See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/305683830

Research Methods in Communication

Book · July 2011

CITATIONS READS

14 45,630

1 author:

Mahmoud Eid
University of Ottawa
74 PUBLICATIONS 489 CITATIONS

SEE PROFILE

All content following this page was uploaded by Mahmoud Eid on 28 July 2016.

The user has requested enhancement of the downloaded file.

10270760834_i-ii_8.5x11_pi-ii 7/13/11 5:13 PM Page i

Edited by

Mahmoud Eid

Research Methods in
Communication
Taken from:
Social Research Methods: Qualitative and Quantitative Approaches,
Seventh Edition
by W. Lawrence Neuman
Doing Survey Research: A Guide to Quantitative Methods,
Second Edition
by Peter M. Nardi
Qualitative Research Methods for the Social Sciences, Seventh Edition
by Bruce L. Berg
Methods: Doing Social Research, Fourth Edition
by Winston Jackson and Norine Varberg
Basics of Social Research: Qualitative and Quantitative Approaches,
Second Canadian Edition
by W. Lawrence Neuman and Karen Robson
Research Methods, Design, and Analysis, Eleventh Edition
by Larry B. Christensen, R. Burke Johnson, and Lisa A. Turner
10270760834_i-ii_8.5x11_pi-ii 7/11/11 6:09 PM Page ii

Cover Art: Courtesy of Photodisc/Getty Imges.

Taken from: Methods: Doing Social Research, Fourth

Edition by Winston Jackson and Norine
Social Research Methods: Qualitative and Varberg
Quantitative Approaches, Seventh Edition Copyright © 2007, 2003, 1999, 1995 by
by W. Lawrence Neuman Pearson Education Canada, Inc.
Copyright © 2011, 2006, 2003 by Pearson Published by Prentice Hall, Toronto, Ontario
Education, Inc.
Published by Allyn & Bacon Basics of Social Research: Qualitative and
Boston, Massachusetts 02116 Quantitative Approaches, Second Canadian
Edition by W. Lawrence Neuman and Karen
Doing Survey Research: A Guide to
Robson
Quantitative Methods, Second Edition by
Copyright © 2012, 2009 by Pearson
Peter M. Nardi
Canada, Inc.
Copyright © 2006, 2003 by Adobe
Published by Allyn & Bacon, Toronto, Ontario
Systems, Inc.
Published by Allyn & Bacon
Research Methods, Design, and Analysis,
Boston, Massachusetts 02116
Eleventh Edition by Larry B. Christensen, R.
Qualitative Research Methods for the Social Burke Johnson, and Lisa A. Turner
Sciences, Seventh Edition by Bruce L. Berg Copyright © 2011, 2007, 2004, 2001, 1997
Copyright © 2009, 2007, 2004, 2001, 1998, by Pearson Education, Inc.
1995, 1989 by Pearson Education, Inc. Published by Allyn & Bacon
Published by Allyn & Bacon Boston, Massachusetts 02116
Boston, Massachusetts 02116

Copyright © 2011 by Pearson Learning Solutions

All rights reserved.

Permission in writing must be obtained from the publisher before any part of this work
may be reproduced or transmitted in any form or by any means, electronic or mechanical,
including photocopying and recording, or by any information storage or retrieval system.

All trademarks, service marks, registered trademarks, and registered service marks are the
property of their respective owners and are used herein for identification purposes only.

Pearson Learning Solutions, 501 Boylston Street, Suite 900, Boston, MA 02116
A Pearson Education Company
www.pearsoned.com

Printed in Canada

1 2 3 4 5 6 7 8 9 10 XXXX 16 15 14 13 12 11

000200010270760834

MHB

ISBN 10: 1-256-28123-9

ISBN 13: 978-1-256-28123-8
10270760834_iii-xi_90983 06 163-197 r4 sp 7/16/11 11:25 AM Page v

Contents

Page - Ethical issues in preparing the

research report . . . . . . . . . . . . . . . . . . . . . 41
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . ix - References . . . . . . . . . . . . . . . . . . . . . . . . 43

About the Editor . . . . . . . . . . . . . . . . . . . . xi II. RESEARCH DESIGNS

CHAPTER 3
Historical-Comparative Research . . . . . . 47
I. INTRODUCTION
W. Lawrence Neuman
- A short history of historical-
CHAPTER 1
comparative research . . . . . . . . . . . . . . . . 48
Introduction to Communication
- Research questions appropriate for
Research . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
historical-comparative research . . . . . . . . 49
Mahmoud Eid - The logic of historical-comparative
- Importance of communication research . . 3 research . . . . . . . . . . . . . . . . . . . . . . . . . . 50
- Research designs . . . . . . . . . . . . . . . . . . . 5 - Steps in conducting a historical-
- Data collection methods and comparative research project . . . . . . . . . . 58
sampling strategies . . . . . . . . . . . . . . . . . . 9 - Data and evidence in historical context . . 61
- Data analysis . . . . . . . . . . . . . . . . . . . . . . 11 - Comparative research . . . . . . . . . . . . . . . .69
- Writing reports and making - Equivalence in historical-
presentations . . . . . . . . . . . . . . . . . . . . . . 13 comparative research . . . . . . . . . . . . . . . . 82
- References . . . . . . . . . . . . . . . . . . . . . . . . 14 - Ethics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
- References . . . . . . . . . . . . . . . . . . . . . . . . 88
CHAPTER 2
Ethics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 CHAPTER 4
Larry B. Christensen, R. Burke Johnson & Unobtrusive Measures in Research . . . . . 93
Lisa A. Turner Bruce L. Berg
- Research ethics: What are they? . . . . . . . 15 - Archival strategies . . . . . . . . . . . . . . . . . . 94
- Ethical dilemmas . . . . . . . . . . . . . . . . . . . 19 - Physical erosion and accretion:
- Ethical guidelines . . . . . . . . . . . . . . . . . . 23 Human traces as data sources . . . . . . . . 105
- APA ethical standards for research . . . . . 27 - References . . . . . . . . . . . . . . . . . . . . . . . 107
- Ethical issues to consider when
conducting research . . . . . . . . . . . . . . . . . 30
- Ethical issues in electronic research . . . . 40
10270760834_iii-xi_90983 06 163-197 r4 sp 7/16/11 11:25 AM Page vi

