and rating scale procedures were

used as an adaptation from the Yamauchi (1989) study that introduced the prediction

component in a study of Japanese mothers and their children about mathematics

performance. While the Yamauchi (1989) study asked only children to predict their

mother's causal attributions for the children's mathematics success and failure, in this

study mothers were asked to predict the attributions their children would give for the

children's mathematics success and failure, and children were asked to predict the

attributions their mothers would give for the children's mathematics success and failure.

Procedures

After volunteers agreed to participate, the studies were explained and informed

consent obtained, mothers were tested on the first visit and their children were tested on

the second visit in their homes. All study materials were read aloud while the subjects

read silently.

Mothers were not present when their child was "tested", and the child was not

present when their mother was "tested."

Sample A: The Rating Scales Procedure. While looking at their children's report

card, mothers were first asked to judge how well their children had performed in 6th, 7th

or 8th grade mathematics on their most recent report card relative to other members of

their class. (Therefore, success and failure are considered in relation to other class

members and is referred to as relative success and relative failure--Holloway & Hess,










1982; 1984.) The mothers were to indicate the children's level of performance on a 6-

point scale.

Mothers: relative success condition. In this condition, the objective was to help

the mother think of her child as a "relative success." For the relative success condition

the interviewer first noted the response made by the mothers. If a mother rated her child's

performance as a 1 or 2, the interviewer said: "You rated your child's mathematics

performance as better than some of the other children in his/her class. Why do you think

your child did this well?"

If a mother rated her child's performance as a 3 or 4, "better than many" was

substituted into the above sentence. If a mother rated her child's performance as a 5,

"better than most" was read in the sentence. If a mother rated her child's performance as

a 6, the interviewer stated: "You have rated your child's mathematics performance as the

very best in the class. Why do you think your child did this well?"

The interviewer then gives the mothers rating scales forms which ask them to rate

each attribution (reason) for their children's performance by circling the number that best

indicates how much a reason causes their child's mathematics performance. The scale

ranged from 1 to 9; 1 indicated the reason was definitely not a cause to 9 indicating the

reason was definitely a cause.

Mothers: relative failure condition. An analogous procedure was used for

questions about relatively low performance (relative failure). In this condition the

question was, "You indicated that your child wasn't doing the very best in the class. Why








38

do you think he/she isn't doing even better in mathematics?" The purpose was to have the

mother think of her child as a "relative failure."

If the mothers reported that the children were doing the very best in the class, the

interviewer suggested that the children might not be the best in the school district and

asked why the children were not doing even better.

Prediction. In the prediction portion of the study, a mother was asked to predict how her

child would rate his or her mathematics performance based on the most recent report

card. The mothers were to indicate their predictions about their children's rating on a 6-

point scale.

Mothers: relative success condition. For the relative success condition the

interviewer first noted the response made by the mothers to the prediction question. If the

mothers predicted their children would rate their performance as a 1 or 2, the interviewer

said, "You predicted your child would rate his/her mathematics performance as better

than some of the other children in his/her class. Why do you think your child thought

he/she did this well?"

If the mothers predicted their children would rate their mathematics performance

as a 3 or 4, "better than many" would be substituted into the above sentence. If the

mothers predicted their children would rate their mathematics performance as a 5, "better

than most" was read in the sentence. If the mothers predicted their children would rate

the children's performance as a 6, the interviewer stated, "You have predicted that your

child would rate his/her mathematics performance as the very best in the class. Why do

you think your child thinks he/she did this well?"










The interviewer next gave the mothers the rating scale forms, indicating the

attributions (reasons) for the children's performance and then asked the mothers to

complete the 9-point scales by circling the number that best indicated the reason the child

would give for the child's mathematics performance.

Mothers: relative failure condition. An analogous procedure was used to ask

mothers to predict how their children would respond about relatively low performance.

In the Rating Scale procedures each mother provided a total of 22 responses.

Mothers rated their children's level of performance in mathematics (1 rating). Then the

mothers rated five success and five failure attributions (10 ratings). Mothers predicted

how their children would rate their level of mathematics performance (1 rating). They

predicted their children's attributions for their mathematics performance by rating their

prediction of their child's response giving five success and five failure attributions (10

ratings).

Rating Scale Procedures for children. In the Rating Scale Procedure, the

procedures designed for children were identical to those procedures for mothers except

for word changes that made the procedures appropriate for children. There were a total of

22 ratings for each child.

Sample B: Chip Distribution Procedure. The procedures for the Chip

Distribution portions of the study were identical to the Rating Scale procedures except

that the 5 attributions were written on 5 cards and 10 chips were given to subjects to

weight the importance of each attribution as a cause of their child's mathematics

performance.










Variables Under Investigation

There were three independent variables under investigation in this study: (1)

informant: mother/child, and (2) Sex: male/female and (3) grade (6th, 7th, and 8th), in

each outcome situation (success and failure). The five dependent variables in each

success and failure outcome were causal attribution scores for ability, effort, training at

school, training at home, and luck as measured by the rating scale procedure and the chip

distribution procedure.

Research Design

The design for this study is a 3 (grades) x 2 (male/female) x 2 (mother/child) x 5

(attributions) within subjects design with repeated measures on the last two factors. This

design was used for Sample A and Sample B.

Data Analysis

The data from this study in each outcome were analyzed using analysis of

variance (ANOVA). Analyses were computed separately for mothers' and children's

predictions. To determine attribution agreement within the family, correlations were

performed on parent and child pairs within the same family. Scores of parents and

children were correlated with Pearson product-moment correlations for each of the 5

success and 5 failure attributions in each sample.












CHAPTER 4

RESULTS



The primary purpose of this study was to compare the causal attributions that low

SES African American mothers and their children give for mathematics performance. A

secondary purpose was to examine whether the rating scale instrument and the chip

distribution instrument yield similar results when measuring causal attributions given by

low SES African American mothers and children for mathematics performance.

The results presented in this chapter are based on the data collected from 6th, 7th,

and 8th grade students and their mothers in six separate government-subsidized housing

projects in Alachua county, Florida. Attribution questionnaires were administered to 200

participants, including 100 parents and 100 children who volunteered to participate in the

study. The sample pool was divided in half. Fifty mothers and 50 children were

administered the rating scale instrument. Fifty mothers and 50 children were

administered the chip distribution instrument.

For both samples, the data analysis included a grade (3) x child sex (2) x

respondent (2) by type of attribution (5) with repeated measures on the last two factors.

Separate analyses were conducted for success and for failure attributions, and

additionally, to examine mothers' and children's success and failure predictions. Further,

Pearson-product moment correlations were used to explore the relationship of mothers

41










and children's beliefs within individual families and of mothers' and children's

performance evaluations of the children's mathematics grade.

Mother's and Children's Success and Failure Attributions for Mathematics Performance:

Rating Scale Method

Mothers and children were asked to evaluate on a six-point scale the mathematics

grade from the children's most recent report card. The six-point scale ranged for 1 (not

doing as well as most) to 6 (doing the very best in the class). Mothers were then asked by

the interviewer "Why is your child doing this well?" and children were asked "Why are

you doing this well?" Mothers and children rated on 9-point scales the relative

importance of five attributional items. The scales ranged from 1 (definitely not a cause)

to 9 (definitely a cause). The items were ability, effort, training at school, training at

home, and luck for both success and failure conditions. Therefore, relative success in this

study is actually mothers' and children's evaluation of the children's current level of

mathematics performance. For relative failure, mothers and children were asked by the

interviewer, "Why did your child not do even better?" (for mothers), and "Why aren't you

doing even better?" (for the children). This condition examines relative failure as a

judgement in contrast to the current level of performance.

