An Achievement History of Chapter 220 Wisconsin
Part #2
Copyright by Pamela Jane Sampson-Malone, 1998
All Rights Reserved
SOCIOECONOMIC STATUS, FAMILY CULTURE, AND ACADEMIC ACHIEVEMENT: A STUDY OF BLACK AND WHITE PUPIL PERFORMANCE AT AN INTERRACIAL SCHOOL by:
Pamela Jane Sampson-Malone
E-mail:
Web site:
Phone: 414-934-9804
FAX: 414-934-9878
-------------------------------------------
Page 51
TABLE 2.1
CHAPTER 220 MINORITY TRANSFERS, 1976-1977 3
District 220 Transfers Percent Enrollment
_
Brown Deer 96 3.3
Fox Point-Elayside 8 .8
Maple Dale-Indian Hills 8 1.0
Nicolet and Feeder Schools 27 1.2
Glendale-River Hills 32 2.3
Greendale 35 .8
Shorewood 64 2.8
Whitefish Bay 58 1.7
TOTALS/AVERAGES 328 1.8
52
TABLE 2.2
1977-1978 SUBURBAN MINORITY ENRoLLMENl44
Total Total Total Black Minority Black
Enroll- Trans- Minori- Enroll- % of % of
District meets fers ties ment Total Total
Brown Deer 2,921 116 315 275 10.7 9.40
Cudahy* 4,364 0 132 3 3.0 .07
Franklin* 2,648 0 32 3 1.2 .11
Greendale 4,123 77 176 75 4.3 1.82
Greenfield* 4,214 0 103 12 2.4 .28
Nicolet ~ Feeders 5,240 124 371 272 7.1 5.19
Oak Creek 4,739 33 187 47 3.9 .99
St. Francis* 1,629 0 36 0 2.2 .00
Shorewood 2,425 95 187 118 7.7 4.87
So. Milwaukee 4,112 10 94 11 2.3 .27
Wanwatosa 8,599 105 311 189 3.6 2.20
West Allis*-
West Milwaukee 10,795 0 146 11 1.4 .10
Whitefish Bay 3,184 68 122 77 3.8 2.42
Whitnall 2,549 17 37 1.5 .63
_ _ _ _ ~
SUBURBAN TOTALS 61,542 645 1,798 1,109
*-non-participating districts
Milwaukee students must apply to be accepted to Nicolet. Prior school records regarding disciplinary problems, grades, and attendance are reviewed. Acceptance to the program is not predicated on grades alone. Applicants are accepted based on an evaluation of whether they
3.7
,
53 have the necessary skills to perform the coursework and ability to adapt (if necessary) to the social and educational environment. In other words, can they "succeed" at Nicolet? There is no standard for-mat for these evaluations. The ''nulti-ethnic" counselor, Ms. Barbara Hamilton, who reviewed the applications, views parent involvement as one of the key elements in dete m ining whether students will succeed. She felt that parents of 220 students, many of whom she had met with, tended to have a high sense of self-esteem and self-direction. The applica-tion process itself is indicative of her assessment. Parents must call the individual school to first find out how to apply. In some cases, prior scholastic performance may be indicative of a student's ability to successfully complete high school at Nicolet. In other instances, a student's own desire to succeed, even though his/her prior grades may have been average, is the basis for acceptance.
Nicolet only accepts 220 students as freshmen. Siblings of Nicolet students and graduates are given first priority. The school's intentions are to accept thirty new students each year.
SOCIAL CLASS AND ACAD~EC ACME
'"Family background,' t '"social class,'' and "economic status" are often used interchangeably by social scientists. This interpretation of these complex concepts often poses serious problems in terms of the comparability and interpretation of research findings. The term "social class" has been defined in a variety of ways. Disputes about
54
the legitimacy of this term have been many and heated.45 Some scholars deny the existence of social classes in America. Jencks (1972) stated that:
The te m family background can itself be somewhat mis-leading, since differences between families derive not just from differences between neighborhoods, regions, schools, and all other experiences that are the same for children in the same family.46 Socioeconomic status also has various interpretations. Definitions of SES will be discussed in greater detail shortly.
By "social class" Sexton meant the position which families occupy in society and the very strong tendency for children to be molded into the same position by influences in the home, the community, and the school.47 The Marxian view of social class, on the other hand, is predicated upon the objective classification of an aggregate of people with reference to their similar reactions to the means of production and the concept of struggle which ultimately embodies the ideology of class consciousness.48
By "family background," Jencks (1972) referred to all the environ-mental factors that make brothers and sisters more alike than random individuals. "Some of these factors are economic, while some are not."49 He further stated that there can be great variation in "family background" among children who come from the same social or economic class.
Like definitions of SES, family background, and social class, conceptualizations of academic achievement vary widely. White (1982) stated that academic achievement refers to a broad range of social
55
process variables including class rank, grades, IQ scores, and standard-ized test scores. Many studies of student performance also examine the amount of education received at the post-secondary and college levels. Curriculum placement and academic awards have also been con-sidered indicators of student achievement.
Four approaches to the study of the relationship between students' home environments and their academic performance were identified by Iverson and Walberg (1982). The first is the sociological survey, which focuses on the effects of socioeconomic status variables. Warner, Meeker, and Eell's (1949) Index of Status Characteristics and Hollinghead's Two Factor Index of Social Position are two of the most well-known measures of SES, but are by no means the ones most frequently used. O.D. Duncan's Socioeconomic Index and the North/ Hatt ONORC) Index are the most widely employed measures of SES and are generally considered to be superior to other measures for most survey and large-sample situations. The U.S. Census Socioeconomic Status Scores and Alba M. Edwards' Social-Economic Grouping of Occupations are two other well-known social status scales. Single indica-tors of SES like family income, education, and occupation, are some-times used. Of the single indicators, occupation, because of its high correlation with both education and income and the fact that occupa-tional prestige ratings are highly stable, is considered the best single predictor of socioeconomic status. Frequently, the effects of status-related characteristics such as family size are examined in sociological surveys.
56
The effects of birth order, spacing of children, and related family structural characteristics are investigated in Constellation studies. Family SES and Constellation studies are less expensive proxies for aspirations, conditions, and processes in the home which are conducive to learning and less valid for predictive validity and psychological theory than the other two approaches.50
Dissertations by Dave (1963) and Golf (1564) provided the basis for the analysis of the parental and parent-child interactive behaviors that are the focus of the Chicago School. Bloom's assumption that the total environment surrounding an individual is composed of a complex network of forces and that a subset of this total network of forces is related to each human characteristic , was the basis for selection of the process variables chosen for investigation by Dave and Wolf. Wolf used thirteen process characteristics to identify 'dress for achievement motivation" "press for language," and "provisions for general learning." The three process variables accounted for 49% of the total variation in intelligence.51
Dave used twenty-one process characteristics to define six press variables believed to determine "the educational environment of the family."52 The six press variables were: achievement press, language models, intellectuality in the home, academic guidance, activeness in the family, and work habits in the family. Dave found different press variables to be related to various aspects of achievement. Family edu-cational environment was found to be related to over 50% of the variation in scores for problem solving in reading, arithmetic, and word
57
knowledge, but only accounted for 31% of the variance in arithmetic computation scores. Earlier Chicago school research focused on all of these presses. Later studies, however, concentratedon the presses for academic guidance, achievement (both for the parent and child), and activeness of the family because these aspects of the home environment are most easily influenced by intervention programs. Language and intellectuality are less subject to change, involve parental status more than behavior, and appear to be less associated with academic achieve-ment.53 The works of Kelvin Majoribanks (1963 and 1972), and Herbert Walberg (1976) characterize Chicago school research.
Studies within the British school are differentiated from those within the Chicago school in that they concentrate on parental ex-periences, attitudes, and material conditions in the home rather than on behavioral processes and are often concerned with parental attitudes towards and experiences with school practices. The home assessment measures used,such as the Survey of Parents of Primary School Children, typically include questions such as, ''bat do you feel about the ways teachers control the children at (present school)?" and '~as the head teacher, or any other teacher, talked to you about the methods they use at (present school)?"54 Typical questions asked in the Index of Edu-cational Environment, a commonly used instrument employed by Chicago school researchers are, 'fib you read to the child?" and '~ho makes the plans (for family vacations)?"55 Studies by Claeys and DeBoerk (1976), Schaefer (1977), and Frazier (1959 ) who used reading habits of parents as a home environment measure, are classified as British school
58
stories.