vi CONTENTS

CHAPTER 5 - Communication components . . . . . . . . . 213

Experimental Research . . . . . . . . . . . . . 111 - What to count: Levels and units of
W. Lawrence Neuman & Karen Robson analysis . . . . . . . . . . . . . . . . . . . . . . . . . 214
- Category development: Building
- Random assignment . . . . . . . . . . . . . . . 113
theory . . . . . . . . . . . . . . . . . . . . . . . . . . 214
- Experimental design logic . . . . . . . . . . . 115
- Discourse analysis and content
- Internal and external validity . . . . . . . . .124
analysis . . . . . . . . . . . . . . . . . . . . . . . . . 218
- Results of experimental research:
- Open coding . . . . . . . . . . . . . . . . . . . . . 218
Making comparisons . . . . . . . . . . . . . . . 130
- Coding frames . . . . . . . . . . . . . . . . . . . . 220
- A word on ethics . . . . . . . . . . . . . . . . . . 131
- Stages in the content analysis
- References . . . . . . . . . . . . . . . . . . . . . . . 132
process . . . . . . . . . . . . . . . . . . . . . . . . . . 224
- Strengths and weaknesses of the
CHAPTER 6 content analysis process . . . . . . . . . . . . 225
Qualitative and Quantitative - Computers and qualitative analysis . . . . 226
Research . . . . . . . . . . . . . . . . . . . . . . . . . 133 - Qualitative research at the speed of
W. Lawrence Neuman light . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
- References . . . . . . . . . . . . . . . . . . . . . . . 231
- Strategies of research design . . . . . . . . . 133
• Triangulation . . . . . . . . . . . . . . . . . . 134
• Qualitative and quantitative
orientations toward research . . . . . . . 135 III. DATA COLLECTION
• Qualitative design issues . . . . . . . . . 144 METHODS & SAMPLING
• Quantitative design issues . . . . . . . . 148 STRATEGIES
- Qualitative and quantitative
measurement . . . . . . . . . . . . . . . . . . . . . 165
CHAPTER 8
• The need for measurement . . . . . . . . 166
Qualitative Research Methods . . . . . . . 237
• Quantitative and qualitative
measurement . . . . . . . . . . . . . . . . . . . 167 Winston Jackson, Angela Gillis & Norine Verberg
• The measurement process . . . . . . . . 168 - Entering the world of qualitative
• Reliability and validity . . . . . . . . . . . 174 research . . . . . . . . . . . . . . . . . . . . . . . . . 237
• A guide to quantitative - Ethnography . . . . . . . . . . . . . . . . . . . . . 243
measurement . . . . . . . . . . . . . . . . . . . 184 - Grounded theory . . . . . . . . . . . . . . . . . . 244
• Scales and indexes . . . . . . . . . . . . . . 189 - Phenomenology . . . . . . . . . . . . . . . . . . . 247
- References . . . . . . . . . . . . . . . . . . . . . . . 205 - Data-collection methods in
qualitative studies . . . . . . . . . . . . . . . . . 251
- Two qualitative approaches to data
CHAPTER 7
analysis . . . . . . . . . . . . . . . . . . . . . . . . . 257
An Introduction to Content Analysis . . 209
- Methods of evaluating qualitative
Bruce L. Berg research . . . . . . . . . . . . . . . . . . . . . . . . . 262
- What is content analysis . . . . . . . . . . . . 209 - Advantages and limitations of
- Analysis of qualitative data . . . . . . . . . . 209 qualitative research . . . . . . . . . . . . . . . . 264
- Content analysis as a technique . . . . . . . 211 - References . . . . . . . . . . . . . . . . . . . . . . . 266
- Content analysis: Quantitative or
qualitative? . . . . . . . . . . . . . . . . . . . . . . 211
10270760834_iii-xi_90983 06 163-197 r4 sp 7/16/11 11:25 AM Page vii

CONTENTS vii

CHAPTER 9 C H A P T E R 12
Quantitative Research Methods . . . . . . 269 Analysis of Quantitative Data . . . . . . . . 379
W. Lawrence Neuman & Karen Robson Peter M. Nardi
- The logic of survey research . . . . . . . . . 270 - Presenting data: Descriptive
- Constructing the questionnaire . . . . . . . 271 statistics . . . . . . . . . . . . . . . . . . . . . . . . . 379
- Types of surveys: Advantages and • Presenting univariate data . . . . . . . . . 379
disadvantages . . . . . . . . . . . . . . . . . . . . 289 • The normal curve and z-scores . . . . . 388
- Interviewing . . . . . . . . . . . . . . . . . . . . . 292 - Analyzing data: Bivariate
- The ethical survey . . . . . . . . . . . . . . . . . 297 relationships . . . . . . . . . . . . . . . . . . . . . 395
- References . . . . . . . . . . . . . . . . . . . . . . . 300 • Presenting nominal and ordinal data
in tables . . . . . . . . . . . . . . . . . . . . . . . 396
• Testing bivariate relationships . . . . . 399
C H A P T E R 10
- Analyzing data: Comparing means . . . . 411
Qualitative and Quantitative
• T-tests . . . . . . . . . . . . . . . . . . . . . . . . 412
Sampling . . . . . . . . . . . . . . . . . . . . . . . . . 303
• Analysis of variance (ANOVA) . . . . 418
W. Lawrence Neuman • Differences not strength . . . . . . . . . . 421
- Reasons for sampling . . . . . . . . . . . . . . 303 - Analyzing data: Multiple variables . . . . 421
- Sampling strategies . . . . . . . . . . . . . . . . 305 • Elaborating relationships:
- References . . . . . . . . . . . . . . . . . . . . . . . 336 Control variables . . . . . . . . . . . . . . . 422
• Multiple relationships . . . . . . . . . . . . 428
- References . . . . . . . . . . . . . . . . . . . . . . . 435
IV. DATA ANALYSIS
AND PRODUCTION C H A P T E R 13
The Research Report . . . . . . . . . . . . . . . 437
C H A P T E R 11 Winston Jackson & Norine Verberg
Analysis of Qualitative Data . . . . . . . . . . 341 - General orientation . . . . . . . . . . . . . . . . 437
W. Lawrence Neuman - Organization . . . . . . . . . . . . . . . . . . . . . 439
- References . . . . . . . . . . . . . . . . . . . . . . . 448
- Comparison of methods of data
analysis . . . . . . . . . . . . . . . . . . . . . . . . . 342
- Coding and concept formation . . . . . . . 344 I N D E X . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
- Analytic strategies for qualitative
data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
- Other techniques . . . . . . . . . . . . . . . . . . 365
- References . . . . . . . . . . . . . . . . . . . . . . . 375
10270760834_iii-xi_90983 06 163-197 r4 sp 7/16/11 11:26 AM Page viii
10270760834_iii-xi_90983 06 163-197 r4 sp 7/16/11 11:26 AM Page ix

Preface

Communication research is an extremely rich field experience of communication students by provid-

of study that merits much more attention from com- ing them with current, accurate, and essential infor-
munication researchers than it currently receives. mation about communication research. As a result,
Research Methods in Communication is a compre- readers of this book are able to find clarification and
hensive book designed to equip communication stu- answers to many of their questions surrounding
dents with the essential tools to effectively conduct research methods in communication.
and accurately understand multi-research-designs This book takes an innovative, multi-
in communication. It is a unique compilation of research-design approach to understanding and
works on communication research that offers advice conducting communication research. For
from some of the most well-known and widely pub- example, the quantitative-qualitative convergence
lished authors in the field. Designed to meet a vari- in this book stems from an in-depth understand-
ety of needs, it provides inclusive material with ing of how communication research methods are
easy-to-follow explanations that span the full range interrelated. There is a common misconception
of research methods and approaches. This model among communication researchers in terms of the
gives freedom and flexibility to students and pro- distinction between quantitative and qualitative
fessors working at all levels of concentration, in research approaches. It is true that each approach
order to draw upon the required and relevant infor- has its own language and tools that challenge this
mation that guides them through their communica- inter-relation. Moreover, some researchers use
tion research courses. As a valuable handbook, it both approaches when conducting a communica-
equips them with the knowledge to design research tion study as a way of improving its quality, or in
projects, collect and analyze quantitative and qual- an attempt to prove the interconnection. Although
itative data, compile the report, and present research there is some logic behind this methodology, it is
findings. nevertheless misleading.
Research Methods in Communication is both a The fact of the matter is that quantitative and
theoretical and practical guide to the world of com- qualitative methods are complementary to one
munication research. It draws on the work of senior another; not in the sense of their joint utilization in
scholars, and has been edited by an academic who one study or research project, but rather that they
has extensive professional and academic experience cannot be entirely separated as they share many
in conducting qualitative and quantitative research commonalities. For example, can any communica-
projects and in teaching communication research tion researcher claim that there is no reasoning or
methods. To ensure the highest possible calibre of critical thinking in using a qualitative approach? In
the book, strict criteria were applied concerning the fact, logic and reason are the overarching premise
contents of the chapters; these include the quality for mathematics under which numbers and statis-
and depth of the subject matter covered, and their tics fall. On the other hand, can any communication
practical application in both academic and profes- researcher claim that there are no observations of
sional environments. The objective is to satisfy a human subjects, or that the analysis of words,
broad range of needs and maximize the learning images, and behaviours does not exist in the quan-
10270760834_iii-xi_90983 06 163-197 r4 sp 7/16/11 11:26 AM Page x

