Role Models, Mentors, and Sponsors: The Elusive Concepts

Author(s): Jeanne J. Speizer

Source: Signs , Summer, 1981, Vol. 6, No. 4 (Summer, 1981), pp. 692-712
Published by: The University of Chicago Press

Stable URL: https://www.jstor.org/stable/3173738

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/3173738?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend
access to Signs

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
 REVIEW ESSAY

Role Models, Mentors, and Sponsors

The Elusive Concepts

Jeanne J. Speizer

Role models, mentors, and sponsors are concepts whose time has com
Professionals must have had one, been one, or be seeking one if they a
to advance their careers. Articles in the popular media and in p
fessional journals continually declare, as did a recent title in the Harva
Business Review, "Everyone Who Makes It Has a Mentor."1 Senior pro
fessionals who look back over their lives assure us that they owe the
success to having had one; middle-level professionals say with pride t
not only have they had one but they are one; and junior professional

This paper was originally prepared as a background document for the Research Co
ference on Educational Environments and the Undergraduate Woman sponsored by Higher
cation Resource Services and the Center for Research on Women at Wellesley Coll
September 1979. The author was supported in part by grants from the Fund for
Improvement of Postsecondary Education (grant G007701346) and the National Insti
of Education (grant G790058). I wish to thank Dr. Pamela J. Perun for her careful read
and editing of this paper and for her suggestion to prepare the tables and Barbara R. L
for her creative ideas and constant support.
1. G. C. Collins and P. Scott, "Everyone Who Makes It Has a Mentor," Harva
Business Review 56, no. 4 (1979): 89-101.

EDITORS' NOTE: From their history women are learning to be wary of tho
who would tell them, even with the best intentions and often with ambigu
intention, what women want or what women need. JeanneJ. Speizer's careful l
at the concepts of role models, mentors, and sponsors shows how important it
be to question all counsel, examine all premises, search out the roots of ideas.
[Signs: Journal of Women in Culture and Society 1981, vol. 6, no. 4]
? 1981 by The University of Chicago. 0097-9740/81/0604-0006$01.00

692

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
Signs Summer 1981 693

are constantly worrying that they will not advance unless th

Women who also must have one, especially a female one,
around and realize that the lack of a role model, mentor, or
be another explanation, in a long list of explanations, for why
will not advance.
The idea that a role model, mentor, or sponsor is a prerequisite fo
success has achieved the sudden recognition that makes it appear self-
evident. At such a moment it is particularly important to ask whether
validity has indeed been demonstrated. Are role models and mento
truly critical to professional advancement? Or are they merely anoth
contemporary rationalization for women's lack of professional progre
What immediately becomes apparent on a review of articles about the
subject is that the concepts are new to the literature: The term "r
models" first appears as a category in most fields in 1973.2 To da
mentor and sponsor are not listed as separate categories and must
found by other descriptions.
This report is divided into three sections: (1) an analysis of resear
efforts that explore role models in the school environment; (2) a study
materials relating to both mentors and sponsors, since the two terms
often used interchangeably; and (3) a final section that summarizes th
findings to date and discusses further research needs. Highlights of t
major findings related to role models, mentors, and sponsors will
discussed in the text. For specific details about the major studies
reader should refer to the tables.

Role Models

Belief in the necessity for role models appears to be based on devel-

opmental theories of identification and modeling in childhood,
specifically, social learning theory and cognitive development theory.3
Kemper described a role model as a person who "possesses skills and
displays techniques which the actor lacks ... and from whom, by obser-
vation and comparison with his own performance the actor can learn."4
Most of the studies related to role models have been focused on young
children and their parental models. Several studies have examined par-
ents as role models for college students and explored the effects of
2. The abstracts and indexes reviewed were Psychological Abstracts; Current Index to
Journals in Education; Business Periodicals Index; Women's Studies Abstracts; and Sociological
Abstracts.

3. A. Bandura, "A Social Learning Theory of Identification Process," and L.

Kohlberg, "Stage and Sequence: The Cognitive-Developmental Approach to Socialization,"
in Handbook of Socialization Theory and Research, ed. David A. Goslin (Chicago: Rand Mc-
Nally & Co., 1969).
4. T. Kemper, "Reference Groups, Socialization and Achievement," American
Sociological Review 33 (1968): 31-45.

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
694 Speizer Review: Role Models

mothers (working and nonworking) and fathers as models on the career

choice, commitment, and aspirations of students.5 The results of these
studies are ambiguous. In fact, the variations in findings are often more
striking than the agreements, with a range of results from those who
found fathers to be key figures6 to those who found working mothers to
be the key,7 confounded by Tangri, who found no evidence that college
women who aspire to nontraditional careers identified with either parent
in preference to the other.8
As socialization by parents has been primarily accomplished by the
time students are in their late teens, this review will not explore studies
related to parents as role models but will concentrate rather on the
effects of role models outside the family, particularly in school settings
where most of the studies have been conducted. Since the presence of
role models in the learning environment has been considered an im-
portant aspect of women's colleges, they receive attention in a separate
section. The value of role models in other settings and for other popula-
tions has often been asserted by authors, but there have been few studies
on other environments or groups.