It was hypothesized that low SES African American mothers would not agree

with their children about the causes of the children's success and failure in mathematics.

For the success condition, means and standard deviations for each of the

attributional items are presented in Table 4-1. The results of an ANOVA of mothers' and

children's mathematics success attribution means are presented in Table 4-2. The results










in Table 4-2 include a significant Respondent x Type of Attribution interaction and a

significant main effect of attribution, indicating some disagreement between mothers and

children about the relative importance they place on the five types of attributions for

success in mathematics. Mean attributions as a function of respondent and type of

attribution are reported in Table 4-3. These means describe the nature of the

disagreement between mothers and children about the relative importance of the five

types of attributions. Mothers and children are in agreement about the relative

importance of the five types of attributions as causes of success. On average, both

mothers and children rate effort as the most important, followed by school, home, ability,

and luck. However, mothers put less emphasis than do their children on each type of

attribution. The most striking difference between mothers and children is the luck

attribution. Children rate luck almost as important as ability; mothers rate luck as less

important than do their children and as less important than ability. Therefore the largest

difference between mother's and children's beliefs about the causes of success in

mathematics is the way luck is perceived.

In Table 4-2, there is a significant Child's Gender x Type of Attribution

interaction. Means for interpreting this interaction are reported in Table 4-4. Each mean

in Table 4-4 is an average of a mean for mothers and the mean for children, with the

means calculated across grades. Thus the means in Table 4-4 reflect both children's and

mothers' attributions for success. The results in Table 4-4 indicate that mother--son

dyads tend to agree with mother--daughter dyads about the relative importance of the five

causes: effort is most important followed by school, home, ability and luck. Mother--








44

daughter dyads rate effort, school, and luck as more important than do mother--son dyads,

whereas mother--son dyads rate ability and home as more important than do mother--

daughter dyads.

Examination of Table 4-4 analysis shows that in comparison to mother--daughter

dyads, mother--son dyads believe success in mathematics depends more heavily on

ability. Mother--son dyads also believe effort is less responsible for success than do

mother-daughter dyads in mathematics. Mother--daughter dyads believe that school

training is more important for mathematics success than do mother--son dyads. Mother--

son dyads believe home training is more important for mathematics success than do

mother--daughter dyads. Finally, mother--daughter dyads depend more on the luck

attribution for mathematics success than do mother--son dyads. For mother--daughter

dyads, luck and ability are of equal importance for successful mathematics achievement.

The absence of a Respondent x Child's Sex x Type of Attribution interaction

implies that the pattern of differences between boys' and girls' mean success attribution is

similar to the pattern of gender related difference between the mean success attributions

made by boys' and girls' mothers. As a result the means in Table 4-4 represent the pattern

of responding by both children and mothers. However, to avoid relying unduly on the

nonsignificant interaction, the means by child's sex for mothers and children are reported

separately in Table 4-5. Most of the generalizations that emerge from an inspection of

Table 4-4 also emerge from an inspection of Table 4-5. The exceptions are that in this











Table 4-1 Ratings


Mother s and Children s mvathernatic xucs nuunb rd n-g


Grade Sex

6 Female






Male






7 Female


Male


8 Female


Male


Subject Statistic

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD


Ability

4.1

2.4

5.4

2.8

6.2

2.1

5.8

3.2

2.1

2.0

1.6

1.5

6.6

3.0

5.2

2.7

4.7

3.4

4.4

2.6

4.7

2.8

4.3

3.0


Effort School Home Luck

8.9 8.6 6.5 5.1

0.4 0.7 1.5 3.1

8.3 8.3 7.1 3.4

1.4 1.5 1.9 3.1

8.1 8.4 8.0 3.4

2.0 1.3 1.7 3.7

8.4 5.9 5.8 1.6

0.9 2.4 2.1 1.7

8.9 5.6 3.9 3.4

0.4 1.7 2.5 2.4

8.0 7.9 4.4 2.0

1.9 2.0 3.6 1.9

7.1 6.9 5.6 4.3

2.8 1.8 2.3 3.4

7.2 6.7 6.0 1.9

1.6 2.1 1.6 1.8

8.7 8.3 5.7 5.3

0.8 1.9 3.3 4.1

8.1 6.6 5.6 3.3

2.3 3.0 2.8 3.1

7.2 7.1 6.5 5.0

2.1 2.3 2.0 3.6

6.6 5.9 6.3 1.7

2.3 2.6 2.2 1.3


.i- -- -. -. 'l L '' .










Table 4-2 Ratings

Summary ofANOVA for Success Attributions


df SS MS F p


Source
Between

Grade (G)

Sex (S)

GxS

Error

Within

Respondent(R)

RxG

RxS

RxGxS

Error

Attribution (A)

AxG

AxS

AxGxS

Error

RxA

RxAxG

RxAxS

RxAxGxS

Error


100.06 50.03

0.02 0.02

55.98 27.99

396.91 9.02


63.14 63.14

9.89 4.95

14.57 14.57

5.05 2.52

299.22 6.80

1368.73 342.18

40.19 5.02

143.24 35.81

58.34 7.29

1041.65 5.92

64.68 16.17

42.03 5.25

14.66 3.67

22.26 2.78

739.46 4.20


5.55 0.0071

0.00 0.9651

3.10 0.0549




9.28 0.0039

0.73 0.4889

2.14 0.1504

0.37 0.6921


57.82 0.0001

0.85 0.5611

6.05 0.0001

1.23 0.2828


3.85 0.0050

1.25 0.2726

0.87 0.4816

0.66 0.7241










Table 4-3 Ratings

Success Means as a Function of Respondent and Type of Attribution


Respondent Ability Effort School Home Luck

Child 4.7 8.1 7.5 6.0 4.4

Mother 4.4 7.8 6.8 5.8 2.3


Table 4-4 Ratings

Success Mean Ratings as a Function of Sex and Type of Attribution


Sex Ability Effort School Home Luck

Boy 5.4 7.4 6.8 6.3 3.0

Girl 3.7 8.4 7.5 5.5 3.7


sample, girls mothers' rate school as more important than do boys' mothers, but girls do

not rate school as more important than do boys.

Table 4-6 represents a main effect for grade. Examination of Table 4-6 means

indicate that mothers and children in sixth grade place more emphasis on the five types of

attributions than do mothers and children in either the seventh or the eighth grade.

Seventh grade mothers and children tend to place less emphasis on the five types of

attributions.

Pearson-product moment correlation coefficients presented in Table 4-7 indicate

that mothers and children within the same family do not agree about the causes of

mathematics success. Mothers and children were also not in agreement about the










Table 4-5 Ratings
Success Mean Ratings as a Function of Respondent. Child's Sex. and Type of Attribution
Respondent Sex Ability Effort School Home Luck

Child Boy 5.8 7.5 7.4 6.7 4.2

Girl 3.6 8.8 7.4 5.3 4.6

Mother Boy 5.1 7.4 6.1 6.0 1.7

Girl 3.7 8.1 7.5 5.7 2.8


Table 4-6 Ratings

Success Attribution Means by Grade


Grade Means

6 6.3

7 5.2

8 5.8


children's performance in mathematics. Of the six correlations, only two coefficients

were significant: home training, r = .42; luck, r = .29. Even these were fairly small.

Overall, mothers and children within the same family were in disagreement about the

causes of the children's successful mathematics performance and about how well the

children were actually performing in the class.

