THE COMMUNICATION PATTERNS AND
THE STRUCTURE OF SOCIAL RELATIONSHIPS
AT A LARGE UNIVERSITY

Thesis for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY
R. LANCE SHOTLAND
1970

..__._ "an”?

LIBRARY
Michigan State
University I";

v
7" V'— p—y

l.‘

2"

 

This is to certify that the
thesis entitled

The Communication Patterns and the Structure
of Social Relationships at a Large University

presented by

R. Lance Shot land

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Psychology

 

 

‘ . H
BINDING av V

IIIIAB & SIJIIS'

BIIIIK BRIBERY INC.

.L‘IPRARY BINDERS

 

 

‘ IIIIIIIIIIIIIIIIIIIIIIIIIIILIIIIIIII9IIII9IIII9IIII9IIIII

3 1293

 

 

 

 

ABSTRACT

THE COMMUNICATION PATTERNS AND THE
STRUCTURE OF SOCIAL RELATIONSHIPS
AT A LARGE UNIVERSITY

by R. Lance Shotland

Within the literature produced by several student
movements some very specific complaints pertainhng to the
social structure of the university appear. Two student
movements on two different campuses were viewed with re-
gard to complaints about the social structure of the uni-
versity. The activist students complained that they were
socially separated from the faculty, from the administra-
tors and from other students.

It was hypothesized that students would be connected
to other students, faculty members and administrators by
the longest informal communication channels. On the basis of
Leavitt's (1958) study, it was also hypothesized that
administrators would have the shortest informal communi-
cation channels to other‘administrators, faculty and
students.

The technique used in the present study to measure
the length of informal communication channels was first
used by Milgram (1967). Milgram called the technique the
“Small World Method.” Using the Small World Method, two
sets of individuals are selected. One set of individuals

is designated the starter persons, a second set of

If“
.Ir

 

R. Lance Shotland

individuals is designated the target persons. A starter
person is asked to try to pass an instructional booklet

to the target person by only passing the booklet to peOple
they know according to a certain criterion (e.g., knowing
the person on a first name basis, etc.). If the starter
person does not know the target person according to the
criterion the starter person is then instructed to pass the
booklet to an acquaintance he does know according to the
criterion, who has a better chance of being acquainted with
the target person. The number and characteristics of the
intermediary persons between the starter and target serve
as the dependent variables.

Student, faculty and administrators were randomly
selected to serve as starter and target persons from the
'population of a large university. Each starter person was
asked to start two booklets to student targets, two book-
lets to faculty targets and two booklets to administrator
targets. Each target person was asked to receive a possi-
ble two booklets from student starters, two booklets from
faculty starters and two booklets from administrator
starters. The starter and target persons were randomly
paired.

The results confirmed the hypotheses. Students had
the longest informal communication channels while the ad-
ministrators had the shortest communication channels. Thus,
in Leavitt's terminology administrators may be said to be

the most centra1_group while the students are the most

 

 

 

 

 

hi. .II a

.F4.4‘-I -

R. Lance Shotland

peripheral group within the university. The results were
discussed in terms of the peripherality of the students and
their contentment with the social structure of the university.
Suggestions were made for the modification of the social

structure of the university.

Approvm? $9 I‘:

Committee Co-Chairman

g Conunittee Co-C Waiman

Date: 30% t0\’\0
Q \

 

 

Thesis Committee:
Robert H. Davis, Co-Chairman
John E. Hunter, Co-Chairman
Charles Hanley I
Lawrence A. Messe

THE COMMUNICATION PATTERNS AND THE
STRUCTURE OF SOCIAL RELATIONSHIPS

AT A LARGE UNIVERSITY

‘k By

4
50%
R. Lance Shotland

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Psychology

l970

..»—

Copyright by

ROBERT LANCE SHOTLAND

1970

Our youth today have luxury. They have bad man-
ners, contempt for authority, disrespect for older peeple.
Children nowadays are tyrants. They contradict their
parents, gobble their food and tyrannize their teachers.

Socrates

There is a time when the operations of the machines
become so odious, make you so sick at heart, that you can't
take part, you can't even tacitly take part. And you've
got to put your bodies upon the wheels, upon the levers,
upon all the apparatus, and you've got to make it stop.

And you've got to indicate to the people who run it, to
the peOple who own it, that unless you're free the machine
will be prevented from working at all.

Mario Savio

DEDICATION

To Joseph Katz and Hans Toch - through their classes

I became interested in the relationship between peOple.

iv

 

 

R II"

ACKNOWLEDGMENTS

I would like to thank Dr. Robert H. Davis and the
Learning Service of Michigan State University for support-
ing this research. Without this early sign of confidence
this research would never have been accomplished. I
would like to thank Dr. John E. Hunter for providing me
with invaluable assistance both with computer programming
and the Markov model within this dissertation. I would
also like to thank Doctors Charles Hanley and Lawrence
Messé, who with Doctors Davis and Hunter, provided me with
many thoughtful suggestions which improved the quality of

this work.

PREFACE

The research to be presented developed from some very
specific complaints by activist students on two different
college campuses between the years of 1964-1966. These com-
plaints concern the informal communication networks and a
feeling of a lack of community within a large university.

It is precisely these complaints that will be explored with-
in this dissertation. The author, however, takes a much
broader scape in reviewing two student movements in the
following section. The complaints of a lack of community
within the academic community are placed in the context of

a larger set of complaints in which they occurred. It is
hoped that the presentation of these specific complaints
within the larger context of the background of the student
movements will enable the reader to view these complaints
with perspective.

The two student movements selected for review were
chosen for several reasons. One, both student movements
occurred within several months of each other. Two, both
student movements occurred on campuses with relatively large
student bodies. Three, the Free Speech Movement is the best
known student movement and thus many of the pamphlets pro-
duced by the movement have been published. Four, the study

to be described was carried out at Michigan State University,

vi

the campus where the Committee for Student Rights took
place.

It should also be mentioned that the results of this
study, which support student complaints about the informal
communication channels and social structure, is not the en-
tire cause of student dissent. Other factors are involved
and interact with the student's place in the social struc-
ture which may result in student activism. This point is

explored within the discussion section.

vii

TABLE OF CONTENTS

DEDICATION O O O O O O O O O O O O O O O O I O 0

ACKNOWLEDGMENTS O O O O O O O O O O O O O O O O O

PREFACE . .

LIST OF TABIES O O O O O O O O O O O O O O O O O

LI ST OF FIGURES O O O O O O O O O O O O O O O O 0

LIST OF APPENDICES I O O O C O C O C O O C O O 0

Chapter

I INTRODUCTION 0 O O O O O O O O O O O O O

The Free Speech Movement . . . . . .
The Committee for Students Rights. . .
Summary of Student Movement Complaints
The Diffusion of an Idea . . . . . . .
The University of a Stratified

Social System. . . . . . . . . . . .
The Advantages and Disadvantages to

Viewing the University as a

Stratified Social System . . . . . .
Objectives of the Current Research . .

II METHOD 0 O O O O O O O O O O O O O O O O

The Technique. .
The Instrument .
Subjects . . . .
The Design . . .
The Setting. . .

III THERESULTS...............

The Validity of the Small World Method
The Social Distance Between Faculty,
Students and Administrators. . . . .

viii

Page

. xii

.xiii

. 12
. 19
. 19

35
41
45
47
48

Chapt

III

 

 

BI

AP

 

 

TABLE OF CONTENTS (Continued)

Chapter Page

III (Continued)

The Relative Size of the Groups and

Its Effect on Chain Lengths . . . . . . . . 63
The Actual Chain Lengths, An Estimate

of the Parameters . . . . . . . . . . . . . 65
The Permeability of Faculty, Student

and Administrator Role Groups . . . . . . . 74
A Faculty Starter to a Faculty Target . . . . 76
A Student Starter to a Faculty Target . . . . 79
Administrator Starter to a Faculty

Target. . . . . . . . . . . . . . . . . . 81
A Faculty Starter to a Student Target . . . . 84
Student Starter to a Student Target . . . . . 85
Administrator Starter to Student Target . . . 88
Faculty Starter to Administrator Target . . . 90
Student Starter to Administrator Target . . . 91
Administrator Starter to an Administrator

Target. . . . . . . . . . . . . . . . . . . 92

IV DISCUSSION. . . . . . . . . . . . . . . . . . . 93

A Partial Summary of the Results. . . . . . . 93
What Do Chain Lengths Indicate? . . . . . . . 95
The Social Structure and the

Diffusion of an Idea. . . . . . . . . . .100
Centrality and the Degree of Contentment

With the Social Structure . . . . . . . . .102
Modifications of the Social Structure

to Produce Shorter Chain Lengths. . . . . .104
Summary . . . . . . . . . . . . . . . . . . .106

BIBLIOGRAPHY. O O O O C O O O O O O O O O O O O O O O O 109

APPENDICES C O O O O O C I C O O O O O O O I O O O O C O 113

ix

Table

LIST OF TABLES

Page
The Proportion of the Instructional
Booklets started, the Proportion That
Were Completed, and the PrOportion Started
That Were Completed for All Nine Starter-
Target Combinations . . . . . . . . . . . . . 52

The PrOportion of a Given Status Group
That Cooperated With the Study. . . . . . . . 53

The Frequency of Residing With a Peer
Across Undergraduate Class Levels . . . . . . 55

The Conditional Probabilities That an
Undergraduate From A Particular Class or

Status Will Send the Instructional Booklet

to Another Undergraduate of a Particular

Class or Status Across A11 Nine Starter-

Target Combinations . . . . . . . . . . . . . 56

The Means and Standard Deviations of the

Number of Intermediaries Required to Link

the Starter to the Target for Each of

the Nine Starter-Target Combinations. . . . 59

The Expected Chain Lengths [E(CL1)]

Without Compensating for Missing Chains

in Comparison to the Observed Chain

Lengths . . . . . . . . . . . . . . . . . . . 72

The Mean Number of Intermediaries Required

to Link the Starter to the Target for

Each of the Nine Starter-Target Combina-

tions Assuming A11 Chains Reached Their

Targets . . . . . . . . . . . . . . . . . . . 73

The Conditional Probabilities That a

Faculty Member of a Particular Rank or

Status Will Send the Instructional

Booklet to Another Faculty Member of a

Particular Rank or Status . . . . . . . . . . 78

Table

10

ILl

.12

113

14

LIST OF TABLES (Continued)

The Relationship Between Faculty
Initiated Contact With Administrators
and Faculty Rank . . . . . . . . . . . . .

Student Class Level and the
Initiation of Contact With
The Faculty. . . . . . . . . . . . . . . .

The Relationship Between Administrator
Initiated Contact With the Faculty and
Facu1ty Rank 0 O O O C C O O O I O O O O O

The Relationship Between Faculty
Initiated Contact With Undergraduates
and Faculty Rank . . . . . . . . . . . . .

The Conditional Probabilities of Passing
the Instructional Booklet to a Member of
the same sex 0 O O O O C O O O O O O O O O

The Relationship Between Student Class
Level and Administrator Initiated Contact.

xi

Page

79

81

82

85

88

90

Fig}

fi|111-;tlii 1

 

LIST OF FIGURES
Figure Page
1 The University Viewed as a

Stratified Social System . . . . . . . . . . 31

2 The Experimental Design. . . . . . . . . . . 49

xii

Appendix

LIST OF APPENDICES

The Instrument . . . . . . . . . . . . . . . 113
The Conditional Probability Tables

Representing the Nine Different
Starter-Target COIllbinatiOnS. o c o o o o o c 119

xiii

CHAPTER I

INTRODUCTION

[A university is often called an ”academic community,"
implying that students, faculty, and administrators share
common goals and objectives. During the middle 1960's stu-
dent disturbances began to upset the normal functioning of
the larger, more prestigeous institutions, showing a clash
of objectives and goals, and thus calling the concept of an
"academic community" into question;

Eeterson (1966) performed a national survey to deter-
mine what issues were related to student unrest. His data
can be interpreted to indicate that 62% of the disturbances
were in response to objections of the students' role in the
university.{j¢Within the literature produced by several
student movements some very specific complaints pertaining
to the social structure of the university appear. The fol-
lowing two sections will concern themselves with the events

on two campuses.

 

¥The 62% is composed of 28% of protests over living group

arrangements, 18% over student participation in campus
policy making, 9% over rules regarding “controversial" ,\
visitors to campus, and 7% over curriculum inflexibility;

One campus, The University of California at Berkeley,
has a history of political activism among its student pOpu-
lation and had the first student protest of the 1960's that
gained national publicity. This protest, perhaps more aptly
referred to as a revolt, was named the Free Speech Movement
(FSM). The second set of events took place at Michigan State
University, a campus with little history of political acti-
vism among its student population. A series of incidents
will be discussed.leading up to The Committee for Student
Rights (CSR), Michigan State University's first student

movement of any size or consequence.
The Free Speech Movement

The revolt at The University of California at Berke-
ley which began in September of 1964 is possibly one of the
best known student disturbances and perhaps the disturbance
with the widest consequences. Berkeley has been described
as the model for future public education in the United
States; the revolt brought it to the edge of collapse and
called the basic premise of the modern university into
question.

The FSM pamphlet writers concentrated on two issues
during the course of the movement, i.e., "Free Speech" from
which the movement took its name and the.relationship of
the student to the rest of the university.

The t0pic.of Free Speech as an issue of protest was

not a new one at Berkeley. In October of 1934 several

students at the University of California at Los Angeles
were suspended for supposed Communist activities. In addi-
tion, the editor of the student newspaper at the University
of Santa Clara was replaced as a result.of a story condemn-
ing the R.O.T.C. program. A one-hour student strike at
Berkeley was-scheduled.and held in protest over the denial
of free expression?“ In the late 40‘s with the close of the
war and the passage of the G.I. Bill there was a preponder-
ance of veterans on.campus whose attitudes seemed to dampen
any attempts at political activity. The early.l950's were
not any more eventful. Starting in the late 1950's politi-
cal protest once more arose.2
Robert G. Sproul, a prior president of the University
of California sought to control the influx of public speakers
on campus. It was under his administration.that Rule 17 was
passed. Rule 17 stated in effect that it was the perogative
of the university (the administration) and the university
alone to decide what speakers would be permitted on campus.
In 1957 Sproul, under.pressure from the political reawaken-
ing of the students relaxed Rule 17. Under this revision,
unrecognized off-campus groups composed entirely of Universi-
ty of California students were permitted to use campus
facilities if the events were judged of interest to the

total student body-, In 1958 and 1959 there were public

2M. Heirich & S. Kaplan, "Yesterday's Discord," in S. Lipsit
& S. Wolin (Eds.) The Berkeley Student gevolt: Facts and
Interpretations.(Garden City, N.Y.: 1965) pp. 10-37.

 

 

expressions of disapproval (a Daily Californian editorial

 

and a student government statement) over what was the re-
mainder of Rule 17. In addition to this there were objec-
tions to the relocation of the Sather Gate Free Speech area
to a new location at Telegraph and Bancroft Streets. The
administration, at this time, was attempting to liberalize
the rules in addition to searching for a formula that would
make it clear that student and faculty activities off campus
did not speak for the university.3

On October 23, 1960 Clark Kerr, the president of the
University of California, issued what was later to be called
the Kerr directives. These rules stated that:

(l) The preamble of the student.government consti-
tution on each campus shall be changed to make
it clear that the student governments are direct-
ly responsible to the apprOpriate chancellor's
office,

(2) student governments are forbidden to speak on
off-campus issues,

(3) amendments to student government are subject to
the prior approval of campus officials,

(4) to be recognized, student organizations must
have an active advisor who is a faculty member
or a senior staff member; such groups must de-
clare their purposes to be compatible with the
educational objectives of the university; they
must not be affiliated with any partisan politi-
cal or religious group; and they must not have
as one of their purposes the advocacy of posi-
tions on off-campus issues.

 

3Heirich & Kaplan, p. 20.

4Heirich & Kaplan, p. 21.

Kerr, at the same time, modified Rule 17 so that it
was no longer exclusively a "university responsibility" to
rengage qualified speakers on important societal issues but
in addition it.also was to be a student's responsibility to
obtain these speakers. Rule 17 now read that the university
recognized the intellectual value of discussion of public
issues on campus. In.short, on October 23, Kerr made two
moves: One was to attempt to control the amount and vari-
ety of politica1.advocacy on university campuses that he
felt would reflect on the university. The second was to
liberalize "free speech" by public speakers on campus.
Students all over the university system protested the Kerr
amendment.

In February, 1961, Kerr modified regulations to
allow the distribution of non-commercial literature on
campus. This change came about as a result of a law suit
brought against the university by a UCLA student.

In August of 1961, another statement was made by
Kerr relating to the political rights of students. These
new rules prohibited political action groups from establish-
ing headquarters on campus and denied the use of the uni-
versity's name in describing themselves.

The following academic year (1961-62) a debate took
place in the Qailngalifornian between Kerr and two members
of SLATE (the Berkeley “Activist" Party). Kerr replied that
it was not true that previously held student rights had been

denied; if the students believe that the.Kerr directives are

less liberal than the previous rules, they could ask for a
return to Rule 17.

During the next academic year (1962-63), Kerr an-
nounced that off-campus political groups could use University
facilities provided their meetings were not used to plan
political action.

In June of 1963, the regents voted to lift the Com-
munist speakers ban as a result of a suit brought against
the University by four Riverside students.

During the 1963-64 school year, "radical“ students
began to spend increasing amounts of time in civil rights
activity in the San Francisco Bay area learning the tactics
of non-violent protest.

Up to this point in the Free Speech controversy,
Berkeley students attempted to reach a peaceful solution
amenable to.both themselves and the administration.5

On September 16, 1964, Dean of Students, Katherine
Towle, sent a letter to all "presidents, chairmen, and
advisors” of students activities stating that henceforth
.no tables used by political groups to collect money and
recruit would be allowed in the Bancroft and Telegraph en-
trance, and that the dissemination of unapproved literature

fl

5Heirich & Kaplan, p. 22.

and other activities on off-campus political issues would
be prohibited.6'7

On September 17, twenty organizations formed a United
Front to protest the new rules. The groups ranged in inter-
est from radical-socialist groups, religious groups, Young
Democrats and Republican Clubs, including Youth for
Goldwater.8

Interested students at first tried to reason with

the university and.expressed their Opinion that the new

ruling was unjust. As the university was unwilling to yield,

 

6Editors of the California Monthly, "Chronology of Events:

Three Months of Crisis, Lipset &.Wolin, p. 100.

7There has been much controversy over the reasons for this

attempt at reducing political activity on campus. Most
speculation, however, seems to center around the Oakland
Tribune, a newspaper owned by former California Senator
WiIIiam Knowland. The fact that the newspaper was picketed
for allegedly discriminatory hiring practices during the
summer of 1964 by Campus CORE certainly was not looked on
with an approving eye at the "Tribune," not to mention the
Civil Rights activity in the San Francisco Bay area the
preceding fall. The final straw came, however, at the Re-
publican Convention in July of 1964. Scranton supporters
organized a pro-Scranton rally originating on campus and
taking place at the convention without getting official
permission from the university. Senator Knowland, a Gold-
water supporter asked why this had been permitted in an
editorial. FSM.supporters felt that this editorial caused
pressure to be brought to bear on the university.

8Hal Draper, Berkeley: The New Student Revolt (New York,

1965) . 32.

It shogld be noted that once the tactics of the FSM turned
to civil disobedience the political right groups tended to
leave the FSM as they disagreed with these tactics.

the students' next move was to disobey the university rul-
ing. They set up their tables and handed.out unapproved.
literature against the university regulation. Five students
were suspended and legal charges were placed against one of
these students. The FSM was born out of these actions.

Thus one issue of the FSM was very clearly that of
"free speech." Yet a great many pamphlets and flyers of
the FSM complained about education at Berkeley, i.e., the
students' role in the university in comparison to the role
of other segments of the university. The FSM members, as
put forth through their literature, believed that the two
issues of free speech and the quality of education were
inseparable:

”In contrast to this tendency to separate the

issues, many thousands of us, the Free Speech

Movement, have asserted that politics and edu-

cation are unseparable, that the political

issue of the first and fourteenth Amendments

and the educational issue cannot be separated.

In place of 9great university“ we have said

"impersonal bureaugracy," "machine" or I'know-

ledge factory" ...
From the point of View of the FSM, the issues were not
separable because it was the administration's use of coer-
cion that violated their constitutional rights and this
same use of coercion which hindered their education.

“We get a four-year-long series of sharp

staccatos: eight semesters, forty courses,

one hundred twenty or more "units," ten to
fifteen impersonal lectures per week, one to

 

9The Free Speech Movement, "We Want A University," Lipset

and Wolin, p. 211.

thrl
led
Ove
rec
tio
tes
gur
twe

The FSM pa;
to divisio
munity.
University
faculty an:
He
fes:
pubj
fes:
ser‘

Again we ge

9111f they :

00.1

haw

 

 

three oversized discussion meetings per week
led by poorly.paid graduate student "teachers."
Over a period of four years the student-cog
receives close to forty bibliographies; evalua-
tion amounts to little more than pushing the
test button, which results in one hundred re-
.gurgitations in four years; and the writing of
twensy to thirty-five ”papers“ in four years,

The FSM pamphlet writers implied and made direct references
to divisions between various segments of the academic com-
munity. For instance a FSM pamphlet entitled "We Want A
University" complained of the divisions between students and
faculty and between faculty and faculty:

He (the student) loses contact with his pro-
fessors as they turn more to research and
publishing and away from teaching. His pro-
fessors lose contact with one another as they 1
serve a discipline and turn.away from dialogue.