High School and Undergraduate Years

Teachers as role models at the high school, undergraduate, and

graduate levels have all been the subject of study. There have been few
studies that explore the effect of administrators as role models, although
the paucity of female role models in administrative positions at the sec-
ondary school and college level has been posited as an explanation for
the continuing lack of women in leadership positions.9
The findings related to faculty who serve as role models are variable
and sometimes contradictory. The variations range from a study by
5. E. M. Almquist and S. S. Angrist, "Role Model Influences on College Women's
Career Aspirations," Merrill-Palmer Quarterly 17 (1971): 263-79; B. B. Seater and C. L.
Ridgeway, "Role Models, Significant Others, and the Importance of Male Influence on
College Women," Sociological Symposium 15 (Spring 1976): 49-64; S. S. Tangri, "De-
terminants of Occupational Role Innovation among College Women," Journal of Social
Issues 28, no. 2 (1972): 177-99.
6. Margaret Hennig and Anne Jardim, The Managerial Woman (New York: Doubleday
& Co., 1977).
7. Almquist and Angrist, pp. 271-73.
8. Tangri, p. 184.
9. K. D. Lyman and Jeanne J. Speizer, "Advancing in School Administration: A Pilot
Project for Women," Harvard Educational Review 50, no. 1 (Winter 1980): 25-35; N. P.
Cummings, "Women in Higher Education Administration," in Women in Higher Education
Administration: A Book of Readings, ed. M. C. Berry (Washington, D.C.: National Association
for Women Deans, Administrators, and Counselors, 1979); T. Antonucci, "The Need for
Female Role Models in Education," in Women and Educational Leadership, ed. S. K. Biklen
and M. B. Brannigan (Lexington, Mass.: D. C. Heath & Co., 1980).

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
Signs Summer 1981 695

O'Donnell and Anderson,10 in which women students r

could not identify specific faculty members who served a
for their choice of majors, to a study by Basow and Howe
female teachers were found to be important for female c
in their career decisions. The Basow and Howe study al
example of contradictory findings related to role mo
searchers conducted two experiments; in the first exp
found that female teachers had little effect on the career choices of
female and male students. Basow and Howe then repeated the exp
ment, using only senior students and balancing the number of fem
and male students in each group. They then found the opposit
sponse; female teachers are important as career-choice role models f
female students. Basow and Howe suggest that the differences betw
the two findings are related to their modifications in design. Howev
the different findings may reflect the type of problems that are inhe
in role model studies and that need to be addressed.
One approach to determining the effects of teacher role models
students has been to survey the behavior of women students as
related to the availability of female faculty models. Female students h
been found more likely to major in departments where there are fe
faculty members.'2 Students also appear more likely to choose th
courses on the basis of the sex of the teacher and to respond dif
entially depending on the sex of that teacher.13 However, Casser
found that in twelve high schools where high percentages of girls w
enrolled in advanced placement programs in math and science, the se
of the teacher was secondary in importance to the teacher's active re
cruitment of female students. This last study is consistent with Kante
theory'5 that the sex ratio of the class rather than the sex of the teac
affects class selection by females. There are no studies that provide
formation about what actually causes girls to choose one class o
another and that tell how their success is affected by those choices.
Another way to determine whether students have role models is

10. J. A. O'Donnell and D. G. Anderson, "Factors Influencing Choice of Major

Career of Capable Women," Vocational Guidance Quarterly 26, no. 3 (1978): 215-21.
11. S. A. Basow and K. G. Howe, "Role-Model Influence: Effects of Sex and Sex-Role
Attitude in College Students," Psychology of Women Quarterly 4, no. 4 (Summer 1
558-72.
12. S. S. Angrist and E. M. Almquist, Careers and Contingencies: How College Women
Juggle with Gender (New York: Dunellen Publishing Co., 1975).
13. S. H. Sternglanz and S. Lyberger-Ficek, "Sex Differences in Student-Teache
Interactions in the College Classroom," Sex Roles 3, no. 4 (1977): 345-52.
14. P. L. Casserly, An Assessment of Factors Affecting Female Participation in Advanced
Placement Programs in Mathematics, Chemistry and Physics, National Science Foundation Re-
port, grant no. GY-11325 (Washington, D.C.: National Science Foundation, 1975).
15. R. M. Kanter, Men and Women of the Corporation (New York: Basic, 1977).

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
696 Speizer Review: Role Models

ask them about the existence of such a person and, if one is identified,
about the effect that the role model has on them. Studies with this
approach have not, to date, been very successful in helping to clarify th
presence or need for a faculty role model. For example, Tangri found
that female faculty in the student's major field, as well as female colleg
friends, provided some role support for college women who chose non
traditional occupations; but a tolerant or supportive boyfriend appeare
as the most important factor.16 Almquist and Angrist found that wom
students were influenced by a teacher or occupational role model, but
the sex of the role model did not matter. 7 Stake and Granger found tha
female and male high school students were more likely to have hig
scores in science-career commitment, if they chose as a role model
same-sex teacher whom they perceived to be attractive.'8 The authors
determined attractiveness by asking students to indicate on a seven-poi
scale the extent to which they would like eventually to resemble their
same-sex science teacher (model). As attractiveness often depends on t
eyes of the beholder, it is difficult to generalize from this study abou
how one might choose "attractive" teachers to serve as role models for
students.
Seater and Ridgeway found that undergraduate women students
who could identify a female faculty role model (44 percent) had
significantly higher degree expectations and were significantly more
likely to have plans to enroll in graduate school than the women students
(56 percent) who could not identify role models.'1 To be chosen as a role
model, a female faculty member had to be perceived as an achiever and
as a woman who had male approval. The authors then assessed the
effects of encouragement by male faculty and found that those female
and male students who had higher degree expectations and more
graduate school plans than their peers also reported encouragement by
male faculty members; however, the researchers point out that the re-
lationship between encouragement and graduate plans for females did
not "quite" reach statistical significance (P < .10).20 There is no way to
ascertain whether the female students who could identify female faculty
role models-and thus had higher expectations-were the same students
who had higher expectations because of male faculty encouragement.
The study becomes even less clear when the authors report that female
students perceive male faculty members as less likely than their female

