The Influence of High School Racial Composition on the Academic Achievement and College Attendance of Hispanics

    The Influence of High School Racial Composition on the Academic 

Achievement and College Attendance of Hispanics 

Rita E. Mahard 

The University of Michigan 

Robert L. Crain 

The Rand Corporation 

and 

The Center for Social Organization of Schools 

December 1979 

 



  

The Influence of High School Racial Composition on the Academic 

Achievement and College Attendance of Hispanics 

Abstract 

Using data from the National Longitudinal Survey of the high school 

graduating class of 1972, this paper examines the impact of high school 

racial composition on the academic achievement, college attendance and 

college retention of three Hispanic minority groups. Data were obtained 

from 120 Puerto Ricans, 654 Mexican=-Americans and 145 "Other" Latins in 

298 schools. Regional regression analyses indicate that Puerto Rican and 

Southern "Other" Latin achievement and college attendance is higher in 

predominantly Anglo schools. Other Latins in the North and West show a 

positive effect of school percentage Anglo on achievement. The largest 

sample, Mexican-Americans, shows a zero effect on all dependent variables. 

 



® » : 

  

Public attention has been focused almost exclusively on school 

desegregation as it relates to black and Anglo students, even though the 

nation's second largest minority group, Mexican-Americans, has long been 

waging a similar battle for equality of schooling. The advent of desegre- 

gation in cities such as Los Angeles where large numbers of Mexican- 

Americans and other Hispanics reside suggests that expanding our perspective 

to include these minorities would be sensible. 

Research seems to indicate that blacks benefit from desegregation; 

achievement test scores frequently go up after desegregation, and Northern 

blacks who attend desegregated high schools are more likely to succeed in 

college. But it does not seem obvious that a different ethnic minority 

will be affected by desegregation the same way. Given the diversity among 

Hispanics, there also seems to be little justification for expecting a 

consistent effect across different Hispanic groups. 

The present paper examines the influence of high school racial 

composition on the achievement, college attendance and college survival 

of Hispanics. There is almost no research on the effects of desegregation 

on educational outcomes of Spanish-speaking minorities. 

METHOD 

The National Longitudinal Survey (NLS) of the High School Class of 

1972, is a longitudinal study of 23,451 high school seniors drawn from 

1,318 high schools. This study uses data from the baseline survey and 

the first two follow-up surveys conducted in October 1973 and 1974. The 

follow-ups have completion rates of 937% and 947%. 

We used this sample to look at the relationship between high school 

racial composition and three dependent variables: achievement test 

 



performance, the probability of attending college, and for those who go to 

college, the probability of remaining there rather than dropping out. 

The NLS sampled 919 Hispanic students from 298 schools. Seventy-one 

percent of our respondents identified themselves in the baseline question- 

naire as Mexican-American or Chicano, 13 percent as Puerto Rican and 16 

percent as ''other Latin-American. The majority of these students 

77 persent) attended high school in the South or the West. 

For analysis purposes, samples in the North and North Central regions 

were collapsed, creating nine ethnicity-region groups. Puerto Ricans from 

the South and the West and Northern Mexican-Americans were eliminated at 

the onset due to small sample sizes. The three Other Latin samples and the 

Puerto Rican sample contained too few college attenders to permit an analysis 

of college retention rates. The achievement and college attendance analyses 

are thus restricted to six of the original nine ethnicity-region groups, 

and College retention analyses are done only for Mexican-Americans. 

Achievement is measured by a test taken during the senior year of 

high school. It is standardized to a theoretical mean of 50 and standar- 

dized deviation of 10. 

The college attendance variable was constructed from student reports 

and is the percentage attending college, either full or part time, at any 

time between high school graduation and the administration of the Second 

Follow-Up Questionnaire (Fall 1974). (Attendance at vocational, technical 

and other types of institutions is not counted.) 

The college retention variable is, for those who attended college 

and completed the second follow-up questionnaire, the percentage who 

were juniors three years after high school graduation. This is an extremely  



stringent measure of college ''success,' since it requires full-time, 

uninterrupted progress toward a four-year degree. 

High school percentage Anglo serves as the independent variable. 

The measure was constructed for about 907% of our schools, from data in 

DHEW's 1972 Directory of Public Elementary and Secondary Schools. These 
  

data are for the fall immediately after our sample graduated, but this 

difference in time introduces only a small error. 