x PREFACE

titative approach? In questionnaires, open-ended in Communication. Courses at the basic level of

questions provide us with qualitative content that each approach should be taught at the undergradu-
can be analyzed, with an option of doing so quali- ate level for second and third year students in
tatively. As well, face-to-face interviews allow a departments of communication, media studies, and
researcher to report information similar to a quali- journalism; the advanced level courses should be
tative participant observation method of data col- taught to fourth year students and at the graduate
lection. In order to accurately understand the level (M.A. and Ph.D.). In addition, introductory
distinction between the two approaches, one must courses in communication research methods should
acknowledge that each approach should be studied also be taught to first year students in order to pro-
in light of the other; in so doing, each receives the vide them with an overview of both approaches. At
attention that it deserves as a rich and comprehen- the graduate level, more specialized instruction
sive approach of communication research, but not should be organized, involving in-depth discussions
in the sense that it is entirely distinct from the other at the most advanced level of both quantitative and
approach. Indeed, communication researchers qualitative approaches. These courses could be
should appreciate the strengths of each approach, divided into various levels that gradually increase
and also acknowledge their respective limitations; in intensity, according to the year of study.
in this way, when designing the methodology of a Research Methods in Communication is
communication research project, researchers can divided into four parts: Introduction; Research
benefit from the advantages and minimize the lim- Designs; Data Collection Methods and Sampling
itations of each approach according to the topic Strategies; and Data Analysis and Production. Pre-
under investigation and the objectives of their sented in a logical order consistent with the process
research project. An appreciation of their mutual of conducting communication research, each part is
benefits ultimately improves the understanding of a composition of a group of chapters, each of which
the communication phenomena being investigated. is designed to provide students with exclusive mate-
The materials provided in this book aid profes- rials that cover a specific stage in the process.
sors in offering instructions to students in both qual-
itative and quantitative approaches at basic and
advanced levels, i.e. Basic Quantitative Research Dr. Mahmoud Eid
Methods in Communication and Advanced Quan-
titative Research Methods in Communication, and Associate Professor
Basic Qualitative Research Methods in Communi- Department of Communication
cation and Advanced Qualitative Research Methods University of Ottawa, Canada
10270760834_iii-xi_90983 06 163-197 r4 sp 7/16/11 11:26 AM Page xi

About the Editor

Mahmoud Eid (Ph.D., Carleton University) is an ment: International Media Ethics and Rational Deci-
Associate Professor at the Department of Commu- sion-Making (2008), editor of Communication and
nication, University of Ottawa, Canada. Dr. Eid pre- Media Studies: An Introduction (2010) and co-edi-
viously taught in the University of Regina’s School tor of The Right to Communicate: Historical Hopes,
of Journalism in Regina, and in Carleton Univer- Global Debates and Future Premises (2009).
sity’s School of Journalism and Communication in Dr. Eid has presented numerous papers at
Ottawa. He has been teaching research methods global conferences and has contributed chapters to
courses in Canadian and Egyptian universities since several books published by Hampton Press, Peter
1994, and has worked as a research manager in sev- Lang, Rodopi, Rowman & Littlefield, Kendall/
eral international research agencies and centers for Hunt, Pearson, and Oxford University Press, among
15 years. His teaching experience, research inter- others. In addition, he has published articles in var-
ests, and publications concentrate on quantitative ious refereed journals, including The Journal of
and qualitative research in international and polit- International Communication; The European Jour-
ical communication, media ethics and effects, pub- nal of Communication Research; International
lic opinion, and social development. Journal of the Humanities; First Monday; INFOR-
Dr. Eid is the editor of the Global Media MATION; Journalism Ethics for the Global Citizen;
Journal--Canadian Edition, and serves on the edi- Corporate Ownership and Control; Journal of Inte-
torial boards of several academic journals and as an grated Marketing Communications; International
organizing committee member for various inter- Journal of Technoethics; and Middle East Journal
national conferences. He is the author of Interweave- of Culture and Communication.
10270760834_C01_90983 06 163-197 r4 sp 7/11/11 4:17 PM Page 2
M01_NEUM5964_07_SE_C01.qxp 6/13/11 3:14 PM Page 1

I. Introduction
M01_NEUM5964_07_SE_C01.qxp 6/13/11 3:14 PM Page 2
10270760834_C01_90983 06 163-197 r4 sp 7/11/11 4:17 PM Page 3

CHAPTER 1
Introduction to
Communication Research
Mahmoud Eid

I M PORTANCE OF All scientific explanations are considered tentative

COM MU N ICATION RES EARCH and are subject to continued testing through the pro-
cess of empirical observation. These explanations
Everyday thinking or inquiry differs from scientific
are continually refined, elaborated, and even
thinking (Nardi, 2006). Everyday experiences are
replaced by more effective explanations. In fact, the
often based on methods that can lead to problematic
production of scientific knowledge is an ongoing
decisions with outcomes that can seriously affect our
cycle: creating an explanation, identifying the con-
lives. We may come to conclusions and act on them
cepts in that explanation, performing meaning
with limited information. Everyday thinking is
analysis to derive concrete indicators of these con-
somewhat less sophisticated than scientific thinking,
cepts, making empirical observations to test
which we would expect from policy makers and
hypotheses about the relationship between them,
others who are in control of our well-being; it is often
engaging in empirical analysis of the measured con-
characterized by biased questions, limited sampling,
cepts to refine the definitions and measurement of
selective attention, perception and retention, and
these concepts, and refining the theoretical expla-
inaccurate generalizations. On the other hand, sci-
nation that links these concepts. This cycle repeats
entific thinking enables us to make accurate and reli-
with every new investigation. The ultimate depend-
able conclusions about human behaviour. It is
ence on observation and experience holds true for
characterized by empirical observations or data, sys-
all communication research. If two communication
tematic and deliberate methods, as well as objective,
researchers, who differ widely in their choice of
intersubjective, and replicable procedures.
abstractions, follow the scientific norm, they accept
In conducting scientific investigations,
specific observations in the same light. Findings
researchers must make many basic decisions in the
must stand the test of inter-observer agreement and
interest of enhancing the growth of knowledge or
reproducibility, meaning that observers using the
making contributions to it (McLeod & Tichenor,
same procedures should obtain the same result. For
2003). These decisions are related to defining
example, regardless of their models concerning the
research problems, constructing explanations, mak-
consequences of televised advertising about kids’
ing observations, and testing hypotheses, as well as
toys, they would ordinarily be expected to agree on
developing research programs. Identifying a
the fact that in a given experiment, say, 85% of the
research problem is a process of abstraction in its
children displayed toy purchasing behaviour as
own right. From the complexity of events that cause
operationalized by the researcher. There may be dif-
concern in a particular area of interest, an investi-
fering interpretations of the motivations for this
gator selects and formulates a problem in terms that
behaviour, but the observed, empirical facts are the
give it some generalizability and make it amenable
same for all in the norms of science. However,
to systematic study. The ultimate goal of abstrac-
observation alone is not a sufficient criterion for a
tion in scientific inquiry is to develop and examine
scientific procedure. The observations must be rel-
explanations. Creating explanations for a specific
evant to a theory, meaning that for a specific case
phenomenon involves attempts to answer a number
they must rest on one or more hypotheses.
of related questions, such as: what are the
Researchers should also strive to develop a research
antecedent and consequent factors and processes?
10270760834_C01_90983 06 163-197 r4 sp 7/11/11 4:17 PM Page 4