16. Tangri, p. 197.

17. Almquist and Angrist, p. 275 (see n. 5).
18. J. E. Stake and C. R. Granger, "Same-Sex and Opposite Sex Teacher Model
Influences on Science Career Commitment among High School Students,"Journal of Edu-
cational Psychology 70, no. 2 (1978): 180-86.
19. Seater and Ridgeway, pp. 54-56 (see n. 5).
20. In studies of this type, findings are usually considered to be significant at the
probability level (P) of .05 or below.

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
Signs Summer 1981 697

counterparts to support their graduate plans, but also consid

couragement of female faculty members as directed towa
support only and not toward specific educational aspirations.
Ridgeway have reported each of the above results as if th
relation to each other-although each finding is based on the
dents answering one questionnaire at one point in time. T
left to wonder whether there are any correlations among the
example, a profile of that 44 percent of the women students
female faculty role models must take account of their need
agement from male faculty and document as well the type o
they received from female faculty. The authors must clarify t
before they can be used for our further understanding
portance of female faculty role models for women students.
The study of interactions between teacher and student w
classroom has been another way to assess role-modeling e
observed that gifted high school girls in special accelerated m
are likely to achieve a higher math level if the course is t
woman in an all-girls class or in a class where there are at le
(undefined) number of girls relative to that of boys.21 If in
ated math courses the number of girls compared to that of bo
the girls, Fox reports, tend to drop out. Fox points out that h
tions were not part of a true experiment and, therefore, sho
as tentative hypotheses rather than as conclusions. Experime
be conducted that ask which of two factors have greater
girls: the presence of a female faculty role model or a sufficie
of other girls in the class.
In classes where male students were in the majority and t
was male, female college students were found to initiate inter
often and end interactions sooner than male students.22 In f
students tended to dominate all interactions with female or male
teachers, whether the male students were in the majority or the m
The authors surmise that female students might be less inhibite
class taught by a female instructor because of her effect as a
Since, however, only 18 percent of the teachers and 38 percent o
students were female in the study, there was little chance to rec
interactions of female students in female-led classes or in classes where
the majority of the students were female. Thus, the behavior recorded
may have more to do with the scarcity of females in the setting, as Kante
would suggest, than with the value of a role model.
In 1979 Sternglanz replicated the above study and found, again,
21. L. H. Fox, "Sex Differences: Implications for Program Planning for the Academi-
cally Gifted," in The Gifted and the Creative: A Fifty-Year Perspective, ed. J. C. Stanley, W.
George, and C. H. Soleno (Baltimore: Johns Hopkins University Press, 1977), pp. 127,
128.
22. Sternglanz and Lyberger-Ficek, p. 349.

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
698 Speizer Review: Role Models

that male students tended to dominate classroom interactions.23 They

then reexamined the data from the two studies to ascertain whether
there were situations where male students did not dominate.
found that in seven of the eight classes taught by female professo
where the department was not a traditional place for females, the in
actions were either egalitarian or dominated by female students. T
in the sixteen female-taught classrooms, seven (44 percent) pro
environments where female students interacted equally with male
dents or dominated the interactions. We do not know if this difference in
pattern of interactions is due to the behavior of the teacher, the number
of female students in the class, the subject matter, or factors as yet
unidentified. In short, the findings on the gender of role models are
contradictory.
Consider another example. Male faculty members report that they
identify with students of their own sex to a greater degree than with
students of the opposite sex.24 The male faculty members in Tidball's
study were relatively insensitive to issues that affected their women stu-
dents and were important as well to their female colleagues; these male
faculty members reported, moreover, that they saw female students as
sexual beings who were less serious and capable than the male students.
In a survey of women students at the University of California at Berke-
ley, Heyman confirmed Tidball's finding.25 Among the frequent com-
plaints of the female students were: the absence of collegial relationships
with faculty members; fewer opportunities than those given male stu-
dents for collaborative work with faculty; and a sense that their academic
aspirations were less positively reinforced than those of male students.
Sternglanz and Lyberger-Ficek, however, found no differential treat-
ment of male and female students by teachers of either sex, despite the
differential interaction pattern of the female students.26 Clearly, addi-
tional research is needed for the proper interpretation of these findings.
Perhaps interactions in a classroom are related to sex-role attitudes and
beliefs not easily observable but powerful in their effect on the student
nonetheless, both during college and after graduation; or perhaps inter-

23. S. H. Sternglanz, "Sex Differences in Student-Teacher Classroom Interactions"

(paper presented at the Research Conference on Educational Environments and tie
Undergraduate Woman, Wellesley College, September 13-15, 1979). (Based on work done
with S. Lyberger-Ficek and M. Liss at the State University of New York at Stony Brook.)
24. A. R. Hochschild, "Inside the Clockwork of Male Careers," in Women and the Power
to Change, ed. Florence Howe (New York: McGraw-Hill Book Co., 1975); M. E. Tidball,
"Of Men and Research: The Dominant Themes in American Higher Education Include
Neither Teaching Nor Women,"Journal of Higher Education 47, no. 4 (1976): 373-89.
25. I. M. Heyman, "Women Students at Berkeley: Views and Data on Possible Sex
Discrimination in Academic Programs," mimeographed (University of California, Berke-
ley, June 1977).
26. Sternglanz and Lyberger-Ficek (see n. 13).