Most of the literature discussed in this review can be broadly classified as sociological surveys and Chicago and British school re-search. Distinctions between types of studies, however, should not be taken too far. Many investigations of the relationship between characteristics of students' families and academic achievement examine both the effects of SES and family background on student performance as well as relationships between socioeconomic status and character-istics of the home environment.
The works of Warner, Havighurst, and Loeb (1944), Hollingshead (1949), Sexton (1961) and Bossard and Boll (1966) provide insight into how social class influences student achievement. Hollingshead's ana-lysis indicated that the effects of SES are mediated through a number of factors and processes including aspects of family background, the attitudes and expectations of teachers and administrators, and direct parental influence in school practices. He believed the motivation of adolescents towards high school and education in general to be directly related to the grades they received and that educational motivation was derived from students' experiences in their classes and family environment:
The class I and II boys and girls know that high grades are necessary if they are to achieve the educational goals set for them by their family and class. Parents, friends of parents, brothers, sisters, and relatives who have been outstanding students in Elmtown High have set precedents they are urged to follow; for most, high school is merely a preparatory step for college. Then too, parents and relatives who have achieved prominent positions in the community, expect them to be leaders. Stimulated by this interest and these
59
examples, thief generally responded by aiming for greater achievement.
The assumption that the majority of upper class students achieve at higher levels than lower class pupils because they are more motivated is not totally accurate. While class related aspects of the home en-vironment including intellectuality of the home have been found to influence achievement motivation, "lower class culture" cannot, by any means, be held totally responsible for low achievement. Warner, Havig-hurst and Loeb argued that students from the lower classes did not have all of the educational opportunities they or their parents desired. They cited the frequency with which "lack of money" was given as a reason for quitting school, the sharp increase in high school and college en-rollment that resulted from the stablishment of the National Youth Ad-ministration student-aid program in 1935, and the fact that parents incurred a substantial out-of-pocket cost even when their children at-tended "free" schools as evidence in support of their position S7 Low economic status affects student achievement not only because pupils in this strata do not have access to better schools as discussed in Chapter I, but because the fact that many students, historically and presently, have worked in order to assist their families financially. Hollingshead stated that many parents in the lower classes actually encouraged their children to quit school and work.
A by no means negligible element is a teacher's expectation that the class I and II child will "make good"; and she helps him realize this goal, for, after all, his parents may "help" a teacher or cause "trouble" very easily. These factors react in subtle ways to produce
60
high grades and leadership in extracurricular activities in classes I and II.58
Warner, Havighurst, and Loeb (1944) and Bossard and Boll (19663 also provided evidence that teacher expectations are influenced by students' social class backgrounds. In Old City, the setting of the Warner, Havighurst and Loeb study, teachers placed students into cur-riculum sections based supposedly on their assessments of students' abilities. When asked whether there was "much class feeling in the school," a junior high school teacher said:
Oh, yes, there is a lot of that. We try not to have it as much as we can, but of course we can't help it. Nbw, for instance, even in the sections we have, it is evident. Sec-tions are supposed to be made up just on the basis of records in school but it isn't and everybody knows it isn't. I know right in my own section A I have children who ought to be in B section, but they are little socialites and so they stay in A. I don't say there are children in B who should be in A, but in the A section there are some who shouldn't be there. We hang discussed it in faculty meetings but nothing is ever done. The principal expressed a belief that children from the "best" families usually did the best work because their families are capable and edu-cated and they enter schools with knowledge that students from other social classes do not have. Based on statements by principals and teachers and the distribution of students in sections by social class, Warner, Havighurst, and Loeb concluded that the ranking of students in classrooms was clearly influenced by status considerations.
Rejecting the use of a single factor index, Hbllingshead's analysis was based on community rater's classifications of families according to five general criteria. These criteria were (1) the way a family lived, which included place of residence, type of dwelling, and furnishin (2) income and material possessions; (3) whether parents participa in community affairs, politics, religious activities--were "civic," "radical," "conservative," "a good community man," didn't give a d.. about education"; (4) family background, including ancestry, kin and national origin; and (5) reputation or prestige.60
Families were divided into five classes. Class I families ha' highest incomes in the community, which were reported to be at lea' $5,000 annually, with some in the S25,000 to $35,000 per year brac} Their wealth had for the most part been attained through investment land, securities, and ownership of businesses. The status and weal of Class I members was usually inherited, not achieved through thei efforts. The inheritance of abilities and characteristics was stro emphasized and acquired traits were believed to represent the outwa pression of "hereditary qualities.''61 Formal education was not high yarded as a tool for a professional career or for the acquisition o: knowledge in the traditional sense. This attitude toward education was no doubt due to the relative unimportance of a college educatior the acquisition of wealth and social mobility in this group. Fewer one-half of Class I members over 60 years of age had attended colle' universities. Nearly all of the middle aged persons graduated from school, but relatively few graduated from college. The younger peop completed high school, but only about one-third of the women and one half of the men completed college.62 All Class I families were home owners and lived in the most exclusive residential areas. Class I
62
individuals were not as active in community affairs or politics as Class II members but they exercised behind the scenes control of the political structure. Almost all Class I families belonged to the country club and most of the women also belonged to the exclusive "Friday Morn-ing Club." Leisure activities took up the majority of this exclusive two percent of the community's time.
Nearly one-half of Class II families attained their positions through their own efforts. The remaining families inherited their posi-tions, but did not have sufficient "breeding" to be accepted into Class I. Annual family incomes ranged between $3,000 and $10,000 in 1941. Income was earned by male heads of families who either owned their own businesses, were salaried executives for enterprises owned by Class I families, were salaried professionals in a public office, or were inde-pendent professionals, such as attorneys, medical doctors, and dentists. According to Hbllingshead, Class II men concentrated on "the aggressive manipulation of economic and political processes." Prestige in Class II was as much dependent upon civic leadership as economic success. Members of Class II were very active in community affairs and participated in organizations such as the Chamber of Commerce, Masonic dodge, and the Women's Christian Temperance Union. Seventy-seven percent of the leader-ship positions in civic organizations were held by members of Class II. Six percent, twelve percent, and five percent of these positions were held by individuals in Classes I, III, and IV respectively. Class II members viewed education as the principal vehicle for obtaining success and emphasized the need for their children to pursue a college education.
63
The adults in this group were the most highly educated in Elmtown. Four out of five parents completed high school and one-half of them at-tended college. One-half of Class II families lived in the "best" residential areas and 90% were home owners.
In Class III, annual family incomes ranged from S2,000 to $4,000. Forty-two percent of these families owned farms,small businesses, or were independent professionals. The remaining 58% derived their in-comes from employment in local mills, mines, banks, offices, and public service. One-sixth of the mothers were employed outside of the home. Two-thirds of Class III families owned their own homes and 25% lived in the best residential areas. Only one male in Class III graduated from college and only 23% completed high school. The women in this class were considerably more educated than the men. Sixty-three percent completed high school and 10% held college degrees.
While Class III parents' aspirations for their children's future education were not as high as those of Class I and II parents, evidence suggested that they at least wanted them to complete high school and that a significant percentage of Class III parents wanted their children to attend college. Twenty-seven percent of Class III high school stu-dents were enrolled in college preparatory courses.
Class III families were socially and politically active. Hollings-head reported that membership and active participation in many organiza-tions conferred prestige within this class. Women and men belonged to social clubs, church groups, patriotic societies, and civic betterment groups among other organizations. The majority of Class III individuals
64
belonged to clubs and organizations dominated by members of that class. A few, however, belonged to the more prestigious Country and Rotary Clubs. Class III members were more active in politics than the other classes, but members of Classes I and II were the policy makers.
Annual family income for Class IV families ranged from S800 to $2,700 and was earned for the most part from employment in the mines, mills, and shops. Thirty percent of the mothers worked outside of the home. The higher classes characterized Class IV individuals as poor but honest hard workers who pay their taxes, raise their children pro-perly, but never seem to get ahead financially. Only 35' of Class PV families owned their own land.
Class IV members were active in fraternal organizations such as the Eagles Club and their auxiliaries. For the most part, however, civic and community organizations were outside of their experience. Their lack of participation in community activites was the result of a number of factors including lack of leisure time, the fact that Class IV women's "place" was in the home or on the job, and insufficient financial resources to participate in costly activities.