4 CHAPTER 1 / MAHMOUD EID

program that links their various investigations into rough sense of what is happening on a particular
a larger whole. The links that bind studies together topic for which we do not yet have sufficient infor-
to form a research program may be shared concepts mation. Descriptive research provides basic infor-
and/or theoretical propositions. mation describing the topic and the respondents
In conducting social and behavioural research, involved. Explanatory research, based on some
the investigator must go beyond the general stan- descriptive information, is intended to explain
dards of scientific inquiry, and must first define a relationships and uncover reasons why and how
problem in a way that makes it amenable to study. some social phenomena occur among respondents.
In conceptualizing that problem, the investigator Evaluation research is interest in understanding
must address key issues, including: epistemology, the causes of human social behaviour and people’s
induction versus deduction, causation versus pre- opinions on a variety of issues; or sometimes eval-
diction, and micro versus macro models. Episte- uating specific outcomes and providing explana-
mology refers to the origin of knowledge and the tions as to why and how a particular result
routes used to establish it. Induction and deduction occurred.
are two commonly used types of logic or reasoning As a social science, communication is very dif-
whose nature is an important concern in the con- ficult to define, as is the case for many other sci-
struction and testing of theory. In deductive reason- ences. For example, if we ask a senior mathematics
ing, a hypothesis or conclusion is implied in the scholar who has spent his/her entire academic life
premise. One proceeds from the general to the par- studying and teaching mathematics what the defi-
ticular; i.e. from a set of assumptions one deduces nition of mathematics is, it might be very difficult,
a specific hypothesis. In contrast, inductive reason- if not impossible, to provide a comprehensive
ing means that one proceeds inferentially from par- answer. However, it might be easier for that scholar
ticular to more general statements. Induction is to answer the question: how do you define or under-
commonly used both in defining problems and in stand mathematics? And it would be easier still if
the interpretation of results. Some scientists and phi- we ask for a definition of a mathematical theory, an
losophers point out that science is both causal and approach, or a tool, or an explanation of a mathe-
predictive; however, the distinction between the two matical equation or dilemma. The fact that it is dif-
is important. Researchers may be able to use X to ficult to provide one definition of communication
predict Y, although little is known about the causal upon which all can agree confirms that it is a very
relationship between them. On the other hand, there rich discipline, whose theories, approaches and
might be a well-investigated causal relationship tools are as deep as the discipline itself. As such,
between two variables; however, one cannot predict conducting communication research is an enormous
any related influences on other variables. Investi- field of study that requires very skilled and knowl-
gators also face a choice between microscopic and edgeable researchers who are aware of its many
macroscopic conceptual approaches. At the micro- areas, approaches, designs, methods, tools, etc. In
scopic level, the emphasis is primarily on the indi- general, a distinction is made between basic com-
vidual as a unit of analysis, while at the macroscopic munication research, which is designed to test and
level, a holistic approach is taken, and the unit of refine theory, and applied communication research,
analysis tends to be a collective or larger social unit, which is designed to solve a practical problem.
such as an organization, a community, or a social However, both types are interrelated, and an inte-
system. grated model of both basic and applied communi-
The most common purposes of scientific cation research is still valid (Frey, Botan & Kreps,
research on human social behaviour and attitudes 2000). Communication researchers study varied
are to explore, describe, explain and evaluate in and complex phenomena. They design their
order to arrive at an in-depth understanding of an research questions and statements in ways that help
issue, as well as to make decisions and predictions. them describe communication behaviour and relate
Exploratory research is conducted to obtain a it to other variables.
10270760834_C01_90983 06 163-197 r4 sp 7/11/11 4:17 PM Page 5

CHAPTER 1 / INTRODUCTION TO COMMUNICATION RESEARCH 5

Communication research involves people as decision-making tool may be a strategy for devel-
researchers, respondents, and interested observers; oping a personal judgment about the ethical char-
people whose interactions are not always ethically acter of a research activity. Moreover, the
neutral. Communication researchers are confronted communication researcher’s ethical responsibilities
by ethical dilemmas and issues arising from human go beyond human subjects. They also have respon-
subject research; thus, there is a need to find strat- sibilities towards society at large, given the fact that
egies to resolve these issues (Greenberg, Eastin & communication research has vast social implica-
Garramone, 2003). They also have responsibilities tions. Society has the right to expect communica-
towards society, which both supports and is influ- tion researchers to investigate important topics, is
enced by their efforts. Communication research that interested in knowing about the conduct and report-
involves observations of human beings must ensure ing of research, and has concerns about the appli-
that their rights are not violated. However, given the cation of scientific knowledge.
fact that one individual’s rights may conflict with
those of another, and that an individual’s rights may
RES EARCH DES IGNS
conflict with what is beneficial to society, it may not
be always possible for the researcher to simultane- Having decided on the problematic and defined the
ously respect the rights of all involved. Therefore, main concepts and objectives, the next key stage in
the best alternative is to demonstrate respect by ask- a research project is to choose the most effective and
ing research participants to forgo certain rights, a relevant design for the topic under investigation.
procedure that is generally referred to as giving The research design can be one or a combination of
informed consent. Debriefing is also appropriate the following: historical-comparative, unobtrusive,
whenever subjects’ perceptions or feelings about experimental, qualitative, quantitative, and content
themselves or about others may have been altered analysis research. The research design dictates the
by participating in a study. Researchers are obliged next phases: the data collection methods and tools,
to attempt to return their subjects to their pre-par- the choice of the sampling strategy, the data analy-
ticipation state, i.e. desensitizing them from the sis techniques, and finally, making presentations
experience. Privacy is another issue that researchers and writing reports.
should avoid violating, whether through the effects Historical-comparative research is a powerful
of data collection methods (e.g., participant obser- design for addressing extensive questions: what fun-
vation) or the disclosure of these data (e.g., sensitive damental features and techniques of advertising are
personal information). As a result, researchers have common to most cultures? How have major Cana-
developed techniques to protect subjects’ privacy, dian media conglomerates developed and changed
such as offering guarantees of anonymity and con- over the last two decades? How has media policy
fidentiality, as well as presenting questions in a way been increasingly developed and implemented to
that shields respondents. protect Canadian media content since the 1960s?
In addition, a variety of strategies for resolu- Historical-comparative research can strengthen
tion have been developed, including: professional conceptualization and theory building (Neuman,
codes, federal regulations, and individual ethical 2011). By looking at historical events or diverse cul-
analysis. Professional codes of ethics provide con- tural contexts, researchers can generate new con-
siderable direction regarding the treatment of cepts and broaden their perspectives.
human subjects. An institutional review board is a Generally, conducting historical-comparative
system used by many institutions (e.g., universities) research does not involve a rigid set of steps or
that requires that all research conducted at the insti- complex techniques; however, there are some
tution conforms to specific ethical guidelines, usu- exceptions. Researchers usually draw on four
ally referred to as ethical clearance. However, both types of historical evidence: primary sources, sec-
codes and regulations are not often entirely com- ondary sources, running records, and recollec-
prehensive, and perhaps the most useful ethical, tions. While traditional historians rely heavily on
10270760834_C01_90983 06 163-197 r4 sp 7/11/11 4:17 PM Page 6