Again we get.a glimpse through the eyes of the FSM into a
gulf they felt divides the students and the faculty.

...the overwhelming majority of faculty members
have not been permanently changed, have not
joined our community, have not really listened
to our voices--at this late date. For a moment
on December 8th, eight hundred and twenty-four~
professors gave us all a glimpse--a brief,
glorious vision--of the university as a loving
community. If only the Free Speech Movement
could have ended that day.

FSM participants not only saw a gulf between faculty and

students, and one between faculty, but they also saw an

 

10The Free Speech Movement, Loc. cit., p. 211.

llHeirich & Kaplan, p. 214.

12Heirich & Kaplan, p. 210.

10

isolation of the student from other students. Gerald
Rosenfield, a pamphlet writer for the FSM, describes the
feeling of communality obtained from a civil rights demon-
stration and then related it to the emotions derived from
the FSM. Rosenfield states:

"...we stayed and lived together that night and
through the next day, and when it was over we
were no longer strangers to one another For
twenty-four hours we were a community."

Again, in another pamphlet the same issue of the students'
isolation from his fellow student is found.

“Although our issue has been free speech, our
theme has been solidarity. When individual
members of our community have acted, we joined
together as a community to jointly bear the re-
sponsibility for their actions. We have been
able to revitalize one of the most distorted,
misused, and important words of our century:
comrade. The concept of living cannot be separ-
ated from the concept of other people. In our
practical fragmented society, too many of us have
been alone. By being willing to stand up for
others, and by knowing that others are willing

to stand up for us, we have gained more than
political power, we have gained personal strength.
Each of us who has acteg, now knows that he is a
being willing to act."

Another possible.division of the academic community exists
between the administration and the students. It is the ad-
ministration that is seen as the enemy by the FSM.
"The University's power structure is explicitly
modeled after that of the corporation. We have

a Board with final and total authority; a Presi-
dent and Chancellors responsible only to it; and

 

13Gerald Rosenfield, “Generational Revolt and the Free

Speech Movement." Paul Jacobs &.Saul Landau, The New
Radicals: A.Report With Documents (New York: 1966) p.
215.

14The Free Speech Movement, p. 208.

11

a mass of students and faculty with no rights

except those they can extract by the threat of

direct action."1
Again, we find this same complaint of the administration
holding all the power to the exclusion of other segments
of the university.

“At the present time, University regulations

governing the form of expression on the cam-

pus are promulgated by the administration,

while other segments of the University com-

munity are limited to a purely advisory

capacity.“

If the administration has full power there might be little
need for contact.with students in order to run the uni-
versity. Thus there might be little contact between ad-
ministrators and students.

The FSM membership thus felt they had two inter-
related problems; the problem of a stifling of free speech
by the administration and the problem of the students' lack
of power and isolation within the academic community in
comparison with the faculty and administration. These pro-
blems were felt to be so severe that Mario Savio, a leader
of the FSM, stated from the steps of the Administration
Building:

"There is a time when the Operations of the

machine become so odious, make you so sick at

heart, that you can't take part; you can't
even tacitly take part. And you'vegot to

‘

15Marvin Garson, "The Free Speech Movement,“ In Draper,

p. 220.

1
6ACLU "The Campus And The Constitution," In Draper, p. 240.

 

12

put your bodies upon the gears and upon the
wheels, upon the levers, upon all the appara-
tus, and you've got to make it stop. And
you've got to indicate to the peOple who run
it, to the people who own it, that unless
you're free the machine will be prevented
from working at all."17

The Committee for Students Rights

Prior to the academic year of 1962-1963, there was
almost no radical activity on the East Lansing campus of
Michigan State University. There appeared to be only one
radical or leftist organization, "The Young Socialist
Alliance.”

As of 1962—1963, however, some Michigan State stu-
dents became vitally interested in university rules and
regulations. The first sign of political activity occurred
in response to the "Speaker Rule" then in effect on campus.
The Speaker Rule stated that no non-university person
could be brought on campus for the purpose of giving a
public speech unless he was first cleared by the “Speaker
Committee." During the week of October 18, 1962 the Campus
Club Conference (CCC) was organized, composed of the Presi-
dents of the:

A11 University Student Government (AUSG)

The Young Socialist Alliance

The Young Democrats

The NAACP

The International Club
The Campus United Nations

 

17Paul Jacobs and Saul Landau, The New Radicals: A Report

With Documents (New York: 1966) p. 61.

13

The Forensic Union

The National Student Association
The reason for the formation of the CCC were twofold accord-
ing to Jim Garrett, President of the Young Socialist A1—
liance:

“(1) the Speaker Committee was in direct viola-

tion of freedom of speech; and (2) because it

denied the student the right to judgg for him-

self, what and whom he could hear.
On October 18, 1962 the CCC held a gathering off-campus at
which three unapproved "Student Non-Violent Coordinating
Committee“ (SNCC) leaders gave speeches in violation of the
Speaker Rule. Dean of Students, John Fuzak, Head of the
Faculty Committee on Student Affairs then asked the committee
to investigate.the reasons for the CCC violation of this
university regulation. As a result of the committee in-
vestigation, those students involved in the CCC were placed
on strict disciplinary probation. Bob Howard, the President
of AUSG was removed from office by the Faculty Committee on
Student Conduct after he refused to resign.

Judging from the letters to the editor column of the
State News, Michigan State University's newspaper, and the
reaction of the Student Congress, the students were dis-
tressed by this act. In spite of this free speech issue
no major active protest occurred along the lines of the FSM./:

Rather, on October 31, 1962 approximately ten to twelve pro-

testors picketed a lecture by the Director of the Honors

 

18C. Chiri, “Ignored Permits For Meeting," State News

(East Lansing, Michigan), October 19, 1962, p. l.

14

College in addition to twenty students picketing the ad-
ministration building and the football stadium prior to a
football game. .Petitions circulated the East Lansing
Campus condemning the University's actions and asking for
the reinstatement of Bob Howard. Seventeen hundred names
were signed to this petition. The head of the Campus
United Nations resigned along with the Vice President of
AUSG and another AUSG functionary. In addition a flyer was
circulated on campus (authors unknown) which said:

"...tell your MHA (Men's Housing Association),

WIC (WOmen's Inter-Dormitory Council), or Pan-

Hel representative to get off campus as Bob

Howard, the AUSG President has done and let's

stop students from betraying students."19
This flyer foreshadowed by approximately two years the FSM
complaint of a segmentation of the student body and the
feeling of a lack of community among students.

The reason, as far as can be ascertained, for the
lack of active protest over the issue of free speech,
an issue that caused a tremendous stir on the Berkeley cam—
pus two years later, apparently was the desire of the stu-
dent to effect the change through established channels.
For example, on November 2, 1962, an editorial appeared in
the State News entitled, "Let's Fight Through Legal Channels."
The editorial said in effect that Howard was wrong for dis-
obeying the rules even.if the rules were wrong. If students

are to protest, they should do it through established legal

channels.

19The flyer was not titled or dated.

15

Spring term and the following school year the first
activist group deeply concerned with students' rights on
campus surfaced and asserted itself. This group was first
known as the Byzantine Anarchist Party and later as the
Basic Action Party (BAP). BAP, according to one member,
was composed of members of the "liberal left, radical
right and moderate middle" ends of the political spectrum.
Robert Mazess, the Co-President of BAP, indicated it had
50 adherents and 200 sympathizers and had the following
three general aims:

"(1) increase student awareness, (2) to return

control of the University to students and faculty,

and to (3) make AUSG an organisation important

and significant to students.

BAP members also stated one of their objectives was the
elimination of en loco parentis, a cry that was going to
be echoed by CSR.

Spring term elections were held to select the AUSG
president and congress for the next coming school year.
Bob Kerr, the winning candidate, campaigned on the asser-
tion that if he could not make AUSG effective as a student
organization he would do his best to dissolve AUSG. This
had a fairly strong appeal to certain segments of the stu-
dent body at this time as a result of the impotence of

student government as demonstrated by the seemingly arbitrary

suspension and removal from office of Bob Howard, the past

 

20Anonymous, “Byzantine Anarchists Hit Administration Con-

trol," State News (East Lansing, Michigan), May 15, 1963,
p. 2.

16

AUSG president. Bob Kerr's campaign manager during this
period of time was Bob Hencken, the President of the “Young
Democrats” who was suspended with Howard. A number of

charges were printed in the State News to the effect that

 

Kerr was a member of BAP and that Hencken was an extreme
leftist both of which were denied by Kerr as he supported
Hencken for the "Speaker's Post" in the student congress.

During the academic year 1964-1965, a new group
concerned with student rights came into being. This was
the Committee for.Student Rights (CSR). CSR did not apply
for a charter which would license it as a legitimate student
group in the eyes of the student government and the adminis-
tration. CSR develOped out of a group with an extremely
short history, the Federation for Student Rights (FSR). FSR
applied to AUSG for a charter but was turned down because at
that time there.was a law in the AUSG.Constitution forbidding
the granting of a charter to two groups with the same pro-
fessed function. As BAP was still officially active, a
charter for.FSR was refused.

CSR concerned itself with such campus issues as
en loco parentis, women's dormitory hours, requirements for
living in off-campus housing, disciplinary procedures, the
distribution of unapproved literature, the abolishment of
dormitory Resident Assistants reports on students, Civil
Rights, and communication channels between students and

other segments of the academic community.

17

CSR issued an undated flyer entitled, ”A Descriptive
Outline of The Committee for Student Rights (CSR)." The
flyer, in stating the reasons for the existence of CSR,
listed as its first point a complaint about student partici-
pation in the university and as its second point a complaint
about the current communication channels.

"CSR grew from a group of students committed

to certain.principles; they were dissatisfied

with the lack of student participation in the

university community and the ineffectiveness

of the present "channels" for voicing student

Opinion..."

Again, we hear the same complaint about poor communication
channels from an irregularly published series of pamphlets-
entitled Logos (CSR's official publication): “Official

21 was printed in bold

Channels Cannot be Secret Channels"
face.

CSR, like the FSM, and like a previous flyer printed
after the removal of AUSG President Bob Howard, complained
of a lack of community among students:

"The future status of students at MSU will

largely depend on whether students stand up

for other students who have been treated

unjustly."22

CSR, like the FSM, sees the administration as the
enemy and feels isolated from other segments of the academic

community, i.e., the faculty and other students:

 

21P. Schiff (Ed-): Logos, East Lansing, Michigan: Mimeographed,

Volume 1, Number 4 (March 30, 1965)

22F. Schiff (Ed.), Logos, East Lansing, Michigan: Mimeographed,

Volume 1, Number 7 August 3, 1965)

18

“The prevailing feeling at M.S.U. (and for
good reason) is that once the administration
acts--that's it: no one cares, no one will
stand up for you if you've been wronged.

But when a portion of the faculty comes alive
with discussion and action 'the times they
are a changing'--for the better."

Like the FSM, the CSR also found the university to
be coercive and impersonal:

"What is a multiversity? ... it is imperson-

ality, it is 300 students in one lecture hall

with one professor; it is vocational training

instead of the search for knowledge; it is an

IBM card and a student number instead of a

student; it is paternalism and en loco parentis

instead of responsibility and freedom; it is

production of graduates instead of education

of students; it is courses and credits instead

of learning; and finally it is a business run

by administrators instead of a community of

scholars run by scholars. It is gnfortunately,

the modern American university."2

Another similarity between the CSR.and the FSM was
the issue of free speech. Paul Schiff, a former graduate
student at Michigan State, was denied re-admission as a re-
sult of writing a Logos editorial advocating the right of
students to distribute unapproved literature and denying that
the university had a right to prohibit such action. Schiff
sued the university. After several months, Court sessions,
university committee meetings on the subject, Schiff was
again denied re-admission and then was finally re-admitted
after making moves to resume his suit.

The similarities between the two student movements

ends at the comparison of the extent to which students took

 

23Stuart Dowty, "The American University: Is Democracy

Possible,“ Organon, East Lansing, Michigan: Mimeographed,
Number II (February 1966). Organon is a CSR publication
in magazine form which was published after the printing
of LOgos was terminated.

part. The
bad betwee
The FSM he

estimated

Sums o

In
dent demor
out the Ur
movements

ments prot

 

 

&Sked is

“Hillel-sit,

19

part. The largest demonstration in which CSR played a part
had between 300 and 400 demonstrators with 59 arrested.
The FSM had 800 arrested at one demonstration and had crowds

estimated at several thousand at their rallies.

Summary of Student Movement Complaints

In summary, during the early and middle 1960's stu-
dent demonstrations began to appear on many campuses through-
out the United States. In viewing two separate student
movements at two large university campuses, the two move-
ments protested common issues:

(1) free speech

(2) students felt isolated from the faculty

(3) students felt isolated from other students

(4) students felt isolated from the decision

making processes, i.e., the administration.

(5) students seemed to be unhappy with the

education at large universities; they found
it to be coercive and depersonalizing.

(6) students at Michigan State University

specifically complained of poor communica-
tion channels.
This research will be concerned with complaints number 2,
3 and 4 only. It will not be concerned with the rest of

the complaints listed above.

The Diffusion of an Idea

A university is an institution concerned with the
communication of ideas. One question then that could be
asked is given that an idea exists somewhere within the

university, in what directions would the idea disseminate

20

through the social structure? Would the idea be more
likely to spread from students to faculty or from students
to administrators? For example, if a group of freshmen
are in favor of a Pass-No Credit grading system, how would
it most likely spread through the academic community?

Social scientists within many disciplines have con-
cerned themselves with the diffusion of an idea, or more
specifically an innovation, through a given social system.
Rogers (1962) has reviewed over 500 articles concerning the
diffusion of an innovation in an attempt to unify this
literature. The kind of innovations covered in this re-
view ranged from hybrid corn to new medical drugs. The
diffusions were attempted in many diverse social systems
ranging from a community in India to a group of medical
doctors in the American Midwest.

Rogers presents the traditional conceptualization of
the process an individual passes through in adopting an in-
novation. This view consists of five stages.' First the
individual becomes aware of the existence of the innovation.
Secondly the individual becomes actively interested in the
innovation and seeks new information about it. Thirdly the
individual attempts to evaluate the innovation and come to
a decision of whether or not to try the innovation. The
fourth stage in the process is the stage at which the in-
.novation is tried on a.temporary basis.with the fifth stage

being the adoption of the innovation.

21

Rogers and Shoemaker (in press) recently arrived at
an alternative to the traditional five stage process of the
adoption of an innovation. This model, like the previous
one, represents the mental process through which an indi-
vidual passes in making the decision to adopt or reject an
innovation. It is not a description of how the innovation
is diffused through a social system. According to Rogers
and Shoemaker, among the advantages of the new model are
that it does not imply that the process always ends in an
adoption decision and it does not state the stages which
have to occur in a specified order.

The new conceptualization is more.explanatory than
descriptive when compared to the traditional five stages.
Several discreet psychological mechanisms such as "selective
perception" and "dissonance reduction" often.play a role in
the new formulation. The individual within the process is
afforded an active decision making role. The new "concep-
tualization consists of four functions or stages:

(1) Knowledge--where the individual is exposed to
the innovation's existence, and gains some
understanding of how it functions.

(2) Persuasion-~where the individual forms a favor-
able or unfavorable attitude toward the
innovation.

(3) Decision-—where the individual engages in ac-
tivities.which lead to a choice to adOpt or
reject the innovation.

(4) Confirmation--where the individual seeks re-
inforcement-for the innovation decision he has
made, but where he may reverse his previous

decision if exposed to conflicting messages
about the innovation."

22

Rogers indicates that several studies have shown
that the adOption rate of an innovation by individuals
within a social system approximates a normal distribution.
Therefore an individual's rate of adoption can be expressed
as a normal score. Rogers has given names to groups of in-
dividuals that adopt the innovations at different times and
thus fall on different areas of the normal curve. The group
of individuals two standard deviations below the mean or the
first 2 1/2% of the pOpulation to adOpt the innovation are
called "innovators." ,The next two standard deviations or
the next 47 1/2% of the population to adopt the innovation
up to the mean are called the "early majority.“ The popula-
tion of adopters falling in the first standard deviation
after the mean are called the "latemajority'I and account
for 34% of the adopter.population.‘ The remaining 16% of
the population of adopters are classified as “laggerds.”

Many different variables affect both the manner and
rate with which an individual passes through the ad0ption
.process and determines whether an individual will be among
the "innovators” or."laggards“ in the frequency distribu-
.tion of adopters. For example an individual's innovative-
ness depends on a modern rather than a traditional orienta-
.tion and "varies directly with the norms of his social
.system on innovativeness.“ Another important variable is
.the individualls position in his informal social structure

and the nature of the informal social structure itself.

Col
tionship b
and the ph
Coleman, 6
of social
in four di
of using a
on the inc'
started tc
records of
files.

One
was hOSpit
at a hOSpi
Celeman, e
at hOSpita
With lower
those doct

the local

 

SIOWer to

adopt the

     
  

 

Ma. '

23

Coleman, Katz and Menzel (1966) explored the rela-
tionship between the time of first prescribing a new drug
.and the physician's integration into the medical community.
Coleman, et. al., investigated several different networks
of social and professional relationships among physicians
in four different cities. This research had the advantage
of using a "hard" dependent variable. Instead of relying
on the individual physician's memory of when he first
..started to use the new drug Coleman, et. al., used written
records of behavior by utilizing pharmacists' prescription
files.

One professional relationship that was investigated
Awas hospital affiliation. A doctor may.have.various statuses
.at a hospital from a.full appointment to courtesy privileges.
Coleman, et. al., found that doctors with full.appointments
at hospitals tended to introduce the drug sooner than those
.with lower level appointments. They also discovered that
those doctors who attended three or fewer staff meetings at
the local hospital (in a two-month period) were 3.6 months
slower to try the new drug and were much less likely to
adopt the new drug than the doctors who attended more
meetings.

Many physicians shared offices with other doctors or
were part of private clinics. Again it appears that those
doctors with medical contacts, i.e., office or clinic part-
ners, used the new drug sooner than those doctors without

.these contacts.

24

Coleman, et. al., interviewed the physicians in
order to obtain sociometric data. They asked questions
.designed to explore “Advice and Information Networks,“
"Discussion Networks,“.and “Friendship Networks," in which
the doctors are immersed.

To investigate the "Advice and Information Networks"
doctors were asked: "When you need information or advice
about questions of therapy, where do you usually turn?“
From this answer a network of relationships were mapped
which indicated that most physicians‘are connected to one
.another either directly or indirectly. Coleman, et. al.,
.found that doctors named as advisorswtried.the new drug
earlier than his colleagues who were not named as advisors.
They concluded that.the early use of.the.new.drug was dis-
.proportionately located at the centers of the "Information
and Advice Networks.“

To explore the “Discussion Networks“ physicians were
asked: ”Who are the three or four physicians with whom you
.most often find yourself discussing cases or therapy in the
course of an ordinary week - last week for instance?“ The
"Discussion Networks“ while resembling the “Information and
AAdvice Networks“ were distinct from it. The "Discussion
Networks“ showed many.more reciprocating relationships than
-did the “Information.and Advice Networks" which were pri-
marily one way (a doctor who is chosen as an advisor rarely
.names as an advisor the doctors who have chosen him).

Coleman, et. al., found as hypothesized that the more a

25

doctor was chosen as.a discussion partner, the sooner was
his first trial of the new drug. The authors conclude
that "day to day shop talk" among doctors is important to
‘the decision to try the new drug.

The doctors were next asked: "Would you tell me
who are your three friends whom you see most often socially?"24
This question enabled the authors to construct sociograms
representing the "Friendship Networks.". This network is
, similar to the “Discussion Networks" in structure and
.in the relationship between the amount one is chosen as a
.friend and the time of the first trial of the new drug.

Thus the results for all the networks indicates
.that the doctor who is integrated into the medical com-
munity is more likely to try the new drug earlier than
those doctors who are not integrated. 4Coleman, et. al.,
~indicate that while their results are correlational a
—causal relationship can be inferred. The cumulative curve
-of new drug adoptions by integrated doctors is S shaped
-indicating new drug.adoption in each month.appears to lead,
according to the authors, to a larger number of adaptions
-in the following month; i.e., a contagion process is at
~‘work. S-shaped curves.in previous diffusion literature
Ihave been alternately interpreted as the result of this

“contagion process” or of a normally distributed “readiness

»

24If fewer than three physicians were named.the doctor was

asked "which three fellow physicians he saw most often
socially?“

26

to adopt.“ It is only the integrated physician that showed
this S-shaped function in the drug diffusion research.
Coleman, et. al., suggest that the preceding fact argues
in favor of the contagion hypothesis and.thus indicates
.-the integration variable is causal with the time of first
-trial of the new drug.being the effect.

Coleman, et. al., found that the different networks
. tended to have their effect on the adaption process se-
quentially. First, interpersonal influences on the ad0p-
..tion of the new drug Operated through professional ties or
.professional relationships between doctors. .Secondly, the
-socially oriented-relationships such as the.'Friendship Net-
.works" exhibited their.influence in.the adOption process.
‘In addition the authors found that different.communication
.channels have their effect during different stages of the
."traditional" adoption process previously cited from Rogers.
.-The "detail man“ (the drug company.salesman) and drug com-
..pany mail dominate the “Awareness Stage.‘' .The “Interest
. Stage“ is effected by many different.communication channels.
“During the "Evaluation.Stage" which leads to the decision to
.try the new drug, the informal and the more.professional net-
Vworks influence the doctor's decision (in addition to profes-

. sional journals).25

 

25The authors presented communication channels.for only the

first three stages.of the five-stage process. It is likely
that the “Trial Stage” and "Adoption Stage" use the same
channels of communication (plus their own experience) as

in the third stage.as further evaluation would be taking
place.

