Methods in social research

395 Pages·1952·8.172 MB·English

by  William Goode, Paul Hatt

McGRAW-HILL SERIES IN
Sociology and Anthropology
RICHARD T. LAPIERE
Consulting Editor
Methods in Social Research
McGRAW-HILL SERIES IN SOCIOLOGY
Baber MARRIAGE AND THE FAMILY
Bergel SOCIAL STRATIFICATION
Bergel URBAN SOCIOLOGY
Blalock SOCIAL STATISTICS
Dornbusch and Schmid A PRIMER OF SOCIAL STATISTICS
Freedman, Whelpton, and Campbell FAMILY PLANNING, STERILITY, AND
POPULATION GROWTH
Goode and Hatt METHODS IN SOCIAL RESEARCH
LaPiere A THEORY OF SOCIAL CONTROL
LaPiere SOCIAL CHANGE
Lemert SOCIAL PATHOLOGY
Schneider INDUSTRIAL SOCIOLOGY
Tappan CONTEMPORARY CORRECTION
Tappan CRIME, JUSTICE AND CORRECTION
Tappan JUVENILE DELINQUENCY
Thompson and Lewis POPULATION PROBLEMS
Vernon SOCIOLOGY OF RELIGION
Methods in Social Research
J.
WILLIAM GOODE
Associate Professor, Department of Sociology
Columbia University
PAUL K. HATT
Northwestern University
NEW YORK TORONTO LONDON
McGRAW-HILL BOOK COMPANY, INC.
I
1952
METHODS IN SOCIAL RESEARCH
Copyright, 1952, by the McGraw-Hill Book Company, Inc. Printed in the
United States of America. All rights reserved. This book, or parts thereof,
may not be reproduced in any form without permission of the publishers.
Library of Congress Catalog Card N'Umber: 52-8317
xv
23755
PREFACE
This volume was written out of our teaching experience with the stu·
dents of Princeton University, Wayne University, Ohio State University,
Northwestern University, and Pennsylvania State College and is an at
tempt to meet some of the problems which were part of that experience.
'Its primary debt must therefore be to those students who have suffered
with us in seminars, classes, and projects over the past ten years. The book
attempts to steer a course between a simple anecdotal account of re
search studies with their technical details, and a set of broad generaliza
tions about research methodology. Its aim is to make both the elements
of basic logic and the research procedures of modern sociology under
standable at the undergraduate level, where the student should first be
introduced to them. It has been a working assumption in our classes that
such an understanding of research techniques is indispensable to the
student, whether or not he wishes to become an active researcher. In the
changing field of sociology, the person who is unwilling to learn how
conclusions are reached cannot judge properly whether or not the con
clusions he reads are correct.
It has also been a working assumption of our teaching that no amount
of lecture material, summarizing in general propositions what has been
learned about research techniques, can communicate to the student the
meaning of research or excite in him any of the fascination which the
social process possesses. We have, then, attempted to present, where pos
sible, accounts of specific research experiences, taken in many cases from
recent unpublished work.
In addition, as teachers, we have tried to see that .all the students in
such classes obtain actual field experience. This serves two important
objectives: (1) it gives some concrete meaning to the general rules and
allows the student some basis for learning when these rules do not apply;
(2) equally important, it teaches the student who is easily able to criticize
published work that it is much easier to criticize on a lofty level than it
is to conduct good research. The objection has been made to putting
students at work fu. research that such research is not "good," and that
the student therefore has an idea that it is easy. It has been our experi-
v
vi PREFACE
ence that no student comes out of such a field experience with a smug
feeling that now he is a "trained researcher." Although field experience
requires that the teacher devote considerable energy to his course, the
difficulties of supervision are reduced somewhat if the outside work is
made into a genuine group project.
Perhaps a third advantage to early field experience can be added to the
two previously mentioned. This advantage applies more specifically to
students who go on to do research at the graduate level. It is preferable
for the student to make his blunders at a phase in his development
when it costs him much less, than later to make the same blunders in his
own research, often after a great investment in time and energy.
Because the complaints, comments, and difficulties of our students have
contributed so much to this book, we wish to record here our gratitude
to these students for their patience and their enthusiasm. In addition, of
course, we wish to thank our many colleagues at other universities who
have read and criticized this manuscript, in whole or in part. Many con
ferences with Melvin Tumin of Princeton University have added greatly
to a clearer understanding of the problems which students face on first
coming into contact with the difficulties of research. Read Bain of Miami
University has helped us by pungent reminders that this idea or that
procedure was not clearly expressed. Edward A. Suchman of Cornell
University contributed by his detailed criticism of a substantial portion
of this manuscript and gave special attention to the chapters on scaling.
Several of our graduate students have worked on specific sections of the
manuscript in seminars, adding to its usefulness. Among these may be
mentioned Nicholas Babchuk, Arnold S. Feldman, Leonard Moss, Irving
Rosow, and John Tarini.
As the teacher will recognize, this volume ow~s much to the thinking
of Robert K. Merton and Paul F. Lazar~feld, and to the Bureau of Applied
Social Research, Columbia University, with which they have done so
much creative work. Peter H. Rossi and Alice S. Rossi, formerly of the
Bureau and now of Harvard University, contributed several useful sug
gestions which were drawn from their own research.
William ]. Goode
Paul K. Hatt
NEW YORK, N.Y.
EVANSTON, ILL.
July, 1952
CONTENTS
Preface v
1 The New Sociology 1
2 Science: Theory and Fact 7
3 Values and Science 18
4 Science: Pure and Applied 29
5 Basic Elements of the Scientific Method: Concepts 41
6 Basic Elements of the Scientific Method: Hypotheses 56
7,. Design of Proof: Testing the Hypothesis 74
8 Further Problems in Research Design 92
r
9 Use of the Library (by Joseph S. Komidar) / 103
10 Observation 119
11 Constructing a Questionnaire 132
12 The Mailed Questionnaire 170
13 The Interview 184
14 Probability and Sampling 209
15 Scaling Teclmiques: The Basic Problem 232
16 Scaling Teclmiques: Social Distance, Sociometric, and
Rating Scales 243
I
17 Scaling Teclmiques: Ranking, Internal Consistency, and
Scalogram Scales 261
18 Reseafch in Population 296
19 Some Problems in Qualitative and Case Analysis 313
20 The Analysis.Qf Data 1!41
,
21 Preparing th&. Report 359
Index 377
vii