Only about 51 of fathers and slightly less than 10% of mothers in this class graduated from high school. Mbst of the children in Class IV had aspirations to complete high school, but parents did not en-courage this and often actually took their children out of school in order for them to work. The average Class IV student only completed two years of high school.
Ninety-two percent of employed Class V fathers were either unskilled
65
or semi-skilled laborers or machine operators. Fiftyifive percent of the mothers worked outside of the home as waitresses, dish-washers, washerwomen, cleaning women, cooks, janitresses, and unskilled domestic workers. Between 1937 and 1941, 53% of these families received supple-mental relief. Annual family income ranged from $500 to $1,500. Only 19% of these families owned their own homes. The majority lived in di-lapidated dwellings. Fifty-six percent of the children in Class V families were from broken homes.
Sixty-seven percent of Class V parents left school before comple-tion of eighth grade. The children, for the most part, only attended elementary school. Members of Class V were almost totally uninvolved in organized community activities.
Several indicators, including grades and failures, were used to measure academic achievement. As the following table shows, students from the higher classes received the best grades. TABLE 2.3
PERCENT WITH MEAN GRADE oF:63
CLASS 85-100 70-84 50-69
I and II S1.4 48.6 00.0
III 35.5 63.2 1.3
TV 18.4 69.2 12.4
V - 8.3 66.7 25.0
TOTAL 23.8 66.3 9.9
66
Class related differences in student performance were also evident in the student failure rate. When the grade records of 495 adolescents who had completed at least one semester of high school were analyzed, it was found that Class II had only one failure. Five and one-half percent, 27% and 89% of the Class III, IV, and V students respectively, failed one or more courses. Academic awards and honors tended to go to students from the higher classes even when pupils from the lower classes outperformed those from the upper classes. IQ was found to be significantly related to class position; however, the degree of associ-ation was not high enough to attribute the high grades of Class I and II students to their intellectual capacity as measured by IQ. This finding led Hbllingshead to conclude that the higher achievement of upper class students did not simply result from differences in innate intelligence.
Coleman (1966), like Hollingshead, included both socioeconomic
status and family background variables in his analysis of the relation-
ship between family characteristics and student achievement. The family
background characteristics used were clustered into eight variables
which were:
(1) Urbanism of background, which was based on the community in which
the student and mother grew up; and migration, which was based on the
student's and mother's place of birth.
(2) Mbther's and father's education.
(3) Structural integrity of the home; which referred to whether both
parents were present in the home.
67
(4) Smallness of family (family size).
(5) Items in the home; which included televisions, telephones, record
players, refrigerators, automobiles, and vac w m cleaners.
(6) Reading material in the home; which referred to dictionaries, en-
cyclopedias, daily newspapers, magazines, and books.
(7) Parents' interest; which was based on whether parents talked with
their children about school and whether anyone read to the children
when they were small.
(8) Parents' educational desires; which included students' perceptions
of how good a student their parents wanted them to be, how far in school
parents wanted them to go, and parents' attendance at PTA meetings.
Overall, background characteristics accounted for between 10 and 25
percent of the variance in individual achievement.
TABLE 2.4
PERCENT OF VARIANCE IN VERBAL ACHIEVEMENT ACXXl~DrED FOR AT GRADES 12,
9 AND 6 BY SIX AND EIGHT BACKGROUND FACTORS FOR BLACKS AND WHITkS.64
GRADE 12 GRADE 9 GRADE 6
Eight- ~ Eight- Eight-
Six Eight Six Six Eight Six Six Eight Six
Blacks 13.48 15.14 1.66 12.15 14.44 2.84 14.01 14.62 .61
Whites 14.71 23.03 8.32 17.81 23.28 5.47 16.20 17.64 1.44
The first column for each of the three grades shows the variance ac-counted for by the first six objective background characteristics. The second col D gives the variation accounted for by all eight background
i
68
factors. The third column gives the added variation explained by parents' interest and aspiration for their children.
As Table 2.4 shows, the six measures of objective conditions as-count for more variation in achievement in earlier grades than later ones. Coleman attributed this phenomenon to the fact that the family's impact on the child has its greatest effect in earlier years and con-cluded that family to family differences in achievement should decline after the beginning of school. He rejected the argument that the family's impact on the child affects his receptivity to later experiences, thus family to family differences in achievement would increase as the child progresses in school. As previously stated, however, there is evidence to indicate that achievement gaps widen with increase in grade level. This does not necessarily mean that Coleman's conclusion was incorrect. Factors other than family background characteristics may account for this increase in achievement differences.
Secondly, the relationship between subjective home conditions and achievement shows an opposite trend over grades. Their relation to achievement increases with grade level.65 The indication is that either older children perceive their parents' interest more accurately than younger ones, or parents' interest has more impact on achievement in later years. Thirdly, Coleman found a difference in the effects of sub-jective home conditions for different groups. Parents' interest accounted for much more variation in achievement for whites and Asian Americans than it did for any other groups. Upon investigation of the reasons for these differential effects, Coleman considered the possibilities that
69
either (a) Negro, Mexican American, Puerto Rican, and American Indian children fail to perceive their parents' interest or lack of interest in their schooling as fully as do whites and Oriental Americans; or (b) the parents of these minority group children are less able to translate their interest into effective support for the child's learning than either Oriental American or whites parents, were considered. The evi-dence supported both of these positions.
Coleman concluded that: (1) in the sixth grade, economic level has the highest relation to achievement for all minority groups, while parents' education has the highest relation for whites; (2) in later years, parents' education has the highest relation to achievement for nearly all groups; (3) for blacks at grade 12, the length of time in an urban environment and family size (in terms of smallness) have ap-proximately the same effect on achievement as parents' education; and (4) the structural integrity of the home (the father's presence or absence) shows very little relation to achievement for blacks. The relative importance of background variables was found to differ for each group. The importance of educationally related attributes of the home (parents' education, reading matter), is greater for white children than for minority children. Armor stated that:
A family's life style--their stress on education, their eco-nomic well-being, their child-rearing practices and other similar factors--is likely to have an important effect upon the ctild's interest in and ability to carry out academic work. OD
Armor (1972) called this complexity of factors family life-style. As a result of the controversial nature of two of Coleman's conclusions,
70
a new analysis of the complete elementary school data from the Coleman report was performed. The first conclusion that A`,.~r examined was that the facilities in schools attended by blacks and whites were basic-ally equal. This finding contradicted the generally accepted belief that one of the major causes of low black student achievement was that blacks attended inferior schools. The second conclusion, which is the one that is more pertinent to this research, was that compared to the effects of family background, the effects of school staff and facilities on achievement are of minimal importance. In other words, improving the quality of schools attended by blacks alone will not reduce the gap between black and white achievement.67
Armor used five major indices to measure family life-style. The first four indices, which were (1) percent of students with fathers in shite collar occupations, (2) average education of both parents, (3) per-cent of students living with both natural (or legal) parents, and (4) the average number of household items, were characteristics of students' fami-lies. Both mothers' and fathers' education were divided into three cate-gories: less than high school, high school including post-high school business or technical training, and some college or college graduation.68 The household items were television sets, telephone, record player or hi-fi, dictionary, refrigerator, encyclopedias, vacuum cleaner, automobile, and daily newspaper. The fifth index, which Armor considered a com-munity index, was percent of blacks in a school. A black school was defined as one in which more than 50 percent of the pupils were black. School percent black was included because Armor felt that black families
may have different attitudes toward education than white families with similar socioeconomic characteristics resulting from the previous caste position of blacks within the structure of status attainment. Family attitudes toward education were not directly measured. Indices were formed for each student from questionnaire data and averaged over all of the students in a given grade. Indices thus represented aggregate family characteristics for the school community. Schools then, not individuals, were the units of analysis. Six grade verbal achievement test scores were employed as the measure of academic achievement.
Armor used the term "community inputs" to describe the aggregated characteristics of students' families. It is important to note that "community input" measures combine both family and neighborhood effects. These inputs may only reflect the composition of schools' immediate neighborhoods. Moreover, school level analysis does not allow for differentiation of the effects of individual family characteristics on their own children from those of families in general on children who are not their own.69
Family Structure
White Collar
Household Items
Parents' Education
Percent Black
TABLE 2.5
CORRELATIONS OF 6TH-GRADE VERBAL ACHIEVEMENT WITH CCMbi£NITY INPUT FACToRS70
Black Achievement in Black Schools White Achievement in White Schools
.14 Family Structure .31
.16 White Collar .51
.60 Household Items .67
.52 Parents' Education .66
.04 Percent Black -.22
72
As Table 2.5 shows, the effects of the five background character-istics are different for blacks and whites. The most obvious differ-ence is that socioeconomic status characteristics have a much greater impact on white student achieve,,~llt than black student achievement. The weak correlation between family structure and achievement contra-dicts Mbynihan's conclusion that family structure is a strong predictor of black or white achievement. Armor's reanalysis supports the find-ings of Coleman that were reexamined.