27

Coleman, et..al., concluded that while the formal or
institutional communication networks Of the medical community
.are determinants.of the speed Of drug adOption it was not as
important as the informal communication structure. The in-
.formal communication network Of a large university is the

topic Of investigation Of this dissertation.

The University as a Stratified Social System

 

The study Of the university also Offers an Opportunity

to investigate a simplified stratification system. A uni-
-versity, unlike some other social systems,.has only three
.primary functions to be performed, or roles tO be played;
.that of student, faculty and administrator.. Within the

. pOpulation playing.the student role, it is possible to make

status distinctions between individuals at the freshman,

SOphomore, junior, senior and graduate student levels.

. Within the population.p1aying the faculty role status dis-

tinctions can be made between individuals at the Instructor,

. Assistant Professor, Associate Professor and Professor
levels.

Sociologists, according to Tumin (1967), differen-
tiate three paths tO the attainment of a status. A status
..can be ascribed (assigned or inherited),.achieved (gained

.by one's own effort).or Obtained through maturation (reach-
..ing a certain age,-e.g., voting age). Tumin states that

.while there are three distinct methods for attaining a

28

status,in no society is there only one method Of acquiring
a given status.

At a university a status change among students, i.e.,
.a change Of class levels, is accomplished through the stu-
-dent's classroom achievements. When a student's records
Hindicate that he has enough academic credits to become a
sophomore he gains the status and rights Of a SOphomore.

Among the faculty the criterion for changing rank or
status is not as clear as it is for undergraduates. Again
it would appear as if the major road to the attainment Of a
higher status is again achievement. If a junior faculty mem-
sber exhibits academic success represented by scholarly publi-
-cations, carries out his administrative responsibilities, and
is reputed to be a good teacher (or at least not a bad tea-
cher), he is a likely candidate for promotion. Maturation,
.however, in addition to achievement can also play a role in
.the attainment Of a higher status for a faculty member. A
gfaculty member's age, date Of receiving his doctorate, the
amount Of time since his last promotion, etc., can influence
the decision to promote the faculty member.

According to Tumin the essence of social stratifi-
cation is the assignment Of members Of,a social group to a
hierarchy Of positions that are unequally rewarded with
,"power' (the ability to Obtain one's goals even against
.Opposition), “prOperty“ (“rights over goods and services“)
.and “evaluation“ (the "societal judgment that a status is

more prestigious and honorable than others“). At,a

29

university the inequality Of rewards in terms Of property
.is striking. Undergraduates do not receive a salary for
playing the role of a student, in fact they pay the uni-
versity tuition for the right to be a student.26 Faculty
members and administrators on the other hand do receive
salaries for fulfilling their functions. Again, an in-
equality Of rewards is noted, for individuals fulfilling
different functions in the university. Students are not
allowed to drive on campus during the day and must keep
their cars in specially designated lots. Graduate assist-
ants On the other hand,.are permitted to drive on campus
and park in any lot on approximately half the campus.
While faculty and staff.are permitted to drive on campus
and park in any lot they desire.

Not only are.rewards distributed unequally tO indi-
-viduals playing different roles or fulfilling different
functions, but rewards are distributed in an unequal manner
to different strata within a role. For example, freshmen
and SOphomores are Officially not allowed to declare a
major until they are juniors or seniors. Seniors are given
the first opportunity to select football tickets, followed

by juniors, SOphomores and freshmen.27 In.viewing the

 

26Students, of course, receive an education from the uni-

versity in exchange for their tuition.

27The rewards being discussed are indigenous to Michigan

State University with many Of these rewards being idio-
syncratic to this campus. The author feels, however,
that at other campuses the reward structures are similar.

30

faculty, the same step progression in the distribution Of
“rewards is recognizable. For example, in looking at faculty
.pay guidelines it is.found that Instructors are the lowest
.paid faculty members, followed by Assistant Professors,
Associate Professors, with Professors being the most re-
warded segment Of the faculty. Administrators, on the
average, are probably at least as well rewarded financially
(as are the Professors.28
It would appear that administrators, in some matters,
have more power tO exert than do either students or faculty.
"For example, students, in order to change university regu-
slations, must seek the approval Of either some branch of the
.administration or the faculty. A faculty member also has
.more power than a student. A faculty member gives grades,
.assigns work to be performed by the student, etc., while
the student has little effect on the teaching methods Of
this faculty member...The faculty in regard to some matters,
are dependent on the administration. For example, adminis-
trators have control of the purse strings. Academic departe
ments can only request certain fundings; however, Deans and
sother administrators have some latitude in rewarding funds.
A Dean could decide to strengthen one department at the ex-
pense of another. A Dean also has a rarely used Option to

approve promotions.and.pay raises among the faculty of his

 

28See the Ss section for the definition of administrators

as used In this study.

31

college thus possibly overriding a decision of the tenured
faculty. In addition, faculty members are elected to the
"Academic Council," a governing body Of the university,
while Deans and some central administrators are members via
their administrative position.

In conclusion, it would appear that when the roles
played in an academic community are ordered according to
their status administrators are on tOp, the faculty second
and the students third. If the strata within each of these
roles are viewed they would appear as diagrammed in Figure 1.

FIGURE 1. The university viewed as a stratified social
system

 

 

I Administrators29

. 1) Professors
——(2) Associate Professors
II Faculty ~—{3) Assistant Professors
4) Instructors

 

 

1) Graduate Assistants
2) Seniors
III Students (3) Juniors
4) SOphomores
5) Freshmen

 

 

The Advantages and Disadvantages to Viewing the University
As a Stratified Social System.

The advantages to viewing the university as a strati-

fied social system are relatively clear cut. Unlike a small

 

29Administrators are treated as one undifferentiated group

in the present study despite the fact that administrators
also are differentiated into statuses. Administrators

were treated as an undifferentiated group.as the hypotheses
to be presented does not call for finer distinctions.

32

city the various strata Of a university are well defined by
the social system itself. There is no question Of whether

an individual is a sophomore or a junior, his records clearly
indicate his status. In contrast, the usual classification
scheme used in a city, i.e., lower middle class, upper mid-
dle class, etc., are not "clear” and are determined by a
number of variables such as income, occupation, etc. In
addition the university has relatively few roles, and those
roles that do exist are clearly interrelated in regard to
function.

There are dangers and disadvantages in viewing the
university as a stratified social system. The university
is 921y_an upwardly mobile system. The failures do not
lose status, they leave the social system. There also
could be a danger in generalizing from a simple system to
one that is more complex. Many important variables may not
exist in the simplified system that dO have an effect in a

more complex system.

Objectives Of the Current Research

The purpose of the current study is to explore the
informal social structure and communication channels that
exist within a stratified social system that is a large uni-
‘versity. The social structure will be viewed and considered

as to how an idea would be diffused through it.

33

Specifically it is hypothesized, as a result Of the
investigation of the complaints during the FSM at Berkeley
and the CSR at Michigan State University that:

(1) Students are not integrated into the academic
community.

(a) students are socially separated from
faculty members in comparison with
faculty members and administrators.

(b) students are socially separated from the
administrators in comparison with faculty
members and administrators.

(c) students are socially separated from
other students.

The measure Of a student's integration into the academic
community or its inverse, the student's separation from
other segments Of the academic community, will be measured
by determining the length of the informal communication
channels linking students to other students, faculty mem-
bers, and administrators. These hypotheses will be further
clarified in the following section in terms Of the technique
used to test the hypotheses and previous research performed
with the technique.

It is further hypothesized that administrators will
have shorter communication channels to other segments of
the academic community than do either the students or the
faculty. Leavitt (1951) studied four different patterns
of communications in five-man groups that had been given a
common problem to solve. The problem was for each group to
determine what symbol was held in common by its membership

from a variety Of symbols held by individuals in the group.

34

The communication pattern, i.e., the arrangement Of Open
communication channels with which one group member may
communicate to another member (not all communication chan-
nels were Open) during the period in which a solution was
sought were fixed by Leavitt. The groups were given sev-
eral trials in which to establish a problem solving proce-
dure in order tO clarify the role each member was to play
in reaching the solution. Leavitt found that the most
"central“ individual in the pattern, i.e., the individual
with shorter communication channels to other group members,
generally made the decisions for his group. He also found
that the individual who was most central was recognized as
the leader most Often by the other members Of the group.
Thus as individual administrators have broader decision
making responsibilities than do other segments of the
academic community. It is expected that they will have
shorter communication channels to the other segments Of

the academic community.

CHAPTER II

METHOD

The Technique

 

The investigating technique used in this study was
originated and first used by Milgram (1967, 1969). He
called it the "'Small World' Technique," a phrase which
according to Milgram (1969) has long been a common expres-
sion of speech, but was first utilized in the social sci-
ences by IthdelPOOl. The following illustrations should
help explain the question the Small World Technique at-
tempts to answer and yield some insight about the method.

The population in a given social unit such as the
United States may be viewed as approximately 200,000,000
points; each point representing one individual from the
200,000,000 individuals residing in the United States.
These people, represented as points, are not isolated, they

have acquaintances.1 Each acquaintanceship between any two

 

1In fact the average individual has approximately 500 ac-

quaintances according to Gurevitch (1961). Gurevitch
asked a variety Of individuals to keep a record Of the
people they came in contact with.during a 100 day period.
Surprisingly, the average person came in contact with
roughly 500 persons during this period.

35

36

individuals Of this population can be represented by a line
linking the two individuals that are acquaintances. If
each acquaintanceship was linked by these lines, the re-
sulting diagram would appear to be a very complex network
of acquaintances. If this was diagrammed, one would readily
Observe that no individual is directly connected to every
other individual via an acquaintanceship. However the vast
majority Of individuals can be indirectly connected, i.e.,
connected through an acquaintance, to a vast number Of peo-
ple they do not know personally. For example, Albert is
directly acquainted with John as they spend Saturdays to-
gether at the town dump shooting rats. John is directly
acquainted with Richard as on Tuesday nights Richard and
John go bowling together. Albert and Richard have never
met. Thus they are not directly acquainted. However, one
may say they are indirectly acquainted as Albert knows
Richard through John. If Albert had to deliver a message
to Richard and.was restricted so that he could only hand
the message to someone he personally knew, he could hand
the message to John who could hand it to Richard. Thus it
may be said that Albert is indirectly acquainted to Richard
through John. This is the principle on which the Small
World Technique is based.

In the Small World Technique an individual, who
Milgram calls the "Starter Person,II is given a packet Of
instructions which includes the name.of a second individual

who Milgram calls the “Target Person.“ “The Starter Person

37

is asked to transmit the packet Of instructions to the
Target Person. However, the Starter Person is re-
stricted so that he can only hand the booklet to an indi-
vidual he knows according to a certain criterion; e.g.,.
he must know the person to whom he will give the packet Of
instructions on a first name basis. Failing to meet the
criterion so as to pass the packet Of instructions to the
Target Person directly, the Starter Person is requested
to select an individual from his pool Of acquaintances,
that he does know according to the criterion who would have
a better chance Of knowing the Target Person. He is asked
tO pass the instructional packet along to this individual.
The second person in this chain, providing that he is not
the Target Person, is asked to follow the.same procedure
as are all the other people involved in the chain. This
process lasts until a person is found who can pass the in-
structional packet directly to the Target Person accord-
ing to the criterion or the booklet is discarded or lost.
The Target Person is requested to hold the instructional
packet for the experimenter. To summarize.then, if the
instructional packet reaches the. Target Person it can
be stated that the .Starter Person is indirectly acquainted
with the Target Person ‘through the group Of intermediaries
that transmitted the instructional packet from the Starter
Person to the Target Person.

Thus through this technique it is possible to trace

a.chain of people that link the Starter Person to the

38

Target Person. Both the number of and characteristics of
the intermediaries that link the Starter Person tO the
Target Person are the major dependent variables of inter-
est. TO Obtain these data each individual who takes part
in a study using this technique is asked to fill out and
return a postage paid business reply card.

Several studies have been carried out by Milgram
and his colleagues using the Small World Technique. In
all these studies Milgram and his associates asked the S8
to follow the same criterion:

"IF YOU KNOW THE TARGET PERSON ON A PERSONAL

BASIS, MAIL THlerOLDER DIRECTLY TO HIM (HER).

DO this only if you have previously met the

target person and know each other on a first

name basis.

“IF YOU DO NOT KNOW THE TARGET PERSON ON A

PERSONAL BASIS DO NOT TRY TO CONTACT HIM DI-

RECTLY. INSTEAD, MAIL THIS FOLDER (POST CARDS

AND ALL) TO A PERSONAL ACQUAINTANCE WHO IS

MORE LIKELY THAN YOU TO KNOW THE.TARGET PERSON.

You may send the folder on tora friend, relative

or acquaintance, but itzmust be someone you know

on a first name basis."l
They placed in the instructional packet the Target-Per-
.son's name and information pertaining to him.such as his
address, occupation, organizational membership and his busi-
ness address.

Travers and Milgram (1970) utilized a stock broker
.residing in Sharon, Massachusetts and working in Boston as

the Target Person. Three different ”Starter“ populations

y

ZQuoted from Milgram (1969)
One form Of the instrument is reproduced on pp. 110-111
Of this source.

39

were used. One Starter population was randomly selected
from the Boston area; i.e., proximal tO the “Targets" home.
The second and third pOpulations came from the state of
Nebraska (distal to the Target's home); the second pOpu-
lation was randomly selected while the third population was
"systematically chosen" for ownership Of blue-chip stocks.
Through this design Travers and Milgram hOped to assess the
relative effects of the Target Person's occupational con-
tacts and the geographical distance between the Starter
Person and the Target Person. The authors found that
.29% Of the instructional packets that were started reached
the Target Person taking a mean Of 5.2 intermediaries to
complete the chain. When the completed chains were separ-
ated by their approach to the "Target," i.e., either
through the Target's business contacts or through his
residence, it was found that the business contacts provided
a shorter route than did the residence approach (a mean Of
4.6 intermediates compared to 6.1 intermediarieS). The
authors also found a significant difference between the
locations Of the ”Starters.“ Boston Starters had rela-
tively shorter numbers Of intermediaries than did the Nebraska
Starters that were randomly selected (4.4 intermediaries
<compared tO 5.7 intermediaries). There was no significant
difference between the Nebraska Starters that were ran-
JSOmly selected and the Nebraska Starters with stock
lorokerage connections. Travers and Milgram.found that as

'the chains converged on a target they tend to pass through

40

common channels, i.e., the chains share intermediaries.
Travers and Milgram report that the 64 instructional packets
that reached the target were sent by a total of 26 peOple.
In fact, 25% of these instructional packets reached the
target through just one neighbor. The authors also report
that they could find no differences between the gs that co-
operated by participating in the study and those that did
not. Thus the authors tentatively concluded that the re-
fusal to OOOperate with the study is a random event. This
point will be Of interest in later sections.

Korte and Milgram (in press) had white Starter Per-
sons in Los Angeles attempt to 'reach' one pOpulation Of
white and a second population Of Negro Target Persons re-
siding in New York. The Objectives of the study, according
to the authors, "was to see what happens to acquaintance-
ship chains as they are impinged upon by social structure."
Korte and Milgram found no significant difference between
the mean number Of intermediaries needed to complete "white-
target" chains (5.5) in comparison to "Negro-target" chains
(5.9). They did find, however, that Negro-target chains
were less successful in reaching their target (13%) in com-
jparison to white-target chains (33%). When the chains
‘were viewed without reference to the race of the Target
IPerson the results were comparable to previous studies.
Trhe mean number of intermediaries between the Starter Per-
son and the Target Person was approximately 5 to 6 with

22% Of the instructional booklets reaching their designated

41

targets. Korte and Milgram report that 72% of the Negro
participants cooperated in unsuccessful Negro-target
chains compared to 82% cooperation for whites in these
same chains. This raises some doubts about the previous
conclusion that the refusal to COOperate with the study
is a random event.

Thus, in general, it would appear that Milgram
(1967) has additional support for the conclusions made
from his pilot study. The "Small World Method“ is work-
able and on the average only about five intermediaries are
needed to link any two individuals chosen at random in the
United States.

In terms of the.objectives of the present study,
then, if the mean number of intermediaries between any two
groups (students, faculty and administrators) approaches
five the two groups will be considered socially distant
from each other and the academic community not well inte-

grated.

The Instrument

 

The instrument used in the present study is similar
in design to the one used by Korte and Milgram. It con-
sisted of a passport style, plastic covered, 3 3/8“ x
6 1/4'I booklet. The name of the Target Person was placed
within each booklet in addition to the status Of the indi-
vidual (SOphomore, Associate Professor, Assistant Dean,.

etc.), his area of.interest (his major if he is a student,

his dePar
office if
dress, a:
her or a<
Each ind

steps:

Ll]

 

0f the (

and his

 

only pa

basis."
from wh
reciPie
ter10n

vidual

42

his department if he is a faculty member, or his college or
Office if he is an administrator), his campus mailing ad-
dress, and his residential address (if he is a faculty mem-
ber or administrator) or his home town (if he is a student).
Each individual taking part was asked to take the following
steps:

(1) to add his name to the booklet roster. This
step allows each participant to see who has
previously taken part, and to prevent the re-
cycling Of the booklet through previous
participants.

(2) to detach, fill out and return a business
reply card. This step enables the experi-
menter to keep track Of the booklet as it
approaches the Target Person and to gather
information about each participant. On each
business reply card was printed the booklet
number from which the card originated and a
sequence number so as tO easily 'place' a
link in the chain.

(3) to send the booklet to the Target Person
if he was known tO the individual on the basis
stated in the criterion. If he did not know
the Target Person as suggested by the cri-
terion, he was asked to hand the booklet to
an individual he did know according to the
criterion who was more likely to know the
Target Person according to the criterion.

The criterion used in the present study is a variant
of the one utilized by Milgram and his associates. Milgram
and his associates, as previously stated, asked their gs to
only pass the booklet to peOple they knew on a “personal
basis." The gs were given a relatively 'hard' criterion
from which to judge whether or not they knew a prospective
recipient Of the booklet on a personal basis. This cri-

terion was, are you on a first name basis with the indi-

.vidual selected to receive the booklet? If the answer was

43

yes, the individual could pass the booklet tO that person
and was judged to know that person on a personal basis.
If this criterion Of knowing the individual on a first
name basis was utilized in an academic community, it is
reasonable to assume that few, if any, undergraduates
would be able to pass the booklet along to a faculty member
or administrator. Social convention at a university dic-
tates that an undergraduate, even if he knows a faculty
member, still addresses him as Doctor or Professor. Thus
what was needed was a criterion that was comparable in
meaning to the one used by Milgram and his associates but
did not utilize the first name basis criterion. What was
needed was a criterion that implied a more intimate form of
contact than the contact provided solely by a relationship
Of complimentary roles, e.g., student and teacher.

The criterion selected was having previously had an
informal conversation with the recipient of the booklet.
It was felt that if a person was acquainted with another
on a first name basis he has had informal conversations
with the person. The criterion as written in the booklet
appears below.

If you do not know the target person on an in-

formal basis, i.e., if you have never previously

entered into an informal conversation with the

target person, then dO not try to contact the

target person directly. Rather pass this booklet

to an individual you know on an informal basis

who has a better chance of passing the booklet
to the target person.

44

Three rules that should be followed in the
selection of the next individual to receive
this booklet are:

(1) THE PERSON YOU PASS THIS BOOKLET TO MUST
KNOW YOU AS A "PERSON" NOT JUST AS A
STUDENT, FACULTY MEMBER, OR ADMINISTRATOR.

(2) DO NOT PASS THIS BOOKLET TO AN INDIVIDUAL
WHERE YOUR CONVERSATIONS ONLY CONCERN "SMALL
TALK" (E.G., HOW ARE YOU TODAY?).

(3) PASS THIS BOOKLET TO AN INDIVIDUAL WITH
WHOM YOU.HAVE DISCUSSED PERSONAL PROBLEMS,
YOUR SOCIAL LIFE, FOOTBALL OR ANYTHING ELSE
OF MUTUALthTEREST TO THE TWO OF YOU. SE-
LECT AN INDIVIDUAL WITH WHOM YOU ARE ABLE
TO DISCUSS MOST ANY TOPIC RELATING TO THE
UNIVERSITY OR TO INDIVIDUALS WITH WHOM YOU
CAN CARRY ON AN HONEST EXCHANGE OF VIEWS.

Some examples Of legal and illegal passes are:

You should NOT pass it to a faculty member if

all your contact concerned tests, grades, course

material for the course he teaches himself. You

MAY pass this.booklet to an individual if your

conversations with him have concerned grading

procedures, or similar general tOpics. Likewise,

an administrator should not.pass the booklet to a

faculty member if all their previous communica-

tions were concerned exclusively with budgets,
course material, etc.

The entire instrument booklet is reproduced in
Appendix A.

The informal.communication network was chosen as
the structure of study for several reasons.. One, the in-
formal network is the only way some types Of communications
.are disseminated. For.example, how well a faculty member
sconducts his class is a type Of communication that is only
.disseminated both to students and faculty on an informal
basis. Two, student complaints that most interested the

author were those concerning the informal social structure

45

of the university. Three, Coleman, et. al., found the
informal communication networks to be important in the
dissemination of a new idea. Four, Rogers reports that
informal communication is very important at the evalua-
tion stage of the adoption process. Five, using in-
formal communication patterns parameters Obtained from
the present study could be compared to those Obtained

by Milgram and his colleagues.

