Neuman Chapter 6

Strategies of Research Design
Triangulation Qualitative Design Issues

Qualitative and Quantitative Quantitative Design Issues
Orientations Toward Research Conclusion
Substantive problems must thus be translated into the vocabulary of social

inquiry. . . .Working out a way of thinking through the choices and some
appropriate sequence of tasks will allow you to answer a research question.
—Robert Alford, The Craft of Inquiry, p. 25
In 1995 more than 700 people died in a few days in a Chicago heat wave. News reports
and officials lacked answers about why it happened. Public and media discussions of
the disaster disappeared shortly after it happened. Klinenberg (2002) conducted a
“social autopsy” of this “extreme event” in a study using the tools of sociological
inquiry—ethnographic field work, interviews, examination of archival documents
(newspapers, statistical reports, various records, maps), and analysis of statistical data.
The study was designed to answer a question: why and how so many died so quickly.
He used social research to dissect the event and reveal its underlying social, political,
and economic causes. The study informs us about why and how the disaster occurred.
It shows how to design a social research study that answers a significant question
(reasons for the unnecessary deaths of hundreds of people in a few days) but that had
remained unanswered or ignored.
This chapter focuses on issues involved in design- qualitative study, you try to immerse yourself fully
ing a study and developing a strategy to guide in a range of data while being very alert to new
you during the research process. Your strategy for insights throughout the process of gathering data.
designing and conducting a study will vary depend- You may ask how you can best capture the richness,
ing on whether it is primarily quantitative or qual- texture, and feeling of dynamic social life. Of course,
itative. You need to plan a quantitative study in you can mix the features of quantitative and quali-
detail before you collect or analyze the data. tative studies to build on their complementary
You may ask how you can best create a logically strengths. Mixing approaches has advantages but
rigorous design that defines and measures all vari- adds complexity and is more time consuming. We
ables precisely, select a representative sample, col- can see the advantages in triangulation, which is
lect data, and conduct statistical analysis? For a described in the next section.
From Chapter 6 of Social Research Methods: Qualitative and Quantitative Approaches, 7/e. W. Lawrence Neuman.
Copyright © 2011 by Pearson Education. Published by Allyn & Bacon. All rights reserved.
STRATEGIES OF RESEARCH DESIGN
TRIANGU LATION
Surveyors and sailors measure distances between
objects by taking observations from multiple posi-
tions. By observing the object from several differ-
ent angles or viewpoints, the surveyors and sailors
can obtain a good fix on an object’s true location Object
(see Figure 1). Social researchers employ a similar
process of triangulation. In social research, we
build on the principle that we learn more by observ-
ing from multiple perspectives than by looking
from only a single perspective.
Social researchers use several types of trian-
gulation (see Expansion Box 1, Example of Four
Types of Triangulation). The most common type is
triangulation of measure, meaning that we take FIGURE 1 Triangulation: Observing from
multiple measures of the same phenomena. For Different Viewpoints
example, you want to learn about a person’s health.
First, you ask the person to complete a question-
naire with multiple-choice answers. Next you con- test results. Your confidence that you have an accu-
duct an open-ended informal interview. You also rate picture grows from the multiple measures you
ask a live-in partner/caregiver about the person’s used compared to relying on just one, especially if
health. You interview the individual’s physician and each measure offers a similar picture. Differences
together examine his or her medical records and lab you see among the measures stimulates questions
as well.
Triangulation The idea that looking at something
Triangulation of observers is a variation on the
from multiple points of view improves accuracy. first type. In many studies, we conduct interviews
or are the lone observer of events and behavior.
EXPANSION BOX
1
Example of Four Types of Triangulation
TOPIC Theory: Compare how a feminist, a functional,

The amount of violence in popular American films and a symbolic interaction theory explains the forms,
Measures: Create three quantitative measures of causes, and societal results of violence that is in pop-
violence: the frequency (e.g., number of killings, ular films.
punches), intensity (e.g., volume and length of time Method: Conduct a content analysis of a set of
screaming, amount of pain shown in face or body ten popular films, as an experiment to measure the
movement), and level of explicit, graphic display (e.g., responses of experimental subjects to violence in
showing a corpse with blood flowing, amputated each film, to survey attitudes toward film violence
body parts, close-ups of injury) in films. among the movie-going public, and to make field
Observers: Have five different people indepen- observations on audience behavior during and imme-
dently watch, evaluate, and record the forms and diately after showing the films.
degrees of violence in a set of ten highly popular
American films.