For the failure condition, means and standard deviations for each of the

attributional items are presented in Table 4-8. The results of an ANOVA of mothers' and

n's mathematics failure attribution means are presented in Table 4-9. The results for










Table 4-7 Ratings

Pearson Correlation Coefficients Between Mother's and Children's Success Attributions
Within the Family


Ability Effort School Home Luck Evaluation

.19 .21 -.01 .42* .29* .11

* =p<.05

children's mathematics failure attribution means are presented in Table 4-9. The results

in Table 4-9 include a significant Respondent x Type of Attribution interaction and a

significant main effect of attribution. These results indicate disagreement between

mothers and children about the relative importance they place on the five types of

attributions for mathematics failure. Mean attributions as a function of respondent and

type of attribution are reported in Table 4-10. These means show the nature of the

disagreement between mothers and children about the relative importance of the five

types of attributions as causes of mathematics failure. Mothers and children were not in

agreement about the relative importance of the five types of attributions. On average,

mothers rated effort as most important followed by school, home, ability, and luck.

However, children rated home training as most important followed by effort, luck, and

school training (which were rated equally,) then ability.

The largest discrepancy between mothers' and children's attributions for

mathematics failure is how they view effort. Mothers put more emphasis on the effort










Table 4-8 Ratings
Mother's and Children's Mathematics Failure Attributions by Grade and Sex

Grade Sex Subject Statistic Ability Effort School Home Luck


6 Female Child


Male






7 Female






Male


8 Female


Male


M 4.0 5.1 2.6 3.5 2.5


SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD


2.9 3.6 2.4 2.1 2.5

3.4 6.6 2.3 3.5 2.3

2.7 3.5 2.3 2.9 2.4

2.9 3.4 3.9 4.3 4.4

2.8 3.4 3.6 3.8 3.6

2.6 5.4 3.7 3.6 1.2

2.0 3.8 3.0 2.1 0.7

2.0 2.3 3.4 4.4 3.0

1.9 2.2 3.6 1.9 2.2

1.6 7.0 1.6 3.9 1.0

1.5 3.1 1.5 3.0 0.0

2.4 3.6 3.6 3.1 3.3

1.9 3.0 2.7 2.8 2.7

3.2 5.0 5.4 3.1 1.9

2.5 3.8 3.3 2.2 1.5

2.3 3.3 3.9 3.6 3.3

2.2 3.0 3.7 3.8 3.9

2.9 7.1 2.9 1.9 1.0

3.1 3.0 3.1 1.6 0.0

4.5 5.3 3.1 5.0 4.1

3.4 3.1 2.7 3.0 3.2

3.3 6.8 4.1 2.8 1.4

2.3 2.7 3.4 1.9 1.0










Table 4-9 Ratings

Summary of ANOVA for Failure Attributions


df SS MS F p


Source

Between

Grade (G)

Sex (S)

GxS

Error

Within

Respondent (R)

RxG

RxS

RxGxS

Error

Attribution (A)

AxG

AxS

AxGxS

Error

RxA

RxAxG

RxAxS

RxAxGxS

Error


13.34 6.67 0.42 0.6624

21.72 21.72 1.35 0.2508

15.90 7.95 0.50 0.6126

705.90 16.04


2.13 2.13 0.19 0.6636

9.57 4.79 0.43 0.6530

1.25 1.25 0.11 0.7392

8.78 4.39 0.39 0.6761

489.27 11.12

389.59 97.40 14.38 0.0001

34.59 4.32 0.64 0.7445

29.90 7.48 1.10 0.3562

65.01 8.13 1.20 0.3014

1191.71 6.77

268.49 67.12 11.81 0.0001

18.06 2.26 0.40 0.9211

46.27 11.57 2.03 0.0916

45.57 5.70 1.00 0.4364

1000.60 5.69










Table 4-10 Ratings

Failure Means as a Function of Respondent and Type of Attribution


Respondent Ability Effort School Home Luck

Child 3.0 3.8 3.4 4.0 3.4

Mother 2.8 6.3 3.3 3.1 1.4




Table 4-11 Ratings

Pearson Correlation Coefficients Between Mother's and Children's Failure Attributions
Within the Family


Ability Effort School Home Luck Evaluation

.20 .19 .14 -.01 .09 .11

* =p<.05


attribution than do children. Although children view effort as important in mathematics

failure, they tend to place slightly more emphasis on home training. Another major

difference between mothers and children regards the luck attribution. Children rate luck

equally as important as school training while mothers rate luck as the least important

cause of their children's mathematics failure.

Pearson-product moment correlation coefficients presented in Table 4-11 indicate

that mothers and children within the same family were not in agreement when evaluating

the children's mathematics performance and were also not in agreement about the causes

of mathematics failure. Of the six correlations, none of the coefficients were significant.








53

Mothers' Predictions of Their Children's Mathematics Success and Failure Attributions:

Rating Scale Method

Mothers were asked to predict the attributions their children would choose for

mathematics success and failure. Mothers predicted on a 6-point scale how their children

would evaluate the mathematics grade on their most recent report card. For predicted

success, the interviewer asked mothers "Why do you think your child thinks he or she is

doing this well?" For predicted failure, the interviewer asked "Why do you think that

your child thinks he or she is not doing even better?" Mothers predicted on 9-point scales

the relative importance their children would place on the five attributional items. The 9-

point scales ranged from 1 (definitely not a cause) to 9 (definitely a cause). The

attributional items were: ability, effort, school training, home training, and luck. It was

hypothesized that African American mothers will not accurately predict their children's

mathematics success and failure attributions. For the predicted success condition,

Pearson product-moment correlation coefficients between children's actual mathematics

success attributions and mothers' predictions of their children's mathematics success

attributions within the family are presented in Table 4-12. The results of Table 4-12

indicate that mothers cannot accurately predict their children's mathematics success

attributions within the family. Of the six correlations, four of the coefficients were not

significant. The two coefficients that were significant were home training, r= .39; and

luck, r= .29, and these were relatively small. Therefore, mothers were not very accurate

in predicting their children's mathematics success attributions within the family. Mothers










were also unable to accurately predict how their children would evaluate their

mathematics performance.

Means and standard deviations for children's actual mathematics success

attributions and mothers' predicted success attributions are presented in Table 4-13. The

results of an ANOVA of children's actual mathematics success attributions and mothers'

predicted success attributions are reported in Table 4-14.

Table 4-12 Ratings

Pearson Correlation Coefficients Between Children's Success Attributions and Mothers's
Predictions of Their Children's Success Attributions Within the Family


Ability Effort School Home Luck Evaluation

.13 -.01 .23 .37* .29* .16

* =p<.05

There are two types of effects in Table 4-14, effects involving respondent and

effects that do not involve respondent. The latter are analyses of the sum of a mother's

predictions and a child's ratings, and consequently, are not meaningful effects. The

former are comparisons of mothers' predictions and children's ratings, and are meaningful

effects. Consequently, this section will focus on effects involving respondent.

The results in Table 4-14 include a significant Respondent x Type of Attributions

interaction and a significant main effect of attribution suggesting some disagreement

about the relative importance the children place on the five types of attributions and the

mothers' predictions of their children's attributions. Mean attributions as a function of

respondent and type of attribution are reported in Table 4-15. These means show the








55

nature of the disagreement between mother's predicted success attributions and children's

actual success attributions when rating the importance of the five types of attributions.

Mothers' average predicted success attributions tended to agree with children's average

actual success attributions about the relative importance of the five types of attributions

as causes of mathematics success. On average, both mothers and children rate effort as

the most important, followed by school training, home training, ability, and luck.

However, mothers' average predicted success attributions were not accurate in predicting

the emphasis their children would place on the five types of attributions. The most

obvious difference between mothers' average predicted success attributions and children's

actual success attributions involved the ability and luck attributions. Mothers tended to

overestimate the importance their children would place on ability, and underestimate the

importance their children would place on luck. Nevertheless, on average, mothers'

predictions about the emphasis their children would place on effort, school training, and

home training were relatively close to their children's actual mathematics success

attributions.