Subjects

 

E5 were chosen from the three major segments of
the university; i.e., the faculty, students and adminis-
trators.

Two hundred students were randomly selected from
a list of full time undergraduates enrolled at Michigan
State University during the fall quarter of 1968. They
were chosen without regard to class level, sex, grade
point average, etc. Half Of the 200 students were ran-
domly designated potential Starter Persons with the
second half designated as potential Target Persons.
Letters were then mailed to the potential Starter Persons
and Target Persons explaining the functions of the project,

the roles they were being asked to play, and soliciting

46

their cooperation. When more than 110 students volun-
teered (55 Starter Persons and 55 Target Persons) stu-
dents were randomly eliminated from the samples to achieve
this number.

Two hundred faculty members were randomly selected
from a list Of faculty members that were currently em-
ployed at Michigan State University during the academic
year of 1968-1969. They were chosen without regard to
rank (Instructors to Full Professors were selected) or
departmental assignments. Half Of the 200 faculty mem-
bers were randomly designated potential Starter Persons
with the second half designated as potential Target Per-
sons. Letters were then mailed to the potential Starter
Persons and Target Persons explaining the functions Of
the project, the roles they were being asked to play, and
soliciting their OOOperation. When more than 110 faculty
members volunteered (55 Starter Persons and 55 Target
Persons) faculty members were randomly eliminated from
the samples to achieve this number.

The Administration, as used in this study, con-
sisted of the 'Officers Of the University' (the President,
the Provost, the Vice President for Student Affairs, etc.),
the Deans (from Assistant Dean to full Dean), the Chair-
men Of academic departments, and the Directors of On-Campus

Housing, the Placement Bureau and the Union.

47

The Administrators were not samples as they are
relatively small in number. Rather, half Of the approxi-
mately 300 administrators were randomly designated po-
tential Starter Persons with the second half designated
as potential Target Persons. Letters were then mailed
to the potential Starter Persons and Target Persons ex-
plaining the functions of the project, the roles they
were being asked to play, and soliciting their OOOpera-
tion. When more than 110 administrators volunteered
(55 Starter Persons and 55 Target Persons) administrators
were randomly eliminated from the samples to achieve

this number.

The Design

If a university population is viewed as a com-
position of individuals playing three separate but inter-
related roles, then there are nine different Starter
Person - Target Person combinations Of these roles. They

are:

48

 

 

Starter Person Target Person
(1) Faculty members Faculty members
(2) Faculty members Students
(3) Faculty members Administrators
(4) Students Faculty members
(5) Students Students
(6) Students Administrators
(7) Administrators Faculty members
(8) Administrators Students
(9) Administrators Administrators

All nine Starter Person - Target Person combinations

are used in the present study. Each Starter Person is
asked to start six instructional booklets toward their
targets. Each set of six booklets will contain two ran—
domly chosen booklets with student targets, two booklets
with randomly chosen faculty targets, and two booklets
with randomly chosen administrative targets. Each Target
Person was told to expect to receive a maximum Of six in-
structional booklets; two from student starters, two from
administrative starters, and two from faculty starters.
Thus a total of 990 instruction booklets were used in the

present study. Figure 2 illustrates this design.

The Setting

 

The study was performed at Michigan State University
during the last two weeks of November and the month of
December 1967.

Michigan State University is a Morrill Act Land
Grant institution and a state supported school. During

the 1960's the university's student body, faculty and

49

 

muwaxoom omm

mo Hmuou
a How numaxoom w
m>eoomu .manm mm

muoaxoom omm

mo annoy
w you mumaxoom m
w>Hmoou mucoosum mm

muwaxoom omm

mo Hmuou
o How muoaxoom m
O>Hooou auaoomm mm

 

muoaxoom omm

mo Hmuou o
How muoaxoon m
Odom .mEUd mm

muoaxoom OHH
mpmHROOh N
o>wooou .mEOd mm
muoaxoon m
Odom .mfiom mm

muoaxoom oaa

mumaxoon m
F>floomu mucoooum mm
mumaxoon m
Odom .mEO¢ mm

muonoom OHH
muoaxooo m

O>Hooon muasomm mm
muoaxooo m
Ocom .mEOd mm

 

numaxoom omm

mo Hmuou
o How muoaxoon m
Odom mucoosum mm

muwaxoom OHH

muoaxoon m
O>Hooou .m96< mm
nuoaxoon m
comm mucoooum mm

mumaxoom OHH

nuoaxooo N
m>flooou mucoosum mm
muoaxoon a
pawn mucoosum mm

mumaxoom oaa

muoaxoon N
o>flmoou muasomm mm
mumeOOh m
Odom mucmosum mm

 

muoaxoom omm

mo Hmuou w you

nuoaxoon m
comm auaoomm mm

 

mumaxoom oaa
mumaxoon N
m>eooou .mEO< mm

muoaxoon N
comm muasomm mm

muwdxoom OHH
mumsxoon m
v>aoomu mucmosum mm

muoaxoon m
Odom auasomm mm

 

 

muoaxoom OHH
nuoaxoos N
o>wooou auaoomm mm

muoaxoon m
noon avasomm mm

 

 

muoumuumeceao<

cmflmoo Housmafluomxm one

mucmosum
mBmUm¢B

muHsomm

.N HMDUHh

muoumuu
Imacwao<

mflgmfiugflm

muasomm

 

SHSIHVIS

50

physical facilities grew at a tremendous rate. During the
school year 1967-1968 there were approximately 300 adminis-
trators, 2,300 faculty members (instructors to full profes-

sors), 32,300 undergraduate students and 7,700 graduate

students.

CHAPTER III

THE RESULTS1

The level Of S cooperation was excellent. From the
990 instructional booklets used in the present study
some 69% reached their assigned target. This completion
rate is approximately twice the level Obtained by Milgram
and his associates. Table 1 shows the prOportion Of the
instructional booklets started, the prOportion that were
completed, and the prOportion started that were completed
for all nine starter-target combinations.

Approximately 94% of the individuals who were handed
the instructional booklet cooperated by passing it along
tO another person. This compares to the best rate Of 82%
reported by Korte and Milgram. Table 2 shows the proportion
Of the SS that OOOperated by passing the instructional

booklets tO another individual.

 

1Many tables Of conditional probabilities will be presented
within the result section. These tables are only segments
of larger tables tO be found in Appendix B. The complete
tables will not be presented in the text for several rea-
sons: (1) Because Of space limitations, (2) rarely will
all Of a single table be needed to make a point, and (3)
to keep the conditional probabilities pertinent to the
discussion proximal to the area Of the text where they are
being discussed.

51

52

TABLE 1. The prOportion Of the instructional booklets
started, the prOportion that were completed, and
the proportion started that were completed for
all nine starter-target combinations

 

 

 

 

The The The Number
PrOportion Proportion Completed
Started Completed The Number
Started
Faculty-Faculty .955 .891 .933
Student-Faculty .809 .527 .652
Administrator-

Faculty .982 .927 .944
Faculty-Student .946 .518 .548
Student-Student .773 .373 .482
Administrator-

Student .982 .591 .602
Faculty-

Administrator .982 .855 .870
Student-

Administrator .764 .527 .691
Administrator-

Administrator .982 .955 .972

TOTAL .908 .685 .754

 

COOperation, or_its inverse a failure to cooperate is not a
random event. Table 2 indicates that undergraduates do not
COOperate as Often (.914) as do the faculty (.961) (Z=4.80,

p <.001) or administrators (.965) (Z=5-10, p <.001).

The Validity Of the Small World Method

Before exploring the results in detail it would be
wise to first examine the validity of the Small World
Method and as.a consequence the validity of the findings

tO be reported in subsequent sections.

53

TABLE 2. The proportion Of a given status group that
cooperated with the study

 

 

 

The Total N on Which
Proportion the Proportion
is based

Freshman .916 308
Sophomores .914 429
Juniors .917 364
Seniors .920 386
Graduate Students .922 244
Instructors .949 98
Assistant Professors .954 280
Associate Professors .954 197
Full Professors .971 419
Administrators .965 818

TOTAL .939 3543

(from Appendix B, Table 13)

 

The concept Of validity will be separated into two
forms for the purposes Of this discussion. The first form
Of validity, the 'traditiOnal' form, refers to the ability
of the instrument to measure along a certain.specified di-
mension the variable or set Of variables it was designed to
measure. The second form of validity refers to the general-
izability Of the results of specific.studies to other situa—
tions or populations. This form Of validity is referred to
as 'external validity' by Campbell and Stanley (1963).

A measure Of 'concurrent validity' (Cronbach and
Meehl, 1955) was Obtained to test the first form Of validity

described above. An independent criterion was selected and

54

was compared with the results from the Small World Method.
The independent criterion used in this validation was the
selection of roommates by undergraduates. This was com-
pared to the undergraduates selection Of the next individual
to receive the instructional booklet on its road to the
Target Person (using the Small World Method).

For the purposes of the criterion both male and fe-
male wings Of a campus dormitory in addition to an Off-
.campus apartment complex were randomly chosen for interviews.
A total of 88 living units were asked for interviews with
100% of those living units COOperating. The students that
answered the door were asked the following questions:

(1) Are you a Freshman, SOphomore, Junior or Senior?

(2) Is your roommate (or roommates) a Freshman,
SOphomore,-Junior or Senior?

(3) Did you select your roommate or was he as-
signed to the same living unit through the
university or did you find each other through
a newspaper.advertisement?

From the total Of.88 living units.61.units housed roommates
that had selected each other to live with. It is these 61
units that are of interest. Table 3 represents the fre-
quency Of undergraduates residing with a peer across class
..levels, i.e., a Freshman living with another Freshman, a
Sophomore with another SOphomore, etc.

From viewing Table 3 it is quite clear that students

generally select their-peers to live with. Ax2 was com-

puted to test the null hypothesis that these results had

55

TABLE 3. The frequency of residing with a peer across
undergraduate class levels

 

 

 

 

Peer Non-Peer
Freshman A ll 0
Sophomore_ 17 4
Junior 13 3
Senior 8 5
occurred by chance. A X2 = 103.82 was obtained yielding a

p value less than .001 (3 d.f.). Thus the null hypothesis
was rejected.

TO compare the results from the Small World Method
to the criterion presented above a problem.must first be
solved. In selecting a roommate a student can choose any
individual he desires to live with. The Small World
Method does not allow the participants this range of choices.
Rather, the participant is asked to find an individual from
his pool of acquaintances that has the best chance of send-
ing the instructional booklet on to the Target Person
(if he knows the Target Person he is further restricted
.as he is instructed to pass the booklet to the Target Per-
son). This places a "teleological strain,“ a strain that
is target directed, on the choice being made. To counter-
balance this strain no single one of the nine starter-
target combinations (e.g., a faculty starter to a student
target) was used as this creates maximum strain toward
one target type. Rather, a table of conditional probabili-

ties was computed.from all nine starteretarget combinations

56

so as to balance the strain by maximizing it for all tar-
gets. Thus there is no greater strain toward one target
type than for either of the other two target types.

The results using the Small World Method, as
Table 4 illustrates, shows that undergraduates selected

their peers more Often than other groups to receive the

instructional booklet.

TABLE 4. The conditional probabilities that an under-
graduate from a particular class or status will
send the instructional booklet, to another
undergraduate of a particular class or status
across all nine starter-target combinations

 

 

Receivers
Freshmen Sophomores .Juniors Seniors N

 

2 Freshmen .464 .192 .140 .049 308
3 Sophomores .100 .422 .182 .084 429
e Juniors .099 .170 .319 .170 364
: Seniors .018 .109 .200 .389 386

(from Appendix B Table 13)

 

Freshmen passed the instructional booklets to other fresh-
men with a greater probability than to sophomores (Z = 7.56,
p <.001), juniors or seniors.2 Sophomores passed the in-
structional booklets to other SOphomores with a greater

probability than to juniors (Z 2 8.00, p <.001), freshmen

 

2The standard.error of the estimate.is largest when a pro-

bability is at .50 and decreases as the probability deviates
from .50. The two probabilities closest to each other were
tested. These probabilities were closest to .50. Thus the
numerator (the difference between the probabilities) of the
Z test was at its minimum and the denominator (the standard
error of the estimate) was at its maximum. Any subsequent
test of a peer selecting another peer versus a non-peer-has
to produce a bigger Z.than the one reported.

57

or seniors. Likewise juniors passed the instructional
booklets to other juniors with a greater probability than
to either sophomores or seniors (z = 4.81, p<.001), or
freshmen. Again, with seniors it is found that seniors
passed the instructional booklets to other seniors with
a greater probability than to juniors (Z = 5.91, p<.001),
sOphomores and freshmen. Table 4 also indicates that the
.tendency for students to select a peer as a roommate is
stronger for freshmen than for seniors. Likewise the ten-
dency to select a peer to receive the instructional book-
let in the .small world situation is weaker for freshmen
than for seniors (Z = 2.03, p<.05). Thus as the results in
the criterion situation yield similar relationships to the
.results from the Small WOrld Method the Small World
Method may be said to be valid.

The results.of the present study also appear to be
externally valid or generalizable to other student pOpula-
tions on other.campuses. Lundberg and Beagley (1943)
studied student social.structures at.a small.eastern women's
college using a standard sociometric device; they asked:
"If it were possible for you to keep in touch with only
three students after.you leave college, which three would
you choose?" They discovered that all students, regardless
of class level, selected their peers to members Of other
classes. Lundberg, Hertzler and Dickson (1944) reported
similar attraction patterns among the residents of four

.women's dormitories at.a '1arge university.‘ Smucker (1946)

58

reported similar results on yet another campus as did Priest
and Sawyer (1967).

As a sample of the results from the present study
are in agreement with a validation criterion and also in
agreement with the results of several other studies per-
formed with different student populations, the Small World

Method appears to be a valid technique.

.The Social Distance Between Faculty,
-.Students and Administrators

The Small World Method yields two possible de-
pendent measures that reflect the structure of a social
,system. The first measure is the number Of intermediaries
required to link the starter person 'to the target per-
son. The second measure is the probability that a member
.Of a given social status will pass the instructional book-
.let to members of the same social status or to members of
different statuses. The first dependent measure described
tabove is the Operational definition Of (social distance as
defined in this.study.and will now be discussed.

Table 5 illustrates the mean number of intermediar-
ies required to link the Starter and Target for each of
,the nine starter-target combinations. As one can readily
Observe from Table 5, the hypotheses were confirmed.

Students are isolated from all segments of the aca-
.demic community including their fellow students. In fact,

the mean number of intermediaries required to link students

59

TABLE 5. The means and standard deviations Of the number Of

intermediaries required to link the
the target for each of the nine

starter to
starter-target

 

 

 

 

 

combinations
STARTER TARGET TYPE
TYPE Administrators Faculty Students
Administrators Yéo.93 iél.31 §é4.26
S =0.67 S =0.69 S =2.70
1 l l
N=55 N=55 N=45
Faculty Yél.64 §s2.23 225.55
Sl=0.87 Sl=l.34 Sl=3.72
N=54 N=53 N=40
Students ié3.69 ié4.26 224.11
Sl=2.04 Sl=2.36 Sl=2.62
N=46 N=40 N=33

 

 

 

 

as starters or targets

to all three segments of the uni-

versity is 4.4 with the lowest number of intermediaries (3.7)

occurring with student

starters

to administrator targets

and the largest number of intermediaries (5.6) occurring

with faculty starters

to student targets.

The mean num-

ber of intermediaries needed to link students, either as a

starter or target to the three.segments of the academic

community appears to approach the number of intermediaries

found by Milgram and.his associates to be required to link

any two members of the U. S. taken at random (approximately

5.0 intermediaries).

However, differences in the patterns

Of the non-completed chains within the studies performed by

. Milgram and his associates and the present study may change

the results of this comparison.

.the following section.

This will be discussed in

60

It was also hypothesized that students were not only
separated from faculty members, administrators, and their
fellow students on an absolute basis (as shown above) but
also require more intermediaries to complete their chains
with faculty members or administrators than do the other
two groups. As hypotheses were made, t-tests for matched
samples were used as a planned comparison method. Individ-
ual a was set at p < .01 to minimize the level of overall a.

The hypothesis was confirmed. Student starters re-
quired more intermediaries to reach either a faculty target
or administrator target than did administrator starters or
faculty starters. Student starters required 4.26 intermedi-
aries to reach a faculty target while administrator starters
required 1.31 intermediaries (t=7.30, df-39, p < .001) and
faculty starters required 2.22 intermediaries (t=4.72, df=39,
p < .001). Student starters required 3.69 intermediaries to
reach an administrator target while faculty starters required
1.63 intermediaries (t=9.30, df=45, p < .001) and adminis—
trator starters required .92 intermediaries (t=9.12, df=44,
p < .001).

Student targets required more intermediaries than
did faculty targets or administrator targets to be reached
by both faculty starters and administrator starters. Facul-
ty starters required 5.56 intermediaries to reach a student
. target in comparison to 2.22 intermediaries (t=5.49, df=38,
p < .001) to reach a faculty target or 1.63 intermediaries

(t=6.83, df=38, p < .001) to reach an administrator target.

61

Likewise, administrator starters required 4.26 intermedi-
aries to reach a student target in comparison to 1.31 in-
termediaries (t=6.93, df=44, p < .001) to reach a faculty
target or .92 intermediaries to reach an administrator
target (t=8.01, df=44, p < .001). In conclusion, the re-
sults indicate that it.requires more intermediaries to link
student starters and targets to both the faculty and ad-
ministrators than to link either faculty starters and tar-
gets Or administrator starters and.targets. Given this
evidence and the absolute size of the chains involving
students, it would appear that students.are isolated from
the rest of the academic community.

From Leavitt's experimentation with five-man groups,
it was hypothesized that administrators would have shorter
communication channels (a smaller number of intermediaries)
to faculty, students, and administrators than do either the
faculty or students._ This hypothesis was confirmed in part.
Administrator starters required 4.26 intermediaries to reach
student targets compared to 4.11 intermediaries for student
starters and 5.55 intermediaries for faculty starters.

Thus the administrator starters did.not.have statistically
significant shorter communication channels to student tar-
,gets in comparison.to faculty and student starters. Admin-
istrator starters did,.however, reach faculty targets and
administrator targets requiring fewer intermediaries than
did faculty or student starters. .Administrator starters

required 1.31 intermediaries and .92 intermediaries

62

.respectively to reach faculty targets and administrator
targets while faculty starters required 2.22 intermediaries
and 1.64 intermediaries (t=4.23, df=52, p < .001), (t=4.83,
df=42, p < .001). Administrator starters also required
fewer intermediaries than student starters to reach ad-
.ministrator targets or faculty targets (these comparisons
were made previously).

Administrator targets also required fewer intermedi-
aries than do faculty and student targets to be reached by
administrator and faculty starters. An administrator
starter requires .92 intermediaries to reach an adminis-
trator target, 1.31 intermediaries to reach a faculty
target (t=2.93, df=53, p < .005) and 4.26 intermediaries
to reach a student target (t=8.01, df=44, p < .001).
Likewise, faculty starters required 1.64 intermediaries
.to reach an administrator target in comparison to 2.22 in-
.termediaries to reach a faculty target (t=2.56, df=51,

p < .01) or 5.55 intermediaries to reach a student target
.(t=6.84, df-38, p < .001). Student starters did not reach
administrator targets with less intermediaries than faculty
or student targets.

When the length Of administrator.communication chan-
nels are compared to the communication channels of faculty
and students, it would appear that administrators have
.shorter communication channels to the faculty and other
.administrators. Administrators do not have shorter com-
munication channels to students in comparison to faculty

and students.

63

Leavitt (1958) suggests that communication is ex-
tremely important to the functioning of an organization:

From management's perspective, we can think of

... (an organization) ... as an elaborate set

of interconnected communication channels designed

to collect and collate, analyze, and sort out

information; also as a system for making deci-

sions, acting them out, getting feedback infor-

mation and correcting itself.

If Leavitt (1958) and others (Dorsey, 1957; Thayer,
1967) are correct as to the importance of communications in
organizations, it is interesting to ask if the lack of short
informal communication channels to students mirrors a fault
in the organizational structure of the university.

The Relative Size of the Groups and
Its Effect on Chain Lengths

One question that may be posed is whether the dif-
ferences in the chain lengths reflect relative differences
in the sizes of the three groups or possibly some normative
prohibition of contact between students, the faculty and
administration.

If the relative size of the groups is a major factor,
it would be expected that the target type drawn from the
largest population would show the longest chain; i.e., it
is harder to find a needle in a large haystack than in a
small one. Thus it would be expected that student targets
would require the longest chains followed by faculty targets
with administrator targets showing the shortest chains.

Looking at Table 5, it would appear that this is generally

 

3H. Leavitt, Managerial Psychology (Chicago: 1964), p. 300.