Any limitations of a single observer (e.g., lack of but the approaches have complementary strengths.
skill in an area, a biased view on an issue, inatten- A study that combines both tends to be richer and
tion to certain details) become restrictions of the more comprehensive. Mixing them occurs in sev-
study. Multiple observers bring alternative perspec- eral ways:1 by using the approaches sequentially,
tives, backgrounds, and social characteristics. They first one and then the other, or by using them in par-
thereby reduce the limitations. For example, two allel or simultaneously. In the study that opened
people interact with and observe the behavior of this chapter, Klinenberg mixed a statistical analy-
ten 5-year-old children at a child care center. One sis of quantitative data on deaths with interviews
of the observers is a 60-year-old White male pedi- and document analysis. (see Example Box 1,
atrician with 25 years of experience working in a A Multimethod Study).
large city hospital. The other is a 31-year-old
Hispanic female mother of two children who has
QUALITATIVE AN D QUANTITATIVE
6 years of experience as an elementary school
ORI ENTATIONS TOWARD
teacher in a small town. Each observer may notice
RES EARCH
and record different data. Combining what both see
and experience will produce a fuller picture than In all research, we strive to collect empirical data
relying on either one alone. systematically and to examine data patterns so we
Triangulation of theory requires using mul- can better understand and explain social life, yet dif-
tiple theoretical perspectives to plan a study or ferences between research approaches can create
interpret the data. Each theoretical perspective miscommunication and misunderstandings. They
has assumptions and concepts. They operate as a are mutually intelligible; grasping both approaches
lens through which to view the social world. For and seeing how each complements the other simply
example, a study of work relations in a bank could takes more time and effort. Next we will look at
use conflict theory with its emphasis on power dif- some sources of differences.
ferences and inequality. The study could highlight A first difference originates in the nature of the
the pay and working condition inequalities based data itself. Soft data (i.e., words, sentences, photos,
on positions of authority (e.g., manager versus symbols) dictate qualitative research strategies and
teller). The study reveals relevant differences in data collection techniques that differ from hard data
social backgrounds: a middle-aged White male (in the form of numbers) for which quantitative
manager with an MBA and a young African Amer- approaches are used. Such differences may make
ican female teller with an associate’s degree. Next, the tools for a quantitative study inappropriate or
rational choice theory is applied to focus on decision- irrelevant for a qualitative study and vice versa.
making and rational strategies individuals use to Another difference between qualitative and
maximize personal benefits. This perspective high- quantitative research originates in principles about
lights how the bank manager varies the time/effort the research process and assumptions about social
he devotes to various customers depending on life. Qualitative and quantitative research principles
their loan or savings account size. It also presents give rise to different “languages of research” with
a better picture of how the teller invests her time different emphases. In a quantitative study, we rely
and energy differently with various supervisors, more on positivist principles and use a language of
depending on whether she believes they might help variables and hypotheses. Our emphasis is on pre-
her get a promotion. Each perspective guides the cisely measuring variables and test hypotheses. In
study: It identifies relevant data, provides a set of a qualitative study, we rely more on the principles
concepts, and helps to interpret the meaning and from interpretive or critical social science. We speak
significance of the data. a language of “cases and contexts” and of cultural
Triangulation of method mixes the qualitative meaning. Our emphasis is on conducting detailed
and quantitative research approaches and data. Most examinations of specific cases that arise in the nat-
researchers develop an expertise in one approach, ural flow of social life. Interestingly, more female
EXAMPLE BOX
1
A Multimethod Study
Lee and Bean (2007) mixed quantitative and qualita- with race and ethnicity. They interviewed adults of
tive research approaches in a study of multiracial iden- various mixtures of Asian, White, Latino, and Black
tity in the United States. They observed that social races. Based on the interviews, Lee and Bean found
diversity has increased because of growing immigra- that multiracial Blacks were less likely to call them-
tion since 1970, and for the first time in 2000, the selves multiracial than people of other mixed race
United States census offered the option of classifying categories. This restriction is consistent with the U.S.