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
Signs Summer 1981 699

actions in the classroom have nothing to do with the effect

role model on the postcollege life of a student.
The status of a female faculty member may also affect
ness as a role model. Block points out that in most colleges
professors who are most likely to be judged successful.27 T
student might be forced to choose between identifying with
same-sex role model or with a successful, opposite-sex role
ticularly in a research university where status is attained by
publishing. If status in a university goes to those who do r
who publish-and women spend most of their time in te
the actual status of women, unrelated to their rank, is like
and their ability to serve as a role model for nontraditiona
be impaired. In a women's college, or other liberal arts c
teaching is required and rewarded for all faculty members
female faculty members may be higher and their potential
role models enhanced, but the professional horizon may
ited.

Graduate Years

Role models for graduate students is a topic that has been explored
from several perspectives. Goldstein studied the productivity four years
after graduation of Ph.D. psychology graduates from three universities
and nine areas of specialization and found that graduates whose advisors
were of the same sex published significantly more research than did
graduates who had opposite-sex advisors.28 However, the number of
people studied was too small (between twenty-five and thirty in each of
the four groups) to allow one to control for probable causal factors.
Thus, it is unclear if sex of the advisor was the cause of increased pro-
ductivity. The study needs to be replicated with a large sample so that
comparisons within and among groups can be made to ascertain the
effects of the different subspecialties, the ages and positions of advisors,
and the differences among the three institutions. The interaction be-
tween the productivity of the faculty member and the doctoral student
also has to be studied to ascertain if, as Goldstein suggests, the sex, and
thus role-model status, of the faculty member has an effect on outcome,
and if so, what other factors enhance this interaction.
Within a professional school, female students in the first two years
of medical school reported that they needed more role models, while
those in the second two years reported that they no longer needed role
27. J. H. Block, "Gender Differences and Implications for Educational Policy" (Uni-
versity of California, Berkeley, Institute of Human Development, 1980).
28. E. Goldstein, "Effect of Same-Sex and Cross-Sex Role Models on the Subsequent
Academic Productivity of Scholars," American Psychologist 34, no. 5 (1979): 407-10.

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
700 Speizer Review: Role Models

models.29 As the students studied were in four different years at the

same medical school and as only a small percentage of first-year female
students responded to the questionnaire, one cannot know if the stu-
dents in each year were just very different from the students in other
years or if changes took place within a student or within the school to
cause a related change in the reported need for a role model. A study
that follows the same students for four years is needed to provide in-
formation about whether the need for role models declines by the senior
year in medical school.
Bucher and Stelling followed for a year a group of psychiatry, medi-
cal, and biochemistry trainees and asked them to identify their role
models throughout the year.30 They found that the trainees constructed
a model for themselves based on their observations of positive and nega-
tive aspects of those who trained them. The idea that one constructs
her/his own role model based not on one person but on a variety of
people is certainly intriguing. However, the findings must be taken
cautiously because the authors did not present supporting data for their
conclusions; as a result we can identify neither the sex and race of the
trainees nor the sex and race of the role model from which the "parts"
were taken. Moreover, since the trainees were studied only during
career preparation, one does not know what effect the role-model parts
had on their actual professional careers.

Single Sex Colleges

Colleges designed specifically for women students have been con-

sidered by many educators the best learning environment for female
students. Graduates from single sex colleges are assumed to have had
opportunities for leadership, to have encountered many female role
models, and to be motivated to pursue further education and a career.
Comparing the environment of a single sex college with that of a coed
institution has proven difficult, given the wide range of differing struc-
tural factors within and among each type of college and the variation in
family background among students. Perhaps most difficult has been
finding matched samples in a single sex and a coed college, because
students choose different colleges for varied reasons.
Tidball and Kistiakowsky studied the baccalaureate origins of
women who obtained a doctorate in the period between 1920 and
1973.31 They found that the graduates of the seven elite women's col-

29. N. A. Roeske and K. Lake, "Role Models for Women Medical Students,"Journal of
Medical Education 52, no. 6 (1977): 459-68.
30. R. Bucher and J. G. Stelling, Becoming Professional, Sage Library of Social Re-
search, vol. 46 (Beverly Hills, Calif.: Sage Publications, 1977).
31. M. E. Tidball and V. Kistiakowsky, "Baccalaureate Origins of American Scientists
and Scholars," Science 193 (August 1976): 646-52.