1
CHAPTER
The New Sociology
Within the past 15 or 20 years, courses in methods of social research
have come to occupy an increasingly important role in sociological cur
ricula. It is likely that at the present writing every major university offers
such courses. In part, the increase is a reflection of the growing job oppor
tunities in the field and the consequently greater number of students who
plan to make their careers in sociology.
In addition, however, an interest in research methods is growing
among those whose job interests are not specifically sociological. Any
one who has a serious interest in understanding society must give some
thought to the ways in which social facts can be and are gathered. Thus
we find a range of needs which the study of research techniques may help
to satisfy. The student with a career interest in socipiogy must acquire
the research tools which he will later need in holding his job, and he
must embark upon an intensive training program for that purpose. The
student who hopes, instead, to obtain an administrative position in
government or business must also acquire some of this knowledge. As
he moves upward through the various jobs, he will often be faced with
the problem of evaluating reports. These are likely to be technical sum
maries of studies and research carried out by others, and he must be able
to decide when they ar~ reliable enough to be used as the basis of his
decisions. The market apalyst, the public-opinion expert, the investigator
of communication and propaganda-all are gathering facts for govern
mental and business needs. A knowledge of social research is useful for
interpreting and weighing such reports.
The lay citizen faces a similar responsibility. W~se decisions about
current events are difflcult to make unless he can judge the truth of
published and spoken I reports. Newspaper accounts, radio broadcasts,
summary statements by governmental agencies must all be evaluated.
This is a civilization in, which decisions are increasingly based upon scien
tific fact, and those who cannot understand how the facts are reached
will be unable to separate fact from speculation and wish. Furthermore,
the prestige of the~ "scientific study" is great enough that many reports
are ~iven this label without justification. The consequence is that a
1
2 METHODS IN SOCIAL RESEARCH
growing number of nonscientists wish to know about the methods of
social science.
At a deeper level, however, it is clear that more courses in research
methods are being offered than before as a result of the growth in the
field of sociology itself. The growth of every science has been accom
panied by the development of research techniques in that field. This fact
is not surprising, since the substantive growth, the gathering and ordering
of facts, is based upon these new techniques. Techniques alone do not
guarantee such a substantive expansion, but they are absolutely indis
pensable.
The increasing emphasis upon research method is, then, a sign of
healthy development within the young science of sociology.
THE CONFLIct OVER SCIENCE
The preceding sentence will seem startling to few readers, and this is
also a measure of the acceptance by sociologists of the scientific approach.
Only a few years ago such a statement would have been the signal for a
fierce debate. Those who wanted to model sociology after the image of
the established sciences and those who objected to this view were in
opposition, and there were many accusations and counteraccusations in
this polemic. This conflict has now almost vanished, but we should at
tempt to understand its basis.
The main points at issue can be stated in four dogmatic propositions:
1. Human behavior changes too much from one period to the next to
permit scientific, exact predictions.
2. Human behavior is too elusive, subtle, and complex to yield to the
rigid categorizations and artificial instruments Of science.
3. Human behavior can be studied only by other human observers,
and these always distort fundamentally the facts being observed, so that
th~re can be no objective p:r:ocedures for achievi:J;lg the truth.
4. Human beings are the subject of such preaictions and have the abil
ity deliberately to upset any predictions we make.
These are actually very complex propositions, and many corollaries
can be elaborated from them. However, in one form or another they
represent the center of the strife, and it is clear that if they are true,
then sociology has a weak scientific foundation. It would then become
the study, of infinitely variable, unique, and nonmeasurable situations,
rather than the investigation of repetitive, simplifiable, and observable
behavior.
It is at least clear on the common-sense level, as noted in elementary
sociology texts, that we are all engaged constantly in predicting socia}
behavior. Indeed, if we could not do so, the society could not exist at all.
We abstract various factors from the behavior of other people, and thus