Control variables for the analyses are individual student socioeconoic 

status, predominant language spoken at home, urbanism of the community 

and school size. 

Individual student socioeconomic status is a scale pooling data on 

father's education, mother's education, parents' income, father's occupa- 

tion and household items. The scale ranges from a low of 0 to a high of 

100. 

The predominant language at home is simply the percentage answering 

"yes" when asked, "Is English the language spoken most often in your home?" 

The urbanism variable measures the size of the community in which the 

school is located, and ranges from O in rural areas to 100 in cities of 

over a half million. 

School size is the total enrollment for She 1971-1972 school year. 

Table 1 gives means and standard deviations of dependent, independent 

and control variables for each ethnicity-region combination. For 

comparison purposes we have also included achievement and college attendance 

data for Northern and Southern-educated Anglos and blacks, taken from an 

earlier analysis (Crain and Mahard, 1978). 

(Table 1)  



6 » 3 

  

Achievement test scores are highest for Northern Anglos, lowest for 

Southern blacks, with the six Hispanic groups performing slightly better 

than blacks and well below Anglos. Mexican-Americans in the West and 

Puerto Ricans in the North score slightly below Northern blacks. In 

every region, Other Latins are the highest achieving minority group. 

College attendance rates in the South are the lowest for Mexican- 

Americans at 43 percent and the highest for Anglos at 56 percent. In the 

North and West, blacks have the lowest college attendance PRESS. followed 

closely by Mexican-Americans. Other Latins are about on par with Anglos. 

The small group of Puerto Ricans in the sample have the highest college 

going rates--67 percent. 

Unfortunately, the earlier analysis did not use the came measure of 

college retention, so we cannot compare the results. Among Hispanics, 

Southern-educated, Mexican=-Americans have high rates in the South, at 29 

percent, while Western Mexican-American rates are low. 

Desegregation and Student Outcomes   

Table 2 presents the first regression analysis. Achievement test 

scores are regressed on school percentage Anglo, with SES, home language, 

urbanism and school size entered as controls. The results are inconsistent. 

Mexican-American aclifeveliont Ls not affected by school racial composition, 

while the other four groups show positive effects. Despite the small 

sample size, school percent Anglo is a significant predictor of achieve- 

ment for "Other Latins" in the West and for Puerto Ricans. 

(Table 2) 

The direction of the language effects is somewhat surprising. English 

as the home language is never helpful for Achievement. Puerto Ricans and 

 



% » 5 

  

two of the three "Other Latin'' groups are less likely to do well on 

standardized achievement tests when the home language is English. For 

Mexican-Americans and other Latins in the South, the effect of English is 

negligible. 

Table 3 presents the data on college attendance. High school percentage 

Anglo has little influence on the college attendance rates of four groups. 

The standardized regression coefficients range from -.079 for Northern 

Other Latins to +.085 for Western Mexican-Americans. Puerto Ricans and 

Southern Other Latins do seem to benefit from attending predominantly 

majority tthaols. however. ‘For Puerto Ricans, the standardized regression 

coefficient is a respectiohie .287. For the Southern Other Latin group 

school percentage Anglo is the single strongest predictor in the equation 

(&= .523). 

(Table 3) 

Individual SES has little impact on college attendance. The effects 

are near zero for four groups. The remaining two groups show a negative 

effect, though in each case. the standardized regression coefficient is far 

from significant due to ths felatively small sample size. 

The language results again vary considerably. No effect is seen for 

either Mexican-American groan: English is a mild negative predictor of 

college attendance for: Puerto Ricans and Northern Other Latins. Southern 

Other Latins appear to be the only group which benefits to any degree 

from English as the home language, but the pattern is far from significant. 

Table 4 presents the college retention rates for Western and Southern 

Mexican-Americans. 

(Table 4) 

There is no evidence that either group is positively affected by desegre- 

gation. 

 



it » 7 

  

The present analyses do not provide a clear-cut picture of the likely 

effects of desegregated schooling on Spanish-speaking minorities. Wnat 

we have seen instead is a mixture of positive and negative effects that 

varies by ethnic group, by region and by outcome, which raises more questions 

than it answers. 

For Mexican-Americans, the story is consistent: desegregation is not 

helpful, whether the criterion is achievement, college attendance, or 

college retention. For other Latins in the North and West, attending 

school with Anglos seems to affect achievement, but not college attendance. 