For the predicted failure condition, Pearson product-moment correlation

coefficients between children's actual mathematics failure attributionsand mothers'

predicted mathematics failure attributions within the family. Only two coefficients were

significant. These were school training, r = .32; home training, r = .43 The four

remaining coefficients, ability, effort, luck, and the evaluation were not significant.










Table 4-13 Ratings
Children's Mathematics Success Attributions and Mother's Predictions of Their Children's
Mathematics Success Attributions by Grade and Sex

Grade Sex Subject Statistic Ability Effort School Home Luck


6 Female Child


Male






7 Female






Male


Female


Male


M 4.1 8.9 8.6 6.5 5.1


SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD


1.5 3.1

6.5 3.4

2.8 2.6

8.0 3.4

1.7 3.7

6.7 4.1

1.7 3.1

3.9 3.4

2.5 2.4

5.0 2.7

3.2 2.4

5.6 4.3

2.3 3.4

5.6 2.8

1.0 2.2

5.7 5.3

3.3 4.1

6.9 3.3

2.5 3.9

6.5 5.0

2.0 3.6

6.3 2.5

2.0 1.9










Table 4-14 Ratings

Summary of ANOVA for Children's Mathematics Success Attributions and Mother's
Predictions of Their Children's Mathematics Success Attributions


Source

Between

Grade (G)

Sex (S)

GxS

Error

Within

Respondent (R)

RxG

RxS

RxGxS

Error

Attribution (A)

AxG

AxS

AxGxS

Error

RxA

RxAxG

RxAxS

RxAxGxS

Error


df SS MS F p



2 84.89 42.44 3.65 0.0340

1 2.79 2.79 0.24 0.6262

2 51.74 25.87 2.23 0.1199

44 511.15 11.61


1 23.26 23.26 3.66 0.0623

2 9.36 4.68 0.74 0.4848

1 5.19 5.19 0.82 0.3715

2 7.79 3.89 0.61 0.5467

44 279.81 6.36

4 1026.74 256.68 40.02 0.0001

8 33.18 4.15 0.65 0.7376

4 62.48 15.62 2.44 0.0490

8 48.12 6.01 0.94 0.4868

176 1128.88 6.41

4 66.52 16.63 3.40 0.0105

8 30.22 3.78 0.77 0.6276

4 39.78 9.95 2.03 0.0917

8 21.16 2.64 0.54 0.8246

176 860.77 4.90










Table 4-15 Ratings

Predicted Success Mean Ratings as a Function of Respondent and Type of Attribution


Respondent Ability Effort School Home Luck

Child 4.7 8.1 7.5 6.0 4.4

Mother 5.3 7.8 7.8 6.1 3.1


Table 4-16 Ratings

Pearson Correlation Coefficients Between Children's Failure Attributions and Mothers
Prediction of Children's Attributions Within the Family


Ability Effort School Home Luck Evaluation

-.19 .26 .32* .43* .12 .16

= p<.05

Means and standard deviations for children's actual mathematics failure

attributions and mothers predicted failure attributions are presented in Table 4-17. The

results of an ANOVA of children's actual failure attributions and mothers' predicted

failure attributions are reported in Table 4-18. There are two types of effects in Table 4-

18, effects involving Respondent and effects that do not involve respondent. The only

meaningful effects are those involving respondent. Consequently this section will focus

on effects involving only the respondent. Examination of Table 4-18 indicates a

significant Respondent x Type of Attribution interaction indicating some discrepancy

about the relative importance children place on the five types of attributions and the

mothers' predictions of their children's attributions. Mean attributions as a function of









Respondent and Type of Attribution are reported in Table 4-19. Table 4-19 shows that

mothers' predicted failure attributions of their children's mathematics performance were

not in agreement for the relative importance of the five types of attribution. Also,

mothers predictions were not in agreement with the emphasis their children placed on

each of the five types of attributions. The only attribution mothers came close to

predicting about their children's mathematics failure was ability. It seems mothers knew

their children would not rate their ability as a cause of mathematics failure.

Children's Prediction's of Their Mother's Mathematics Success and Failure Attributions:

Rating Scale Method

Children were asked to predict the attributions their mothers would choose for the

children's mathematics success and failure. Children predicted on a 6-point scale how

their mothers would rate the mathematics grade on their most recent report card. For

predicted success, the interviewer asked children, "Why do you think your mother thinks

you are doing this well?" For predicted failure, the interviewer asked "Why do you think,

your mother thinks you are not doing even better?" Children predicted on 9-point scales

the relative importance their mothers would place on the five attributional items. The 9-

point scales ranged from 1 (definitely not a cause) to 9 (definitely a cause). The

attributional items were ability, effort, training at school, training at home and luck. It

was hypothesized that African American children will not accurately predict their

mother's mathematics success and failure attributions.

For the predicted success condition, Pearson product-moment correlation

coefficients between mothers' actual mathematics success attributions and children's











Table 4-17 Ratings
Children's Mathematics Failure Attributions and Mother's Predictions of Their Children's
Mathematics Failure Attributions by Grade and Sex


Grade Sex Subject


6 Female Child


SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD


Statistic Ability Effort School Home Luck


M 4.0 5.1 2.6 3.5 2.5


2.1 2.5

3.9 2.3

3.0 2.4

4.3 4.4

3.8 3.6

3.4 3.0

2.0 2.8

4.4 3.0

1.9 2.2

2.4 1.0

1.9 0.0

3.1 3.3

2.8 2.7

3.6 2.4

2.1 1.9

3.6 3.3

3.8 3.9

3.7 2.7

2.8 3.1

5.0 4.1

3.0 3.2

3.0 3.3

2.0 2.6


Male






7 Female







Male


8 Female


Male











Table 4-18 Ratings

Summary of ANOVA for Children's Mathematics Failure Attributions and Mother's
Predictions of their Children's Mathematics Failure Attributions


df SS MS F p


2 27.13 13.57 0.66 0.5203

1 26.28 26.28 1.28 0.2632

2 1.50 0.75 0.04 0.9640

44 900.05 20.46


Source

Between

Grade (G)

Sex (S)

GxS

Error

Within

Respondent (R)

RxG

RxS

RxGxS

Error

Attribution (A)

AxG

AxS

AxGxS

Error

RxA

RxAxG

RxAxS

RxAxGxS

Error


1.87 1.87 0.22 0.6393

13.65 6.82 0.81 0.4496

0.43 0.43 0.05 0.8228

13.74 6.87 0.82 0.4471

368.81 8.38

131.45 32.86 4.21 0.0028

28.21 3.53 0.45 0.8882

13.24 3.32 0.42 0.7910

45.82 5.73 0.73 0.6614

1373.57 7.80

84.55 21.14 3.29 0.0124

34.37 4.30 0.67 0.7181

9.09 2.27 0.35 0.8409

61.62 7.70 1.20 0.3017

1129.99 6.42










Table 4-19 Ratings
Predicted Failure Mean Ratings as a Function of Respondent and Type of Attribution


Respondent Ability Effort School Home Luck

Child 3.0 3.8 3.4 4.0 3.4

Mother 3.3 5.0 4.2 3.3 2.4

Table 4-20 Ratings

Pearson Correlation Coefficients Between Mother's Success Attributions and Children's
Predictions of Mothers's Success Attributions Within the Family


Ability Effort School Home Luck Evaluation

.23 .26 .06 .17 .16 .31*

* = p<.05


predictions of their mothers' mathematics success attributions within the family are

presented in Table 4-20. The results of Table 4-20 indicate that children were not able to

predict accurately their mothers' mathematics success attributions within the family. Of

the six correlations, five were not significant. However, a small significant correlation

was found when children predicted how their mothers would evaluate their mathematics

performance.