 

64

the case. However, viewing the starter-target combinations
in which students play the role of either a starter person
or target person it becomes Obvious that students are
approximately equidistant from other students, faculty and
administrators. This appears to be true regardless of
whether the student.is playing a starter or a target role.
This would tend to disconfirm 'size' as an important vari-
able as a student starter should have an easier time reach-
ing a faculty target than vice versa (a faculty target is a
needle in a smaller 'haystack' than a student target). Thus
to test the effect Of the relative size of the groups, com-
parisons should be made between the number of intermediaries
required to complete a student to faculty chain versus a
faculty to student chain, an administrator to student chain
versus a student to administrator.chain and finally a faculty
to administrator chain versus an administrator to faculty
chain. If size.is a major variable, chains that have the
largest group acting as a target should require more inter-
mediaries to complete the chain to that target. This hy-
pothesis was rejected as the administrator starter to stu-
dent target chain did not require significantly more inter-
mediaries to complete than did the student starter to
administrator target chain (t=1.12, df=43). The faculty
.starter to student target chain did not require signifi-
cantly more intermediaries to reach the target than did

the student starter to faculty target chain (t=l.83, df=38).

Administrator starters.required significantly fewer

65

intermediaries to reach faculty targets, the larger target
type, than did the faculty starters to administrator tar-
gets (t=2.18, df=52, p < .05) thus disconfirming the
hypothesis.

It would seem that the relative size of the groups
is not the major determiner of the length of the chains

as judged from the Observed chains.

.The Actual Chain Lengths, An
Estimate of the Parameters

One-problem that should be considered is that the
size of the observed chain lengths reported previously are
shorter than their true or actual size. Observed chain
(lengths shorter than.their actual size should be expected
as a long chain is less likely to reach its target than a
shorter chain. For example, if the probability that an
individual will cOOperate by passing on the instructional
booklet ‘is .90, the probability is .90 that a given book-
let will not be lost with the first pass, with two passes
the probability than an instructional booklet ‘is not lost
is (.90)2 or .81, with three passes the probability that
the booklet is not lost is (.90)3 or .73, etc. A target
person that is one intermediary distant from a starter
person will be more likely to receive the booklet than a
target person two intermediaries away. Thus the observed

chain lengths are biased toward shorter chains.

66

White (1966) designed a three-state probabilistic
model to handle this problem. White assumed that if an
intermediary knows the target person he will send it to
the target person with a fixed probability of 1.00. If
the intermediary does not know the target person he has
two choices. He can send the booklet on with a fixed prob-
ability Of a or he may discard the booklet with a probabil-
ity of 1-0. White defines four terms for use in the model.

Qi = the probability that an intermediary reached at

th

the 1 step of the chain knows the target person.

Ti = the total number of booklets received at the ith

step of the chain.

Ci = the number of booklets received by the target
at the ith step of the chain.

Thus the parameter a can be estimated for each value
of i or step in the chain by:
A Ni + l

a:

 

A
An estimate of the parameter Qi is Obtained by:
g = Ci + l
1 Ni

White stated that the two parameters behaved as ex-

pected when the Travers and Milgram, and Korte and Milgram

 

4The ith step equals the starter person pins the number
of intermediaries required to reach the 1 position.
For example, step 4 equals the starter person plus 3
intermediaries.

67

Qi remains at a negligible level through

data were viewed.
The para-

step 2 and then arose to a plateau by step 6.

meter 0. did not vary as a function Of the step in the chain.

Thus it appears reasonable to assume that the model would

predict chain lengths. To calculate the predicted chain

lengths as if all chains reached their assigned targets,
Target Person

White set: Pi = to the probability that the

is reached at step i. Then,-
i 1 5

= Q
0

Pi+l=Qig:0(l-Qj),fori>0,+Pi

Travers and Milgram used White's model to control

for the now-completed chain problem. The found that the

median chain length shifted upward from an observed five

intermediaries to a predicted seven. However, they con-

cluded that no substantial revision of conclusions drawn

from the raw data was required other than. the estimate of

the chain length parameter.
White's model, while meeting the needs of Travers

and Milgram, do not meet the requirements .Of the present

study for two reasons. First, it does not consider dif-

ferences between subject sub-pOpulations. The model has

target and lost, and only one

two absorbing states,
transitory state, that of being moved toward .the target
through an undifferentiated pOpulation. If .§_s belong to

591 (ti) = the number .of booklets completed at the ith
istep. To Obtain an average chain length (number of inter-
mediaries) ,.multiply the product of Pi (ti).by the step
number and divide by the total number of booklets started.

Then subtract one.

68

different sub-populations, the lumping together of these SS

is based on an unwarranted assumption, i.e., that the §
population is homogeneous. Secondly, the model does not
allow for differences in the cooperation rate (White's a)

as a function of the subject sub-population.
In order to control for the effects of missing data

in the present study, a first order Markov model was util-

 

 

.ized.6 One way to conceptualize the Markov model is to
Like the previous model, the

compare it to White' 3 model.
lost and target.

Markov model has two absorbing states,
However, instead of treating the pOpulation as homogeneous

with one transitory state, the Markov model takes into ac-
count the individual's university status (Freshman ,, Assist-

ant Professor, Administrator, etc-) -with .11 transitory

As was demonstrated previously, students did not

states.
COOperate as readily (91%) as the faculty (96%) or the ad-

The Markov model allows the rate of

ministration (97%).
COOperation to vary as .a function of the .status group to

which the individual belongs instead of using an average

:OOperation rate.
The basic assumption on which the Markov model is

seed is that the conditional probability of selecting a
instructional

.ven type Of individual to receive the

is independent (of a prior pass. For example, if

oklet

Snell, Finite Markov Chains, (Princeton,

. Kemeny & J.
ew Jersey, 1960).

 

69

the target is a faculty member and a freshman has the in-
structional booklet the probability that the freshman will
pass the booklet to another freshman, sophomore, Associate
Professor, etc. , is invariant regardless of from whom the
sender obtained the booklet. A corollary of this assump-
tion is that the probability of absorbing a chain into the
lost state, i.e., the rate of COOperation or its inverse
non-cooperation is invariant across the steps Of the chain.
One method Of testing the independence assumption of
a first order Markov model is to square the one-step condi-
tional probability matrices and compare them to the two step
conditional probability matrices. If the corresponding
squared one step and the two step matrices are comparable
the assumption Of independence is supported by the data.
The two step matrices indicate the probabilities that a
booklet will reach a particular status in two steps given

the status of the individual that presently holds the

booklet.7

Given the selection of an individual to receive the
booklet is independent of prior selections it would then
be expected that the squared one step matrix is equal to

the two step matrix for the following reason. The joint

 

7For example, the probabilities in the two step matrix can
answer the following question. If a freshman currently
holds the booklet, what is the probability that another
freshman holds the booklet two steps later. A freshman
can hold the booklet two steps later in a number of ways.
A freshman can pass the booklet to another freshman
who passes the booklet to another freshman. Or a freshman
can pass the booklet to a SOphomore who passes the booklet

to a freshman, etc.

70

The joint probability for two independent events is found

by multiplying the probability for the first event by the

second. The mechanics of squaring the matrix provides the

correct combinations Of multiplications and additions to

yield the two step matrix providing the independence as-

sumption of the Markov model holds. For example, in squar-

ing a matrix in the present study the two step probability
.of a freshman passing the booklet to another freshman is
computed, in part, byadding the product of .the probability
.of a freshman passing the booklet to a SOphomore multiplied
~by the probability of a SOphomore passing .the booklet back
to a freshman plus the probability of a freshman passing
the booklet to a junior multiplied by the probability of a
junior passing the booklet back to a freshman, etc.

The two step matrices correspond almost exactly to

.the squared one step matrices indicating that the data meet

the assumptionof the model. The largest deviation between

.any pair of corresponding cells within corresponding ma-

-trices was .153 out of 507 comparisons. .The mean deviation

was . 018. 8

 

Looking at the mean chain lengths the data .might appear to
some readers to be Non-Markovian as the length of the

chains started by students is much'longer than those
Hence it might

started by the faculty or administratOrs. _
he argues that the student‘starter effected more than just

the pass the starter was involved in. What .should be re-
.membered is that the data have to .be Markovian in nature
only within each of -the nine starter-targetcombinations

.for the data to be Markovian.

71

Before applying the model to generate the expected
chain lengths to compensate for the lost chains, the model
.can be tested in one other manner. The model can generate
chain lengths without compensating for the lost chains and
.these expected chains can be compared to the Observed chain
lengths. The chain lengths obtained from the model with-
out correcting for lost chains should.approximate the ob-
.served chain lengths.

Let E(CL1) = the expected chain lengths without
compensating for the lost chains

Let I = the identity matrix

Let A = the transition matrix, i.e., that part of
the matrix representing the probabilities
that a given status group will pass the
.booklets to members Of their own or differ-
ent statuses (no absorbing states)

Let B.= the probabilities of each.status group
passing the booklets to target or lost
states (the absorbing states)

Then

x = (I-A)"l B
Y = (I-A)"2 B

And:

_ Yi' _
E(CL1) ‘ YI%

The data again appear to fit the model as the E(CL1)
-approximates the observed chain length closely as Table 6
.indicates.

The model's predicted chain lengths deviated from
the observed chain lengths on the average of .34 intermedi-
aries for 12.2% error. This isva reasonably good fit of

the model to the data.

72

TABLE 6. The expected chain lengths [E(CL1)] without compensating
for missing chains in comparison to the observed chain

 

 

 

 

lengths.
Observed Per Cent of Error of E(CL )'
Starter-Target (E(CLl)“ Chain Observed Chain Lengths-E(CLLL
Types Lengths Observed Chain Length
Faculty-Faculty .l.72 2.22 22.5%
Student-Faculty 4.50 4.26 5.6%
Administrator-

Faculty 1.57 1.30 20.7%
Faculty-Student 4.81 5.55 13.3%
Student-Student 3.82 4.11 7.1%
Administrator-

Student 5.18 4.26 21.6%

Faculty-

Administrator 1.58 1.63 3.1%
Student-

Administrator 4.16 3.69 12.7%
Administrator-

Administrator 0.89 0.92 3.3%

 

Thus to generate chain lengths to compensate for non-
completed chains the probability Of each status group not
cooperating or losing the booklet.is eliminated and the

matrices are reconditionalized.

Let E(CLZ) = the expected chain lengths if all
chains reached their target
Let E(CLZ) = the row sums of (I-A)”l - 1

Table 7 shows the expected chain lengths if all

chains had reached their target.

73

TABLE 7. The mean number of intermediaries required to link
the starter to the target for each of the nine
starter-target combinations assuming all chains
reached their targets.

 

 

 

 

 

STARTER 7 TARGET TYPE

TYPE _h~Administrators Faculty Students
Administrators 1.00 1.79 8.56
Faculty _ 1.78, 1.95 8.16
Students ‘ , 4.88 5.36 6.96

 

The conclusions drawn from the nonéadjusted chain
lengths pertaining to the hypotheses do not-have to be ad-
justed. It still appears that the student is isolated from
all other segments of the university including other students.
Administrators still appear to have the shortest communica-
tion channels to all segments of the university except for
the students.

The relative sizes of the populations, however,-do
seem to play a role in determining chain lengths in which
.students are involved. .When the non-corrected chains of
student starters to faculty targets .were compared to
faculty starters to student targets no.differences in chain
lengths were found.. Likewise when administrator starters
.~to student targets were compared to student starters to ad-
ministrator targets, and administrator.starters to faculty
.targets were compared to faculty starters to administrator
targets differences in the predicted direction were not

found. Hence it was concluded that the relative sizes of

74

the groups did not play a role in determining chain length.
Viewing the corrected chain lengths [E(CL2)] this conclu-
.sion has to be adjusted. Student starters require fewer
intermediaries to reach faculty targets (5.36) than do
faculty starters to reach student targets (8.16). Student
starters require fewer intermediaries to reach administra-
tor targets (4.88) than do administrator starters to student
targets (8.56). As it requires more intermediaries to find
an individual in a large group than in a small one, the
relative sizes of the group does appear to influence the
lengths of the chains. The relative sizes of the adminis-
rtrator and faculty groups does p25 seem to determine the
length of chains between these two groups as the same number
.Of intermediaries are required to link a.faculty starter to
an administrator.target (1.78) as is required to link an

-administrator starter to a faculty target (1.79).

. The Permeability of.Facultyy Student and
.Administrator Role Groups

One indication of the permeability of a role group
.designated as A by members of role groups designated as B
.and C is the degree to.which members.of role groups B and C
.are used as intermediaries when the starter.and.target per-
sons both belong to role group A. For.example, how often
are undergraduates, graduate students, administrators and
.others (secretaries, faculty wives, etc.)-used as inter-
mediaries in a chain with a faculty starter to a faculty

target?

75

One reason for not selecting an intermediary that is
..not a member of the starter-target group is that they are
not known to the holder of the booklet. For example, within
a faculty starter to a faculty target chain students might
not be used as intermediaries because the faculty are not
acquainted with students on a personal basis. A second rea-
.son is that intermediaries from role groups other than the
starter's or target's might not be viewed as being efficient
at forwarding the booklet to the target. For example, with—
in an administrator starter to administrator targetrchain
students might not.be used as intermediaries because the
holder of the booklet feels that handing the booklet to a
.student will extend rather than shorten the length of the
chain. Both reasons lead to the interpretation that the
.role group is relatively impermeable with regard to out-

. group social contact--.For example, if the faculty does not
,know students on a personal basis so that the faculty cannot
.pass the booklets to students, then a logical conclusion is

that there is little social contact between faculty and stu-

,.dents, and thus few communication channels between faculty

.and students.. The faculty might not forward the booklet to
a student because the faculty feels that students would
reach the target less efficiently. This.leads to the in-

.terpretation that the faculty role group is relatively im-

. permeable to students as it was concluded that the student

does not have sufficient social contact within the faculty

role group to be an efficient next step.

76

The faculty role group appears to be the most per-
meable as faculty members passed the booklet to out-group
members 24.5% of the time. Administrators were the second
most permeable group with administrators passing the book-
let tO outegroup members 14.6% of the time (Z = 2.50,

p < .01). The student role group was significantly less
permeable (4.7%) than either the administrator role group
(Z = 3.41, p < .01) or the faculty role group (Z = 6.39,

p < .001). Thus it would appear that the faculty has the
greatest amount of social contact with out-groups followed

by administrators and then students.

A Faculty Starter to a Faculty Target

The major target variable selected by the faculty to
forward the booklet to a faculty target was departmental
affiliation. It would appear that some faculty members
view academic departments as being semi-isolated. The
faculty selected administrators and students to forward
the boOklet to targets in other departments. Thus it would
appear that some faculty members feel that administrators

and students would be more efficient than the faculty at

 

spanning the distances between departments. Administrators

 

appeared to be very efficient at reaching faculty in the
correct department while students did not appear to be. A
faculty member may pass the booklet to a student knowing
that the student takes classes in the target person's de-

partment. However, the student might not.know any faculty

77

in the target's department and thus the student attempts to
pass the booklet to another student that does. The chain
may be passed on by several students before reaching the
faculty again.

Faculty acquaintances with faculty appear to be de-
termined by their status or academic rank. A X2 = 59.29,
P<.001 (df = 9) was obtained when the frequency of passing
the instructional booklet by each academic rank to every
other academic rank was compared to a random model based
on the frequency of each academic rank in.the population of
faculty.9 Table 8 illustrates the probability of each
faculty status passing the instructional booklets to each
faculty status.

Most of the differences represented by the X2 value

2 = 43.46) were found in

reported in Table 8 (over 73%, x
the pattern of passes initiated by full Professors. It
would appear that full Professors tend to pass the instruc-
tional booklet to Instructors and Assistant Professors less
frequently and to full Professors more frequequently than
would be expected by chance. Korte and Milgram found that

there was a tendency for their booklets to be passed to

high status individuals as they have "maximum surveillance"

 

EEO compensate for teleological strain the x2 reported
was performed on the data from the total summary table
(Appendix B, Table 13). The same relationship was found
to hold using the data obtained only from the faculty

st rter to faculty target starter-target combination

(x = 41.36, df = 9, P < .001).

78

TABLE 8. The conditional probabilities that a faculty
member of a particular rank or status will send
the instructional booklet to another faculty
member of a particular rank or status

 

 

 

 

 

 

 

Receivers
Instruc- Assistant Associate Pro-

Senders tors Professors Professors fessors N
Instructors .122 .153 .061 .051 98
Assistant
Professors .065 .175 .115 .175 280
Associate
Professors .071 .137 (.196 .188 197
Professors .043 .091 _.l48 .274 419

 

 

 

 

 

 

of the domain. _This does not appear to be the case in re—
.gard to full Professors as only full Professors pass more
often to full Professors. Lower faculty ranks pass to full
Professors either at a chance level or less frequently than
chance.

There appears to be a relationship between the rank
of faculty members and the initiation of contacts with ad-
ministrators. Table 9 illustrates this relationship.

As Table 9 indicates, full Professors initiated more
contacts with administrators than do Assistant Professors
(Z = 4.58, p < .001) or Instructors (Z = 6.39, p < .001).
Associate Professors initiated more contact with adminis-
trators than did Assistant Professors (2.: 2.82, p < .004)
or Instructors (Z = 4.51, p < .001). Likewise, Assistant

Professors initiated more contacts with Administrators than

79

TABLE 9. Therelationship between faculty initiated con-
tact with administrators and faculty rank

 

 

 

RECEIVE RS
SENDERS , Administrators N
Instructors .071 98
Assistant Professors .147 280
Associate Professors .253 197
Professors .289 419

See Appendix B, Table 13

 

did Instructors (Z = 2.27, p < .03). There was no signifi-
cant difference between the amount Of contact that Professors

and Associate Professors initiated with administrators.

A Student Starter to a Faculty Target

The major target variable selected by the student to
forward the booklet to a faculty target was the target's
departmental affiliation. The student starter usually at-
tempted to forward the booklet to another student taking
courses in the target's department. This was accomplished
.with relatively few steps. The student taking courses in
the target's department typically was not.acquainted with
.any faculty in that department. He would then pass it to
a student majoring in the target's department. The booklet
.then is passed until a studentjs found who knows a faculty
.member affiliated with the same academic department as the

.target. Once the booklet reaches a faculty member of the

80

target person's department, it typically reaches the target
in one step. Clearly, the bottleneck occurs at the point
in the chain where students are attempting to find a stu-
dent 'gatekeeper' that is acquainted with a faculty member
in the target's department.

It would appear that seniors (.281) tend to be gate-
keepers more Often than do juniors (.119) (Z = 2.32,
p < .03), SOphomores (.091) (Z = 2.98, p < .003) or freshmen
(.086) (Z = 3.03, p < .003). There appears to be no differ-
ences in the frequency with which freshmen, SOphomores or
juniors act as gatekeepers to the faculty. Surprisingly,
graduate students (.415) did not act as gatekeepers signifi-
cantly more often than did seniors. The lack of statistical
significance in this.case is due to the small N caused by
-the infrequent use of graduate students as intermediaries.
The same relationships are observed from.the total summary
table. Graduate students contact the faculty a greater pro-
portion Of the time (.197) than do seniors (.101) (Z = 3.21,
p < .002), juniors (.066) (Z = 4.67, p < .001), SOphomores
(.035) (Z = 5.99), p < .001), or freshmen (.036) (Z = 5.86,
p < .001). Seniors contact the faculty more often than do
sophomores (Z = 3.71, p < .001) or freshmen (Z = 3.49, p <
.002). Other differences were not statistically significant.
The relationship between student class level and the initia—

tion of faculty contacts are shown in Table 10.

81

TABLE 10. Student class level and the initiation of
contact with the faculty

 

 

 

 

 

 

 

RECEIVERS
SENDERS Faculty N
Freshman .036 308
Sophomore _ .035 429
Junior .066 364
Senior .101 386
Graduate Students .197 244

 

See Appendix B, Table 13

 

Administrator Starter to a
Faculty Target

The major target variable selected by the administra-
tor to forward the booklet to a faculty target was the tar-
get's departmental affiliation. If the administrator
starter did not know the target person personally, he passed
the booklet to the chairman of the target person's depart-
ment or the dean of the target person's college. From
these positions the booklets were passed to the target in
relatively few steps.

In a previous section it was shown that the initia-
tion of contact with administrators by the faculty was re-
lated to faculty rank (see Table 9). It appears that this
relationship is symmetrical; an important characteristic of

.a sociometric or informal communications trace device.

82

Administrator initiated contact with the faculty also ap-
pears to be related to faculty rank. Table 11 portrays
this relationship.

TABLE 11. The relationship between administrator initiated
contact with the faculty and faculty rank.

 

 

 

 

RECEIVERS
Instruc- Assistant Associate Admin-
SENDERS tor Professor Professor Professor istrator
Admin-
istration .034 .059 .061 .146 .537
N 818 818 818 818 818

 

See Appendix B, Table 13

 

Administrators contact Professors more Often (.146)
than they do Associate Professors (.061) (Z = 5.67, p <
.001), Assistant Professors (.059) (Z = 5.80, p < .001), or
Instructors (.034) (Z = 8.00, p < .001). Administrators

contact Associate Professors more often than they do In-

structors (Z 2.46, p < .003) but not significantly more
Often than they do Assistant Professors (Z = .17). Like-
wise, Administrators contact Assistant Professors more
.Often than they do Instructors (Z = 2.50, p < .003). Not
surprisingly, administrators contact administrators more
Often (.537) than Professors (Z = 18.62, p < .001), Asso-

ciate Professors (Z = 25.05, p < .001), Assistant Profes-

sors (Z = 25.16, p < .001), and Instructors (Z = 26.47,

p < .001).1

83

0

 

10

The results reported were taken from the Total Summary
Table (Appendix B, Table 13). The same results are ob-
tained using the table representing only an administrator
starter to faculty target (Appendix B, Table 3) with the
following exception. Administrators did not pass the
instructional booklet to other administrators significantly
more often than to Professors. This could be a consequence
of teleological strain as the target was a faculty member
(1.31 intermediaries were required to link the starter to
the target). Thus the Total Summary Table was used to
counteract the strain. However, in this case even the use
of the Total Summary Table has a flaw as only one status
of administrator was designated in this study and one of
the nine tables summarized was one representing the admin-
istrator to administrator starter-target combination.