But for Puerto Ricans and Other Latins in the South, a predominantly white 

high school means both higher achievement and higher college attendance 

rates. 

Individual socioeconomic status tends to be associated with higher 

achievement, although the magnitude of the association varies considerably. 

Generally, the, pattern is strongest for the Other Latin groups. This 

seems reasonable, since this group has a higher mean and standard deviation; 

many immigrants from this group are well-educated, and middle-class. 

The language effects are reasonably straightforward, if counterintui- 

tive. English as the predominant language contributes little to successful 

achievement, college attendance or college retention. In only two cases 

is the effect of English language on subsequent attainments positive. 

Southern Other Latins are slightly more likely to attend college and 

Western Mexican-Americans are significantly more likely to remain in college 

when English is the home language. 

 



* $ ° 

  

INTERPRETATION   

Today, desegregation law defines Hispanics, and all racial minorities, 

as minorities entitled to a desegregation remedy. If these data tell us 

anything, it is that blacks and Hispanics are different. Which is not 

necessarily to say they do not both merit desegregation; but we desparately 

need more research on the significance of desegregation for Hispanics. We 

also need research which helps us understand desegregation as a process; 

for if further research agrees with this analysis, and shows that Mexican- 

Americans do not benefit from desegregation, we will need to know why 

blacks benefit and how Chicanos differ from blacks in order to interpret 

our findings. 

It is presumptious to pose any hypothesis as more than speculation, 

but perhaps we can assist the next researcher in this difficult Field by 

stating some possibilities. 

Perhaps the most compelling theory of why blacks benefit from desegre- 

gation is that desegregation simply means the merging of low-income with 

high-income students; achievement goes up because the instructor sets a 

faster pace for the class, and covers more of the material that appears on 

standardized tests; college attendance and retention is higher because 

the school is accustomed to sending its graduates to college and to 4-year 

colleges with low dropout rates rather than to junior colleges or vocational 

schools. The theory may be stated as a structural argument, stressing the way 

the institution adapts to a desegregated student body, or as a social 

psychological theory, stressing ''transfer' of aspirations from majority 

to minority. But such a theory would fit the desegregation of any low- 

income group, black, white or Hispanic. Why does it not fit Mexican- 

Americans? 

 



Gordon's (in Grebler et al., 1970) analysis of Mexican-American 

achievement in Los Angeles, found no such effect. In a secondary analysis 

of the Jordon data, TenHouten et al. (1971) concluded that Mexican-Americans 

were less likely to develop college plans in predominantly majority schools. 

We hypothesize that these transfer effects occur only if either (1) the 

group is not viewed as inferior, or (2) there is strong social pressure 

to counteract such a view. With this hypothesis, one can argue that other 

Latins will benefit from desegregation because they are not a major ethnic 

group sabidet to discrimination, and Blacks and Puerto Ricans will benefit 

from the social pressure to make desegregation work; but Mexican-Americans 

will again be the forgotten minority, ignored, placed in the bottom track, 

or discriminated against in the classroom in the way Jackson and Cosca 

(1974) demonstrate. 

The next research question is, why do other Latins in the North and 

West who attended schools with Anglos not have higher rates of college 

attendance? Again, we can only speculate, but perhaps the key is that 

other Latins in the North and West are never the majority in their school; 

they are usually in schools with Anglos, and when they are not, they are 

likely to be a high-status minority in a black or Chicano school, and 

hence should be encouraged to attend college. 

There is one final peculiarity in the data which begs for further 

analysis. This is the difference between Other Latins in the South and 

those in the North and West. The Other Latins in the South are presumably 

mainly Cuban-Americans in Florida; not only does this group benefit 

from desegregation, their college attendance benefits, unlike every other 

group, from growing up in an English-speaking household. Cuban-Americans  



® oe 

  

are unusual in one other way; they are reputed to be quite successful 

entrepenuers. Drawing on Portes' work, let us consider the possibility 

that for this group, and only this group, there is little social pressure 

to succeed academically, simply because there are ample opportunities for 

success within the barrio, where neither English cognitive skills nor: a 

college education is necessary. Contact with Anglos in a desegregated 

school provides information about a second way to succeed. Of course, 

it is the traditional conservative argument that minorities must be helped 

to assimilate in order that they might escape poverty. It would be a 

delicious irony if further research were to conclude the opposite; _ that 

poor minorities do not need to rub shoulders with the majority group in 

order to realize that their best hopes lie in education, while successful 

minorities need desegregation in order to learn that they do not have to 

stay in the ghetto in order to succeed in life. 