6 CHAPTER 1 / MAHMOUD EID

primary sources, historical-comparative researchers case studies, life histories, and understanding of ide-
often use secondary sources or different data types ology and culture (Berg, 2009). Archival strategies
in combination. A comparative perspective include public archives (e.g., commercial media
exposes weaknesses in research design and helps accounts, actuarial records, and official documen-
researchers improve the quality of their research. tary records) and private archives (e.g., autobiog-
The focus of comparative research is on similar- raphies, letters, and diaries). Therefore, from a
ities and differences between units. In commun- media and organizational communication research
ication studies, these units can be different perspective, unobtrusive measures can be quite
communication or media systems in other coun- helpful. Let us think about researchers who are
tries, different media in one culture, or a combi- interested in studying the potential impact of tradi-
nation of the two. For example, a researcher may tional and new media in Canada on beliefs and atti-
compare the coverage of a specific historical tudes of Canadian citizens. Human traces left in the
terrorist event in two different Canadian news- letters and comments to the editor in the form of
papers, a Canadian radio station and a Canadian newspaper articles, e-mails, and chat forums (fol-
magazine, or Canadian and American TV chan- lowing the broadcast of specific programs) would
nels. Comparative research can eliminate or offer certainly be very efficient means to gauge how the
alternative explanations for causal relationships. public reacts to specific radio, television, news-
It improves measurement and conceptualization, paper, and Internet content.
as concepts developed by those conducting The goal of experimental research is to estab-
research across several settings are less likely to lish cause and effect conclusions by isolating causal
apply to only one specific culture or setting. How- variables (Jackson & Verberg, 2007). In this case,
ever, comparative research is more difficult, more the procedure is to manipulate one or more inde-
costly, and more time-consuming. Comparative pendent variables to determine their effect(s) on a
researchers can apply theory, not test it; and they dependent variable. The experimental design
are only able to make limited generalizations. requires an artificial environment in order to con-
A major concern for any researcher is to ensure trol for all antecedent, intervening, and spurious
maximum reliability and validity throughout the variables. Its strength is its ability to clarify causal
research process. Indeed, how the research is inferences. It has a high level of internal validity
designed and conducted can have a profound influ- because researchers can demonstrate that the
ence on the accuracy of the results. The types of manipulation of the independent variable alone pro-
measures used to test a research hypothesis can be duced changes in the dependent variable. The exper-
threats to its reliability and validity. For example, imental design is therefore intended to control for
the fact that subjects “know” they are being studied the effects of potentially confusing variables.
can influence their behaviour or attitude, which Some natural settings lend themselves well to
threatens the validity of the research. This phe- quasi-experimental designs and field experiments.
nomenon is commonly called the Hawthorne effect. Quasi-experimental designs control some, but not
Sometimes, the researcher can find ways to neu- all of, irrelevant factors. These types of designs have
tralize this biased behaviour by using standardized stronger external validity because the behaviour is
procedures, as is the case with control groups and measured in real life settings. Field experiments
control variables. But when this is not possible, facilitate making causal inferences, have strong
unobtrusive measures prove to be very useful. These validity, and are often fairly inexpensive to com-
involve examining and studying human traces. plete. However, the nature of experimental design
Archival materials, radio and videotape tran- makes it difficult to take into account systematically
scriptions, photographs, autobiographies, diaries, a large number of variables at one time. Also,
and even paint deposits and tombstones are all although experiments provide the clearest explana-
human traces. These can in fact be quite revealing tion of causation, it is not always possible to exper-
for a researcher, as they allow for the creation of iment on all topics.
10270760834_C01_90983 06 163-197 r4 sp 7/11/11 4:17 PM Page 7

CHAPTER 1 / INTRODUCTION TO COMMUNICATION RESEARCH 7

Qualitative researchers are more concerned data, and ask theoretical questions about the data.
with issues of the richness, texture and the feeling The operationalization process for qualitative
of raw data because their inductive approach research often precedes conceptualization. Instead
emphasizes developing insights and generalizations of turning refined conceptual definitions into mea-
from the data collected. They also want to become surement operations, a qualitative researcher oper-
intimate with the details of a natural setting or a par- ationalizes concepts by describing how specific
ticular cultural-historical context (Neuman, 2011). observations and thoughts about the data con-
They often rely on interpretive or critical social sci- tributed to working ideas that are the basis of con-
ence. For example, communication researchers who ceptual definitions and constructs. In this instance,
conduct qualitative research frequently rely on crit- operationalization is a description of how a
ical theory. They apply logic in practice and follow researcher develops working ideas, while making
a nonlinear research path. Qualitative researchers observations and collecting data. It is a description
speak a language of cases and contexts. They stress of how specific observations or data, preliminary
conducting detailed examinations of cases that arise ideas about the data, and struggles to understand the
in the natural flow of social life, and try to present data, become constructs. It is an after-the-fact
authentic interpretations that are sensitive to spe- description more than a pre-planned technique.
cific social-historical contexts. Fewer standardized Reliability and validity are central issues in
procedures or explicit steps are used, and often on- qualitative measurement. Reliability in qualitative
the-spot techniques are devised for one situation or research means dependability or consistency. Qual-
study. Finally, the researchers rarely separate plan- itative researchers use a variety of techniques to
ning and design decisions into a distinct pre-data record their observations consistently. They want to
collection stage, but continue to develop the study become consistent in how they make observations
design throughout early data collection. over time. In qualitative research, validity means
When considering measurement, qualitative truth. Qualitative researchers are more interested in
researchers develop ways to capture and express authenticity than in the idea of a single version of the
concepts using various alternatives to numbers. truth. Authenticity means providing a fair, honest
They often adopt an inductive approach, creating and balanced account of social life from the per-
new concepts as part of measurement. Qualitative spective of someone who lives it every day. They
researchers usually begin with empirical data, fol- are less concerned with matching an abstract con-
lowed by abstract ideas; they relate ideas and data struct to empirical data than with giving a candid
and conclude with a mixture of the two. As qualita- portrayal of social life that is true to the experiences
tive researchers gather data, they use some pre- of the people being studied.
existing ideas to assist with data collection, and then Quantitative researchers focus more on issues
merge the old ideas with the new ideas that are of design, measurement and sampling because their
developed from the data. In the conceptualization deductive approach emphasizes objectivity and
process, qualitative researchers refine rudimentary detailed planning prior to data collection and analy-
working ideas during the data collection and analy- sis (Neuman, 2011). Most quantitative researchers
sis processes. In this case, conceptualization is a rely on a positivist approach to social science. For
process of forming coherent, theoretical definitions example, communication researchers who conduct
as one seeks to make sense of the data, or organize quantitative research often rely on the administra-
one’s preliminary thoughts. As qualitative tive school of thought. They apply reconstructed
researchers gather and analyze data, they develop logic and follow a linear research path. They speak
new concepts, formulate definitions for major con- a language of variables and hypotheses, and empha-
structs, and consider relationships among them. size measuring variables precisely and testing
Ultimately, they link concepts and constructs to cre- hypotheses that are linked to general causal expla-
ate theoretical relationships. Qualitative researchers nations. Quantitative research is often deductive
form and refine constructs as they examine their with a sequence of discrete steps that precede data
10270760834_C01_90983 06 163-197 r4 sp 7/11/11 4:17 PM Page 8