Means and standard deviations for mothers' actual mathematics success

attributions and children's predicted success attributions are presented in Table 4-21. The

results of an ANOVA of mothers' actual mathematics success attributions and children's

predicted success attributions are reported in Table 4-22. As in the preceding analyses










Table 4-21 Ratings
Mother's Mathematics Success Attributions and Children's Predictions of Their Mother's
Mathematics Success Attributions by Grade and Sex


Grade Sex Subject


6 Female






Male






7 Female


Male


8 Female


Male


Statistic Ability Effort School Home Luck


Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD


4.9 8.0 8.3 6.9 5.8

3.3 1.9 1.8 1.6 3.5

5.4 8.3 8.3 7.1 3.4

2.8 1.4 1.5 1.9 3.1

6.8 7.7 7.6 8.1 3.0

1.9 2.5 2.8 1.5 3.2

5.8 8.4 5.9 5.8 1.6

3.2 0.9 2.4 2.1 1.7

4.0 7.7 7.0 6.0 3.1

2.1 1.7 3.1 2.0 2.5

1.6 8.0 7.9 4.4 2.0

1.5 1.9 2.0 3.6 1.9

5.3 7.0 7.7 5.6 2.4

.33 2.0 1.5 3.2 2.2

5.2 7.2 6.7 6.0 1.9

2.7 1.6 2.1 1.6 1.8

4.0 8.6 8.3 7.6 2.9

3.6 1.1 1.9 2.5 2.3

4.4 8.1 6.6 5.6 3.3

2.6 2.3 3.0 2.8 3.1

5.2 6.4 7.7 6.8 4.5

2.8 2.3 1.4 2.8 3.5

4.3 6.6 5.9 6.3 1.7

3.0 2.3 2.6 2.2 1.3












Table 4-22 Ratings
Summary of ANOVA for Mother's Mathematics Success Attributions and Children's
Predictions of Their Mother's Mathematics Success Attributions


df SS MS F p


Source
Between

Grade (G)

Sex (S)

GxS

Error

Within

Respondent(R)

RxG

RxS

RxGxS

Error

Attribution (A)

AxG

AxS

AxGxS

Error

RxA

RxAxG

RxAxS

RxAxGxS

Error


82.89 41.44

5.19 5.19

17.83 8.91

499.24 11.35


60.90 60.90

3.37 1.68

1.98 1.98

11.67 5.84

303.89 6.91

1496.43 374.11

33.91 4.24

97.91 24.48

46.01 5.76

938.77 5.33

32.71 8.18

22.26 2.78

10.49 2.62

59.74 7.47

795.55 4.52


3.65 0.0341

0.46 0.5022

0.79 0.4622




8.82 0.0048

0.24 0.7851

0.29 0.5954

0.85 0.4364


70.14 0.0001

0.79 0.6080

4.59 0.0015

1.08 0.3805


1.81 0.1290

0.62 0.7639

0.58 0.6772

1.65 0.1133














Table 4-23 Ratings
Predicted Success Means by Respondent


Respondent Means

Child 6.1

Mother 5.4


involving predictions, only the effects involving respondent are meaningful, and of these

effects, only the respondent effect was significant.

Table 4-23 shows a main effect for respondent indicating a discrepancy between

children's predicted mathematics success attributions and mothers' actual mathematics

success attributions. Children tended to overestimate the emphasis their mothers would

place on the five types of attributions about the children's mathematics success.

For the predicted failure condition, Pearson product-moment coefficients between

mothers' actual mathematics failure attributions and children's predictions of their

mother's mathematics failure attributions within the family are presented in Table 4-24.

The results of Table 4-24 indicate that children were not able to accurately predict their

mothers' mathematics failure attributions nor were they able to predict how their mothers

would evaluate their mathematics performance. Only one of the six correlations was

significant: effort, r = .29, which does not show a strong relationship.










Table 4-24 Ratings

Pearson Correlation Coefficients Between Mother's Failure Attributions and Children's
Predictions of Mother's Failure Attributions Within the Family


Ability Effort School Home Luck Evaluation

.06 .29* .20 .13 .24 .16

* = p<.05

Means and standard deviations for mothers' actual mathematics failure attributions

and children's predicted failure attributions are presented in Table 4-25. The results of an

ANOVA of mothers' actual mathematics failure attributions and children's predicted

failure attributions are reported in Table 4-26. Again, the only meaningful effects are

those involving respondent.

The results of Table 4-26 include a significant Respondent x Type of attribution

interaction indicating some disagreement about the relative importance the children place

on the five types of attributions and the mothers' actual mathematics attributions. Mean

attributions as a function of respondent and type of attribution are reported in Table 4-27.

These means describe the nature of the disagreement between mothers' actual

mathematics failure attributions and children's predicted failure attributions for the five

types of attributions. On average, mothers rated effort as most important, followed by

school, home, ability, and luck. However, children predicted their mothers, on average,

would rate effort as most important, followed by home, ability, luck, and school.

Therefore, children's predictions were not consistent with mothers actual mathematics










Table 4-25 Ratings
Mother's Mathematics Failure Attributions and Children's Predictions of Their Mother's
Mathematics Failure Attributions by Grade and Sex


Grade Sex Subject Statistic Ability Effort School Home Luck

6 Female Child M 3.5 4.8 2.4 3.4 2.4


SD

Mother M

SD

Male Child M

SD

Mother M

SD

7 Female Child M

SD

Mother M

SD

Male Child M

SD

Mother M

SD

8 Female Child M

SD

Mother M

SD

Male Child M

SD

Mother M

SD


3.1 1.8

3.5 2.3

2.8 2.4

4.0 3.2

3.2 3.5

3.6 1.2

2.1 0.7

2.3 3.7

1.7 3.0

3.9 1.0

3.1 0.0

3.0 3.0

2.7 2.8

2.1 1.9

2.2 1.5

2.6 1.6

3.0 1.5

1.9 1.0

1.6 0.0

4.8 2.8

3.0 2.4

2.8 1.4

1.9 1.0












Table 4-26 Ratings
Summary of ANOVA for Mother's Mathematics Failure Attributions and Children's
Predictions of Their Mother's Mathematics Failure Attributions


df SS MS F p


Source

Between

Grade (G)

Sex (S)

GxS

Error

Within

Respondent (R)

RxG

RxS

RxGxS

Error

Attribution (A)

AxG

AxS

AxGxS

Error

RxA

RxAxG

RxAxS

RxAxGxS

Error


13.14

31.77

36.59

768.03


9.07

9.26

4.39

16.48

328.20

494.86

93.83

70.04

22.19

1001.68

184.29

43.22

16.83

43.14

936.16


0.38

1.32

1.05




1.22

0.62

0.59

1.10


21.74

2.06

3.08

0.49


8.66

1.02

0.79

1.02


0.6884

.0.1842

0.3592




0.2760

0.5422

0.4476

0.3402


0.0001

0.0420

0.0176

0.8642


0.0001

0.4259

0.5323

0.4273


6.57

31.77

18.30

17.46


9.07

4.63

4.39

8.24

7.46

123.71

11.73

17.51

2.77

5.69

46.07

5.40

4.21

5.39

5.32












failure attributions in the relative importance they place on the five types of attributions.

Further, children tended to underestimate the emphasis mothers would place on effort for

mathematics failure and overestimate the emphasis their mothers would place on the luck

attribution.