Thus the conditional probability of an administrator pas-
sing the booklet to an administrator in this case should
approximate 1.00 and will be weighted in the Total Summary
Table. Thus both the Faculty Target Summary Table (Ap-
pendix B, Table 10) and the Student Target Summary Table
(Appendix B, Table 11) was viewed as these do not include
the administrator to administrator starter—target combina-
tion pass. Both these tables show that administrators
pass the instructional booklet to other administrators
more often than to Professors. Caution is still recom-
mended, however, for two reasons: (1) Most passes went
across academic units and administrators have the broadest
surveillance. For example, if the Dean of the College of
Natural Science holds the booklet and the target is a
faculty member in a department of Social Science, probably
the best intermediary to pass the booklet to is either the
target's department chairman or the Dean of Social Science.
(2) Administrators are a heterogeneous group composed of
central administrators (Provost, Registrar, President,

. etc.), department chairmen, etc. While it is reasonable

to assume central administrators communicate mostly with
each other, it would seem reasonable that department
chairmen would communicate most often with their own
faculty.

84
A Faculty Starter to a Student Target

There was no major target variable selected by the
faculty to forward the booklet to a student target. Two
different strategies seemed to predominate. One strategy
winvolved passing the booklet to a faculty member in the
student's major academic department. The theory being a
faculty member that is involved in a student's education
should be acquainted with the student. This strategy ap-
peared to be more successful with students enrolled in
colleges such as Engineering that have small classes.

The second strategy used by the faculty was to pass the-
instructional booklet to a student immediately; reasoning
that students are better acquainted with students in com-
parison with the faculty. If the second strategy was used
after the booklet reached a student the chain was similar
in character to a student starter to a student target com-
bination. Of the chains started, 77.9% used one of the
two strategies. The two strategies were not significantly
different from each other in regard to completing the
chain to the target. The chains that were started using
the major strategy completed 61.7% while.the student
.strategy completed 50.0%.

There appears to be a relationship between faculty
.rank and the frequency of faculty initiated.contact with

students. Table 12 diaplays this relationship.

85

TABLE 12. The relationship between faculty initiated
contact with undergraduates and.facu1ty rank.

 

 

 

 

 

 

RECEIVERS
SENDERS Undergraduates N
Instructors V .306 98
Assistant Professors .154 280
Associate Professors .091, 197
Professors .050 419

 

See Appendix B, Table 13

 

Instructors have more contact with students (.306)
than do Assistant Professors (.154) (Z = 2.98, p < .003),
Associate Professors (.091) (Z = 4.22, p < .001), or Pro-
fessors (.050) (Z = 5.33, p < .001). Assistant Professors
have more contact with students than do Associate Professors
(Z = 2.10, p < .005) or Full Professors (Z = 4.33, p < .001).
Associate Professors.did not have significantly more con-
tact with students than Professors. Thus it is found that
the high status faculty have the smallest amount of contact

with students.11

Student Starter To a Student Target

As with the faculty starter to student target com—

bination, there appeared to be no predominant target

 

11Similar results were obtained with the probabilities from

only the faculty starter to student target combination
(Appendix B, Table 5).

86

variable selected to forward the booklet to a student tar-
get. Two different.strategies seemed to predominate. One
1strategy involved passing the booklet to a student who
shares the same academic major as the target. The theory
being a student who takes classes with the target has a
high probability of being acquainted with the target.

The second strategy was to try to reach the target
through his residence. The theory being a student should
be acquainted with his neighbors. The two strategies
(judged by the starter's strategy) accounted for 71.8% of
the passes that were started. The "residence” strategy
had completed 73.5% (N = 34) of the chains to the target,
while the major strategy completed just 44.4% (N = 27)

(Z = 2.39, p < .005). Thus it would appear that student
acquaintances with.other students are structured around
their residence.more so than their major.

Both strategies.p1aced the booklet either with a
student with the same academic major or the same dormitory
residence as the target. Usually this was accomplished
with the first pass, indicating that students.are acquainted
.with other students across a wide range of academic majors
.residing in many different locations.. At this point in
the chain the residence strategy showed to be superior.
The address in the dormitory gave the holder of the book—
let finer distinctions to proceed.by; i-e., which wing of
.the dormitory and which.floor the.target lived on. The

academic major gives no further distinctions. The student

87

using this strategy must turn to other criterion such as
year in school. Thus the booklet tends to wander from one
student with the same major to another until it is either
lost or finds the target. Of the booklets that reached

the target using the academic major strategy usually either
involved targets from a small pool of academic majors or
involved a switching to the resident strategy.

The matrix of conditional probabilities represent-
ing the class levels of undergraduates passing the in-
structional booklet to other class levels of undergraduates
shows the same relationship as was reported in Table 4 of
the validity section. Students tend to pass the booklet to
students of the same class level (peers) (see Appendix B,
Table 5). There is one exception to this correspondence
with Table 4. The tendency to pass the instructional book-
let to peers instead of constantly decreasing as one rises
from class level to class level decreases until the senior
year in which case it rises to approximately the
freshman level again.

Milgram (1967 and 1969) discovered a tendency to
pass the booklet to a member of the same sex. This same
tendency was found within a studentfstarter to student tar-
get chain as Table 13 illustrates.}

Males preferred to pass thexbooklet to other males
(.621) rather than to females (.214) (Z = 9.92, p < .001).
Likewise, females preferred to pass the booklet to females

(.651) rather than males (.192) (Z = 12.41, p < .001).

88

TABLE 13. The conditional probabilities of passing the
instructional booklet to a member of the same

 

 

 

 

sex
RECEIVERS

SENDERS Male Female N

Male .621 .214 243

Female .192 .651 281

 

There was no difference in the tendency of within sex trans-
mission of the booklet between males and females (Z = .714)

(males to males vs. females to females).

Administrator Starter to Student Target

Two different strategies were used in forwarding
80.6% of the chains originating with an administrator and
terminating with a student target. Administrators either
attempted to reach the student through faculty affiliated
with his academic major or attempted to reach the student
through his residence. The residence strategy again ap-
peared to be more effective than the academic major strategy
as 82.6% of the booklets forwarded by residence were com-
.pleted while just 57.8% of the booklets forwarded by major
were completed (Z = 2.48, p < .005). The academic major
strategy was approximately three times more widely used
than the residence strategy.

The residence strategy as used by the administrators
was a variant on the one used by the students as the ad-

ministrators used their informal acquaintances within the

89

formal structure. For example, an administrator using the
residence strategy might pass the booklet to the head of
the dormitory system who would then hand the booklet to
the administrator in charge of the particular dormitory

in which the target resided. The administrator in charge
of the dormitory would then pass the booklet to the stu-
dent acting as the Resident Assistant on the floor the

target resided.12

The Resident Assistant would then pass
the booklet to the target. Students, as one would expect,
did not use the formal structure in the same manner as did
the administrators as few students would be acquainted with
high level dormitory administrators. However, students oc-
casionally did make use of elements of the formal structure
as they tended to use the Resident Assistant structure in
.the dormitory. A student might pass the booklet to his own
Resident Assistant who would forward the booklet to the
Resident Assistant on the target's floor.and hence to the
target.

The academic major strategy was similar to those
previously used with the exception that some administrators

would pull the target's academic records in.order to deter-

mine what classes the target was currently taking. The

 

12A Resident Assistant is an undergraduate who lives within

a certain segment of rooms in the dormitory. He is given
room and board without cost in exchange for acting as a
friend to other residents of his segment in addition to
serving as a liaison between the administration of the
dormitory and the students residing in his segment.

90

administrator then made an effort to pass the booklet to a
member of the faculty he was acquainted with who was teach-
ing a class the target was enrolled in.

Freshmen appear to be gatekeepers to the students
for the administration less often than upperclassmen. Ad-
ministrators initiate less contact with freshmen than with
SOphomores, juniors (Z = 3.40, p < .001) and seniors (Z =

4.60, p < .001) as Table 14 illustrates.13

TABLE 14. The relationship between student class level and
administrator initiated contact

 

 

 

RECEIVERS
SENDERS Freshmen SOphomores Juniors Seniors N
Adminis-
trators .003 .020 .020 .026 818

See Appendix B, Table 13

 

Faculty Starter to Administrator Tagget

The.strategy taken by the faculty to reach an ad-
ministrator target is to pass the booklet upward in rank,
i.e., to an administrator they are acquainted with. The
hypothesis being that administrators are acquainted with

other administrators. A variant of this strategy was to

 

13The same relationships appear within just the administra-

tor starter to student target passes. This relationship
may be indigenous to Michigan State University as Resident
Assistants are composed of SOphomores and more heavily of
juniors and seniors. Also officials in the Office of Off
Campus Housing have contacts restricted to upperclassmen
as university policy requires students to be 21 years old
in order to reside off campus.

91

pass the booklet to a faculty member who should be acquainted
with the administrator; e.g., to a professor of history to
reach the chairman of the department of history.

As reported previously from the total summary table,
it would appear that tenured faculty (Professors and Asso-
ciate Professors) initiate more contact with administrators
than did the non-tenured faculty (Assistant Professors and
Instructors). Analyzing only the data within the faculty
starter to administrator target passes (Appendix B, Table
7), there is a non-significant trend in the above stated

direction.

Student Starter to an Administrator Target

The strategy taken by the student to reach an ad-
ministrator target was to pass the booklet to a relevant
student, i.e., to a student who is majoring in the academic
unit (department or college) the administrator target di-
rects. This student passed the booklet to a faculty member
the student was acquainted with. The faculty member then
tried to reach the administrator target. Another relevant
student was the Resident Assistant who was used both to
forward the booklet to a target within the dormitory manage-
ment system or to central administrators (the Provost,
Registrar, etc.).

In a previous section it was reported that adminis-
trators initiated fewer contacts with freshmen than with

upperclassmen (sOphomores, juniors and seniors). Again,

92

this relationship appears to be symmetrical as freshmen
initiate fewer contacts with administrators (.016) in com-
parison to upperclassmen (.036) (Z.= 2.25, p < .005).14 ’
Possibly the most important point to appreciate from ad-
ministrator student communications are the extremely small
conditional probabilities indicating little direct contact

between these two groups.

Administrator.Starter to an
Administrator Target

No strategy appeared to be.used or necessary in
these chains. The starters either knew the subject directly
,(the starter passed the booklet directly to the target in
33% of the cases) or knew who was acquainted with the tar-

get if they were not.

 

14The data used in the relationship reported above were
derived from the Total Summary Table. ’The.data taken
from only the student starter to administrator target
shows a non-significant trend in the same direction.
Freshmen initiate fewer contacts with administrators
(.055) than do upperclassmen (.119) (Z = 1.83, p < .07).

CHAPTER IV

DISCUSSION

A Partial Summary of the Results

,The complaints of activist students concerning the
social structure Of a large university have been substan-
tiated. The activists complained that there was no aca-
demic community. They complained that students are socially
separated from the faculty, administration, and other stu-
dentsi The present study adds substance to these complaints.
Given that the criterion used in the present study is
roughly equivalent to the criterion used by Milgram and
his associates,[it would appear that-students are separated
from elements of the academic community by as-many inter-
mediaries as separate two people chosen at random from the
population of the United States.

The activists complained that:

He (the student) loses contact with his professors

as they.turn more to research.and publishing and

away from teaching. His professors lose contact

with one another as they serve a discipline and

turn away from dialogue.[

This complaint also was substantiated.. The.resu1ts Of the

present study.show that the tenured faculty (Associate

93

94

Professors and Professors) initiated less contact with
students than did the non-tenured faculty. In fact, the
amount of contact initiated by a faculty member with stu-
.dents is inversely related to their academic rank.‘ The
faculty used out-group.(administrators and students) in-
termediaries to reach faculty targets more often than did
either students attempting to reach student targets or
administrators attempting to reach administrator targets.
This indicates that the faculty feel that administrators
and students may have more contacts across academic de-
partments than the faculty themselves have; i-e., there is
limited faculty contact across academic disciplines. It
also indicates that the faculty have more contacts with
out-group members than do either the students or adminis-
trators.

While the tenured faculty seem to have.less contacts
.with students than do the non-tenured faculty, the tenured
faculty appear to have more administrator contacts. Thus
it would seem that the faculty ranks endowed with the most
power have the greatest amount of contacts with adminis-
trators (or at least uses them most often)- With caution
it might also be stated that administrators communicate
with other administrators more frequently than with other
role groups.or statuses.

It also appears that administrators are the most
"central" role group in that they have shorter communica-

tion channels to other administrators and faculty than

95

does any other role group. Administrators did not have
appreciably shorter communication channels to students.

In fact, administrators had very little direct contact
with students, with freshmen having the least contact.
Student initiated contact with faculty was directly related

to the student's class level.

What do the Chain Lengths Indicate?

 

Previously, it was stated that chains involving stu—
.dents as either.starters or targets were approximately equal
in magnitude to a chain linking any two individuals selected
at random in the United-States. Milgram.and.his associates

‘dobserved a mean value of five intermediaries-and seven in-
termediaries adjusting for incomplete chains to link any two
persons selected at random from the pOpulation-Of the United
States. In the present study mean values approaching those
found by Milgram.and his associates were discovered in
chains involving students as starters or targets.

One question that may be posed is whether.a parameter
estimate of five or seven intermediaries has the same mean-
ing within a.population.in excess Of 200,000,000 persons
as it does in a pOpulation less than 50,000 persons? The
answer is that the estimate of the.social distance para-
,meter is the same in both cases but the implications for

the social structure may be quite different.

96

When a starter person in Nebraska attempts to reach
a target person in Massachusetts he searches through his
aset of acquaintances.for the individualswith.the best Oppor-
tunity of knowing the target. This individual has the
widest set of acquaintances that possibly.could be relevant
to the task. He has a broader selection of useful criterion
to be used to pass the booklet and hence the largest set of
relevant acquaintances. For example, the starter can sel-
ect a friend that resides in the target's home community.
The starter's.friend residing in the.targetls.community no
longer considers residing in the targetis home town as a
.relevant criterion.but rather as a limiting factor. It no
longer is profitable just to pass.the.booklet.to anyone in
the town as the booklet is present in-the town and further
selections on this basis amount tO.1ittle-more than random
passing within the community. Residing.in the target's com-
munity is now a.limiting factor as the search.for the next
intermediary within a set of acquaintances will be limited
to those residing within the community, i-e., an intermediary
is unlikely to select an acquaintance residing outside the
.community and automatically eliminates them from considera-
tion. Travers.and.Milgram give some qualitative support
for this position:

Chains which converge on the.target principally

by using geographic information reach his home

town or the surrounding area readily, but once

there Often circulate before entering the target's
circle of acquaintances.

 

1J. Travers and S. Milgram, “An Experimental Study of The

Small WOrld Problem." Sociometgy, 1969, 32, p. 432.

97

The university chains are started in.an analogous
position to the residence chains of Traver's and Milgram's
after they have first entered Sharon, Massachusetts. The
situations are analogous in that the university chains are
started within the community the target person holds mem-
bership and are generally not sent beyond.the confines Of
the community byra starter or intermediary.

Given that.the two situations are approximately
analogous: (1) why should faculty starters require 5.55
intermediaries to reach a student target.and.(2) why should
the length of corrected chains thattinvolve students as
starters or targets approach or surpass Traver's and Mil-
gram's across country corrected chain length (seven inter-
mediaries)?

The approximate equivalence of the chain lengths
reported berilgram.and.his associates and the university
chain lengths (involving students) is.not as surprising as
one might first think. Travers and.Milgram.reported a mean
of 6.1 intermediaries for chains that approached the target
through his home town. Travers and.Milgram also reported
that chains reach.the.home town area readily, but once
there circulate before reaching the target. Assuming
”readily" means between one and two intermediaries, it
would appear that if the booklet was started in the target's
home town the chain lengths required to link the starter to
the target approximates the Observed university chain

lengths involving students. Hence it is not extremely

98

surprising that the two sets of chain lengths can be com-
parable. This, however, is not the entire answer for several
reasons. One, the Observed chain lengths were compared.
Thus when adjustments are made for lost chains the results
might change. Second, the faculty starter to student tar-
get chains (observed) cannot be explained in this manner as
the chain is too long. Three, the faculty starter to student
target and the administrator starter to student target chains
are both longer than those reported by Travers and Milgram
after both sets of chains are adjusted for the lost chains.
One reason for long chain lengths within the universi-
ty community may be a greater amount of inbreeding of ac-
quaintanceship networks in comparison to other communities;
i.e., individuals within a university community may have
more mutual acquaintances. The greater the inbreeding of
acquaintances the longer the chains. For example, assume
that every person within the university has only five ac-
quaintances. Hence the starter has.five acquaintances which
we will assume does not include the target. The starter
passes the booklet to Al, one of the starter's acquaintances.
If the starter and A1 share four of the same acquaintances
(counting themselves), Al has only one acquaintance left
that could be the target as it was determined previously
that the starter was not acquainted with the target. If
Al has two acquaintances not known to the starter, A1 has
twice the chance of knowing the target.and.thus terminating

the chain by passing the booklet to the target. Hence the

99

more inbreeding Of acquaintances the longer one would ex-
pect the chains to be.

Another reason for long chain lengths involving stu-
dents as either starters or targets is the extremely rapid
turnover of the undergraduate pOpulation. The academic
plan for the Bachelor's degree calls for approximately a
25% turnover of undergraduates per year due to graduation
and new admissions. Hence within four years time there
should be approximately a 100% turnover of undergraduates.
There is additional student turnovers due to withdrawals
and transfers. The faculty also is mobile with new faculty
arriving and others leaving for different jobs. It has
been shown that upper classmen have the most.contact with
faculty and administrators. Thus the students who are most
.integrated into the academic community are.the ones that
leave. These turnovers cause disruption of.communication
andacquaintanceship networks and hence longer chains.2

As the student role group was shown.to be relatively
impermeable to contact by faculty and administrators and
the chains started by these groups to students were rela-
tively long, a more efficient strategy to use to reach

students may be an indirect approach.3 SHence a direct

 

2 . . . .

As shown preViously, the relative size of groups is one
determinant of chain lengths. It is not discussed in this
context as the question being answered is.how can chain
lengths be approximately equivalent within a pOpulation of
200,000,000 and.one of less than 50,000.

3It should be remembered that the term.impermeable refers to
a lack of efficient contacts by one role group within a se-
cond role group. One role group may still be impermeable to
another while.having contacts with the second.role group. An
anology of this situation is that water can come in contact
with skin but still does not penetrate the skin.

100

confrontation with the barrier separating students from
other groups is avoided. For example, in one of the short-
est faculty starter to student target chains, the starter
sent the booklet to a friend in the target's home town.

The town was an extremely small farming community. The
first intermediary passed the chain to the target's mother
who then sent the chain to the target. Hence this indirect
approach used few intermediaries. Possibly the most rele-
vant criterion to be used when one is faced.with a choice of
variables to forward the booklet is the size of the member-
ship group. The smaller the group the more.efficient the
chain.

Of course another pgssible reason exists for the
equivalence of the university chain lengths and those chain
lengths obtained by Milgram and his colleagues. It is pos-
sible that the criterion used by Milgram and his colleagues
is not equivalent to the one used in this.study as the
author has assumed. VHence the comparison would not be valid.
It should be remembered that this comparison is not necessary
to the forthcoming discussion in the section entitled the
"Centrality and the Degree of Contentment With the Social

Structure."

The Social Structure and the
Diffusion of an Idea

Coleman, et. al., found that the simple spreading of

information about a certain idea was not itself sufficient

101

to cause the adOption of the idea. What was also needed
were discussions about the pros and cons of the idea.

Within the result section it was shown that some
statuses and roles communicate with certain other statuses
and roles more so than with others. When communications
between one role or status and a second role group occurs
less frequently than between the second role group and a
third role group, it should be expected that an idea will
spread less quickly from the first role or status to the
third. For example, an idea should spread to undergraduates
more quickly from non-tenured faculty than from tenured
faculty as non-tenured.faculty have more contacts with
undergraduates. The summary section of the discussion is
a summary of these results. Assuming that each status
.group is equally susceptible to the idea, predictions can
be made by this method concerning how an idea would spread
through the university. Appendix B, Table 13 is the best
description of how an.idea would spread through the uni-
versity. As these predictions follow directly from the
conditional probabilities already presented it will not
be repeated. To make meaningful predictions, however,
the results of this study should be weighted with the
susceptibility levels of each status group.

It has been noted by the author that it appears to
be the non-tenured faculty rather than the tenured faculty
who are most sympathetic to student activists. It is not

suggested that the greater amount of contact with

102

.undergraduates by non-tenured faculty made them more sym-
pathetic. Variables such as the shorter interval between
the present and their own years of schooling and their
proximity in age to students in comparison to the tenured
faculty are more powerful explanations. However, it seems
reasonable that the greater amount of contact with under-
graduates by the sympathetic non-tenured faculty might be
reinforcing for both those sympathetic faculty members and
activist students. It is also interesting to ask if the
tenured faculty and administrators' greater amount of con-

tact might also reinforce a particular set of their ideas.