8 CHAPTER 1 / MAHMOUD EID

collection: narrow the topic to a more focused ques- data and test the empirical hypotheses. These tests
tion; transform nebulous, theoretical concepts into are logically linked back to a conceptual hypothe-
more exact variables; and develop more hypotheses sis and causal relations in the world of theory.
to test. Reliability and validity are also central issues
Quantitative researchers treat measurement as in quantitative measurement. In terms of quantita-
a distinct step in the research process that occurs tive research, measurement reliability means that the
prior to data collection, and they develop special ter- numerical results produced by an indicator do not
minology and techniques. As they adopt a deduc- vary because of the characteristics of the measure-
tive approach, they begin with a concept, and then ment process or the measurement instrument itself.
create empirical measures that precisely capture it Measurement validity refers to the compatibility of
in a form that can be expressed in numbers. Quan- the conceptual and operational definitions; the bet-
titative researchers begin with an abstract idea, fol- ter the fit, the greater the measurement validity.
lowed by a measurement procedure, and finish Validity is more difficult to achieve than reliability.
with empirical data that represent the ideas. As a Researchers cannot be absolutely confident about
quantitative researcher develops measures, the con- validity, but some measures are more valid than
structs become more refined and clear, and as the others; this is because constructs are abstract ideas,
researchers apply the measures to gather data, they whereas indicators refer to concrete observations.
often adjust the measurement technique. The mea- When a researcher systematically and objec-
surement process for quantitative research follows tively identifies characteristics of messages, s/he is
a straightforward sequence: first is the conceptual- using a procedure of data analysis defined as con-
ization, followed by the operationalization, and then tent analysis. As messages are a key component of
the application of the operational definition or communication, one can understand the importance
measurement to collect the data. Quantitative of this procedure for communication scholars. For
researchers have developed several ways to rigor- example, a researcher might want to assess the
ously link abstract ideas to measurement procedures changing images of women in the Canadian media
that produce precise quantitative information about from the 1970s to the year 2010, or that same
empirical reality. In the process of quantitative mea- researcher might aim at examining Canadian news-
surement, there are three levels to consider: con- paper orientations towards candidates throughout an
ceptual, operational and empirical. At the most election campaign. When we think about it, virtu-
abstract level, the researcher is interested in the ally all data gathered require a certain level of con-
causal relationship between two constructs, or a tent analysis. This is particularly true for data
conceptual hypothesis. At the level of operational collected through interviews, field notes, observa-
definition, the researcher is interested in testing an tions, case studies, and specifically for any unobtru-
empirical hypothesis to determine the degree of sive data. For example, archival materials, videotape
association between indicators. This is the level at transcriptions, photographs, newspaper articles or
which correlations, statistics, and questionnaires are focus group discussions must all be condensed in a
used. The third level is the concrete empirical world. systematic and rigorous manner in such a way that
If the operational indicators of variables are logi- some kind of manifest pattern can emerge in the light
cally linked to a construct, they capture what hap- of previous research, theories, and hypotheses.
pens in the empirical, social world and relate it to the There are certain steps and concepts involved
conceptual level. The measurement process links in the procedure of content analysis: from the data
the three levels together, moving deductively from collected that are transformed into a text or docu-
the abstract to the concrete. A researcher first con- ment, to the construction of categories and the cod-
ceptualizes a variable, assigning it a clear concep- ing process, to the patterns resulting from the coding
tual definition. Next, s/he operationalizes it by (Berg, 2009). In content analysis, the process of
developing an operational definition or set of indi- specifying the content characteristics being exam-
cators. Finally, s/he applies the indicators to collect ined interact with the application of unequivocal
10270760834_C01_90983 06 163-197 r4 sp 7/11/11 4:17 PM Page 9

CHAPTER 1 / INTRODUCTION TO COMMUNICATION RESEARCH 9

rules for identifying and recording those character- examining their answers. When writing survey
istics. These are essential to the validity and relia- questions, researchers should be aware of the prin-
bility of content analysis. For example, from the ciples of good writing, including the knowledge of
very beginning of the process, researchers must what to include or avoid. There are various types of
establish clear selection criteria concerning which survey research, including mail, telephone inter-
parts of the document or text will be included in, or views, face-to-face interviews, and web surveys,
excluded from the analysis. In other words, each of which has advantages and disadvantages. In
researchers need to specify the characteristics of the general, researchers should be aware of the fact that
sample chosen: is the focus of the analysis on words, there is no one perfect type of survey; instead there
phrases, sentences, paragraphs, sections, chapters, are ideal types depending on the research design and
books or subject topics? The subsequent steps topic of investigation. A major ethical issue in sur-
involve the conceptualization and operationaliza- vey research is the invasion of privacy. Researchers
tion of categories from which parts of the text or should treat all respondents with dignity and do
written document have to be rigorously coded. The everything possible to reduce their discomfort.
construction of these unequivocal rules can be done Researchers are also responsible for protecting the
inductively, deductively, or by a combination of the confidentiality of data.
two approaches. There are logical steps to conducting a sur-
Content analysis can be very useful when a lon- vey: developing hypotheses, deciding on the type
gitudinal study is being conducted (e.g., assessing of survey, writing survey questions, deciding on
the changes in English Canadian newspapers’ ori- response categories, designing the layout, plan-
entations towards immigration from the 1960s to ning how to record data, pilot-testing survey tools,
the 1990s). As the level of validity and reliability deciding on the target population, getting the
generated from longitudinal studies is high, such sampling frame, deciding on the sample size,
strength cannot be underestimated. However, one selecting the sample, locating the respondents,
major limitation of content analysis (as is the case conducting interviews, recording the data care-
with some other research designs) is its ineffective- fully, entering the data into computers, rechecking
ness in testing causal relationships. When a topic or data and running statistical data analysis, and con-
theme seems to reappear frequently in a text or doc- cluding with describing the methods and the pre-
ument, the researcher can only conclude that there senting findings in a research report.
seems to be a trend or an emerging pattern, but no In constructing a questionnaire, there are vari-
causal explanations should be attached to these, as ous issues that need to be handled carefully and pro-
only experimental designs allow for cause and effect fessionally. These may include: the principles of good
conclusions. question writing, aiding respondent recall, obtaining
honest answers, the design of open versus closed
questions, etc. There is no proper length of a ques-
DATA COLLECTION M ETHODS AN D
tionnaire or an interview; it depends on the survey
SAM PLI NG STRATEGI ES
format, the objectives of the study, and the respon-
After selecting and planning the research design, dents’ characteristics. Survey researchers should also
researchers should decide on the data collection give careful consideration to the order and sequence
method(s) and the sampling strategy and tech- of the questions to be asked by the interviewer, as
nique(s). Researchers use a variety of quantitative well as the context in which they are posed. A large-
and qualitative data collection methods, and conse- scale survey requires the hiring of several interview-
quently utilize the relevant sampling strategy and ers; an interview of professional quality demands the
technique(s). careful selection of highly-trained individuals. Sur-
Survey research is the most widely used quan- vey researchers proscribe interviewer behaviour to
titative research method (Neuman, 2011). It is a pro- reduce bias. Ideally, the actions of a particular inter-
cess of asking many people the same questions and viewer do not affect the way in which a respondent
10270760834_C01_90983 06 163-197 r4 sp 7/11/11 4:17 PM Page 10

10 CHAPTER 1 / MAHMOUD EID

answers, and responses do not vary from those given typically consists of six to twelve individuals who
to any other interviewer. are asked to discuss topics suggested by a modera-
In qualitative research, there are three traditions tor. The researcher observes the interactions among
that move across a continuum in the size of the unit focus-group members and solicits their attitudes,
being studied as they have different research goals: opinions, and solutions to problems. In all data col-
ethnography, grounded theory, and phenomenology lection methods, qualitative researchers should
(Jackson, Gillis & Verberg, 2007). Ethnography is develop and ensure high standards.
concentrated on larger aggregations of people and When the population is small enough, you can
focuses on describing a culture. It is a qualitative easily survey every element. A population is the
research method that attempts to understand human total collection of units or elements that a researcher
behaviour in the cultural context in which it is wants to analyze. If researching the entire universe
embedded. Grounded theorists typically examine or population of study is difficult or impossible,
smaller aggregations and focus on basic social pro- selecting a representative sample is the only alter-
cesses. The goal of grounded theory is the develop- native. The sampling strategy is dependent on the
ment of a theory that explains underlying social design, and it is strongly linked to the methods of
processes. The researcher is interested in under- data collection; hence the different approaches to
standing how people interact, take action, or engage sampling taken by qualitative and quantitative
in the process in response to some phenomenon. researchers.
Phenomenological studies tend to be concentrated Qualitative researchers focus less on a sample’s
on individual experiences and perceptions, and representativeness or on detailed techniques for
focus on discovering the essence of the lived expe- drawing a probability sample (Neuman, 2007).
rience. Phenomenologists seek to reveal the mean- Instead, they focus on how the sample, or small col-
ing of a lived experience from the perspective of the lection of cases or units, illustrates key features of
participant. a universe; the goal is to collect cases, events or
Qualitative researchers usually use more than actions that clarify and deepen understanding.
one data collection method in a single study. They Therefore, qualitative researchers tend to use non-
choose among the various data collection methods, probability or non-random sampling, which means
including the three major ones: participant obser- that each unit of analysis in the population does not
vation, in-depth interviews, and focus group dis- have an equal chance of being selected for the
cussions. The participant observation method sample. The unit of analysis is the element about
involves an intensive examination of a culture, com- which a researcher is observing and collecting data,
munity, organization or group by a researcher who such as a person responding to a questionnaire, a
has joined it for an extended period of time. In so school, an editorial or a local business. Qualitative
doing, researchers face multiple challenges: gain- researchers rarely determine the sample size in
ing entry into the group, developing and maintain- advance and select cases gradually, with the specific
ing rapport, developing a method for taking field content of a case determining its inclusion. There
notes, and integrating the data collection and data are various types of nonprobability samples, each
analysis. In-depth interviews provide a method of of which has its own principle. A haphazard sample
collecting respondents’ perceptions of their world. means obtaining units in any manner that is conve-
Typically, they are done to solicit people’s descrip- nient. A quota sample means fulfilling a preset num-
tions and explanations of events taking place in their ber of units in each of several predetermined
own environment. Quotations from the interviews categories that reflect the diversity of the popula-
are used to illustrate points the researcher wants to tion, using haphazard methods. A purposive sample
make in his/her report. In the communication field, means obtaining all possible units that fit particular
researchers use this method when interviewing criteria. A snowball sample means acquiring cases
experts or members of the social, economic or polit- using referrals from one or more cases, and then
ical elites who are difficult to meet. A focus group referrals from those cases. A deviant case sample
10270760834_C01_90983 06 163-197 r4 sp 7/11/11 4:17 PM Page 11