Table 4-27 Ratings
Predicted Failure Means as a Function of Respondent x Tvne of Attribution


Respondent Ability Effort School Home Luck

Child 2.9 3.9 2.7 3.3 2.8

Mother 2.8 6.3 3.3 3.1 1.4


Mothers' and Children's Success and Failure Attributions for Mathematics Performance:

Chip Distribution Method

Mothers and children were asked to evaluate on a six-point scale the mathematics

grade from the children's most recent report card. The six point scale ranged from 1 (not

doing as well as most) to 6 (doing the very best in the class). Mothers were then asked

by the interviewer "Why is your child doing this well?", and children were asked "Why

are you doing this well?" Mothers and children distributed 10 plastic chips on five cards

labeled with five attributions to determine the relative importance of each attribution.

The five attributions were ability, effort, training at school, training at home, and luck.

This procedure resulted in five attribution variables, each with possible ranges of 0 to 10









for both the success and failure conditions. For example, if all ten chips are placed on

one attribution there will be no chips left to place on any of the other attributions.

Relative success in this study is actually mothers' and children's evaluation of the

children's current level of performance. For relative failure, mothers and children were

asked by the interviewer, "Why did your child not do even better?" (for mothers), and

"Why aren't you doing even better?" (for the children). This condition examines relative

failure as a judgment in contrast to the current level of performance.

It was hypothesized that low SES African American mothers will not agree with

their children about the causes of their children's success and failure in mathematics.

For the success condition means and standard deviations for each of the

attributional items are presented in Table 4-28. The results of an ANOVA of mothers'

and children's mathematics success attribution means are presented in Table 4-29. The

chip distribution method has no between subject effects because there were only 10 chips

available for distribution. Therefore, only within subject effects will be reported in this

section. The results in Table 4-29 include a significant main effect of attribution and a

significant Respondent x Type of Attribution interaction indicating some discrepancies

between mothers and children about the relative importance they place on the five types

of attributions for success in mathematics.

Mean attributions as a function of respondent and type of attribution are reported

in Table 4-30. These means show the nature of disagreement between mothers and

children about the relative importance of the five types of attributions. Mothers and

children differed slightly about the relative importance of the five types of attributions as










causes of mathematics success. On average, mothers weighted effort as most important,

followed by school and home training which were weighted equally, followed by ability,

then luck. Children on average, weighted effort as most important, followed by school

training, home training, luck and ability. Mothers and children also differed in the

emphasis they placed on the five types of attributions. Mothers placed more weight than

the children on the effort attribution, while children placed more emphasis on the luck

attribution for mathematics success. Also, children believed that school training was

more important for mathematics success than their mothers. Mothers and children were

similar in their weightings of ability and home training as causes of mathematics success.

Overall, mothers and children agreed that effort was most important for mathematics

success. However, mothers and children differed in the emphasis placed on effort. Also

mothers and children differed in their beliefs about the importance of school training and

luck as causes of mathematics success.

Pearson-product moment correlation coefficients presented in Table 4-31 indicate

that mothers and children within the family do not agree about the causes of mathematics

success. Mothers and children were also not in agreement about the children's

performance in mathematics. Of the six correlations, only one coefficient was

significant: ability r = .50, indicating a moderate relationship. However, there was very

little agreement for the remaining attributions.











Table 4-28 Weightings
Mother's and Children's Mathematics Success Attributions by Grade and Sex


Grade Sex Subject


6 Female Child


Male






7 Female






Male


8 Female


Male


Statistic Ability Effort School Home Luck


M 1.0 3.3 2.5 1.8 1.5


SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD


0.8 1.0 0.6 1.3 1.3

0.5 5.3 1.5 2.8 0.0

1.0 2.2 1.7 1.5 0.0

1.0 3.3 2.8 1.2 1.7

1.7 1.5 0.8 1.2 1.5

1.7 5.2 1.5 1.2 0.5

3.1 2.9 1.4 1.5 1.2

0.8 4.0 3.0 1.0 1.2

1.1 1.3 1.5 1.7 1.6

0.7 5.0 2.2 1.8 0.3

1.2 2.4 1.6 1.5 0.8

1.2 4.2 1.8 2.0 0.9

2.0 2.4 1.7 1.9 1.1

1.8 3.7 2.2 2.1 0.3

3.1 2.7 1.6 2.1 0.8

0.5 4.8 2.3 1.5 1.0

1.0 1.3 1.7 1.7 2.0

0.8 8.0 0.8 0.5 0.0

1.5 2.4 1.5 1.0 0.0

1.2 3.9 2.9 1.3 0.7

1.3 2.3 1.5 1.3 1.1

1.4 3.2 2.6 2.6 0.1

1.7 1.7 1.4 1.7 0.4











Table 4-29 Weightings
Summary of ANOVA for Success Attributions


Source df SS MS F p

Within

Attribution (A) 4 697.97 174.49 40.74 0.0001

AxG 8 14.34 1.79 0.42 0.9089

AxS 4 35.30 8.82 2.06 0.0880

AxGxS 8 54.41 6.80 1.59 0.1313

Error 176 753.85 4.28

RxA 4 57.77 14.44 5.17 0.0006

RxAxG 8 16.01 2.00 0.72 0.6768

RxAxS 4 22.23 5.56 1.99 0.0980

RxAxGxS 8 25.16 3.14 1.13 0.3480

Error 176 491.61 2.80


Table 4-30 Weightings

Success Means as a Function of Respondent and Type of Attribution


Respondent Ability Effort School Home Luck

Child 0.9 3.9 2.5 1.4 1.2

Mother 1.1 5.0 1.8 1.8 0.2










Table 4-31 Weightings

Pearson Correlation Coefficients Between Mothers' and Children's Success Attributions
Within the Family
Ability Effort School Home Luck Evaluation

.50* .19 .05 .12 .10 .18

=p<.05

For the failure condition, means and standard deviations for each of the

attributional items are presented in Table 4-32. The results of an ANOVA of mothers'

and children's attribution means are presented in Table 4-33. The results in Table 4-33

include a significant Respondent x Type of Attribution x Grade interaction and a main

effect for attribution, indicating disagreement between mothers and children in each grade

level about the relative importance of the five types of attributions as causes of

mathematics failure. Means as a function of respondent, type of attribution, and grade are

reported in Table 4-34. These means describe the nature of the disagreement between

mothers and children in each grade level. On average, sixth grade children believed the

home was the most important cause of mathematics failure, while seventh and eighth

grade children believed effort was the most important cause. Sixth, seventh, and eighth

grade mothers, on average, all agreed that effort was the most important cause of their

children's mathematics failure. However, these mothers disagreed about the importance

of the remaining types of attributions. Eighth grade mothers and children were more in

agreement about the causes of mathematics failure than were sixth and seventh grade

mothers and children. Sixth and seventh grade mothers tended to weight home training











Table 4-32 Weightings
Mother's and Children's Mathematics Failure Attributions by Grade and Sex