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
Signs Summer 1981 701

leges and two coeducational institutions (Cornell and the Univ

Chicago) went on for doctoral degrees in far greater proporti
women who received their bachelor's degrees from other e
institutions. As the college of choice for many of the outstand
students during the time under study was one of the nine coll
to produce the largest number of female doctorates, one
termine if the selection process or the learning environme
greatest effect. The study needs to be replicated for the
women have begun to graduate from the elite male colle
graduates from women's colleges continue to obtain doctorate
numbers, then one can surmise that the environment of an all
college is an important factor for producing women scholars.
In another study, Tidball examined a group of women
who appeared in Who's Who in American Women, 1966-1971 to
what effect could be ascribed to the type of undergraduate c
attended.32 She found that women who graduated from w
leges appeared in far greater numbers than women who grad
coed institutions. Graduates of selective and less selective women's col-
leges were at least twice as likely to become achievers as women from
coed institutions, matched for selectivity. Tidball compared the num
of women faculty to the number of women students and found that
the ratio of women faculty to women students increased, so did
number of women achievers. She postulates that women faculty mem
bers serve as role models for women students and are thus a critical
ingredient in a college environment which seeks to produce tal
women.

Oates and Williamson sought to explore further the r

between the type of college education received and subse
decisions and success for women students.33 They studied w
appeared in Who's Who in America, 1974-1975, rather than t
the index used by Tidball, as they believed an index th
women and men would be more selective and would better reflect "suc-
cess" in American society. Oates and Williamson discovered that t
greatest achievement was demonstrated by the graduates of the e
"Seven Sisters" colleges rather than by graduates of all women's colleg
as Tidball reported: A result that they related to the vigorous selectio
process in such colleges rather than to the experience of students
college. They suggest that socioeconomic status, as well as admiss
standards, can also explain the differential outcomes for students fro
the elite single sex colleges. Oates and Williamson found no support f

32. M. E. Tidball, "Perspective on Academic Women and Affirmative Action," Educ

tional Record 54 (1973): 130-35; M. E. Tidball, "The Search for Talented Women," Ch
6, no. 4 (May 1974): 51, 52, 64.
33. M. J. Oates and S. Williamson, "Women's Colleges and Women Achievers," Sig
Journal of Women in Culture and Society 3, no. 4 (Summer 1978): 795-806.

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
702 Speizer Review: Role Models

the argument that female faculty role models in a college environment

promote nontraditional career aspirations for women students. Yet, be-
cause Tidball's study differs greatly from Oates and Williamson's in
methodology (see table 1), each study must be examined separately, and
the two findings should not be considered comparable.
A direct relationship between the effects of attending a women's
college and the presence of female role models within women's colleges
has yet to be conclusively demonstrated. Certainly the overall statistics
about the number of women faculty members and administrators at
women's colleges support the theory that more women are to be found
to serve as role models in women's colleges than are present in coed
institutions. In 1976 women made up 55 percent of the faculty in wom-
en's colleges (vs. 25 percent at coed institutions); 42 percent of all ten-
ured professors (vs. 10-12 percent tenured professors); 71 percent of
the academic deans (vs. 15 percent); and 60 percent of the presidents (vs.
6 percent).34 The presence in such large numbers of women faculty
members and administrators to serve as potential role models has been
used to explain the finding that in women's colleges two to three times
the national average of women major in math, chemistry, and biology.35
However, the number of women faculty members and administrators is
not evenly distributed among the women's colleges, and the number of
women faculty role models has not been shown to correlate directly with
the number of women achievers. For example, Radcliffe, one of the elite
Seven Sisters colleges that produced a large number of women achiev-
ers, never had a faculty of its own. Radcliffe students were taught by
Harvard professors either in individual classes made up entirely of
women students or in classes with Harvard male students. Radcliffe
students had very few female faculty role models, and yet they wen
in great numbers to obtain doctorates and to become achievers.
After a review of the literature on role models, one must agree w
Elizabeth Douvan when she notes: "We must take it on faith that know-
ing a woman academician and being close enough to see something of
the reality of her life and action will help the young intellectual women
to concretize her own role conception, invest her aspirations with greater
reality and perhaps offer her some useful clues about ordering her
spheres of action. It seems plausible. But we must take it on faith because
identification and modeling have [not] been studied systematically ex-
cept in pre-school children."36
The systematic proof which Douvan sought is still missing. There
have been no studies designed specifically to test the hypothesis that role

34. M. K. Sharp, "Women's Colleges: Equity and Optimum," College Board Review 111
(Spring 1979): 18-20.
35. Ibid.
36. E. Douvan, "The Role of Models in Women's Professional Development," Psych
ogy of Women Quarterly 1, no. 1 (Fall 1976): 5-20.

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
 Table I

Major Role-Model Studies

Author/Year Source Number Sex Selection Criteria Resea

Tangri 1972 ..... College seniors in 1967 200 F Selection by occupa- Retrospective q
at large, public, coed, tional choice: one-third naire in
midwestern university role innovators, one-
third moderates, one-
third traditionals

Tidball 1973, 1974 1,100 be- F Random selection of

College graduates Outcome
tween 1910 and 1940 college graduates
compar
listed in Who's Who in 1'. ty
American Women, tend
1966-1971 2. num
fac
of
ates

Almquist &
Angrist 1971 ..... Undergraduates 110 F From 188 studied in Retros
(1964-68) at women's freshman year,
naire9i
colleges within a univer- plus 15 sophomor
sity transfers
Angrist & Alm-
quist 1975 ....... Undergraduates 87 F From 188 studied in Retros
(1964-68) at women's freshman year,
naire 8
i
college within a univer- who completed a
sity tionnaires
Seater & Ridgeway
1976 ............ Undergraduates (202); 269 F (112) Not specified Question
graduates (67) at large, M (157)
urban midwestern uni-
versity