Kenneth Wilson (1979) investigated the effects of integration and class on black educational attainment. Wilson used data fro n Bachman's tl970) Youth in Transition Project which investigated the effects of social environments on factors such as high school graduation, college completion, employment, and educational attainment. Bachman's study analyzed data from a national sample of 2213 male high school students in 87 public schools,of which 256 were black. Wilson used both an average of black students' grade point averages in the ninth and tenth grades and pacing to measure academic performance. Pacing was added because academic performance had been found to be a poor predictor of black educational attainment by scholars including Porter (1974) and Porter and Wilson (1976). Pacing refers to the frequency of students either being held back a grade, staying in the usual age cohort, or being advanced to a grade cohort one year ahead of his or her initial cohort.71
An index of six equally weighted items was constructed to determine socioeconomic level. The items were: (1) fathers' occupational status,
73
(2) fathers' education, (3) mothers-' education, (4) possessions in the home, (5) number of books in the home, and (6) number of rooms per person in the home. Home possessions were similar to those looked at in Coleman's research.
Wilson was unable to draw any firm conclusions concerning the re-lationship between social class and academic achievement; however, he did observe that youths from upper class backgrounds followed attainment sequences that differed significantly from those of blacks from lower class backgrounds.
SOCIOECONKh0C STATUS AND ACADEMIC ACHIEVEMENT
Numerous studies show a strong relationship between socioeconomic status and academic achievement. Kelin (1971) reported that the esti-mated correlation between SES and science achievement for elementary school pupils was .802. A correlation of .755 between SES and Stanford Achievement test scores among suburban elementary school students was reported by Dunnell (1971). Stephenson and Mares tl973) found a cor-relation of .865 between SES and standardized achievement test scores in big city schools. St. John (1971) stated that the relationship between SES and almost any type of school behavior was so well documented that it had become axiomatic to social scientists.
Other studies have found a moderate to weak association between SES and academic achievement. Knief and Stroud (1959) reported a cor-relation of .340 between SES and the composite score on the Iowa Test
74
of Basic Skills for 344 fourth graders. A correlation of .263 between SES and reading achievement and .284 with math achievement was reported by Fetter (1975) from the National Longitudinal Study of the class of 1972. Wright and Bean (1972) found correlations of .124, .089, and .072 between SES and verbal and quantitative scores on the Stanford Achievement Test and Grade Point average respectively. Hennessy (1976) reported a correlation of .136 between SES and verbal achievement as measured by the Comparative Guidance and Program test battery.72
Warner, Havighurst and Loeb (1944) discussed several earlier stud-ies of the relationship between socioeconomic status and student per-formance. One was a 1936 study conducted by Harlan Updegraff for the Pennsylvania State Department of Public Instruction and the American Youth Commission. Socioeconomic status and educational histories were obtained for a group of 910 students with IQs of 110 or above. Students were divided into two subgroups on the basis of SES. Ninety-three per-cent of students graduated from high school and 57% attended college in the upper SES group. In the lower socioeconomic group, 73% gradu-ated from high school and only 13% attended college. These findings indicate that even though the two groups were basically equal in in-tellectual ability, the higher SES group had significantly more educa-tional opportunity.73 A second study conducted by Helen Hoetsch (1940) of 1,023 students who graduated from high schools in Milwaukee, Wiscon-sin in 1937 and 1938, drew a similar conclusion. All of the students had IQs of 117 or above. The results showed that parents' income was directly related to college attendance. As parents' income increased,
73
(2) fathers' education, (3) Smothers-' education, (4) possessions in the home, (5) number of books in the home, and (6) number of rooms per person in the home. Home possessions were similar to those looked at in Coleman's research.
Wilson was unable to draw any firm conclusions concerning the re-lationship between social class and academic achievement; however, he did observe that youths from upper class backgrounds followed attainment sequences that differed significantly from those of blacks from lower class backgrounds.
SOCIOECONoMIC STATUS AND ACADEMIC ACHIEVEMENT
Numerous studies show a strong relationship between socioeconomic status and academic achievement. Kelin (1971) reported that the esti-mated correlation between SES and science achievement for elementary school pupils was .802. A correlation of .755 between SES and Stanford Achievement test scores among suburban elementary school students was reported by Bunnell (1971). Stephenson and Mbres (1973) found a cor-relation of .865 between SES and standardized achievement test scores in big city schools. St. John (1971) stated that the relationship between SES and almost any type of school behavior was so well documented that it had become axiomatic to social scientists.
Other studies have found a moderate to weak association between SES and academic achievement. Knief and Stroud (1959) reported a cor-relation of .340 between SES and the composite score on the Iowa Test
75
the proportion of students who attended college increased.74
Sexton (1961) conducted a study of the relationship between edu-cation and income in a large midwestern city. The data examined this relationship for both elementary and high school students. IQ and achievement test scores were available for elementary school pupils but not for high school students. Trade, vocational, and technical schools were excluded from the study.
Social class was measured in aggregated units. First Sexton de-termined average family income levels in each school area. Schools of the same or similar income levels were then grouped together. School groups were ranked from the lowest to the highest based on income. The first group of elementary schools consisted of 33 schools with an aver-age family income range in 1957 of $3,500 to $4,857. The second group consisted of 107 schools with an average family income of between S5,300 to $6,695. Eighty-five schools were included in Group III. Average family income ranged from S7,100 to S8,500. Group TV, the highest class group, contained 20 schools with an average family income of over $9,ooO.75
Academic achievement was measured by scores on the Iowa Achieve-ment Test. The test is divided into five sections: language skills, work skills, arithmetic skills, reading, and vocabulary. The test was administered to fourth, sixth, and eighth graders. In all schools in which the income level was above $7,000, with the exception of eighth graders, students achieved at above average grade levels. In schools in which the average family income was below $7,000, average student
76
achievement was below grade level. The average achievement in schools in the highest income group was two whole years above that of students in schools in the lowest income group. Without exception, achievement scores rose with family income levels.76
Sexton stated that upper income groups were more successful in school because of their proficiency in language, reading, and verbal skills. Relatively speaking, lower income groups did well in arith-metic and work skills, while upper income groups performed poorly in these two areas in relation to their composite scores. Upper income groups did quite well on the reading section of the test. Lower income groups performed more poorly in reading than in all other areas. The fact that language, reading,and verbal skills are extremely important to overall academic success would at least in part account for the lower achievement of lower income groups.
"Social class is very apparent at the high school level in a num-ber of significant ways. Social class distinctions are also present within each high school, much more than within each elementary school."77 At the high school level, Sexton observed that failures were more pre-valent in the lower income groups. In Group I, 42.2t of students failed one or more subjects, whereas in Group V, 28.61 of pupils failed one or more subjects. Social studies and English were required courses in Big City schools. English was the more troublesome area for Group I students. There were 10.2% more failures in Group I than Group V. This finding substantiates White's (1982) and Bernstein's (1977 ) argument that the language patterns
of upper and lower class students
77
are different and that those of upper class students are more conducive to scholastic success. There was also a significantly greater represen-tation of upper SES groups in the higher curriculum levels compared to lwoer SES groups.
Sexton also addressed the race issue. The-income of blacks in Big City was significantly lower than that of whites. In 1958, the differ-ence between mean black and white income was $2,600. Some may claim that low income groups perform poorly in school because there are so many blacks in these groups. Sexton, hwoever, argued, that in low income
groups, students in predominantly black schools were achieving at approxi mately the same level as students in predominantly white schools. This
pattern was similar in upper class schools, as is illustrated by the fol-lowing statement:
Of course the social class position of Negroes is generally much worse than that of whites and much more "frozen" by pre-judice which limits their job opportunities and income levels, while providing inferior segregated education and housing in overcrowded, neglected ghetto areas. For Negroes, the class system has almost the rigidity of a caste system, and indeed could accurately be called a semi-caste system. Restricted by class walls, Negroes inevitably have special problems in school. Yet, it has been clearly demonstrated that, when the walls are scaled or broken down, Negro school achievement levels have improved remarkably.78
Sexton's point was that the lower achievement of blacks is a class prob-lem, but one that is seriously aggravated and complicated by racial discrimination.