 



  

MEANS AND STANDARD DEVIATIONS FOR INDEPENDENT AND DEPENDENT VARIABLES 

  

Table 1 

    

  
  

  

  

  
  

  

  

  

  

  

  

  

Mexican- Mexican- Other Anglo-American 

Americans Americans Latins 

(West) (South) LU ANorth) (North- (South) 

. i < = pe West) - 

PET. Lah ey po Nis Le X 

Achievement (mean) 42.6 6.72 44.1 6.99 Li 2 6.26 52.3 51.2 

College Attendance (7%) 57 50 43 50 54 51 59 56 

College Retention (%) 16 37 29 46 -- -- -— -— 

SES (mean) 32 14 31 14 45 14 -— -- 

English at Home (7%) 59 49 45 50 46 51 -— -— 

Urbanism (mean) 35 42 43 40 78 36 -- -- 

School Size (mean) 1680 766 1737 876 2904 1427 ee a 

School 7 White (mean) 50 27 32 22 53 39 94 81 

N (376) (240) (65) 

Other Other Puerto Blacks 

Latins Latins Ricans 

(West) (South) (North) (North & (South) 
- . fe i - - West) 
X G X G X q - 

2 X X 

Achievement (mean) 46.0 8.41 Ly, 2 6.90 42.8 5.36 43.7 41.6 
College Attendance (7%) 63 48 48 51 67 47 56 48 

College Retention (%) -— —- -— -—- 19 40 -- on 

SES (mean) 52 20 Lb 18 29 12 a -- 

English at Home (%) 68 48 57 51 52 50 -- -- 

Urbanism (mean) 4L2 37 68 4L2 88 30 -- -- 

School Size (mean) 1961 606 2575 1479 2889 1472 -- -- 

School 7 White (mean) 71 24 52 36 19 25 39 43 

N (42) (38) (38) 

  

 



  

THE INFLUENCE OF HIGH SCHOOL RACIAL COMPOSITION ON ACHIEVEMENT 

  

Puerto Ricans Other Latins Mexican-Americans 

  

  

North North South 

RB r RB r 8 r 

Individual SES .164 .136 404% 401 240% .236 

English at Home -.240 -.187 -.269 -.097 030 .108 

Urbanism «371 1% 021 -.152 -.040 -.005 056 

School Size -.047 -.002 .119 -.079 -.016 046 

Percentage Anglo 432% «153 «201 «224 -.054 010 

2 142 L243 .058 

N (60) (37) (149) 
  

  

Other Latins Mexican-Americans Other Latins 

  

  

South West West 

B r B r 3 r 

Individual SES 155 .360 127 .162 419% .607 

English at Home 004 .169 014 .078 -.187 -.057 
Urbanism 128 .100 -.262% -.086 -.218 011 

School Size 142 -.,008 225% .068 479% 430 

Percentage Anglo 449 «372 095 .178 404% 428 

yj 

rr .235 .077 .641 

N (29) (254) (23) 
  

* 
p<.05 

 



  

Table 3 

THE INFLUENCE OF HIGH SCHOOL RACIAL COMPOSITION ON COLLEGE ATTENDANCE 

  

Puerto Ricans Other Latins Mexican-Americans 

  

  

North North South 

B r BR r RB r 

Individual SES 062 .166 -.175 .187 040 120 

English at Home -.128 -.202 -.158 -.329 003 056 

Achievement 304% .391 326 .170 408% 414 

Urbanism 099 041 368 «322 -.009 032 

School Size 330% .343 250 412 -.024 .003 

Percentage Anglo 287 «116 -.079 -.433 007 .000 

.306 L404 174 

N (46) (24) (116) 
  

  

Other Latins Mexican-Americans Other Latins 

  

  

South West West 

B r BR r 8 T 

Individual SES -.272 .090 -.039 061 -.090 1.193 
English at Home .320 +313 010 022 ~-.357 -.218 
Achievement .356 402 319% 310 220 «347 

Urbanism -.027 -.034 261% 226 283 273 

School Size .489 -.194 004 213 303 +387 

Percentage Anglo .523 411 085 101 017 .130 

> 339 «165 334 

N (23) (193) (19) 
  