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
 Table 1 (Continued)

Author/Year Source Number Sex Selection Criteria Resea

Tidball & Kistia-
kowsky 1976 ..... Baccalaureate recipients
F 352,423
All baccalaureate recipi- Analys
1910-69 who attained (46,805) ents of each sex for 137 calaureat
doctorates 1920-73 M institutions; 25 institu- recipient
(305,618) tions graduating largest
number of women and
men who went on for
doctorates
Bucher & Stelling
1977 ............ Postgraduate, profes- 48 Unspeci- Of 62 trainees in the Three uns
sional trainees in 1965 fled four programs in 1965 depth int
at four different institu- the 48 who had com-
tions pleted three interviews
were selected
Roeske & Lake
1977 ............ Medical students in 145 F Of the 192 women Cross-sec
1976 at midwestern naire145
medical students,
medical school (76%) returned the
questionnaire
Sternglanz & Ly-
berger-FicekUndergraduate
1977 students 2,285 F (870) Random selection of 60 Observat
(year unspecified) at M (1,415) classes with enrollment sion of 6
large, public, coed, east- Female of 50-100 students each teacher-s
ern university teachers tions
(11)
Male
teachers
(49)

Sternglanz Undergraduates
1979 .. (year Unspeci- Unspeci- Random selection of 10 Two obs
unspecified) at large, fied and fied and nonscience classes with at start
public, coed eastern 10 teach- female enrollment of 25-50 10 classes
university ers teachers students each teacher-s
(5) tions
Male
teachers
(5)

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
 Table 1 (Continued)

Author/Year Source Number Sex Selection Criteria Res

Oates & William-
son 1978 ......... F All1,735
Baccalaureate recipients women listed who Outc
1920-59 listed in Who's were college gradu
com
Who in America, 1.
1974-1975, 38th ed.
2.

O'Donnell &
Anderson 1978 ... Undergraduates at 20 F 10 students selected Stru
large, public, coed, from traditional m
western university and 10 from nontradi-
tional major

sI Stake & Granger

Z 1978 ............
~14 Science students (year Stude
nts F (141) 20 science teachers (10 Outc
unspecified) from Mo. (270) M (129) female and 10 male) re- com
high schools Teach ers Sex of quested to administer for:
(14) teacher and return attitude sur- 1.
unspeci- vey to students; 14
fied (70%) administered
survey 2.

Goldstein 1979 ... Doctoral recipients in110 F (55) Random selection from Publ
psychology 1965-73 M (55) four groups: female doc
from psychology de- Ph.D.'s with female fou
partments in three N.Y. advisors (25); female tion
universities (nine areas Ph.D.'s with male advi-
of specialization) sors (30); male Ph.D.'s
with male advisors (29);
male Ph.D.'s with female
advisors (26)

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
 Table 1 (Continued)

Author/Year Source Number Sex Selection Criteria Resea

Basow & Howe
1980:
Experiment I ... Undergraduates in two 62 F (31) Students in two classes Students
psychology courses M (31) who completed ques- three g
tionnaires and were on attitu
younger than 23 women
females
analyze
Ox
respons
of role

Experiment II .. 300 college seniors ran- 60 F (30) From the 96 (30%) se- Analysi
domly selected from M (30) niors who completed sex-role
master list were mailed questionnaires, 30 fe- on attitu
two questionnaires males and 30 males women
were picked by attitudemodel (
score and subdivided of role
into three equal groups
of 10 females and males
in each

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
Signs Summer 1981 707

models exist. It is not surprising, therefore, that findin

tradictory and variable. Between and sometimes within studi
often inconsistencies that result from such factors as: me
flaws in the design of the study;37 ascription of causal reasons
without consideration of all the possible confounding factor
findings;38 or presentation of findings without the needed s
data.39
The popularity of the concepts of role models and mentors has, in
many cases, provided respondents with a plausible explanation for their
present situation. For example, 237 women microbiologists surveyed by
Kashket et al.40 reported that, although they worked as long and hard as
the men for the same reasons and remained in their positions as long as
the men, they felt handicapped by a lack of encouragement and of
proper role models. The women were probably right in their judgment
that they had been handicapped, but one wonders if the lack of role
models is a satisfactory explanation of the situation. In another study,
faculty members reported themselves as more satisfied with their jobs if
their research administrators were, among other things, good role mod-
els.41 However, productivity, the factor that determines promotion and
tenure, was not related to feelings of satisfaction or the possession of a
good role model. A further example of the way in which accepted ideas
on the subject may influence reaction to it is a study conducted in a
northeastern elite college where the junior faculty were asked if they had
mentors.42 The women junior faculty reported that they did not have a
mentor and that they doubted whether they would ever have one; how-
ever, they were certain that the male junior faculty had mentors and that
as a result it would be easier for the men to get tenure. The men re-
ported few thoughts about mentors and did not appear to believe that
their chances of success were tied to the support of a mentor. Thus, it
would appear the women were ascribing a great deal of importance to
mentors, while in their striving for tenure the men were turning their
attention elsewhere.
Thus far the study of role models has not been successful in an-
swering many of the basic questions about the concept. Most of the
studies have been focused on white female students; none has examined
minority students, and those studies that include male students often
37. Seater and Ridgeway (see n. 5).
38. Tidball, "The Search for Talented Women."
39. Angrist and Almquist (see n. 12); Bucher and Stelling.
40. E. R. Kashket et al., "Status of Women Microbiologists," Science 183 (1974):
488-94.
41. S. Coltrin and W. F. Glueck, "The Effect of Leadership Roles on the Satisfac
and Productivity of University Research Professors," Academy of Management Journal 20
1 (1977): 101-16.
42. M. H. Bonz, "Assessment of Junior Faculty Experience at Dartmouth Colleg
(research report prepared for the president, Dartmouth College, April 1978).