Grade Sex

6 Female






Male






7 Female


Subject Statistic

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD


Male


8 Female


Male


Ability Effort School Home Luck

1.8 2.3 2.5 2.5 1.0

1.3 2.6 1.0 0.6 1.2

1.3 5.0 2.8 0.0 1.0

1.3 0.8 2.1 0.0 1.4

2.2 1.7 1.7 1.8 2.7

4.0 2.3 2.0 2.5 2.5

1.0 4.3 1.7 1.3 1.7

1.7 4.6 4.1 2.8 2.6

1.7 2.8 1.1 3.0 1.4

1.6 3.3 1.4 4.0 2.0

3.1 3.3 1.8 1.1 0.7

2.8 3.2 1.8 1.4 1.3

1.5 4.0 2.0 1.3 1.2

3.0 3.8 2.1 2.0 1.9

0.9 6.0 1.5 1.2 0.4

1.7 3.0 1.9 1.9 1.4

1.5 5.0 2.5 0.0 1.0

3.0 5.8 5.0 0.0 2.0

1.8 0.3 3.5 2.0 2.5

1.7 0.5 3.9 1.8 2.5

0.6 4.7 1.2 1.9 1.6

1.0 3.0 1.7 1.9 2.8

1.3 5.6 1.5 1.4 0.2

1.8 3.1 1.5 1.6 0.6










Table 4-33 Weightings

Summary of ANOVA for Failure Attributions


Source df SS MS F p

Within

Attribution (A) 4 318.53 79.63 9.01 0.0001

AxG 8 26.76 3.34 0.38 0.9310

AxS 4 50.99 12.75 1.44 0.2219

AxGxS 8 73.16 9.15 1.04 0.4115

Error 176 1555.04 8.84

RxA 4 20.53 5.13 0.93 0.4502

RxAxG 8 92.03 11.50 2.08 0.0405

RxAxS 4 41.73 10.43 1.88 0.1156

RxAxGxS 8 61.50 7.69 1.39 0.2050

Error 176 975.61 5.54


as the least important cause while eighth grade mothers believed home training was an

important cause of their children's mathematics failure. Eighth grade children believed

home training was the least important cause of mathematics failure but sixth grade

children believed home training was the most important cause. The most striking

difference between mothers' and children's mathematics failure attributions is the

emphasis on effort. The children's emphasis on effort increases from sixth to eighth

grade while their mothers' emphasis decreases between the sixth and eighth grade.

In summary, mothers and children at all grade levels agreed that effort was the

most important cause of mathematics failure except for sixth grade children. However,










Table 4-34 Weightings

Failure Means as a Function of Respondent. Type of Attribution. and Grade


Grade Respondent Ability Effort School Home Luck

6 Child 1.9 1.9 2.1 2.2 1.8

Mother 1.1 4.7 2.2 0.7 1.3

7 Child 1.6 3.4 1.6 2.2 1.3

Mother 2.1 4.7 1.7 1.1 0.6
8 Child 1.1 4.9 1.9 0.9 1.3

Mother 1.5 2.9 2.5 1.7 1.4


mothers' endorsement of effort decreased as their children progressed through the middle

grades while their children's endorsement of effort increased.

Pearson-product moment correlation coefficients presented in Table 4-35 indicate

that mothers and children within the same family do not agree about the causes of

mathematics failure and were also not in agreement about the evaluation of the children's

mathematics performance. Of the six correlations, only one correlation was significant:

effort, r = .32, which is a relatively small relationship.

Table 4-35 Weightings

Pearson Correlation Coefficients Between Mothers' and Children's Failure Attributions
Within the Family


Ability Effort School Home Luck Evaluation

.19 .32* .25 .13 -.06 .18

* = p<.05








78

Mother's Predictions of Their Children's Mathematics Success and Failure Attributions:

Chip Distribution Method

Mothers were asked to predict the attributions their children would choose for

mathematics success and failure. Mothers predicted on a 6-point scale how their children

would evaluate the mathematics grade on their most recent report card. For predicted

success, the interviewer asked mothers "Why do you think your child thinks he or she is

doing this well?" For predicted failure, the interviewer asked "Why do you think your

child thinks he or she is not doing even better?" Mothers predicted by distributing 10

plastics chips on five cards labeled with five attributions to determine the relative

importance of each attribution. The five attributions were ability, effort, training at

school, training at home, and luck. This procedure resulted in five attribution variables,

each with possible ranges of 0 to 10 for both success and failure conditions.

It was hypothesized that African American mothers will not accurately predict

their children's mathematics success and failure attributions. For the predicted success

condition, Pearson-product moment correlation coefficients between children's actual

mathematics success attributions and mothers' predictions of their children's mathematics

success attributions within the family are presented in Table 4-36. The results in Table 4-

36 indicate that mothers cannot accurately predict their children's mathematics success

attributions within the family. Mothers were also unable to accurately predict how their

children would evaluate their mathematics performance. Of the six correlations, none of

them were significant.










Table 4-36 Weightings
Pearson Correlation Coefficients Between Children's Actual Success and Mothers'
Predicted Success Attributions Within the Family


Ability Effort School Home Luck Evaluation

.33 .19 -.04 .02 -.10 .10

= p<.05


Means and standard deviations for children's actual mathematics success

attributions and mothers' predicted success attributions are presented in Table 4-37. The

results of an ANOVA of children's actual mathematics success attributions and mothers'

predicted success attributions are reported in Table 4-38.

The results in Table 4-38 include a significant Respondent x Type of Attribution

interaction and a significant main effect of attribution suggesting disagreement about the

relative importance the children place on the five types of attributions and the mothers'

predictions about their children's attributions. Mean attributions as a function of

respondent and type of attribution are reported in Table 4-39. These means show the

nature of the disagreement between mothers' predicted success attributions and children's

actual success attributions when weighting the importance of the five types of

attributions. On average, mothers' predictions that children will place the most emphasis

on effort and school is correct but mothers overestimate the amount of emphasis their

children will place on effort.

On average, mothers predicted their children would weight effort as most

important followed by school, ability, home, and luck. Children's actual success









attributions on average were effort as most important, followed by school, home, luck,

then ability.

Although mothers predicted correctly on the importance their children would give

effort and school training, mothers did not predict accurately on the emphasis their

children would place on these two attributions. Mothers underestimated the emphasis

their children placed on school training and overestimated how their children would

weight effort. Overall, mothers were not very accurate in predicting their children's

success attributions.

For the predicted failure condition, Pearson-product moment correlation

coefficients between children's actual mathematics failure attributions and mothers'

predictions of their children's mathematics failure attributions within the family are

presented in Table 4-40. The results in Table 4-40 show that mothers cannot accurately

predict how their children would evlauate their mathematics performance nor could they

predict their children's mathematics failure attributions within the family. Of the six

correlations, only two correlations were significant: effort, r = .52, indicating a moderate

relationship, and home training, r = .38, indicating a small relationship.

Means and standard deviations of children's actual failure attributions and

mothers' predictions are presented in Table 4-41. The results of an ANOVA of children's

actual mathematics failure attributions and mother's predictions of their children's

mathematics failure attributions presented in Table 4-42 indicate no significant

interactions.











Table 4-37 Weightings
Children's Mathematics Success Attributions and Mother's Predictions of Their Children's
Mathematics Success Attributions by Grade and Sex


Grade Sex

6 Female






Male






7 Female


Subject Statistic

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD


Male


8 Female


Male


Ability Effort School Home Luck

1.0 3.3 2.5 1.8 1.5

0.8 1.0 0.6 1.3 1.3

0.8 6.3 1.8 1.3 0.0

1.5 2.9 1.3 1.9 0.0

1.0 3.3 2.8 1.2 1.7

1.7 1.5 0.8 1.2 1.5

1.7 4.3 2.3 0.8 .83

4.1 3.3 2.1 1.6 1.3

0.8 4.0 3.0 1.0 1.2

1.1 1.3 1.5 1.7 1.6

0.9 4.8 3.4 0.7 .20

1.5 2.1 2.5 1.2 .63

1.2 4.2 1.8 2.0 .92

2.0 2.4 1.7 1.9 1.1

1.4 4.5 2.1 1.3 .67

2.1 3.0 1.6 2.3 1.2

0.5 4.8 2.3 1.5 1.0

1.0 1.3 1.7 1.7 2.0

2.5 6.3 0.5 0.5 .25

5.0 4.8 1.0 1.0 .50

1.2 3.9 2.9 1.3 .71

1.3 2.3 1.5 1.3 1.1

1.6 3.7 2.1 1.8 .86

1.7 1.9 1.6 1.6 1.6












Table 4-38 Weightings

Summary of ANOVA for Children's Mathematics Success Attributions and Mother's
Predictions of their Children's Mathematics Success Attributions