oneself as multiracial. The new diversity contrasts to historical pattern of the public identifying a person with
the long history of single-race categories and a dom- only some Black heritage as being Black. Persons of
inant White-Black dichotomous racial division. Lee and mixed White and Asian or Latino or Latino-Asian
Bean asked whether multiracial people feel free or heritage had more flexibility. Some mixed Asian-
highly constrained when they pick a single racial- White or Latino-White people self-identified as White
ethnic or multiracial identity. They also asked whether because of public perceptions and a narrow stereo-
selecting a multiracial category on the census form is typical definition of proper Asian or Latino appearance.
a symbolic action or a reflection of a person’s multi- Other White-Asian and White-Latino people said that
racial daily existence. In the quantitative part of the they are proud of their mixed heritage even if it made
study, the authors statistically analyzed 2000 census little difference in their daily encounters. People did
data on the numbers and mixes of people who classi- not stick with one label but claimed different racial-
fied themselves as multiracial. In the qualitative part ethnic backgrounds in different situations. Pulling
of the study, they conducted forty-six in-depth semi- together the quantitative and qualitative findings, Lee
structured interviews with multiracial adults from and Bean suggested that racial-ethnic group bound-
northern and southern California. In the interviews, aries are fading faster for Latinos and Asians than for
Lee and Bean asked how and why a person chose to Blacks. They concluded that a new Black versus non-
identify herself or himself as she or he did, whether Black divide is emerging to replace the old White-Black
that identity changed over time or by context, and division but that Blacks are still in a disadvantaged
about language use and other practices associated position relative to all racial categories.
than male social researchers adopt the qualitative The test of a hypothesis may be more than a simple
approach.2 true or false answer; frequently it includes learning
A third difference between qualitative and that a hypothesis is true for some cases or under
quantitative research lies in what we try to accom- certain conditions but not others. In the heat wave
plish in a study. “The heart of good work”— study, Klinenberg asked whether a person’s social
whether it is quantitative or qualitative—“is a class influenced an outcome: being likely to die
puzzle and an idea” (Abbott, 2003:xi). In all stud- during the heat wave. Using quantitative data,
ies, we try to solve a puzzle or answer a question, he tested the relationship between class and death
but depending on the approach, we do this in dif- rate by comparing the social class of the roughly
ferent ways. In the heat wave study that opened this 700 who died with thousands who did not.
chapter, Klinenberg (2002) asked why so many In many qualitative studies, we often generate
people died. But he also asked how they died, and new hypotheses and describe details of the causal
why some categories of people were greatly affected mechanism or process for a narrow set of cases.
but others were not. In a quantitative study, we Returning to the heat wave study, Klinenberg (2002)
usually try to verify or falsify a relationship or tested existing hypotheses about class and death
hypothesis we already have in mind. We focus on rates. He also developed several new hypotheses as
an outcome or effect found across numerous cases. he looked closely into the mechanism that caused
some to die but not others. He learned that high practice includes advice that comes from the prac-
death rates occurred in poverty- and crime-ridden tical activities of doing specific real-life studies
neighborhoods. More males than females died, and more than a set of restated, ideal rules. This logic
more African Americans died than Latinos or relies heavily on “judgment calls” and “tricks of
Whites. By walking around in different low-income the trade” that active, experienced researchers
neighborhoods and interviewing many people first- share. We learn it best by reading many studies and
hand, he identified the mechanisms of urban isola- being an apprentice researcher and from the folk
tion that accounted for very different heat wave wisdom that passes informally among experienced
survival rates among people of the same social class. researchers. It is like cooking without a written
He examined the social situations of older African recipe—adding a pinch of an ingredient here, stir-
American men and discovered the local social envi- ring until something “looks right,” and adjusting
ronment to be the critical causal mechanism. He while cooking until we reach a certain smell or taste.
also looked at larger forces that created the social You can see the reconstructed logic in the dis-
situations and local environments in Chicago in the tinct research methods section of a quantitative
mid-1990s. research report. In contrast, in qualitative research
A fourth difference between quantitative and reports, you may not see the research method (com-
qualitative studies is that each has a distinct mon for historical-comparative research) discussed
“logic” and path of conducting research. In a or find it mixed with a personal autobiographical
quantitative study, we employ a logic that is sys- account of a particular study (common for field
tematic and follows a linear research path. In a research). The absence of a standard method does
qualitative study, the logic arises from ongoing not make qualitative study less valid; however, it
practice and we follow a nonlinear research path. often requires more time and a different style of
In the next section, we examine the logics and thinking for the newcomer to master.
paths of research.