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
708 Speizer Review: Role Models

appear to do so primarily to make a comparison with the female stu-

dents. Further, the methodological flaws of such research are apparent:
(1) the use of small, highly selected, atypical samples often without ap-
propriate controls; (2) the use of cross-sectional information (i.e.,
freshmen and seniors) as though the data were collected longitudinally
from the same group; and (3) the retrospective bias inherent in trying to
assess what caused measured differences. Finally, studies have been
piecemeal rather than systematic, anecdotal rather than observational.
In 1970 Bell suggested that two different processes were involved in
role-model relationships: interaction and identification.43 Few re-
searchers have considered Bell's approach, and thus we do not know if a
role model must interact with a student to be useful or if the sheer
presence of a person is sufficient for positive or negative identification.
Perhaps role models with whom one interacts are different people and
serve different functions than role models with whom one identifies
from a distance. Also, since most studies of role models have been con-
ducted with college students and faculty members as respondents, there
is little information about whether a faculty person can be a proper
model for a career outside academia. Before researchers and the public
alike can accept the concept of role models as a viable consideration to be
taken into account by aspiring professionals, more studies which define
and systematically examine the theory will need to be conducted.

Mentors and Sponsors

Role models have been studied in their effect upon college students,
while mentor and sponsor research has focused on people in the work
arena. In fact, according to the recent business literature, one must find
a mentor or sponsor (perhaps two or three) if one wants to make more
money, have higher satisfaction with career progress, be younger when
one reaches the top, and be a leader.44 The terms "mentor" and "spon-
sor" are often used interchangeably to indicate older people in an or-
ganization or profession who take younger colleagues under their wings
and encourage and support their career progress until they reach mid-
life. The term "sponsor" was in vogue in the 1960s and into 1970,45 and
then appears to have dropped from use or become an alternate term for
43. A. P. Bell, "Role Modelship and Interaction in Adolescence and Young Adult-
hood," Developmental Psychology 2, no. 1 (1970): 123-28.
44. G. R. Roche, "Probing Opinions," Harvard Business Review 57, no. 1 (1979): 14-28;
A. Zaleznik, "Managers and Leaders: Are They Different?" Harvard Business Review 55, no.
3 (1977): 67-78.
45. C. F. Epstein, "Encountering the Male Establishment: Sex-Status Limits on Wom-
en's Careers in Professions," American Journal of Sociology 75 (1970): 965-82; M. S. White,
"Psychological and Social Barriers to Women in Science," Science 170 (October 1970):
413-16.

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
Signs Summer 1981 709

the newly popular "mentor." Thus far, most of the data ab

have been collected by way of surveys or interviews with p
who look back over their careers and identify key peop
served as mentors for them.46 Those people who could n
mentor or sponsor were reported as being less successful an
in their careers. The findings related to male business execu
been the foundation upon which to recommend to wom
acquire mentors.47
Kanter, in her study of a complex business organization,
sponsor or mentor as a person who provided power in thre
"comers" who were on the "fast track": (1) by being in a pos
for the person in question; (2) by helping the person bypas
chy; and (3) by providing reflected power.48 She found mo
and mentees to be male, because business organizations a
nantly male, and she postulated that mentors prefer mente
similar to them. Since sponsorship provides power and sinc
often excluded from such relationships in the informal sys
recommends that an organization develop an "artificial
program" to help women and minorities who traditionally
it." Kanter does not specify how many people in her study
mentors, neither does she show how she reached the conclusion that
mentors are important for men and thus necessary for women.
The recent popularity and acceptance of the concept of mentor has
been primarily based on the work of Levinson who studied the lives of
forty men between the ages of thirty-five and forty-five who were equally
distributed among four occupations-hourly workers in industry, busi-
ness executives, university biologists, and novelists-to ascertain the
adult developmental stages through which men pass.49 The sampling
procedure varied somewhat among the four occupations, but in each
category the men were chosen for their diversity on the assumption that
if a common theme was found within groups or across group lines, it
might be a universal truth for all men (see table 2).
Levinson identifies the need for a mentor as common to all men.
However, since he never specifies how many of the forty men he stud
actually had mentors, the reader cannot assess the universality of th
mentor's role. Levinson also points out that "Many adults give a
receive little mentoring. Despite the frequent emphasis on teamw

46. Collins and Scott (see n. 1); Roche; G. W. Dalton, P. H. Thompson, and R.
Price, "Professional Careers-a New Look at Performance by Professionals," Organizatio
Dynamics 6, no. 1 (1977): 19-42.
47. C. D. Orth and F. Jacobs, "Women in Management: Pattern for Change," Harvar
Business Review 49, no. 4 (1971): 139-47.
48. Kanter, pp. 181-82 (see n. 15).
49. D.J. Levinson et al., The Seasons of a Man's Life (New York: Alfred A. Knopf, In
1978).