Centrality and.the Degree of Contentment
With the SOOia Structure

As hypothesized, administrators were shown to have
the shortest communication channels to other administrators
and the faculty. Administrators did not, however, have the
shortest communication channels to students. .Students, on
the other hand, had the longest communication channels to
the faculty and administrators but not to.other students
(based on the corrected chains). Hence, the administrators
may be said to hold the most central position within the
university, the students the most peripheral position and
the faculty somewhere in between the administrators and
students.

Leavitt (1951), experimenting with five-man groups,

found some indication that the pattern with the most central

103

position was the most efficient. Efficiency in this case
was measured by the single fastest solution of a problem.
Leavitt also found, however, that his gs holding the most
peripheral positions enjoyed their job much 1223 than did
gs holding the most central position who stated that they
enjoyed their job (see Bavelas, 1950). Leavitt suggests

that centrality affects behavior by the ”limits that cen-

 

trality imposes upon independent behavior." Thus it is
not surprising that a segment of the student population is
dissatisfied with the university.

Obviously, not all students are activists. The
question then might be asked, why are those activist stu-
dents activist students? One answer is that these students
have a different self image of themselves in comparison to
non-activists. The activists have been raised to both act
and see themselves as independent beings. Flacks (1967)
investigated the child rearing practices of both the par-
ents of activists and non-activists. He found that activists
rate their parents as "milder,“ "more lenient," and "less
severe“ than do non-activists. Flacks also asked the par-
ents of activists and non-activists how they would respond
in a hypothetical situation. For example, the parents were
asked "what they would do if their son (daughter) decided
to drOp out Of school and doesn't know what he really wants
to do." Or the parents were asked what would you do if your
"child was living with a member of the opposite sex?" The

parents' answers to these questions were then rated on a

104

parental intervention continuum. The fathers of activists
were reported to be much less interventionist than fathers

of non-activists.4

Thus the parents of activists train their
children to be independent and do not generally intervene in
the decisions of their children. The activists then enters
the university where the administration might intervene by
setting living group regulations, etc.

Thus activism on the part of students is seen as

being caused by an interaction of the social structure of

the university and the personality Of the activist student.

Modifications of the Social Structure
to Produce Shorter.Chain Lengths

 

One question that may be asked is, how can the so-
cial distance between students and the remainder of the
academic community be reduced? In other words, how can the
chain lengths be made shorter than they.are currently?

Possibly the most practical variable to manipulate

is that of prOpinquity. Newcomb (1956) states that the

 

4Flacks also reports that activists and their parents have

different values than do non-activists and their parents.
Flacks reports that activists are higher on “romanticism,"
"intellectualism" and "humanitarianism" than non-activists.

For those readers interested in pursuing the differences
between activists and non-activists, see: .E. Sampson,.
"Stirring Out Of Apathy: Student Activism and the Decade
of Protest," The Journal of Social Issues, 1967, 23, 3,
pp. 1-137.

105

shorter the physical distance between two individuals, the
more probable the two individuals will interact and the
more likely they will be attracted to each other. As
stated perviously, the greater the probability that two
peOple will interact or are acquainted with each other,
the shorter the chain lengths.

One possible solution to increase prOpinquity is to
reduce the size of all segments of the university. With
fewer students, faculty and administrators housed on less
acreage it would be more probable that individuals within
the university would know each other. Another similar
possibility is small self-sufficient colleges within the

university.5

Each college could have its own student body,
faculty and administration. Thus it is possible for all
three role groups to have a substantial amount of contact
with each other as a result of small role group sizes.
The greater the contacts across role groups and the smaller
the size of the role groups, the smaller the possibility of
inbreeding of acquaintances and the shorter the chains
across groups and within groups.

Another possible solution is to build recreational
centers and organize activities that will be engaged in by
all segments of the academic community. Instead of building

student unions and faculty club houses a common building

where students, faculty and administrators either engage

 

5Michigan State University has established three experi-

mental colleges similar in nature to the brief description
presented here.

106

in activities together or proximate to each other should
produce more contacts between individuals across role
groups. Instead of building student dormitories and faculty
and administrative offices, buildings housing all units in
common buildings will increase contacts across role groups.
The more contacts an individual has across role groups, the
less inbreeding of acquaintance networks and thus the shor-

ter the chains.

Summary

Within the literature produced by several student
movements some very specific complaints pertaining to the
social structure of the university appear. Two student
movements on two different campuses were viewed with re-
gard to complaints about the social structure of the uni-
versity. The activist students complained that they were
socially separated from the faculty, from the administra-
tors and from other students.

It was hypothesized that students would be connected
to other students, faculty members and administrators by
the longest informal communication channels. On the basis
of Leavitt's (1958) study, it was also hypothesized that
administrators would have the shortest informal communica-
tion channels to other administrators, faculty and students.

The technique used in the present study to measure

the length of informal communication channels was first

107

used by Milgram (1967). Milgram called the technique the
"Small World Method." Using the Small World Method, two
sets of individuals are selected. One set of individuals
is designated the starter persons, a second set of indi-
viduals is designated the target persons. A starter per-
son is asked to try to pass an instructional booklet to

the target person by only passing the booklet to people
they know according to a certain criterion (e.g., knowing
the person on a first name basis, etc.). If the starter
person does not know the target person according to the
criterion the starter person is then instructed to pass the
booklet to an acquaintance he does know according to the
criterion, who has a better chance of being acquainted with
the target person. The number and characteristics of the
intermediary persons between the starter and target serve
as the dependent variables.

Student, faculty and administrators were randomly
selected to serve as starter and target persons from the
population of a large university. Each starter person was
asked to start two booklets to student targets, two book-
lets to faculty targets and two booklets to administrator
targets. Each target person was asked to receive a possi-
ble two booklets from student starters, two booklets from
faculty starters and two booklets from administrator
starters. The starter and target persons were randomly

paired.

108

The results confirmed the hypotheses. Students had
the longest informal communication channels while the ad-
ministrators had the shortest communication channels. Thus,
in Leavitt's terminology administrators may be said to be
the most central group while the students are the most
peripheral group within the university. The results were
discussed in terms of the peripherality of the students and
their contentment with the social structure of the university.
Suggestions were made for the modification of the social

structure of the university.

BIBLIOGRAPHY

BIBLIOGRAPHY

American Civil Liberties Union. "The Campus and The Con-
stitution," in H. Draper (ed.) Berkeley: The New
Student Revolt. New York: Groves Press, Inc., 1965.

 

Bavelas, A. "Communication Patterns in Task-Oriented
Groups," in D. Cartwright and A. Zanders, Group
Dynamics: Research and Theory. New York: Harper
and Row Publishers, 1968, 503-511.

 

Campbell, D., and Stanley, J. Experimental and Quasi-
Experimental Designs for Research. Chicago: Rand
McNally and Company, 1966.

Coleman, J., Katz, E., and Menzel, H. Medical Innovations:
A Diffusion Study. Indianapolis: Bobbs-Merrill,
1966.

Cronbach, L., and Meehl, P. "Construct Validity in Psy-
chological Tests," in D. Jackson and S. Messick (eds.)
Problems in Human Assessment. New York: McGraw—Hill
Book Company, 1967, 57-77}

Dorsey, J. "A Communication Model for Administration,"
Administrative Science Quarterly, 1957, 2, 307-324.

Dowty, S. "The American University: Is Democracy Possible?"
Organon, East Lansing, Michigan: Mimeographed, 2,
l-24.

Draper, H. Berkeley: The New Student Revolt. New York:
Groves Press, Inc., 1965.

Editors of the California Monthwy. "Chronology of Events:
Three Months of Cris1s, in S. Lipset and S. Wolin,
The Berkeley Student Revolt: Facts and Interpreta-
tions. Garden City, New York: Doubleday, I965,
99-199.

 

Flacks, R. “The Liberated Generation: An Exploration of the
Roots of Student Protest," in E. Sampson (ed.) Stir-
rings Out of Apathy: Student Activism and the Decade
of Protest. The Journal of Social Issues, 1967, 23,
3, 52-75.

110

111

The Free Speech Movement, We Want a University, in S.
"‘“’ Lipset and S. Wolin, The Berkele Student Revolt:
Facts and Interpretations. Garden City, New York:
Doubleday, 1965, 2094216.

 

Garson, M. "The Regents," in H. Draper (ed.) Berkeley:
The New Student Revolt. New York: Groves Press,
Inc., 1965, 215-221.

 

 

Gurevitch, M. The Social Structure of Acquaintanceship
Networks. Unpublished Doctoral Dissertation.
CamEridge: Massachusetts Institute of Technology,
1961.

Heirich, M. and Kaplan, S. "Yesterday's Discord," in S.
Lipset and S. Wolin, The Berkeley Student Revolt:
Facts and Interpretations. Garden City, New York:
DoubIedEy, 1965, 10-37}

 

 

Jacobs, P. and Landau, S.. The New Radicals, A Report With
Documents. New York: Vintage Books, 1966.

 

Kemeny, J. G. and Snell, J. L. Finite Markov Chains.
Princeton, New Jersey: Van Nostrand, 1960.

Korte, C. and Milgram, S. "Acquaintance Networks Between
Racial Groups: Application of the Small World Method,"
Sociometry (in Press).

 

Lundberg, G. and Beazley, V. "Consciousness of Kind in a
College Population," Sociometry, 1948, 11, 59-74.

 

Lundberg, G., Hertzler, V., Dickson, L. "Attraction Pat—
terns In A University," Sociometry, 1949, 12, 158-
169.

 

Leavitt, H. "Some Effects of Certain Communication Patterns
on Group Performance,‘I The Journal of Abnormal Social
Psychology, 1951, 46, 38- 50.

 

 

Leavitt, H. Managerial Psychology. Chicago: University of
Chicago Press, 1964.

 

Milgram, S. "The Small World Problem," Psychology Today,
1967, 1, 61-67.

Milgram, S. "Interdisciplinary Thinking and The Small World
Problem." In M. Sherif and C. Sherif (eds.) Inter-
disciplinary Relationships in the Social Sciences.
Chicago: Aldine Publishing Company, 1969, 103-120.

 

112

Newcomb, T. "The Prediction of Interpersonal Attraction,"
The American Psychologist, 1956, 11, 575-587.

 

Peterson, R. "The Scope of Organized Student Protest in
1964-65." Princeton, New Jersey: Educational Test-
ing Service, 1966.

Priest, R. F. and Sawyer, J. “Proximity and Peership:
Bases of Balance in Interpersonal Attraction,‘I
American Journal of Sociology, 1967, 72, 6, 633-649.

Rogers, E. Diffusion of Innovations. New York: The Free
Press, 1962.

 

Rogers, E. and Shoemaker, F. Communication of Innovations:
A Cross-Cultural Approach. New York: Free Press of—
Glencoe, in press.

 

 

Rosenfield, G. "Generational Revolt and The Free Speech
Movement," in P. Jacobs and S. Landau (eds.) The
New Radicals: A Report With Documents. New York:
Vintage BoOks, 1966, 2194220.

Sampson, E. (ed.) "Stirrings Out of Apathy: Student Acti-
vism and The Decade of Protest," The Journal of
Social Issues, 1967, 23, 3, 1—137.

 

Schiff, P. (ed.) Logos. East Lansing, Michigan: Mimeo-
“x/‘l graphed, I, 1965 o

Smucker, O. "Sociographic Study of Friendship Patterns on
a College Campus," Abstract of Doctoral Dissertation,

1946' NO. 500

State News (East Lansing, Michigan) October 23, 1962 -
June 20, 1966.

 

Thayer, L. "Communication and Organization Theory." In
F. Dance (ed.) Human Communication Theory. New
York: Holt Rinehart and Winston, Inc., 1967, 70-115.

Travers, J. and Milgram, S. "An Experimental Study of The
Small World Problem.” Sociometry, 1969, 32, 425-443.

 

Tumin, M. Social Stratification: The Forms and Functions
of Inequality. Englewood Cliffs, New Jersey:
Prentice-Hall, Inc., 1967.

 

White, H. "Search Parameters for Milgram's Small World."
Cambridge: Harvard University, Undated Mimeograph.

 

A P P E N D I X A

THE INSTRUMENT

114

 

.0 omen :0 62:;

0530.50 of oc§0__£ >3 00200 .090. 05
0. 0.03300 as: .695: o. 2 >33 of .0
000. 2:. 4.62... 25 c. 568 .092 05 no
028 2 00060 no; 0...; 363668 0600
-000 0:. .0 .3606 0 .0 120515600 .0
03606 3.301. .6002: 0.0.. 0:0 $2000
.060: 05 0c; :3 30> 0000 :8: 05 :0
3.0 000069. 2 10.0.03 0.21 060.03

.51 5 .63: 2. so :3 :9 .91. 88,1 «I
30> 0. :0 “01002 $5 069.00 >5 >03... ME.
0020 mo... 01 ._ 0000 cc .281 05 c. 3000
-00 060: ....I 0806.330 :0 60: 30> 2
0600 no... 0.03300 25 .05 0020: :i 30>

.20; 50> to. 07.0 03
£0.03! 50 2 60:006. 20 £023 .30:
-330 89.: .0380 8. «3.636600 0.600
-000 50 0.. 080.0 ..0 0000 30: 3.3 nmuucoc
6.03000 0:0 6:05 .0 {020: 50> >60
05...: 08.00 .9: 305. o. .00 30> 0300 .10...
0 .0 50 09.0.0 203 b.5663 0.600000
9: *0 .3606 0 .0 ~60: of : .3068
06.80.. .3“ 9: :0 82:00 .203 do 22
-0: or. 200.32g 2 0060.60 >02“ 60:00
156600 .0336: :0 c. 0.2 50> 000: 03

52.03..
mZOr—(BZSSZOU m1...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

9
3
m.
m.
_ .
o.

o

m

n

o

m

v

m

N

_

.83: ”find £38 1522
.20 18.0.

3.59.
.323 128m

05 c. 022. m. 0.9 0.600000 mE 0:0 8:03
95 30> .50. 0...) 00200 05 *0 060: 9:.

.323 00.03 00:25 05 .5 .0108 0.... 0:.
-0522 3 9: 3.99.5 “2. no ...; 8.1

.9: $00... 0: 50> 2 0.03300 2.: 68 me;
0223 <

23.5.: 02.sz 63
£322: 55 23:10.:
35.3 02.253
520: "20:62:23.8

115

 

v
0.0 ._0_.0.06 02300

900030 :03 30253.0 00:00:00 0.03

000.000.5668 03030.0 .105 :0 t .00606
2.302 0 0. .0380 0;. $00 .0: 030;...
5.0.5.6600 :0 .3300: 00.00. 60:00 .0_
.66 .0 00500005 9.00.0 00500000 0>0£
6.; 1:3 30.60.0600 50> 0 630.20: :0 0.
.0380 mi. .300 >00. 30> :36... 000000.
0: 00.300 0;. .0. .0029: 00.300 0000.0
.200. 00:00:00 .0208 50> __0 0. .00606
3.300. 0 0. 0 $00 .00 0.30.0 :0) 0.0
8300 600:. 0:0 _000_ .0 0.03690 060m

.363 ”_O m02<IUXu .50
.201 2< 20 >~E<U 2<U 30> 20107
IFS? m._<Do_>_oZ_ 0.». 00 >F.mmw>_23
mI._. 0._. 02.5.3.3 0&0... >2< .505.
mmDUmE 0... m..m< mm... 30> §0I>> I..._>>
._<Do_>_oz_ 2< Humdm 30> 00 02:.
mIk 0.. 5%.me 04.30.22 .10 wmd
02_I._.>Z<. m0 _ 290000 m0... 030
-Om «30> mEmDmOmd 0(20mmma ammmDU
-m_o m><I 30> .2010? IFS? .2503
._02_ 2< 0.. E0¥00m m_I.r mm<0 Am.
R><000 30> $2 0901 Om: ..v_._<._.
._ ._ < <4 m: 260200 >._20 m20_h<m
-mm>200 ~50> $01.5 ._<Dn__>_oz_ 2<
0... Fwd—00m 91.0 mm<a ._.02 00 AN.
vm00<mkm_z_<<o< m0 mmmimé b.1501}.
Fzmozhm < m< 53.. ...02 .2033:
< m< 30> 3020. .535. 0._. E0x00m
m_I.—. mm<a 30> 203mm mIk A:
05 00.000
0:... 02000. 0. 63030:. 00: 0;. .0 0000200
0:. :. 0032.2 00 050:0 .05 00.3. 000:.
.0020.— .092 0.: 2 .0302. 0... as
-0000 .0 0002.0 .020... 0 00.. 0...: 0.00.. .0602
-0_ 00 00 300.. 30> .030300. 00 2 .0300.—
25 32. .0000“ .>:00._0 000000 .092 0.:
.02000 2 >2 .00 00 00.: .0020: .092 0...
0:) 0020200000 .0220. 00 22 00.200 >_
$00.00... .0000 000.. 00:10.. .20.. .2220.
00 00 000.00 .092 0.: :00... .00 00 00> =

m

.02....200322
00.00.010.030:

A 1|

 

20m! Bug—(b

 

116

0.00.. 0. 30> 0.00 0; >02

.:0m.00 .00.0. 0... >0 00200
-0. m. 0.000300 0.... ...:3 3000.0 0... .000
-0. :0... ...3 0.0000000 0...... m0>.000. 9.3 :8
-.00 0...... >03 2. :0 0.000300 0... ..0.m 0. m.
.86 2:300 .003 .30 08.00 .00.0. 0:. o.
.00 0. 0:0... .30> 0:0>00 m00.m .900. 00.0.
>06 .. “:0200 .005. 0;. 0.0 8.0.0 .908
060m 0... :. 0>06 >.00>.00:00 0.0.6 30:0.
0.03000 .30> 0:060 003 0.0690 .0“. :0.
-.00 .00.0. 0... 0.030. 0.0.000 0.... 00:0>00
:00 003 08.00 0 m00:0.:.03.000 0:0 0:0...
.3o> 0:060 >...:00. 0. m. 00:0..0:0 .00.
0...... 0200.0 .00.. 0.... :. 0:306 .. ..0.m
:00 30> .000. .0 .30 ._0.3.0: 0. m0: :8
-.00 .005. 0... ......s 00.603000 .... 0:3 0:0
>00 >55. .0: 00 30> .00. >06 30> £030...
.0... :0 00002603000 0:0 0:0... .0
£000 0 >.:0 0:03 :0600 .006. 0... 0.030.
0.000300 0...... 056 0. m. 6.0 0... 00060600

 

\O

.2029“.
hug—(h m1... 302.. Oh 30> 25....
>33...— uxOI m. 0...? .mmZ...mo.30
m1... 0... 02.5.0090 .3020. 30> 20m
-..mm < 0... £00033. 2...... 325.. o<mhm
-2. .>._._.Uu.:o 2.... ._.U<._.ZOU 0... >00
.002 on .mwzfimasw 2.... >0 03.30
-030 m<. .20: .2203: < 20 2023..
.593... mi... 302.. ...02 00 30> ..

 

006.0050
0;. :. 00.m0003m m0 .060 0000 30:. 0:0
:0300 .006. 0.... .06 >.m30.>0.0 0.6.. 30>

._ >.:0 mi. 00 ....m... 2.... 0.. >.:0:
-.o .5039: 21.0 02.1. 32.60.30
5.... Z. amh<hm 93 .298 302093.. <
20 20mm: taxflr ml... 302.. 30> u:

. 08.00 .005. 0... 0.030.
m0>06 .. m0 .02000 0... .0 0.0.00.0 0... .0
0.00.. 0000. 0. 3 30:0 .. ..:0..006. >.0>
m. 0.00.80 0.: 00000: W. 060... 02 .mU_>
-..mm 02.2.33.— m.... 0... .2 2.5.5.. oz<
...30 .... ...—... #93500 2...... “.0 ¥U<m
mi... 20.... 9.00.500 mZO 105—ho

.60.. 0600 .. 0...: >65. ...3 0.000300 ......
m0>.000. 0...) :8.00 .x0: 0... .0... 8 ._ ~03.

20 5.50.. 5.: 0... $402 «30> oo<

>035 2...... 2. ...—<0 3.0... 0.. 30:

 

 

117

m3a§<u
ZO mmexx/>2< MOOmmme SIP
#2625: Oh 3.52 m3a2<8 HUI/mum
$0meme m3a<<<u wI._. mm: >32 30>

 

 

33.5w @5584
bcozogm J twnom

293:3 23>

6233.2... state a. ...... =8» .23..
a 5...? .3332. .2. 35...... :8:

.635 m5 *0 2.32 .2: 9.: *o :82 o
9.39 :3) m: 9 You Son 9: 3.29 vco
€37. m3. 5 8.8058 9..) coax .696

118

 

 

 

 

 

 

 

 

 

 

 

 

 

E“. ‘7 - i
i N? 1101 2 ;
First Class ’
Permit '
No. 188
East Lansing
Michigan
BUSINESS REPLY CARD
Na Postage Stamp Necessary it Mailed In the United States
Postage will be paid by
COMMUNICATIONS PROJECT
Learning Service 1
I7 Morrill Hall 1
Michigan State University ‘
East Lansing, Michigan 48823 i
My‘NomC' Pimv f!” in m» f5.»l,'nwm;j infcrmntu 2. ck, : 3!.