CHAPTER 1 / INTRODUCTION TO COMMUNICATION RESEARCH 11

means acquiring cases that substantially differ from interval unit, e.g., every tenth person of one hun-
the dominant pattern. A sequential sample means dred. A cluster sample means creating a sampling
obtaining cases until there is no additional infor- frame for larger cluster units, drawing a random
mation or new characteristics. sample from them, creating a sampling frame for
The primary goal of quantitative researchers units within each selected sample, and then draw-
is to obtain a representative sample, or a small col- ing a random sample of units.
lection of units or cases, from the universe being
studied so that s/he can research the smaller group
DATA ANALYS IS
and make accurate generalizations about the whole
population (Neuman, 2007). For example, if a One of the crucial steps in the research process is
researcher targets communication students as a uni- being able to understand the collected information.
verse or population of study who all are in the Fac- A researcher often accumulates a great deal of data,
ulty of Arts at the University of Ottawa and who which must be organized in such a way that they can
number 1,000 in total, it might be very difficult, be given practical and substantial meaning.
albeit still possible, for him or her to conduct Researchers try to make sense of the data so that the
research on the whole population. Therefore, numbers can be translated into words from which
selecting a representative sample is a more feasi- predictions and projections can be made.
ble choice. However, it is impossible for a Ph.D. From a communication perspective, we could
student, who conducts research on how Canadian compare this process of analyzing, understanding,
citizens perceive the idea of multiculturalism, to and interpreting data to a conversation between indi-
apply methods of data collection to the whole pop- viduals: each aims at making sense of the other’s
ulation; consequently, selecting a representative underlying message. After all, what would the value
sample is the only choice. Therefore, quantitative of research be without this essential exchange that
researchers tend to use probability or random takes place not only among scholars, but also
sampling, which means that each unit of analysis in between scholars and the public? However, as in
the population has an equal chance of being any conversation, if the codes are not clearly under-
selected for the sample. A well-designed, carefully stood by each individual, serious misunderstanding
executed probability sample produces results that can arise. The same can occur when data are not
are as accurate, if not more so, than trying to reach properly analyzed, interpreted and understood. In
every single person in the whole population. As other words, the rules pertaining to data analysis
well, deciding on the sample size requires have to be clearly comprehensible to the researchers
researchers to follow specific rules. in order to make the conversation flow as smoothly
There are various types of probability samples, as possible.
each of which has its own technique. A simple ran- Communication researchers need to analyze
dom sample means creating a sampling frame for qualitative data, which are in the form of text, writ-
all units and then selecting units using a purely ran- ten words and phrases, or symbols describing or
dom process such as a random number table. A representing people, actions and events. In fact,
stratified sample means creating a sampling frame there are as many kinds of texts as there are com-
for each of several categories of units, and then munication media. Communication texts can be
drawing a random sample from each category, written transcripts of speeches and conversations,
which are then combined into several samples. A written documents (e.g., letters, personnel records,
systematic sample means creating a sampling newspapers and magazines), electronic documents
frame, calculating the sampling interval (which tells (e.g., audiotapes, videotapes, films and CD-ROMs),
the researcher how to select units from a sampling or visual texts (e.g., paintings, photographs, web-
frame by skipping a number of units in the frame sites and architecture).
before selecting one from the sample), choosing a Textual analysis is the method communication
random starting place, and then selecting every researchers use to analyze (i.e. describe and
10270760834_C01_90983 06 163-197 r4 sp 7/11/11 4:17 PM Page 12

12 CHAPTER 1 / MAHMOUD EID

interpret) the characteristics of written, recorded, or bers can be derived. In today’s Information Society,
visual messages (Frey, Botan & Kreps, 2000). It can we are bombarded with statistics more than ever. Not
be used to answer the two major questions posed in a day goes by without our being highly exposed to
communication research: “what is the nature of com- them, especially in the media: for example, we might
munication?” and “how is communication related to read that the latest federal election poll showed that
other variables?” To answer the first question, textual one candidate obtained 38% of the votes, the second
analysts describe the content, structure and functions 28%, the third 24% with 10% undecided; the poll
of the messages contained in texts. However, was conducted with an error or margin error of 1.7%
describing the communication embedded in a text is in a sample of 3,000 participants. The first question
not as easy as it might seem because there is neither that comes to mind to those unfamiliar with statistics
one sole meaning of a text, nor is there a single per- is how to make projections regarding the entire
spective from which to interpret it. Some analysts Canadian population, based on such a small portion
investigate how various factors/variables that pre- of it. There are two dimensions to statistics: descrip-
cede communication are related to the messages tive and inferential (Nardi, 2006). Descriptive sta-
contained in a text, while others study relationships tistics are meant to structure, summarize and
between the communication embedded in a text and describe the characteristics of a set of data. Inferen-
their various outcomes, or lack thereof. Further, tial statistics allow the researcher to move beyond
some textual analysts, such as rhetorical critics, often the set of data and to potentially generalize the
go beyond these two questions to evaluate texts, results to a broader population. In this sense, infer-
pointing out their strengths and weaknesses. ential statistics are used for estimating the charac-
A qualitative researcher analyzes data by orga- teristics of a population based on the sample; they
nizing them into categories on the basis of themes, are also used for hypothesis testing, i.e. to ascertain
concepts, or similar features. The researcher devel- significant differences between groups or significant
ops new concepts, formulates conceptual defi- relationships between variables.
nitions, and examines the relationships among them Different statistics are used to test associations
(Neuman, 2011). Qualitative coding is an integral or correlations between variables, depending on the
part of data analysis. It is guided by the research level of measurement. For example, when a
question and leads to new questions. Most qualita- researcher is looking at the potential positive corre-
tive researchers use techniques of coding, memo lation between the number of hours spent studying
writing, and the search for outcroppings as generic and the final grade for a specific course, s/he should
approaches to the analysis of qualitative data, often be using Pearson R to assess this relationship
combined with a more specific strategy for the because both variables are measured at a ratio level.
analysis of qualitative data. They need to be aware Very often, researchers want to determine if signif-
of these strategies (e.g., the ideal type, successive icant differences exist between two or more groups.
approximation and narrative analysis) and various For example, a communication scholar might want
qualitative analysis techniques (e.g., network analy- to know if men and women (the variable in this case
sis, time allocation analysis, and software for qual- is gender) differ in terms of their level of self-dis-
itative data such as NVivo). In many respects, closure in the context of intimate relationships; or, if
qualitative data are more difficult to deal with than three types of social publicity related to cultural sen-
data in numeric form. Qualitative analysis requires sibility each has a significantly different impact on
more effort by a researcher to read and reread data attitudes of Canadian university students. The
notes, reflect on what is read, and make compar- researcher then needs to learn when to apply differ-
isons based on logic and judgment. ent statistics in order to compare and assess if means
Statistics (or numerical indicators of a set of differ significantly from one another; s/he must also
data) provide us with the tools and the rules to inter- know how to run and correctly interpret a T-test
pret quantitative data in such way that practical and (when comparing two groups) and an ANOVA
substantial meaning that goes beyond just the num- (when comparing three or more groups). When the
10270760834_C01_90983 06 163-197 r4 sp 7/11/11 4:17 PM Page 13

CHAPTER 1 / INTRODUCTION TO COMMUNICATION RESEARCH 13

questions are complex, the researcher needs to be ation (APA) writing style is commonly used in com-
more knowledgeable about data analysis. For munication research. Plagiarism must also be
example, when a researcher wants to study the rela- avoided; this is the unacknowledged borrowing of
tionship between three or more variables, s/he might ideas and words from other authors. Adopting the
turn to multiple correlation; or, to assess the relative APA style is useful in this respect, as it requires
weight of a set of independent variables on depen- researchers to identify their sources of information
dent variables, a multiple regression technique in the body of the text and to include a complete bib-
would be the best option. Another very important liography at the end of the written report.
aspect of quantitative data analysis closely relates to By organizing the report into sections, the
the validity and reliability of the research: the pro- researcher can include relevant material and discuss
cess of controlling for variables that are not part of issues under key headings (Jackson & Verberg,
the hypothesis being tested. Consequently, learning 2007). The introduction should inform the readers
how to perform elaboration techniques with control about the research project and its problematic; it also
variables then prove to be very useful. Moreover, the raises interesting questions and resolves issues that
use of SPSS (Statistical Package for Social Sciences) justify the significance of the research study. The lit-
software is very popular among social scientists, erature review provides readers with an overview of
including communication researchers. A quantita- other research that is relevant to the topic of investi-
tive analysis researcher must gain knowledge about gation. This leads to a section that defines the
the logic behind the techniques of data analysis, as hypotheses or research questions to be examined. At
well as the know-how of these techniques. this point, identifying the research design and the data
collection procedures are required. The sampling
strategy and the description of the intended sample
WRITI NG REPORTS AN D MAKI NG
should be discussed in detail. A thorough explana-
PRES ENTATIONS
tion of the variables, indexes, and measurement pro-
The last stage of conducting communication cedures contributes to an understanding of the key
research is writing the report and presenting the concepts and relationships being investigated. Pre-
qualitative findings and quantitative results. This senting basic results, organizing summary tables, and
phase is extremely important as it demonstrates the creating tables and figures are major sections in the
strength of the design and how well the research report that need to be well planned. The next step
project has been conducted. In fact, the impact of explains the results of testing hypotheses. Following
the research project is highly dependent on the qual- this, a discussion should be devoted to tying the
ity of the organization and the presentation of its whole project together. Finally, conclusions should
material. In general, reports are written for a variety be drawn and suggestions should be made. To ensure
of audiences, and this fact should be taken into a high quality report, it is also important that the
account in their preparation. research report incorporates a review of a final check-
Know your audience! If a report is intended for list. This list may include: the title, abstract, intro-
a professional journal, then it should be organized duction, literature review, statement of hypotheses or
to conform to the style in which other documents research questions, methodology, results, discussion,
are submitted. If the audience is not technical, then conclusions, references, spell check, format, view
the report should avoid technical terminology. If document, print, and proofreading.
researchers know their audiences, they are aware of The process of presenting the qualitative find-
what questions are likely to arise, and they are able ings and quantitative results includes the creation
to address the concerns of their audiences accord- of tables and figures—e.g., graphs, charts, and illus-
ingly. Researchers must explain their points clearly, trations (Cookman, 2003). Tables present actual
and should never assume that the reader is an expert. numbers organized in rows and columns. Graphs
Careful attention should also be devoted to the present numbers with physical markers such as
writing style. The American Psychological Associ- columns, lines, wedges, and data points.
10270760834_C01_90983 06 163-197 r4 sp 7/11/11 4:17 PM Page 14

14 CHAPTER 1 / MAHMOUD EID

Charts diagram relationships, explain theoretical graph is used to segment the universe of study; a
models, schematize hierarchies or flows, and con- column graph represents a count of the same
vey other concepts that can be best understood visu- group(s) at various moments; a multiple column is
ally. Creating these tables and figures is not a used to compare attitudes towards different things
random process; instead, it is professional and fol- or different groups of respondents; a pyramid but-
lows rules and procedures. Most importantly, dif- terfly format is helpful when the relationship
ferent types of tables and figures are selected for between sets of information is critical; a gradual
specific purposes. For example: a single-scale bar graph is used to illustrate attitudes (e.g., agreement,
graph visualizes comparisons at a particular support, and satisfaction); and a line-column graph
moment in time; a bar graph is used to compare is used to examine the influence of an action or
means and rank categories at the same time; a line event (e.g., advertising or media campaign) on
graph links observations to show trends; an area people or things (e.g., image of a company, country,
graph implies volume; a pie chart divides a whole or product). In general, a presentation or report is
into component parts to review relationships and to greatly affe cted by the quality of the tables and fig-
illustrate the relative size of the parts that make up ures used. Creating effective tables and figures is
a whole; a 3-D pie graph gives us an immediate, not difficult, but does require much planning and
intuitive sense of the proportions in a whole; a donut consideration.

REFERENCES
Berg, Bruce L. (2009). Qualitative research methods for the Jackson, Winston, Gillis, Angela & Verberg, Norine. (2007).
social sciences. Boston: Allyn & Bacon. “Qualitative research methods.” In Winston Jackson and
Cookman, Claude. (2003). “Presenting quantitative data.” In Norine Verberg (Eds.), Methods: Doing social research (pp.
Guido H. Stempel, David H. Weaver and G. Cleveland Wil- 145-184). Toronto: Prentice Hall.
hoit (Eds.), Mass communication research and theory (pp. McLeod, Douglas M. & Tichenor, Phillip J. (2003). “The logic
327-349). Boston: Allyn and Bacon. of social and behavioral science.” In Guido H. Stempel,
Frey, Lawrence R., Botan, Carl H. & Kreps, Gary L. (2000). David H. Weaver and G. Cleveland Wilhoit (Eds.), Mass
Investigating communication: An introduction to research communication research and theory (pp. 91-110). Boston:
methods. Boston: Allyn and Bacon. Allyn and Bacon.
Greenberg, Bradley S., Eastin, Matthew S. & Garramone, Gina Nardi, Peter M. (2006). Doing survey research: A guide to quan-
M. (2003). “Ethical issues in conducting mass communica- titative methods. Boston: Allyn & Bacon.
tion research.” In Guido H. Stempel, David H. Weaver and Neuman, W. Lawrence. (2007). Basics of social research: Qual-
G. Cleveland Wilhoit (Eds.), Mass communication research itative and quantitative approaches. Boston: Allyn and
and theory (pp. 299-326). Boston: Allyn and Bacon. Bacon.
Jackson, Winston & Verberg, Norine. (2007). Methods: Doing Neuman, W. Lawrence. (2011). Social research methods: Qual-
social research. Toronto: Prentice Hall. itative and quantitative approaches. Boston: Allyn & Bacon.

View publication stats