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
 Table 2

Major Mentor/Sponsor Studies

Author/Year Source Number Sex Selection Criteria Rese
Dalton, Thompson,
& Price 1977 ..... Professionally trained
550 Unspeci- High- and low-rated One retros
adults in 13 different fled performers (rating scale view
institutions unspecified)

Levinson et al.
1978 ............ 40 Min
10 men aged 35-45 Men who fit within the 5-10 r
1969 in each of four category of worker,
depth
categories from four ecutive, academic bio
each m
companies, two univer- gist, or novelist mont
sities, and specific writ- up in
ing categories later in
Collins & Scott
1979 ............ Upper-level executives 3 M Chief executives of the Retr
at Jewel Tea Co. same company

Roche 1979 ...... Senior executives whose 1,250 F (less All 3,976 men and 28 Retros
appointments were than 1%) women senior execu-
listed in "Who's News" M (99%) tives listed were sent
column of Wall Street questionnaires (a 31%
Journal in 1977 response rate)

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
Signs Summer 1981 711

and loyalty in business organizations, mentoring relationships

the exception than the rule for both workers and managers.
of higher education, though officially committed to fosterin
tellectual and personal development of students, provides
that is generally limited in quantity and poor in quality.
middle-aged men never experience the satisfactions of me
This is a waste of talent, a loss to the individuals involved, an
pediment to constructive social change."50 One must assume f
above quote that Levinson did not find many men in his stud
had a mentoring relationship. Yet Levinson describes the stag
relationship in vivid detail, and it is his description that has
imagination of researchers and the public alike. Accepting
definition of a mentor and his recommendation that all men need a
mentor, researchers have set out to discover how many men have h
mentors and if, in fact, having a mentor promotes career success. S
tematic studies that explore the definition of a mentor and examine
function such a person might perform have yet to be undertaken. T
there is little foundation for the generally accepted belief that all w
and men need a mentor.
One example of how Levinson's concept of mentor has been us
or misused, was a study by Roche.51 Basing his work on Levins
premise that mentors are important during an adult's developm
Roche conducted a mail survey of senior executives who were ne
appointed to their positions in 1977. The executives were asked whet
they had had a mentor or sponsor at any time during their care
One-third of the executives responded to the questionnaire; of th
two-thirds reported that they had had a mentor or sponsor, and one
third of them had had two or more mentors. Most of the responden
did not consider their mentors an important factor in their own succ
There is no way to reconcile Levinson's insistence on the importance
mentors, though few of the men he studied had mentors, with Roch
finding that most men have mentors yet think them unimportant.
As with role-model research, studies of mentors need to be
methodologically more sound. Existing studies are often flawed by
methodological problems. (1) The numbers are too small to allow one to
generalize from the findings. (2) The information collected is ret-
rospective. (3) The concepts of mentor or sponsor are left undefined.
The interest in mentors has been primarily in the business community
where a mentor or sponsor is thought to be an older, successful, male
executive. No studies have explored mentor relationships for other
groups, nor have any ongoing relationships been followed to determine
what accrues to each person who serves as a mentor or mentee. Clearly,

50. Levinson, p. 334.

51. Roche.

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms
712 Speizer Review: Role Models

there needs to be more research if the hypothesized link between a

mentor and professional success is to be documented.

Conclusion

Role models, mentors, and sponsors are concepts which still need to
be defined and studied. Despite their almost universal acceptance, there
is very little supportive evidence for their validity. Until methodologi-
cally sound studies are conducted on large, randomly selected popula-
tions, these concepts should be considered as suggestive rather than
proven.
The first step which researchers must take is to establish accepted
definitions for each concept. Shapiro et al.52 recognized the need for
clearer definitions when they called role models and mentors code words
and panaceas for downtrodden women professionals. Their solution was
to redefine the terms and place them on a continuum that they called
"patron relationships." Unfortunately, the authors did not succeed in
clarifying the "code words"; what they did was to add a new layer of
definitions and terms over an already miry conceptual foundation. An
interdisciplinary approach to the study of role models and mentors will
probably provide the best ground plan, but-to change the metaphor-
scholars need first to search for the roots of these concepts in their own
fields. They must then establish connections between their work on role
models or mentors and other areas of their discipline. Once universally
accepted definitions have been established by scholars within their disci-
pline and perhaps among disciplines, research with different ap-
proaches can be pursued.
It is important to ask why, with so little research foundation, the
concepts of role models, mentors, and sponsors have caught the imagi-
nations of so many people. Perhaps for those who are striving to succeed
in an environment where they are "other," the ingredient needed to
alleviate the pressures of loneliness and tokenism is the presence of a
sufficient number of people like themselves. However, increasing the
number and diversity of women and minorities in work and educational
settings might ease the burden on tokens and allow people to pursue
their goals by trusting in their own competence, unconcerned about
whether they have a role model, mentor, or sponsor.

Higher Education Resource Services

Wellesley College

52. E. C. Shapiro, F. P. Haseltine, and M. P. Rowe, "Moving Up: Role Models,

Mentors, and the 'Patron-System,' " Sloan Management Review 19, no. 3 (1978): 51-58.

This content downloaded from

143.106.202.107 on Tue, 22 Sep 2020 13:35:05 UTC
All use subject to https://about.jstor.org/terms