 

{worsen tn ‘.\I'i:,iT> mu arr: swatisrg tint fold-r

[My Address (Campus)

 

HIS (HON NUITM _
My Role is (Circle One)
His (Hui Avit’Irt

F213“ 50pr Jr. Sr. Grad;

Instructcr Asst. Inst. Asst. Prof

IIIS Pkl l‘(:!|!\‘_L (\n ‘1 C'ij’I".11r ‘It, A 7,. 1’r-Q|
”mac Prof Professor Adminisfmrtr V ,

.' II\ ”(I I. ‘11 . A? I"'3I ‘
”3""0'7 pith, ‘tato

 

‘X'Hct mulvf :7; ..zz-ur best guCS-s c; to H1. rwwwtw‘r rt

 

 

 

t- 1,

1x ""‘I~ ’I‘w’s C‘tiw‘t ”111‘? ",0 thruwflw Yi': 1:9] tiwi- _

;.t ,Ar .in ' 1' arty/:15 ‘iCUf 110311379 mu"
r— ' *- - ~ _ u

 

A P P E N D I X B

THE CONDITIONAL PROBABILITY TABLES
REPRESENTING THE NINE DIFFERENT

STARTER-TARGET COMBINATIONS

APPENDIX B

Appendix B contains thirteen conditional probabil—
ity tables. The first nine tables represent and correspond
to the nine starter-target combinations used in the present
study. The last four tables are summary tables. Table
numbers 10, 11, and 12 were computed by collapsing the three
tables with the same target role group. Table 13 represents

all nine starter-target combinations collapsed into one

table.
E21
Numbers
1. Freshmen
2. SOphomores
3. Juniors
4. Seniors
5. Graduate Students
6. Instructors
7. Assistant Professors
8. Associate Professors
9. Professors
10. Administrators
11. Others (secretaries, wives, etc.)
12. Lost
13. Target

120

121

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ma
m ooo. ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NH
Ha mom. ooo. «ma. ooo. mod. Hmo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. Ha
mm vmv. mao. mHo. mom. Hma. mao. omo. ooo. mao. ooo. ooo. ooo. ooo. oH
moa mmm. mao. mmo. hma. mmm. moa. mmo. mHo. wmo. ooo. ooo. ooo. ooo. m
me How. «no. ooo. ham. How. moa. mvo. mmo. «No. ooo. ooo. ooo. mmo. m
hm ova. mao. mmo. moo. mmm. ooa. mma. mmo. moo. ooo. mao. ooo. ooo. h
«a mod. duo. Hho. ooo. ooo. mod. mva. vam. ooo. mod. ooo. Hho.. ooo. m
on boa. boo. ooo. ooo. boo. mma. ooo. ooo. oom. mmo. mmo. ooo. ooo. m
m ooo. ooo. ooo. ooo. ooo. ooo. HHH. HHH. ooo. vow. mmm. ooo. ooo. o
m omm. ooo. mNH. ooo. ooo. ooo. ooo. ooo. mma. mNH. omm. mNH. ooo. m
N ooo. ooo. ooo.L ooo. ooo. ooo. ooo. ooo. oom. oom. ooo. ooo. ooo. m
a ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo.a ooo. ooo. H
z ma NH Ha oH m m h o m o m N H mum
m m m > H m U m m lwcmm
.GOHumcHnEoo ummuoutumuumum ummumu huaooou m on Houumum huasomm one .H Manda

122

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooo.HToo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. mH
mm ooo. Too.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NH
vH omm. _va. mvH. Hho. ooo. ooo. moH. ooo. ooo. mvH. ooo. ooo. ooo. HH
mH mom. meH. ooo. Hmm. who. ooo. who. ooo. ooo. who. ooo. ooo. ooo. oH
oH moo. ooo. ooo. ooo. mmH. ooo. moo. ooo. moo. ooo. ooo. ooo. ooo. m
mH mmm. mmH. boo. boo. mmH. ooo. oom. ooo. ooo. ooo. woo. ooo. ooo. m
mm «mm. boo. ooo. boo. moo. woo. woo. moo. mvo. ooo. ooo. ooo. ooo. b
oH ooo. _ooo. ooH. ooo. ooo. oom. ooo. ooH. ooH. ooo. ooH. ooo. ooo. o
Ho vow. ooo. ooo. «No. vmo. moo. moo. moo. mmo. moo. moo. omo. ooo. m
wo oHo. woo. woo. ooo. who. woo. woo. moo. mmH. «mm. mmH. woo. woo. o
mm omo. moH. «mo. vmo. ooo. «mo. wmo. ooo. «mo. omm. Hum. NoH. Hmo. m
mo HHo. moo. vmo. omo. HHo. HHo. wmo. mmo. ooo. moo. mom. Ham. ooo. m
Ho «Ho. moo. mmo. ooo. mmo. mHo. moo» ooo. hmo. moo. ooo. mmH. woo. H
z mH NH HH oH m m WM” o MTi w m m H mum
m m m > H m U m m Ivcmm
coHumcHnfioo uamnmuiuounmum ummnmu huHsomw ou Houumum ucooaum one .m mnmda

123

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. MH
o OOH. ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NH
oH ooo. hoH. ooo. hoH. com. boH. ooo. ooo. ooo. ooo. ooo. ooo. ooo. HH
NoH mHm. mmo. mmo. mom. oHN. mvo. hmo. mmo. ooo. ooo. ooo. ooo. ooo. 0H
om ooo. ooo. ooo. omo. ovH. 00H. ooo. ooo. ooo. ooo. ooo. ooo. ooo. m
vH Hbm. ooo. me. Hho. Hho. ooo. ooo. ooo. moH. ooo. ooo. ooo. ooo. m
0H ooo. 00H. ooo. 00H. ooo. ooo. OCH. OCH. ooo. ooo. ooo. ooo. ooo. h
m com. ooo. ooo. ooo. ooo. ooo. oom. ooo. ooo. ooo. ooo. ooo. ooo. o
v ooo. ooo. ooo. omm. ooo. omm. ooo. ooo. 0mm. 0mm. ooo. ooo. ooo. m
H ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. w
00 ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. m
00 ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. N
00 ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. H
2 MH NH HH OH m m h o m o m N H
who
m m m > H m U m m lficmm
.coHumcHnfioo ummuounuouumow ummumu wuHuoom ou Hmuumum HoumuumHGHEoo one .m mqmme

124

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. mH
no ooo. ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NH
NH ooo. mHH. mmN. omo. mmo. mmo. mHH. ooo. mHH. mHH. moo. omo. ooo. HH
mm mHH. omo. ooo. moN. omo. omo. omo. mHH. ooH. mho. omo. mHo. mHo. oH
mm moo. omo. omo. ooH. omN. Hmo. Hmo. omo. oHH. omo. HHo. HHo. mNo. m
om moo. omo. omo. HHH. moo. noH. ooH. hoH. ooo. oNo. mNo. ooo. mNo. o
om mNo. moo. ooo. mho. mmH. hmo. ooH. omo. mNH. NHH. ooH. mNo. mHo. h
mN NhH. omo. omo. moo. omo. omo. omo. moH. hoN. moo. NhH. ooo. omo. o
oo Hmo. HoH. woo. moo. ooo. boo. Hmo. oHo. HmN. omH. HoH. Hmo. oHo. m
on ooH. oHH. ooo. ooo. ooo. ooo. moo. ooo. moo. HNN. oom. oHH. oHo. o
mm mNN. moo. HHo. omo. ooo. ooo. NNo. ooo. HHo. ooH. mom. ooH. moo. m
on ooH. HoH. ooo. mHo. ooo. ooo. ooo. ooo. ooo. ooH. HoH. HHo. omo. N
HN moH. moo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. omN. mmN. mmm. H
z mH NH HH oH o o h o m o m N H mum

mmm>Hmomm _1ocmm
coHumcHnEoo ummnmuiuounmum ummumu ucoosum ou Hmuuoum manomm one .o mqmda

125

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooo.H ooo. wOH. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. mH
om ooo. ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NH
o ooo. ooo. omN. omN. ooo. ooo. ooo. ooo. ooo. ooo. omN. ooo. omN. HH
mH omH. omH. oo. Noo. ooo. ooo. ooo. ooo. who. who. who. ooo. ooo. oH
o ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. m
o ooH. ooo. ooo. ooo. ooo. ooo. oom. ooo. ooH. ooo. ooo. ooH. ooo. m
h moH. ooo. ooo. omN. ooo. omN. ooo. moH. ooo. moH. ooo. ooo. ooo. b
N oom. ooo. oom. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. o
mH omH. ooo. who. ooo. ooo. ooo. ooo. ooo. mom. omH. omH. who. ooo. m
mm omo. mmo. ooo. HHo. ooo. HHo. HHo. ooo. HNo. moo. HNN. oHH. HHo. o
om HmH. NmH. ooo. ooo. ooo. oHo. OHo. 0H0. oNo. HHH. mNm. HmH. Hwo. m
bmH ooo. mho. ooo. mHo. ooo. mHo. ooo. ooo. mHo. ooo. NmH. moo. oHH. N
moH omo. Nmo. ooo. moo. ooo. ooo. moo. ooo. ooo. moo. HOH. moH. omo. H
2 mH NH HH 0H m m h o m o m N H mum

oemm
m m m > H m U m m
GOHumoHafioo ummuouiuounoum ummnmu unmooum ou Houuoum pooosum one .m wands

126

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. mH
No ooo. ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NH
NN omH. ooo. mom. mom. ooo. ooo. ooo. ooo. NoH. omH. ooo. ooo. ooo. HH
NNN moo. omo. omo. Nmo. oNH. omo. ooo. omo. omo. omo. omo. Nmo. ooo. oH
oo omH. moo. Hmo. omH. omH. .me. Hmo. moo. moo. moo. ooo. ooo. ooo. m
Nm ooo. moo. omo. mmH. mNH. moo. omH. omo. moo. omo. ooo. ooo. Hmo. o
Ho ooH. moo. ooo. ooo. ooo. omo. HoH. ooH. moo. omo. omo. omo. omo. 5
mm omo. moo. ooo. ooo. omo. ooo. moH. mmH. mom. moo. ooo. NoH. omo. o
mo moo. mmo. HNo. Hmo. Hmo. Hmo. HNo. ooo. mom. ooH. ooH. moo. Noo. m
on oHH. moH. ooo. mHo. mHo. mHo. omo. mHo. omo. Hmo. oHH. moH. mHo. o
om ooH. omo. ooo. omo. ooo. ooo. omo. omo. omo. ooo. ooN. mNH. ooH. m
Nm omN. mmH. ooo. ooo. ooo. ooo. ooo. ooo. mHo. omo. mmH. omm. ooo. N
HN ooH. moo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooH. omN. omN. H
z mH NH HH oH m m o o m o m N H mum
m m m > H m o m m fioqmm
comomcHQEoo ummumuiumuuoum ummumu pooosum ou Houumum HouoHuchHEUm one .o mqmoe

127

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooo.H ooo. ooo. ooo. 000. 000. ooo. ooo. ooo. 000. 000. 000. ooo. mH
mH ooo. ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NH
HH omm. NmH. Hmo. Hmo. Hmo. Hmo. 000. 000. Hmo. ooo. ooo. ooo. ooo. HH
Hm Noo. omo. mNo. NNN. omo. NHo. NHo. ooo. NHo. ooo. ooo. ooo. ooo. oH
mo moN. ooo. ooo. Hom. moH. Noo. moo. ooo. omo. NHo. ooo. ooo. NHo. m
Hm omN. moo. Nmo. omN. HoH. omo. Nmo. Nmo. Nmo. ooo. ooo. ooo. ooo. o
Nm omo. mHo. mHH. omN. omN. omo. mHH. ooo. omo. ooo. omo. ooo. ooo. m
m ooo. mNH. ooo. mom. mNH. ooo. ooo. mNH. mNH. mNH. ooo. ooo. ooo. o
HN moH. ooo. moo. moH. moo. ooH. ooo. ooo. mmm. ooo. ooo. ooo. ooo. m
N ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. oom. ooo. ooo. oom. ooo. o
m ooo. ooo. ooo. mmm. mmm. ooo. ooo. ooo. mmm. ooo. ooo. ooo. ooo. m
H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo.H ooo. ooo. ooo. ooo. N
H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo.H ooo. ooo. H
z mH NH HH oH m m m o m o m N H mum

m m m > H m o m m Hocmm
coHumoHnEoo ummumuluouuoum pomuou HoumnuchHaom 0p Monumum muHsomw one .5 mqmoe

128

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. mH
oN ooo. ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NH
o omN. ooo. mNH. mom. mNH. mNH. ooo. ooo. ooo. ooo. ooo. ooo. ooo. HH
mo mmm. ooo. NNo. moN. ooo. ooo. ooo. ooo. NNo. NNo. ooo. ooo. ooo. oH
oH Hoo. ooo. ooo. oHN. Hoo. Hoo. Hoo. ooo. ooo. ooo. ooo. ooo. ooo. o
NH ooo. ooo. ooo. ooH. ooo. ooo. ooo. ooo. ooo. moo. ooo. moo. ooo. o
o mmm. ooo. ooo. ooH. mmm. ooH. ooo. ooo. ooo. ooo. ooo. ooo. ooo. o
H ooo. ooo. ooo. ooo. ooo. ooo. ooo.H ooo. ooo. ooo. ooo. ooo. ooo. o
oH on. omH. ooo. moH. moH. mmo. ooo. ooo. HHN. ooo. ooo. mmo. ooo. m
mo ooo. omo. moo. omo. moo. moo. oHo. ooo. ooo. oom. HoH. Noo. oHo. o
oo ooo. Hoo. oNo. NoH. ooo. Noo. ooo. ooo. Hoo. ooN. moH. ooH. Noo. m
mo HHo. omo. NNo. omo. HHo. HHo. NNo. HHo. omo. ooo. HmH. oom. moo. N
oo oHo. ooH. ooo. Hoo. oNo. ooo. oHo. ooo. oHo. oHo. ooH. mmH. omo. H
z mH NH HH oH o o o o o o m N H mum

o m m > H m o m m ioqmm
COHuoanEoo uomnouuuouuoum uomnou HonoHuoHnHEoo on» on Monmoum unoosum one .m mqmde

129

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. mH
m ooo. ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NH
o oom. ooo. ooo. mmm. ooH. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. HH
ooH Noo. mNo. omo. mmm. omo. mNo. oHo. ooo. HHo. ooo. ooo. ooo. ooo. oH
oH mom. ooo. ooo. omN. mNH. ooo. ooo. ooo. ooo. ooo. moo. ooo. ooo. m
mo ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. o
oo oom. ooo. ooo. omN. ooo. omN. ooo. ooo. ooo. ooo. ooo. ooo. ooo. o
oo ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. o
oo ooo. ooo. ooo. ooo. omN. ooo. omN. ooo. oom. ooo. ooo. ooo. ooo. o
oo ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. o
Ho ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. m
oo ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. N
oo ooo. ooH. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. H
z mH NH HH oH m m o o m o m N H
muo
m m m > H m U m m fonom
noHuoanfioo uomuounuouuoum uomuou HouoHumHnHEoo onu on Houuoum HouoHannHfioo one .o mqmde

130

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooo.H—ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. mH
oo ooo. _ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NH
Hm mNm. mNH. mNH. moo. HoH. moo. moo. ooo. ooo. moo. ooo. ooo. ooo. HH
oNN oom. Hmo. NNo. omN. moH. mmo. mmo. mHo. moo. ooo. ooo. ooo. ooo. oH
ooH ooo. HHo. oHo. HNH. oHN. Noo. omo. HHo. mNo. ooo. ooo. ooo. ooo. a
mo mmm. ooo. ooo. ooH. ooN. ooo. ooo. mHo. ooo. ooo. mHo. ooo. mHo. m
om moN. ooo. mmo. moo. omH. HHH. omH. ooo. ooo. ooo. HHo. ooo. ooo. o
mN mom. omo. moo. ooo. ooo. omH. moH. omH. omo. moo. omo. omo. ooo. o
mo ooN. oNo. ooo. oNo. ooo. mmo. oNo. oNo. mmo. mmo. ooo. mHo. ooo. m
oo oNo. Hoo. Hoo. ooo. ooo. ooo. omo. moo. ooH. omN. moH. Hoo. Hoo. o
oo moo. omo. moo. omo. ooo. omo. omo. ooo. moo. mmN. ooN. ooH. moo. m
om HHo. ooo. mmo. mmo. HHo. HHo. mmo. NNo. moo. ooo. ooN. mmm. ooo. N
No NHo. Hoo. oNo. ooo. oNo. NHo. moo. ooo. omo. Hoo. omo. moH. mmo. H
z mH NH HH oH m m o o m o m N H mHo
omm>Hmomm ocmo
uomuou muHsoom o ou mHouHoum nouoHumHnHEoo ono unoooum .muHsoom one .oH.mHm¢e

131

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. mH
moH ooo. ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NH
mo ooo. ooo. omN. ooH. mNo. mNo. ooo. ooo. ooH. oHH. ooo. mNo. mNo. HH
ooN Noo. omo. oNo. oom. ooH. ooo. moo. omo. moo. moo. omo. oNo. moo. oH
NmH ooo. moo. ooo. ooH. mmN. HNH. Hoo. omo. Hoo. omo. ooo. ooo. mHo. o
oo mmo. omo. ooo. ooH. moo. ooH. moN. NNH. Hoo. omo. oHo. oHo. oNo. o
oNH ooo. moo. ooo. ooo. ooo. ooo. NoH. ooo. mmH. ooH. ooo. mNo. oHo. o
oo oHH. omo. mmo. ooo. mmo. oHo. ooo. oHH. ooN. ooo. oHH. moo. mmo. o
mNH omo. ooH. ooo. ooo. ooo. Nmo. oNo. ooo. on. ooH. oNH. ooo. oNo. m
HoN ooH. ooo. ooo. ooo. ooo. ooo. oNo. ooo. oNo. oom. NHN. NHH. NHo. o
ooN ooH. NoH. ooo. oNo. ooo. ooo. mNo. NHo. oNo. mNH. oom. mmH. ooo. m
moN oHH. ooH. ooo. NHo. ooo. ooo. ooo. ooo. NHo. moo. ooH. NHo. ooo. N
HmH moo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NmH. moN. omo. H
z mH NH HH oH o o o o m o m N H mum

.o m m > H m o m m iocom
uomuou unoosum m on mHouHoum HouoHumHnHfioo ono unoooum .mquoom one .HH mqmde

132

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. mH
oo ooo. ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NH
mN oom. ooo. ooo. ooN. oNH. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. HH
Nom on. omo. omo. HHm. omo. oHo. mHo. ooo. mHo. moo. ooo. ooo. ooo. oH
mHH mom. mmo. ooo. on. mmH. Noo. ooo. ooo. oNo. moo. moo. ooo. ooo. m
oo omo. Noo. HNo.. oNN. ooH. moo. HNo. HNo. HNo. HNo. ooo. HNo. ooo. o
No moH. oHo. omo. NoN. NoN. Hoo. ooo. ooo. ooo. ooo. Nmo. ooo. ooo. o
o ooo. HHH. ooo. mmm. HHH. ooo. HHH. HHH. HHH. HHH. ooo. ooo. ooo. o
oo moN. Hmo. mNo. oHH. oHH. oHH. mNo. ooo. moN. ooo. ooo. mNo. ooo. m
Ho moo. omo. Noo. omo. Noo. Noo. oHo. ooo. ooo. oom. omo. moo. oHo. o
mm mHo. omo. mHo. mHH. oHo. moo. ooo. ooo. omo. ooN. NmH. ooH. moo. m
om HHo. mmo. HNo. ooo. HHo. HHo. HNo. HHo. ooo. moo. ooH. oom. ooo. N
mo mHo. ooH. ooo. ooo. oNo. ooo. mHo. ooo. mHo. mHo. ooH. mmH. mmo. H
z mH NH HH oH m o o o m o m N H mum
m m m > H m o m m wocoo
nomuou HonouomHnHfioo no on Houuoum HouoHumHnHfioo ono unoosum .mUHsoom one .NH mqmde

133

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. mH
omN ooo. ooo.H ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. NH
mm NNN. Hoo. NoH. NoH. Hoo. Hmo. ooo. ooo. Hoo. Hoo. oNo. oHo. oHo. HH
oHo mHm. mmo. oNo. mmm. oHH. mmo. omo. oNo. mNo. oNo. mHo. oHo. Hoo. oH
oHo on. mNo. mNo. HmH. ooN. moo. mmo. oHo. ooo. oHo. moo. Noo. ooo. m
omH ooN. ooo. ooo. omH. omH. Hoo. ooH. omo. omo. mNo. oHo. oHo. mHo. o
ooN omH. ooo. Nmo. oHH. moH. ooo. omH. omo. mmo. omo. omo. HHo. ooo. o
om moH. Hmo. Hoo. Hoo. Hmo. Hmo. Noo. NNH. moH. Hoo. Noo. Hoo. oNo. o
ooN oNH. ooo. mNo. ooo. omo. ooo. mNo. NHo. Hom. ooo. ooo. mmo. NHo. m
oom moo. ooo. oHo. oHo. mNo. oNo. Hmo. oHo. omo. mom. moH. mmo. oHo. o
oom mmH. moo. oHo. omo. moo. oHo. NNo. ooo. omo. moH. moN. mmH. ooo. m
oNo ooo. ooo. oHo. mmo. moo. moo. NHo. ooo. ooo. ooo. ooH. Nmm. ooo. N
oom Noo. ooo. ooo. mHo. mHo. moo. oHo. ooo. mHo. moo. omH. NoH. omo. H
z mH NH HH oH o o o o m o m N H mum

oamo

mmm>Hm0mm

 

 

 

mnoHuoanEoo uomnounuouuoum oan HHo mo oHnou NHoEEdm

.mH 39mm.

HICHIGQN STATE UNIV. LIBFRRIES
iliilmillNW1“WWII"IlililillililllMINIMUM
31293103769935