Jencks (1972) estimated that a family's economic status probably correlates about 0.35 with children's test scores. What this means is
that the test schores of children whose fathers rank in the top fifth
of the occupational hierarchy will on the average be 13 to 15 points
78
higher than children whose fathers rank in the bottom fifth. If family income is used as a measure of SES, the disparity between these two groups will be less than 13 points. Jencks further stated that class differences appear to be greatest for verbal ability and gen-eral information. Test of mathematical skills, reading comprehension, and non-verbal ability are less influenced by economic background.
Epps (1974) studied SES and academic achievement among Northern and Southern blacks. The sample consisted of high school students in four schools in a large Southern city and 4 schools in a large Northern city. Originally, Epps planned to sample students in two biracial schools and two racially segregated schools in each city. This was ac-complished in the Northern city, but information from biracial schools could not be obtained in the South. There were 400 males and 566 fe-males in the Northern sample. The Southern sample consisted of 1,572 students, 721 males and 851 females. All data with the exception of grades was reported by students. Grades were obtained from school records.
A sixty-item version of Miner's (1957) vocabulary test was used as the measure of ability. School achievement was determined by students' grades in English, social studies, science, and arithmetic. Cumulative grade point averages were converted into a five-point scale. Either father's occupation or mother's education were the indicators of SES. Father's occupation was used to measure SES for Northern males and mother's education was the SES measure for all other groups.
Results showed that there was not a very strong relationship be-tween SES and school grades. Only among Southern females was SES
79
significantly correlated with grades. SES was found, however, to be positively related to vocabulary scores for all groups of students. Cbrrelation coefficients ranged from .18 to .2S. The relationship was slightly stronger for females than males, particularly in the South.
Fotheringham and Creal (1971) conducted a study of the academic achievement of 971 third graders who attended public and Roman Catholic schools in southern Ontario County CWelland). These 971 students were classified as above average, average, and below average achievers based on their performance on the Metropolitan Achievement Test QMAT). Thirty-five students from each group constituted the sample for the study. All 105 children were Canadian born, had both parents in the home, were from homes in which English was the principal language, and according to school nurses, were without evident neurological impairment. The study investigated the effects of both socioeconomic and educational and emo-tional atmosphere variables on achievement. The contribution of the SES variables was calculated separately from that of the eight educational and emotional atmosphere variables by performing a stepwise multiple re-gression analysis in which the four SES variables were entered before the family background variables. Family educational and emotional atmo-sphere will be discussed in the following section.
The SES variables were father's schooling, mother's schooling, father's occupation, and family income. For reading achievement, the multiple correlation of SES variables was .53, accounting for 28% of the variance in test scores. The multiple correlation for SES and arithmetic computation was .50, explaining 25% of the variance.
80
Fotheringham and Creal stated, however, that although SES variables related to children's intelligence and achievement scores, they provide little information on how these effects are mediated to the child.
Bond (1981) and Rossi (1961), in agreement with Fotheringham and Creal, argued that while there is widespread agreement on the existence of a socioeconomic status/academic achievement correlation, there is considerable controversy over the reasons for the correlation. Bond placed the explanations for this relationship into four basic categories which were: (1) the genetic argument, t2) the cultural argument, (3) arguments positing unequal educational treatment, and (4) explanations of educational differences as part of class analysis.
The genetic argument posits the genetic inferiority of lower socio-economic groups. Proponents of this position, including Jensen (1969) and Eysenck (1971), maintain that certain groups have low status because they are genetically inferior. According to this position, students from low SES groups who perform poorly in school do so because they lack the genetic ability to do otherwise. Compensatory education, therefore, would have little, if any, effect on the achievement of students from low SES backgrounds.
The cultural argument concentrates on the effects of the different cultural environments of children from various socioeconomic groups. Bernstein (1961), for example, argued that children are socialized to-ward speech codes that control access to either relatively context tied or relatively context independent meanings. "Elaborated codes," which are typical of middle class speech, free speech from its "evoking
81
social structure" and are based in articulate symbols. 'restricted"
codes, which are characteristic of lower class speech, are more tied
79 to a local social structure. Even if working and middle class child-ren possess similar vocabularies and share the same linguistic rule system, they may nevertheless use language differently in specific con-texts. Since schools are predicated upon elaborated codes, the working class student is placed at a disadvantage. The "culture of poverty" thesis also falls into this category of explanations for SES related academic failure. The basic argument postulates a deficient intellec-tual environment for lower class children and will be further discussed in the section on family culture. The cultural argument also takes into account the effects of the physical conditions of poverty such as inade-quate diet, overcrowding, and inadequate living conditions in general.
The third set of explanations for the relationship between SES and academic achievement is predicated upon the argument that lower class students receive inferior treatment from the educational establish-ment, are more likely to have inadequately trained teachers, are more likely to be placed in overcrowded classrooms, and have less money to spend on education than middle and upper class students. Karabel and Halsey (1977) and Leacock (1969) supported this position.
The low achievement of lower class students is often attributed to the role-of the educational system in maintaining class differences. Those adhering to this view of education maintain that as long as so-ciety remains divided on class lines, lower class children will perform poorly in school. Bowles (1977) viewed inequalities in education as
82
part of the evils of a capitalist society and maintained that these inequalities are likely to persist as long as capitalism survives. Education serves to legitimize the socioeconomic structure by deter-mining the position of children in the class structure.80
Two important conclusions were drawn in Bond's review. The first was that socioeconomic status is only one of the many aspects of human diversity contributing to educational achievement. It is, however, an extremely important aspect that should never be overlooked in evalu-ating the most appropriate educational treatment for any particular individual. The second was that the exact nature of the relationship between SES and academic achievement has not been fully researched because various scholars have tended to concentrate on one aspect of the relationship to the exclusion of others. The tendency has been to attribute low achievement to "cultural deprivation," poor schools, attitudes of teachers, children's low expectations, or the evils of the class system in general. All of these factors undoubtedly play a part; however, their relative significance is not clear.8]
Black families living in Gary, Indiana that participated in the Gary Negative Income Tax Experiment were the subjects of Mbrname, Maynard and Ohl's (1980) study. Average family income for the group was $5,200 in 1971, 55t of the average for the U.S. Three-fifths of the families were female headed. The study analyzed data for two samples of children. The first sample consisted of 529 students in grades 3 through 6 during the 1972-73 school year. The second consisted of students in the same grades during the following school year. Scores on the vocabulary sub-
s
test of the Iowa Test of Basic Skills were the measure of student achievement and were used as both a dependent variable and an indicator of prior achievement.
The results indicated that mothers who completed high school were more successful in helping children acquire cognitive skills than mothers who did not complete high school. The relationship between academic achievement and mother's high school completion was statis-tically significant for both the 1973 and 1974 samples.
Two alternative explanations concerning how the effects of this SES variable are mediated to the child were given. The first was that the importance of mother's education may reflect the quantity and quality of child care. The second was that the effects of mother's education may be indicative of genetically transmitted endowments. The authors suggested that since the effects of mother's education greatly exceeded the effects of father's education (in two parent households), the hy-pothesis that the mother's education influences patterns of child care which in turn effects achievement is the most plausible. Evidence did not support the hypothesis that children whose mothers worked outside of the home achieve less than those whose mothers are not employed out-side of the home. Family income was not found to be consistently related to achievement.