Source df SS MS F p

Within

Attribution (A) 4 672.50 168.13 35.64 0.0001

AxG 8 11.85 1.48 0.31 0.9600

AxS 4 19.70 4.93 1.04 0.3859

AxGxS 8 43.55 5.44 1.15 0.3298

Error 176 830.15 4.72

RxA 4 45.11 11.28 3.00 0.0198

RxAxG 8 25.26 3.16 0.84 0.5677

RxAxS 4 14.19 3.55 0.94 0.4393

RxAxGxS 8 12.27 1.53 0.41 0.9147

Error 176 660.82 3.75


Table 4-39 Weightings

Predicted Success Means as a Function of Respondent and Type of Attribution


Respondent Ability Effort School Home Luck

Child 0.9 3.9 2.5 1.4 1.2
Mother 1.4 5.0 2.0 1.0 0.5










Table 4-40 Weightings

Pearson Correlation Coefficients Between Children's Actual Failure and Mothers'
Predicted Failure Attributions Within the Family


Ability Effort School Home Luck Evaluation

.13 .52* .05 .38* .19 .10

=p<.05


Children's Predictions of Their Mother's Mathematics Success and Failure Attributions:

Chip Distribution Method

Children were asked to predict the attributions their mothers would choose for the

children's mathematics success and failure. Children predicted on a 6-point scale how

their mothers would evaluate the mathematics grade on their most recent report card. For

predicted success, the interviewer asked children, "Why do you think your mother thinks

you are doing this well?" For predicted failure, the interviewer asked "Why do you think

your mother thinks you are not doing even better?" Children predicted by distributing 10

plastic chips on five cards labeled with five attributions to determine the relative

importance of each attribution. The five attributions were ability, effort, training at

school, training at home, and luck. This procedure resulted in five attribution variables,

each with possible ranges of 0 to 10 for both success and failure conditions.

It was hypothesized that African American children will not accurately predict

their mothers mathematics success and failure attributions. For the predicted success










Table 4-41 Weightings
Children's Mathematics Failure Attributions and Mother's Predictions of Their Children's
Mathematics Failure Attributions by Grade and Sex


Grade Sex Subject


6 Female






Male






7 Female


Male


Female


Male


Statistic Ability Effort School Home Luck


Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD


1.8 2.3 2.5 2.5 1.0

1.3 2.6 1.0 0.6 1.2

1.8 3.5 2.3 1.8 0.8

1.7 2.6 2.2 2.4 1.5

2.2 1.7 1.7 1.8 2.7

4.0 2.3 2.0 2.5 2.5

1.5 2.7 2.3 2.0 1.5

1.6 2.2 3.9 2.3 1.8

1.7 2.8 1.1 3.0 1.4

1.6 3.3 1.4 4.0 2.0

2.6 2.1 1.5 2.3 1.5

2.8 2.6 1.8 2.7 1.6

1.5 4.0 2.0 1.3 1.2

3.0 3.8 2.1 2.0 1.9

0.9 3.6 3.3 1.6 0.7

1.4 2.6 3.0 1.8 1.4

1.5 5.0 2.5 0.0 1.0

3.0 5.8 5.0 0.0 2.0

1.8 5.5 0.3 1.0 1.5

2.4 5.3 0.5 1.2 1.7

0.6 4.7 1.2 1.9 1.6

1.0 3.0 1.7 1.9 2.8

1.6 4.2 1.8 1.8 0.6

2.4 2.7 2.0 1.6 1.2












Table 4-42 Weightings

Summary of ANOVA for Children's Mathematics Failure Attributions and Mother's
Predictions of Their Children's Mathematics Failure Attributions


Source df SS MS F p

Within

Attribution (A) 4 234.80 58.70 6.29 0.0001

AxG 8 94.88 11.86 1.27 0.2612

AxS 4 7.01 1.75 0.19 0.9445

AxGxS 8 74.64 9.33 1.00 0.4377

Error 176 1641.81 9.33

RxA 4 4.14 1.03 0.19 0.9415

RxAxG 8 26.10 3.26 0.61 0.7681

RxAxS 4 18.27 4.57 0.86 0.4921

RxAxGxS 8 16.13 2.02 0.38 0.9315

Error 176 940.21 5.34


condition, Pearson-product moment correlation coefficients between mother's actual

success attributions and children's predictions of their mothers' mathematics success

attributions within the family are presented in Table 4-43. The results in Table 4-43

indicate that children cannot accurately predict their mothers' mathematics success

attributions within the family nor could the children predict how their mothers would

evaluate their mathematics performance. Of the six correlations, none of them were

significant.










Table 4-43 Weightings
Pearson Correlation Coefficients Between Mother's Actual Success and Children's
Predicted Success Attributions Within the Family

Ability Effort School Home Luck Evaluation

.09 .12 .12 .02 -.03 .23

= p<.05

Means and standard deviations for mothers actual mathematics success

attributions and children's predictions of their mother's mathematics success attributions

are presented in Table 4-44. The results of an ANOVA of mothers' actual mathematics

success attributions and children's predicted success attributions are reported in Table 4-

45.

The results in Table 4-45 include a significant Respondent x Type of Attribution

interaction and a significant main effect of attribution indicating disagreement about the

relative importance the mothers place on the five types of attributions and children's

predictions about their mothers' attributions. There are two types of effects in Table 4-45,

effects involving respondent and effects not involving respondent. Only the effects

involving respondent will be reported in this section.

Mean attributions as a function of respondent and type of attribution are presented

in Table 4-46. These means show the nature of the disagreement between mothers' actual

success attributions and children's predicted success attributions when weighting the

importance of the five types of attributions. Children on average predicted correctly on

the importance their mothers would place on the five types of attributions for predicted










Table 4-44 Weightings
Mother's Mathematics Success Attributions and Children's Predictions of Their Mother's
Mathematics Success Attributions by Grade and Sex

Grade Sex Subject Statistic Ability Effort School Home Luck


6 Female






Male






7 Female


Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD

Child M

SD

Mother M

SD


1.5 3.5 2.8 1.0 1.3

1.0 1.3 0.5 1.2 1.0

0.5 5.3 1.5 2.8 0.0

1.0 2.2 1.7 1.5 0.0

2.3 3.7 2.0 1.8 0.2

1.5 1.2 1.7 1.0 0.4

1.7 5.2 1.5 1.2 0.5

3.1 2.9 1.4 1.5 1.2

0.9 3.9 3.6 1.2 0.4

1.3 1.4 1.7 1.8 1.0

0.7 5.0 2.2 1.8 0.3

1.2 2.4 1.6 1.5 0.7

0.9 4.3 2.8 1.8 0.1

2.1 2.1 2.1 1.7 0.3

1.8 3.7 2.2 2.1 0.3

3.1 2.7 1.6 2.1 0.8

0.3 5.5 3.3 1.0 0.0

0.5 1.3 1.3 2.0 0.0

0.8 8.0 0.8 0.5 0.0

1.5 2.4 1.5 1.0 0.0

1.8 3.2 2.8 1.5 0.6

1.3 1.3 1.3 0.9 1.1

1.4 3.2 2.6 2.6 0.1

1.7 1.7 1.4 1.7 0.4


Male


8 Female


Male




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
INGEST IEID EBEGR7NIU_70WVA0 INGEST_TIME 2013-11-15T22:26:53Z PACKAGE AA00014227_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES