Introduction 1

CHAPTER ONE

Introduction: Approaches
to Data Collection and
Data Analysis
Glynis M. Breakwell

The Purpose of the Book

Doing social psychology research is a fascinating adventure. It is a

route to understanding in a systematic way the individuals and the
social world around you. It encourages you to ask and start to answer
difficult but fundamentally important questions. What threatens iden-
tity? Why do people rebel? How are risks communicated effectively?
How do leaders command? How are sexual relationships developed?
How do minorities gain influence? How do people react under intense
pressure? How do the mass media represent societal crises? How
does the public behave when faced with a terrorist attack? The list of
interesting questions is probably endless and the ones chosen as
the priorities for research vary according to the values of the social
psychologist involved, the availability of funding for the work and
the theoretical preoccupations of the time. It is, however, true to say
that in the main social psychology research is focused on issues that
matter.
The interesting questions are invariably complex, usually even more
complex than they appear at first sight. The prime skill of a social
psychology researcher lies in refining the question and crystallizing
out something that is capable of being answered. The second vital
skill of the social psychology researcher lies in choosing from the
vast arsenal of research methods that are available the one that is
best suited to address the question posed. To develop this second skill,
2 Glynis M. Breakwell

it is necessary to gain a thorough understanding of all the tools that

reside in the arsenal.
There are now many books to help students of psychology learn
about the research methods they might use. This one is specifically
aimed at undergraduates who want to learn about some of the data
collection methods and data analysis approaches that are used in
social psychology. It is not comprehensive in coverage of all the myriad
methods used, but those selected for inclusion are very commonly
used or they are methods that arouse much student interest. The
selection is deliberately broad. The book is not designed to proselytize
any particular epistemological orientation or set of methodological
principles. It is designed to give the reader an opportunity to judge
what a variety of methods can offer. Each is described in an honest
and open way, and a real effort is made to explain its weaknesses as
well as its strengths.
However, it is inevitable that any book on methods in social
psychology will find itself drawn into the epistemological debates
that have enlivened the discipline for at least 30 years. Questions
emerge naturally about the feasibility of hypothesis testing, the value
of experimentation, the limitations of measurement, the objectivity
of data, the ethics of manipulation, the role of reactance to the pro-
cess of research, the implications of sample structure, the pitfalls
of prediction and so on. Here these issues are not extracted and
dissected as formal philosophical problems. Instead, they are addressed
as they need to be, embedded in the consideration of each of the
methods.
This is not a book that is designed to explain the details of the vari-
ous statistical techniques that can be employed. Where a statistical
technique is the most obvious candidate for use, in combination with
a specific research design and form of data collection, it is described
here. However, the descriptions of statistical procedures are not meant
to be stand-alone expositions. Some chapters are more elaborate and
explicit about the statistics to be used than others. These tend to be
the chapters where the technique of data collection employed is less
important for the question posed than the way the data that are
collected are subsequently treated. Essentially, in this book the reader
is introduced to a statistical technique as an intrinsic element in the
research method, rather than as an end in its own right. The chapters
are designed to expose the logic that indicates which statistical test is
best matched to the data collected. Understanding the assumptions
that underlie statistical tests is often best achieved in the context
of their repeated application. This book offers an opportunity to see
 Introduction 3

the diversity of approach to the use of statistics that exists in social

psychology.

The Organization of the Book

Each chapter describes a method and illustrates how it can be applied

to a particular research question drawn from the substantive topics
that are typically included in current social psychology courses. The
topics covered include identity processes, attribution, stereotyping,
attitude change, social influence, communication, and group dy-
namics. The chapters provide a succinct overview of the theoretical
arguments that surround the topic. In the main, this means that each
chapter can be used as a discrete and independent basis for gaining
an understanding of both the method and the social psychological
problem to which it is applied.
The chapters are written by researchers well known for using the
specific techniques they describe. The strategy adopted in inviting
contributors was to seek them from UK institutions that are very
active in social psychology research. Authors were asked to present
the method and outline the design for an exercise that would illus-
trate its use – in the way they might for their own undergraduate
students. This means that there is diversity across chapters in the
approaches taken. Some are more heavily concerned with the philo-
sophical and logical rationale underlying the method they describe.
Others are clearly more concerned with the practicalities of executing
the method rigorously. It is useful for undergraduates to be exposed
to these varieties in emphasis; there are many routes to achieving a
good understanding of a research method.
Most chapters include three broad structural elements:

• Description of the method and its applications – a succinct outline

of the generic features of the method (what it is, how to do it,
what it can be used for, its strengths and weaknesses), its links to
particular theoretical models and exemplars of classic studies that
have used it to good effect.
• Specific exercise – a description of a training exercise that can be
pursued by the individual students or as part of a methods course;
identifying a research question and showing how the method can
be used to address it. It identifies step by step what needs to be
done in order to utilize the method. The forms of analysis that
might be used with the information collected are explained (if the
4 Glynis M. Breakwell

analytical approach is not already obvious because it is intrinsic to

the method).
• Notes for course leaders – a brief description for course tutors who
might wish to use the exercise of the assumptions it makes about
the level of experience of the student, the time it takes, the materials
needed, the preparatory reading required and so on. While this
element of the chapters is directed at course leaders, it may be
valuable for students as well, allowing them to locate the method
within a matrix of real constraints.

However, each chapter takes on a unique shape and style. There is

no excessive standardization of format across chapters since this can
be boring and sometimes undermines the very real differences in
approach included.
Some chapters address fundamentally the same technique (e.g.
interviewing) but they do so from quite different perspectives or with
different emphases. This is a valuable element of a contributed book
on methods. The subtle differences between contributors in their rep-
resentation of a technique highlight the need for students critically to
evaluate assertions that are made about the characteristics of any
method.
Some chapters are targeted at students who have had little previous
research methods training. Others are more appropriate for students
who have already been introduced to the basic techniques. Chapters
are clearly categorized as introductory or more advanced so that the
reader or course leader using them can be selective. The book is
intended to be a useful resource throughout undergraduate studies. It
could be used as the basis for developing expertise in social psycho-
logy research methods in short bursts but in a cumulative fashion
throughout the full period of an undergraduate course. Equally, it
could be used as the framework for a single intensive injection of
methods within a brief but cohesive programme focused on social
psychology.

Data Collection Methods:

Elicitation and Recording

Data collection is in reality about two intimately connected activities:

data elicitation and data recording. Data elicitation is about accessing
the information, opening it up for examination. Data recording is
about codifying the discovered information in a way that allows the
research question to be addressed.
 Introduction 5

What is a data elicitation method? It is a way of getting informa-

tion. So you could gain information by observing what people do
(observation). For example, you could stand on a busy road with a
machine that measures the speed of the motor cars that pass in order
to determine the number of speeding offences on that road. Altern-
atively, you could gain information by asking people questions
about what they do, think or feel (self-report). To extend the earlier
example, you could ask people who drive along the particular
stretch of road whether they ever exceed the speed limit on it and
how often. Those questions could be asked verbally through some
form of an interview or they could be asked in written form through
some form of a questionnaire. The answers people are allowed to
give could be open-ended (without any constraint imposed by you)
or they could be structured to varying degrees (limited in format
by you in advance – e.g. through the use of rating scales). Some
methods of data collection involve no direct contact with the object
of the research. These rely essentially upon archives (i.e. records) or
artefacts as sources of data (archival). So, for instance, you might find
out what people who are long dead were doing in the nineteenth
century by examining archives reflecting their behaviour. For ex-
ample, in order to understand something about family structures,
you might look at parish records to determine how old people were
when they married in the nineteenth century. To continue with the
earlier example in the context of the use of archives, you could
use criminal records of automobile speeding fines on the target
stretch of road to quantify the level of compliance with road use
restrictions.
Observation, self-report and archival methods of data elicitation
dominate in social psychology. Most others are merely variants of
these three prime types. They are essentially different ways of struc-
turing the observation, shaping self-report or accessing the archive.
For instance, face-to-face interviews rely on self-report but so too do
self-completion questionnaires. Equally, the use of focus groups as a
method of discovery relies upon observation but the measurement of
physiological variation is also a type of observation.
The data that observation, self-report and archival methods reveal
can then be recorded in many different ways. All three can be used
to generate either qualitative or quantitative data records (and
sometimes both at the same time). The form of the data is not intrins-
ically dictated by the method of data elicitation used. When you
use observation you can choose to give a qualitative account of
what happens or you can decide to report a quantified breakdown
of what happens. For instance, you could observe the meeting of
6 Glynis M. Breakwell

two people in a railway station in terms of their demeanour, their

movement to a table in a café, what they order to drink, whether
they look happy when they part. Alternatively, you could decide to
report how many times they touch each other and the interval
of time between each such contact. The object in both cases may be
to understand how people behave in public places and in both cases
the data elicitation method is observation. However, one approach
to recording the data is clearly qualitative and the other obviously
quantitative. The dichotomy between qualitative and quantitative
methods is not a division in the approach to data elicitation at the
macro-level but a distinction in the way data are recorded and then
subsequently analysed.
Crucially, the way data are recorded matters. This is not simply
because the way they are recorded will radically influence the way
they can be subsequently treated (i.e. analysed, and this does not
simply involve statistical analysis). It is also because the way data
are recorded reflects the underlying theoretical and epistemological
beliefs of the researcher. Differences in data elicitation methods are
less revealing of these distinctions in belief than differences in data
recording. Both Popperian hypothesis testers and social construction-
ists may use observation within the context of an interview but
they are exceedingly unlikely to do the same things with the data
thus collected. In learning about a method, it is always important
to examine how data elicitation and recording may be related. All
data elicitation techniques allow for a variety of recording methods.
Choosing the best recording method is as important for the question
you are addressing as choosing the most appropriate data elicitation
method.

Research Designs

When learning about research methods it is also worth remembering

that the same data elicitation method can be used in many different
types of research design. Essentially, the label ‘research design’ is
simply shorthand for the overall structure of the study to be under-
taken. It specifies the components that comprise the study. It should
identify the logical relationships between those components. The
researcher should be able to explain how each component in the
research design is necessitated by the research question that is being
addressed. The chapters in this book describe how research designs
are built up from a research question. The various approaches to
 Introduction 7

developing the research design that are included here illustrate that
the key to a good study is a very systematic unpicking of each aspect
of the research question. A thorough understanding of what it is you
really want to ask or actually want to know is the vital basis for good
research. Ambiguity or under-specification in the research question
leads to poor research because it can result in the wrong data being
elicited or data being recorded ineffectively or data analysis being
misdirected. Getting the specification of the research question correct
is the prime precursor of effective research.
It should already be evident that experimental and non-
experimental research designs are not distinguishable on the basis of
the data elicitation and recording methods they may employ. The
logical distinctions between experimental, quasi-experimental and
non-experimental research designs are explained in various chapters
in this book. Suffice it to say here that the distinction hinges not upon
the method of data elicitation or recording but upon the extent
to which the researcher can and chooses to introduce structured
manipulation of the participants in the study. Experimental designs
involve systematic manipulation. Non-experimental designs involve
no manipulation. Manipulation essentially comprises an intentional
intervention on the part of the researcher that is designed to affect
the participants in the study and lead to recordable outcomes.
Manipulation is deemed to be intentional. It should not be confused
with the unintended effects that the researcher might have upon
study participants simply by attempting to elicit data. The chapters in
this book illustrate a range of approaches to manipulation and their
relationships to the research designs that can be used.
Figure 1.1 summarizes the various levels that coexist and can be
described in piece of research. The figure can be used as an aide
memoir when thinking about how to describe the methodological
components of any study.
Each of the chapters in this book can be characterized in terms of
the first three levels described in Figure 1.1. Table 1.1 provides that
summary. Table 1.1 can be used in order for the reader to make a
quick selection of a chapter that provides information relevant to
particular forms of research activity. It can be used by a course leader
to map out the sequence in which chapters might be included in a
methods course. The table illustrates that chapters overlap in the
research skills that they depict. It is consequently useful in planning
comparisons between different approaches to one research tool (e.g.
alternative ways of using self-report data from interviews or quasi-
experimental designs).
8 Glynis M. Breakwell

Level 1: Research design

Experimental Quasi-experimental Non-experimental

Level 2: Data elicitation

Observation Self-report Archival

Each can be done through many media (e.g. written, audio, visual, artefactual).
Each may involve varying degrees of interaction between the researcher and the
subject of the research. Technological developments (e.g. closed circuit televisual
recording, computer-assisted questionnaire administration or online web-based
techniques) are dramatically modifying the nature of the interaction between
researcher and researched.

Level 3: Data recording

Data recording can be structured prior to initiating data elicitation or can have
structure imposed post hoc. Some structuring prior to analysis is inevitable if the
research is to involve more than direct description.
The structure imposed can take many forms. It might be unitary (e.g. indicating
presence or absence of the entity), or frequency-based, or intensity-based (e.g. rating
or scaling approaches) or thematic (i.e. identifying patterns in the information
collected).

Level 4: Data treatment and analysis

Data may be analysed using qualitative or quantitative techniques. The same data
can often be analysed using both types of technique.

Figure 1.1 Methodological levels of a research study

Table 1.1 Chapter coverage

Chapter Research design Data elicitation Data recording

Observation Self-report Archival Pre-structured Post-structured

Chapter 2 Experimental X X
Chapter 3 Experimental X X X
Chapter 4 Quasi-experimental X X
Chapter 5 Quasi-experimental X X X X X
Chapter 6 Quasi-experimental X X
Chapter 7 Non-experimental X X
Chapter 8 Non-experimental X X X
Chapter 9 Quasi-experimental X X X X
Chapter 10 Non-experimental X X X
Chapter 11 Non-experimental X X
Chapter 12 Non-experimental X X
Chapter 13 Non-experimental X X X
Chapter 14 Non-experimental X X
Introduction
9