White (1982) stated that a fairly extensive review of the litera-ture on the relationship between SES and achievement leaves most readers confused. Correlations between SES and various measures of achievement generally range from .100 to .800. Reviews by Cuff (1933), Neff (1938),
84
Loevinger (1940), Havighurst (1961), Lavin (1965), Findley and Bryan (1970), Duncan, Featherman, and Duncan (1971), and Bryant, Glazer, Hansen, and Hirsh (1974) do not provide explanations of why there is so much variation in the magnitude of the correlations between SES and academic achievement reported in the published literature. Addition-ally, they do not attempt to determine the most reasonable estimate of the true or expected correlation between SES and achievement.82 White further stated that the available reviews cite results of from 10 to 20 studies and then discuss issues such as the pros and cons of various methods for collecting SES information, the theoretical causal relation-ship between SES and other variables, the relationship between SES and factors such as intelligence and ethnicity, or inequities that result from an unequal distribution of SES characteristics among the general population. The reader is usually either left with evidence from a relatively small number of nonrepresentative studies or sweeping gen-eralizations from which to draw conclusions about the strength of the relationship between SES and achievement. 3 White used the Fabrication Index, the Current Index to Journals in Education, ERIC documents (obtained by a computerized search), Dissertation Abstracts International, and bibliographies from other studies, to identify 248 studies for possible inclusion in his analysis. One hundred and one studies were actually examined. The remaining studies were excluded either because their content was considered inappropriate for the topic under investi-gation (they dealt only with philosophical issues or instrument develop-ment), or they did not report correlation coefficients or sufficient
85
information to calculate satisfactory estimates of the correlation coefficient.
White used meta-analysis techniques for integrating research findings, originally proposed by Glass (1976), to determine the magni-tude of the relationship between SES and academic achievement and to investigate factors contributing to the variance in previously reported correlations between achievement and SES. In meta-analysis, the re-viewer locates either all studies, or a sufficiently large representa-tive sample of studies, on a given topic. The results of each study are expressed in a common metric and the various characteristics of each study that may have affected its results are then quantified or coded. Common descriptive statistics (means, medians, standard deviations, and standard errors of measurement) and relational statistics (correlations, cross tabulations, multiple regression, and analysis of variance) are used to study the association of these characteristics across all studies with variations in outcomes. Typically obtained results of studies with given characteristics can be estimated because the results are expressed in a common metric.84 The selection of variables included in Whitets review was based on an extensive review of the literature, a pilot test of the coding instrument, and discussions with other scholars. The variables employed were as follows: (1) Unit of analysis used in computing the correlation coefficient, which were either aggregated, confounded, or student. Aggregated units referred to schools or school districts. Both the SES and achievement measures were averaged for the unit. When the unit of analysis was
86
confounded, SES was measured at the aggregate level and achievement
was measured at the individual student level or vice-versa. When both
SES and achievement were analyzed separately for each student, as is
the case in the present research, the student was identified as the
unit of analysis.
(2) Type of achievement measure, which was broken down into verbal, math,
science, composite achievement, IQ and other. IQ may in many cases be
a predictor of achievement, but does not necessarily determine actual
performance. Further, such tests tend to favor middle and upper class
white students (Havighurst and Newgarten, 1975).
(3) Grade level of students used in studies.
(4) SES reporting error, which was an estimate of the potential ac-
curacy of the SES measure. When parents reported this information, for
example, information was considered more accurate than if school of-
ficials estimated how various schools differed in respect to this
variable.
(5) Achievement range restriction, which ranged from no restriction
to substantial restriction. A sample consisting of students in a
given IQ range, for example, was considered as having a substantial
restriction.
(6) SES range restriction, which referred to the SES of the population
under investigation. If a sample consisted entirely of inner-city, low
income students, the restriction was considered substantial.
(7) Percent ethnic minority, which was the percentage of sample students
from a racial or ethnic minority (i.e., black, Chicano, Oriental, and
87
American Indian).
(8) Year of study, which was the year of publication.
(9) Number of items in the SES instrument.
(10) Number of students upon which correlation coefficients were based.
(11) Type of publication in which the literature was reported, i.e.,
journals, books and unpublished materials. Unpublished sources con-
sisted of theses, dissertations, and reports (such as project reports
in government contracts).
(12) Sample, which referred to whether samples were taken from a small
geographic region or were nationally representative. Mbst of the studies
were conducted in the United States, but a few were done in England or
Canada.
(13) Type of SES measure. SES variables were rated on a continuum de-
pendent upon whether they were represented in the instrument or whether
they were of major importance to the research. The SES measures were
as follows: (a) family income, (b) parent's education, (c) occupation
of the head of the household, (d) home atmosphere (parents' attitude
toward education, parents' aspirations for their children, and cultural
and intellectual activities of the family), (e) dwelling value, (f)
school resources, (g) subjective judgement, and Oh) other, including
number of siblings, ethnicity, and mobility of family. Home atmosphere
referred to whether parents created an environment that was conducive
to their childrens' learning. In other words, did parents read to their
children, help them with their homework, encourage them to go to college,
and take them to the library and to cultural events. It is important
88
here to recognize that home atmosphere is not a traditional measure of SES. Hbnff~ver, parents' education, family income, dwelling quality, or combinational measures utilizing one or more of these variables were overwhelmingly the most frequently used in the literature reviewed. (14) Number of SES groups, indicating how the SES variable was divided. (15) Lateral validity of the study. Examples of factors contributing to low internal validity included: (a) only using two extreme groups to compute correlations, (b) using ethnicity as a measure of SES, and (c) estimating the correlation coefficient from multiple T-tests.8S
White's results showed that for the sample of 101 studies, the best estimate of the correlation between SES and academic achievement is .251, indicating that the relationship between SES and achievement is probably much weaker than has been commonly assumed. The findings of high corre-lations between SES and achievement were attributed to the use of aggre-gated units of analysis and type of publication. Robinson (1977), as well as White, found that correlations computed Bated units will almost always be higher than correlations computed using individuals as the unit of analysis.
(1950), Knapp
using aggre-
White's analysis also showed that the more prestigious publications are more likely to publish results that are statistically significant, whereas a number of statistically insignificant but possibly valid findings go unreported in these publications. For studies published in books, the mean correlation between SES and achievement is .508. Studies published in journals and those that were unpublished had mean correla-tions of .343 and .242, respectively.
89
There was a slight trend to find weaker correlations when SES and achievement were restricted in range. The magnitude of the correlation decreased as the number of items in the SES measure decreased. The use of more than one indicator of SES increased the predictive reliability of the SES measure. Of all the traditional measures of socioeconomic status, income was found to be the highest single correlate of achieve-ment. White offered two explanations for these findings. The first was the increased availability of home items such as televisions, to people of all SES groups. Additionally, participation in community ac-tivities is not as restricted to the upper and middle classes as has been the case in the past. Secondly, compensatory education efforts have had a positive effect on reducing the strength of the relation-ship between socioeconomic status and achievement.
White concluded that SES is positively but only weakly correlated with achievement when the individual student is the unit of analysis and traditionally defined measures of SES are employed, which is most frequently the case. "In such situations, measures of SES can be ex-pected to account for less than 5t of the variance in student's achieve-ment. Correlations of that magnitude seriously restrict the utility of SES in most research applications."86 Because of this weak correla-tion, White argued, there is little utility in using SES as a covariate, stratifying variable, predictor, descriptive variable, or causal agent in studies of SES and achievement.
When schools or other aggregated groups are the unit of analysis, traditional SES measures are generally correlated strongly enough with
go
achievement measures to be useful as covariate, predictor, or strati-fying variables.87 In these cases, researchers should specify that grouped data were used and that correlations will be much higher than if individual students are the unit of analysis. It is important to note that the causal effects of SES are more relevant to individuals than to groups so even though correlations are higher when aggregate units are used, it is probably more appropriate to use the individual student as the unit of analysis.88
FAMI W CULTURE AND ACADEMIC ACHIEVEMENT
"Despite the attention which relationships between home environ-ment variables and scholastic ability and attainment have received in the literature, our understanding in the area is not very good."89 Research focusing on material possessions in the home does not provide information about parental behaviors and attitudes that may influence achievement. Although Chicago and British school studies examine some characteristics of parents, they suffer from essentially the same prob-lem. The question of, '~hat characteristics do parents possess?" re-mains largely unanswered. Another factor complicating this type of research is the fact that many of the Chicago and British school studies were performed in countries there the stratification systems, general cultures, and policies toward education are quite different from those of the United States. While these studies sometimes look at the effects of ethnicity, the issue of race is not addressed.
91
In the United States, the "culture of poverty" argument, which postulates a deficient home intellectual environment for lower class children, has had important implications for educational research and government policy. The notion of a "culture of poverty" formulated by Lewis (1966), revolves around commonly held stereotypes about the poor who are believed to maintain the very conditions that consign them to failure. The alleged deficiencies of lower class culture that are thought to influence educational performance include a lack of verbal interaction between mother and child and between family members in gen-eral, a lack of interest in intellectual activities for their own sake, lack of parental involvement in schools, and a lack of emphasis on read-ing. As a result of these deficiencies, the lower-class child is locked into a self-perpetuating cycle of educational failure and poverty.90
The concept of "culture of poverty" is often used as a justifi-cation for not providing equal educational opportunity to lower class and black students. Lewis himself distinguishes between the life-ways of groups which he believed lived by a distinctive poverty subculture, those who are poor but do not possess such a subculture, and those who have gotten rid of the poverty subculture.91 In other words, all poor people do not necessarily possess the same attitudes and values, nor do they always behave in the same manner. Valentine (1971) examined Lewis' propositions concerning the social participation, values, and psychological characteristics of those living in the "culture of poverty." Lewis contended that at the local community level, those who live in the culture of poverty have a minimum of organization past the level of the
nuclear and extended family.92 Valentine stated, however, that the factual situation in the San Juan slum investigated by Lewis and evi-dence from similar communities is consistent with quite a different formulation. Low income districts have definite local social structures which commonly include groups and relationships similar or analogous to those found in communities that are not slums.93 Common analogous ele-ments include community councils, political ward organizations, conven-tional youth groups, block associations, personal social networks, and various human service institutions. Additionally, low income groups participate in social activities that are specialized adaptations to con-ditions of disadvantage and marginality such as civil rights or minority advancement groups, poverty-oriented social-change movements, adult blue-collar criminal associations, and other voluntary associations.94
Valentine also refutes Lewis' argument that those living by a "cul-ture of poverty" do not share the standard values of society. The poor share many or most of the major values associated with the dominant strata in American society. Values shared by the poor and other ele-ments of society probably include the belief that educational achieve-ment is desirable, self-sufficiency is an admirable quality, material comfort is a worthwhile goal, and that competition and cooperation are appropriate in different contexts. Lewis depicted those living in the "culture of poverty" as having a strong feeling of marginality, help-lessness, dependency, and inferiority; weak ego structures; a sense of resignation and fatalism; and being present-oriented. Valentine then argued that these "pessimistic" characteristics of the poor
93
are not personality disorders as suggested by Lewis, but are realistic perceptions and evaluations based on lifelong and intergenerational experiences imposed by the social order. Secondly, the fact that most poor people have managed to function as human beings under conditions of extreme adversity and deprivation is a tribute to the inner strength and flexibility of humanity. This far from indicates a psychological weakness.95
Investigations by Dave (1963), Wolf (1964), Bloom (1974), Reeves (1972) and Walberg and Mbrjoribanks (1976), have produced evidence to the effect that various process charactersitics of the home contribute more strongly to the prediction of children's abilities than do social status or family structure indices. Correlations as high as .70 with intelligence test scores and .85 with school achievement have been re-ported with the index used by Dave and Wolf.96 White stated that in the studies he reviewed, home atmosphere accounted for 4 to 11 times as much of the variation in academic achievement as did traditional measures of SES
Kellaghan (1977) studied the relationship between the home environ-ment and scholastic behavior of disadvantaged 8 and 9 year olds living in central Dublin, Ireland. The sample consisted of 30 girls and 30 boys attending eight different schools. Dave's (1963) questionnaire was used to obtain measures of six environmental process variables, which were: (a) achievement press--parental aspirations for the edu-cation of the child; (b) language model--quality of language usage of parents (e.g., pronunciation, vocabulary); (c) academic guidance--
94
extent to general supervision and suggestions regarding schoolwork; (d) family activeness--variety, frequency, and educational value of the activities of the family; (e) intellectuality of the home--variety and thought-provoking elements in toys and games available to the child; and (f) work habits of the family--degree of structure and routine in home management.97 Scholastic behavior was measured by scores on: (1) the Stanford-Binet Intelligence Scale, a conventional test of general intelligence that yields a deviation IQ score; (b) the Cattell Culture Fair Intelligence Test, Scale I, a test of general mental ability; (c) the Marino Graded Word Reading Scale; (d) the Irish Word Recognition Test; and (e) the Schonell Mechanical Arithmetic Test.
Academic guidance predicted 12% of the variation in Cattell Chl-ture Fair scores. Twenty-six percent of the variance in Stanford-Binet scores was predicted by activeness of the family. Language model predicted 25% of the variation in Cattell Non-Culture Fair scores, and achievement press predicted 30% of the variance in arithmetic scores. Work habits and intellectuality of the home in combination predicted 40' of the variance in English reading scores and 36! of the variation in Irish reading scores. Mbasures of attainment were most closely re-lated to environmental factors.
Iverson and Walberg (1982) analyzed 18 studies of home environ-ment and achievement including those performed by Fraser (1959), Dave (1963), Dyer (1967), Weiss (1974), and Keeves (1972). The studies employed a variety of home environment measures, but the one most com-monly used was the Index of Educational Ehvironment
The studies
(IKE) used by Dave
95
and Wolf. Sample sizes for the studies ranged from 15 to 3,092. Eleven of the 18 studies were conducted outside of the United States and all were classified as either Chicago or British school studies. Like Kellaghan, Iverson and Walberg concluded that home factors are more highly correlated with measures of attainment (i.e., language, reading, arithmetic scores), than with IQ or intelligence. They also stated that Chicago school studies appear to be better measures of edu-cationally relevant home factors than British school studies. Their analyses also suggested that ability and achievement are more closely related to the socio-psychological environment and intellectual stimulation in the home than they are to socioeconomic status variables.
Mh~nane, Maynard, and Ohls (1980) looked at the effects of home goods inputs and time inputs on achievement. Two of their goods inputs measures, the availability of reading materials in the home and home ownership, are pertinent to this research. The subjects of the study were black families living in Gary, Indiana. Average family income for the group in 1971 was $5,200, 55% of the national average, and three-fifths of the families were female-headed. Five hundred and twenty-nine students in grades 3 through 6 during the 1972-1973 school year and 562 pupils in the same grades during the 1973-1974 school year constituted the same. Scores on the vocabulary subtest of the Iowa Test of Basic Skills were the measure of achievement. None of the variables they used describing physical resources in the home, including the availability of reading materials, were found to be consistently re-lated to student achievement. This finding was inconsistent with the
96
majority of research on the subject. does not necessarily mean that goods inputs do not influence achieve-ment and suggested that the fact that home resources are available does not necessarily mean that children derive any particular benefit from them. The effects of these resources, they posited, appear to be medi-ated through the educational level of mothers.
Beeves En 972) viewed the home environment in terms of structure, parent's attitudes, and processes. The structural dimension of the home was determined by (a) father's educational level, (b) father's
They stated, however, that this
occupation, (c) mother's occupation before marriage, (d) stated reli-gious affiliation, and (e) number of children in the family. Fire variables characterized the attitudinal dimension of the home. They were: (a) father's attitudes toward the child's present education, (b) mother's attitudes toward the child's present education, (c) father's ambitions for the child's future education and occupation, (d) mother's ambitions for the child's future education and occupation, and (e) prents' hopes and aspirations for themselves. The process dimension of the home was assessed by determining (a) parents' reports of favor-able relations between home and school, (b) use of books and library facilities (by the student), (c) provision of help with formal school work, and (d) arrangements made for tackling homework assignments. Keeves looked at the relationship between these home environment characteristics and student ac R evement in mathematics and science for Australian 11 and 12 year olds. The measure of ac R evement used was not specified.
97
The results of the path analysis Keeves performed indicated that the major factor influencing final achievement in mathematics was initial achievement, but both the attitudes of the home and initial attitudes of students toward mathematics had a small but statistically significant effect. Initial achievement accounted for 68.4% of the variation in final achievement scores. Only another 5.2% of the varia-tion in mathematics achievement was explained by home, classroom, and peer group environments. It is important to note, however, that ini-tial achievement was related to home, classroom, and peer influences. Initial achievement in sciences, the attitudes and practices of the home, structural characteristics of the classroom, and the interaction between the teacher and student, influenced science achievement. These results support other findings, including those of Dave (1963), which indicate that the home factors influencing achievement differ according to the nature of the subject,
Blau's C1981) analysis of the determinants of intellectual com-petence and achievement of black and white children concentrated more on family cultural variables as defined in the present research than on any of the other research examined. Five hundred seventy-nine black and 523 white mothers of fifth and sixth graders in three Chicago metro-politan area communities were interviewed in 1968. Students were from working and middle class families. The most recent IQ and achievement test scores were obtained for each child. The type of achievement test utilized was not specified.