* 
p<.05 

 



Table 4 

THE INFLUENCE OF HIGH SCHOOL RACIAL COMPOSITION ON COLLEGE RETENTION 

    

Mexican-Americans Mexican-Americans 

South West 
  

B 
  

Individual SES .049 
English at Home -.024 
Achievement «521% 
Urbanism .045 
School Size -.052 
Percentage Anglo -.252 

+ .306 

N - (48) 
  

* 
p<.05 

 



  

REFERENCES 
  

Brischetto, Robert and Tomas Arciniega. 
1973 "Examining the examiners: A look at educators' pers- 

pectives on the Chicano student." In de la Garza, 
Rudolpho 0., Z. Anthony Kruszewski and Tomas Arciniega. 
Chicanos and Native Americans: The Territorial Minorities. 

Englewood Cliffs: Prentice Hall. 

Coleman, J.S., E.Q. Campbell, C.J. Hobson, J. McPartland, A. M. Mood, 

1966 F.D. Weinfeld and R.L. York. Equality of Educational 
Opportunity. Washington, D.C.: Government Printing 

Office. 

Crain, Robert L. and Rita E. Mahard. 

1978a "School racial composition and black college attendance 
and achievement test performance." Sociology of Education 
(April): 81-101. : 

Crain, Robert L. and Rita E. Mahard. 

1978b The Influence of High School Racial Composition on the 
College Attendance and Achievement Test Scores of Black 
Students. Washington, D.C.: National Center for Edu- 
cation Statistics. Forthcoming. 

Felice, Lawrence G. 

1974 The Effects of School Desegregation on Minority Group 
Student Achievement and Self-Concept: An Evaluation of 

Court-Ordered Busing in Waco, Texas. Washington, D.C.: 

National Center for Educational Research and Development. 

Godoy, Charles E. 

1970 Variables Differentiating Mexican-American College and High 
School Graduates. Sacramento: California State Department 

of Education. 

Gordon, C. Wayne, Audrey J. Schwartz, Robert Wenkert and David Nasatir. 

1970 "Mexican-American youth in Los Angeles schools." In Grebler, 
: Leo, Joan W. Moore and Ralph C. Guzman. The Mexican- 

American People: The Nation's Second Largest Minority. 

New York: The Free Press. 

Jackson, Gregg and Cecilia Cosca. 

1974 "The inequality of educational opportunity in the Southwest: 
An observational study of ethnically mixed classrooms." 

American Educational Research Journal (Summer): 219-229. 

Kimball, William L. 

1968 Parent and Family Infiuences on Academic Achievement Among 
Mexican American Students. Doctoral Dissertation. Univer- 

sity of California, Los Angeles. University Microfilms 

No. 68-16, 550. 

 



  

Lopez, David E. 

1976 "The social consequences of Chicano home/school bilin- 
gualism.'" Social Problems (December): 234-246. 

Mayeske, George W. 
1967 Educational Achievement Among Mexican-Americans: A Special 

Report from the Educational Opportunities Survey. 

Washington, D.C.: Government Printing Office. 

National Assessment of Educational Progress. 
1977 Hispanic Student Achievement in Five Learning Areas: 

1971-75. Denver: National Assessment of Educational 
Progress. 

Ovando, Carlos Julio. 

1977 Factors Influencing High School Latino Students' Aspi- 
rations To Go To College: The Urban Midwest. San Francisco: 

R and E Research Associates. 

Rangel, Jorge C. and Carlos M. Alcala. 
1972 "De jure segregation of Chicanos in Texas schools." 

Harvard Civil Rights - Civil Liberties Law Review (March): 

307-391. 

Salinas, Guadalupe. 

1971 ""Mexican-Americans and the desegregation of schools in the 
Southwest." Houston Law Review (May) :929-951. 

TenHouten, Warren D., Tzuen—-jen Lei, Francoise Kendall and C. Wayne 

Gordon. "School ethnic composition, social contexts, and 
educational plans of Mexican-American and Anglo high school 

students." American Journal of Sociology (July) :89-107. 

U. S. Congress, 91st, 2nd session. 

1970 Equal Educational Opportunity Hearings, Part 8--Equal Edu- 
cational Opportunity for Puerto Rican Children. Washington, 

D.C.: Government Printing Office.