One of the parent characteristics Blau looked at was participation
s
98
in voluntary associations. Mother's extrafamilial participation was found to be related to both SES variables and student achievement. The relationship between SES and organizational memberships was stronger for blacks than whites. For blacks, the correlations between extra-familial participation and mother's education, father's education, and mother's occupational status were .43, .32, and .31 respectively. For whites, the correlations between participation in voluntary associa-tions and SES components were .31, .30, and .32, respectively. Child-ren's mean IQ and achievement scores rose with the extent of mother's organizational memberships in both races. The relationship between IQ scores and mother's organizational affiliations was stronger for blacks (r-.21) than whites (r=.13). The relationship between mother's organizational memberships and achievement scores was similar for blacks (r=.l9) and whites r=.17).
Blau used three items to measure ''paternal fatalism," a concept similar to Rotter's (1966) "generalized expectancies for Internal versus External Control of Reinforcement." The items were: (1) children should be taught not to expect too much out of life so that they won't be disappointed, (2) children should learn early that there isn't much you can do about the way things are going to turn out in life, and (3) children should learn that planning only makes a person unhappy, since your plans hardly ever work out anyway.98
Fatalism was found to have an adverse effect on all measures of
positive investment of time and resources in children and was negatively
associated with students' achievement test and IQ scores for blacks
and whites. Early interaction between mothers and children, preschool mastery of skills, the provision of culturally enriching experiences, and the reading of material by experts on children were all found to be negatively associated with achievement.
As Table 3.6 shows, fatalism has a stronger negative association with achievement than IQ for blacks and whites. Blau interpreted this
TABLE 2.6
CORRELATIONS OF FATALISM WITH COMPONENT MEASURES OF INVESTMENT IN CHILD
AND CHILDREN'S ACHIEVEMENT99
Early interaction
Read experts
Mastery
Cultural enrichment
IQ
Achievement
Blacks Whites
-.25 -.18
-.19
-.17
-.22
-.21
-.35 -.12 -.29 -.18 -.29
finding as suggestive of the proposition that the communication of a fatalistic orientation is more damaging to children's progress in school as the tasks they are called on to learn become more difficult.l°°
Mbthers were asked three questions designed to determine their aspirations for their children's educational attainment, scholastic ex-cellence, and occupational attainment. The questions were: (1) What is the least amount of education that you think would be enough for
100
your (son or daughter)? (w) Would you be satisfied if (child) were passing; at the middle of (his/her) class; above the middle; out-standing? and (3) Would (child) be doing well enough if (he/she) had a job like ? OExamples of nonskilled, skilled, clerical, and professional occupations were provided.)l°l Mothers were also asked how far they expected their children to go in school. Performance on IQ and achievement tests was more highly correlated with maternal aspiration level for whites (.36 and .38, respectively) than for blacks (.27 and .26, respectively). This finding can in part be attributed to the fact that black mothers' aspirations for their childrenst edu-cation were as high or higher than those of white mothers. The correla-tion between mother's educational expectations and IQ was .33 for blacks and .36 for whites. For whites, the correlation between educational expectations and achievement scores was .41 compared to .36 for blacks. Blau interpreted these findings as an indication that the educational expectations voiced by black mothers may be more influenced by their children's demonstrated scholastic ability than are their aspirations.
Blau also investigated the relationship between family size and achievement. Family size was found to be a stronger predictor of achievement for blacks than for whites, There was a 10-point differ-ence in the mean IQ scores of black children in hones with one or two children (102) and those in homes with five or more children (92). The corresponding difference for whites was seven points tl09 and 102, respectively). Differences in achievement scores followed a similar pattern for both races.
101
Pbthers' participation in voluntary activities, aspirations for children's education, fatalism, and family size were found to be related either directly or indirectly to some indicator of SES.
Puny studies of the relationship between family characteristics and academic achievement investigate the relationship between parents' aspirations for their children's education. Babcock's (1972) conten-tion that high achieving children generally come from families with high educational expectations for them even though definitions of ex-pectations or aspirations vary, is supported in the majority of research. Crandall, Dewey, Katkowsky, and Preston (1964) defined parents' educa-tional aspirations as parents' minimal standards for children's in-tellectual performance. Their results showed that parents' aspirations were moderately correlated with girl's school achievement. The cor-relation between parents' educational aspirations and achievement was much smaller for boys than for girls. A moderate relationship between parents' aspirations and academic performance was also reported by Woelfel and Haller (1971), Keeves (1972) and St. John (1972). Gigliotti and Brookover (1975) reported a strong relationship between student achievement and parents' educational aspirations. Pugh (1976) concluded that the effects of parents' aspirations on academic performance were negligible for black and white male high school seniors. The number of years of school that parents expected their sons to complete was the measure of educational expectations in the Pugh study.
102
CHAPTER III
THE RESEARCH PROJECT: ITS St-l-lING,
MOTIVATION, POLITICAL BASIS, AND METHODOLOGY
Findings from a 1976 study entitled "Options: School Desegrega-tion," provided the incentive for the present investigation. The re-search was initiated in September of 1975, just a few months prior to the federal court order to desegregate the public schools in Milwaukee, Wisconsin. The Milwaukee community was preparing to implement the mandates of the impending court order. Individuals, civic organiza-tions, the media, business interests, and segments of the University of Wisconsin-Milwaukee community, were among the supportive participants in the desegregation planning process. One of the organizations in-volved, the Coalition for Peaceful Schools, consisted of representatives from various segments of the Milwaukee community. The Coalition was made up of a number of groups including the Greater Milwaukee Conference on Religion and Urban Affairs, the NAACP, the Milwaukee City Council of PTA's, and the Department of Urban Affairs at UW-M. Coalition activities included sponsoring a human relations training workshop at the YMCA, distributing over 950 informational packets listing community resources on the desegregation issue, and assisting local school prin-cipals in the implementation of desegregation related programs.] Milwaukee Junior Bar Association lawyers explained the legal aspects of desegregation to interested groups. The League of Women Voters engaged in a number of desegregation related activities, such as sponsoring a conference designed to assist community interests in the
103
development of communication techniques to disseminate information on desegregation, and submitting policy statements and recommendations to the school board.2
"Options: School Desegregation" was an investigation of the plan-ning processes, specific desegregation techniques employed (such as magnet schools and educational parks), patterns of community response, and problems encountered in desegregating five urban school districts. The declining educational quality in large urban school districts and arguments for and against desegregation were among other issues ad-dressed.
While the project did not directly address the relationship be-tween socioeconomic status, family culture, and academic achievement in interracial schools, several observations indicated that these is-sues are of major importance if equality in educational opportunity is to become a reality. The first observation was that many urban school systems were not providing students with an education that would prepare them for further education and gainful employment in today's market. The quality of education in urban school districts was rapidly deteriorating and desegregation could not be held responsible. In Pontiac, Michigan, for example, the quality of education provided was a problem the district had faced for many years.3 In 1972, a civic organization leader informed U.S. Commission on Civil Rights staff members that desegregation had helped the community to see what a poor job the schools were doing.4 Funds received under the Emergency School
.
104
Assistance Program and Emergency School Act to aid desegregation ef-forts were used to implement a variety of programs including in-service programs for professional and non-professional staff, community rela-tions programs, curriculum development, research and evaluation, and Latino student needs programs.5 These programs were not in operation prior to desegregation and would have been difficult if not impossible to establish without these funds. In Boston, Massachusetts, colleges, universities, and businesses assisted public schools as part of the desegregation effort. New programs of instruction were implemented and existing ones were strengthened. Brandeis Uhiversity worked with the citywide magnet English High & hoof and Harvard professors worked with the staff and students of Roxbury High School. It seemed that in many districts, educational quality improved as a result of de-segregation.
A second observation pertained to social class, but not in the same context as in this research. Several of the studies examined the relationship between social class and acceptance of interracial school-ing. Middle and upper class whites seemed to hold more liberal atti-tudes towards desegregation than did lower class whites. It seemed logical that if upper and middle class whites had what could be con-sidered a "healthy" attitude towards interracial contact in the schools, their general attitudes were most likely conducive to other forms of social and individual progress. On the other hand, the rejection of the idea of interracial schooling by many lower class whites was
-MORE- -MORE- -MORE- -MORE- -MORE- -MORE- -MORE- -MORE-
Part #2
Copyright by Pamela Jane Sampson-Malone, 1985,1998
All Rights Reserved
E-mail: dr.pamsm@sbcglobal.net