Linear and Nonlinear Paths
Reconstructed Logic and Logic in Practice
The path is a metaphor for a sequence of things to
How we learn and discuss research tends to follow do: what you finish first or where you have been and
one of two logics.3 The logics summarize the degree what comes next. You can follow a straight, well-
to which our research strategy is explicit, codified, worn, and marked path that has clear signposts and
and standardized. In specific studies, we often mix is where many others have trod before. Alterna-
the two logics, but the proportion of each varies tively, you may follow a path that meanders into
widely by study. unknown territory where few others have gone. The
A reconstructed logic emphasizes using an path has few signs, so you move forward, veer off
explicit research process. Reconstructed logic has to the side, and sometimes backtrack a little before
been “reconstructed” or restated from the many going forward again.
messy details of doing a real-life study into an ide-
alized, formal set of steps with standard practices
and consistent principles, terms, and rules. You can Reconstructed logic A logic of research based on
think of it as a “cleansed model” of how best to do reorganizing, standardizing, and codifying research
a high-quality study. Following this logic is like knowledge and practices into explicit rules, formal
procedures, and techniques; it is characteristic of
cooking by exactly following a printed recipe.
quantitative research.
Thus, the way to conduct a simple random sample
is straightforward and follows a clear step-by-step Logic in practice A logic of research based on an
apprenticeship model and the sharing of implicit
procedure. knowledge about practical concerns and specific expe-
The logic in practice is messy and closer to riences; it is characteristic of qualitative research.
the concrete practice of doing research. Logic in
When using the linear research path, we fol- of meaning, subtle connotations, or contextual dis-
low a fixed sequence of steps that are like a stair- tinctions can be important (see Figure 2 for a
case that leads upward in one direction. By following graphic representation of each path).
a linear path, we move in a direct, narrow, and
straight way toward a conclusion. This pathway
Objectivity and Integrity
toward task completion is the dominant approach
in western European and North American cultures. We try to be fair, honest, truthful, and unbiased in
It is most widely used in quantitative research. By our research activity, yet, we also have opportu-
contrast, a nonlinear research path requires us to nities to be biased, dishonest, or unethical in all
make successive passes through the steps. We may knowledge production including social research.
move forward, backward, and sideways before The two major research approaches address the
advancing again. It is more of a spiral than a straight issue of reducing difficulties and ensuring honest,
staircase. We move upward but slowly and indi- truthful studies in different ways.
rectly. With each cycle or repetition, we may col- In qualitative research, we often try to acquire
lect new data and gain new insights. intimate, firsthand knowledge of the research set-
People who are accustomed to a direct, linear ting. Thus, we do not want to distance ourselves
approach often become impatient with a less direct from the people or events we are studying. Acquir-
cyclical path. Although a nonlinear path is not dis- ing an intimate understanding of a setting does
organized, undefined chaos, the cyclical path appears not mean that we can arbitrarily interject personal
inefficient and without rigor. People who are used opinion, be sloppy about data collection, or use evi-
to a nonlinear path often feel stifled and “boxed in” dence selectively to support our prejudices. Rather,
by a linear approach. To them, a linear path feels we take maximum advantage of personal insight,
artificial or rigid. They believe that this approach inner feelings, and life perspective to understand
prevents them from being naturally creative and social life. We “walk a fine line” between intimacy
spontaneous. and detachment and place personal integrity and
Each path has its strengths. The linear path is honesty at the forefront. Some techniques may help
logical, easy to follow, and efficient. The nonlinear us walk a fine line. One technique is to become
path can be highly effective in creating an authen- highly sensitive to our own views, preconceptions,
tic feeling for understanding an entire setting, for and prior assumptions and then “bracket” them, or
grasping subtle shades of meaning, for integrating put them aside, so we can see beyond them better.
divergent bits of information, and for switching Instead of trying to bury or deny our assumptions,
perspectives. Each path has its own discipline and viewpoints, and values, we find that acknowledg-
rigor. The linear path borrows from the natural ing them and being open about them is best. We can
sciences with their emphasis on logic and preci- then recognize how they might influence us. We try
sion. A nonlinear path borrows devices from the to be forthright and candid in our involvement in
humanities (e.g., metaphor, analogy, theme, motif, the research setting, in dealing with the people in
and irony) and is suited for tasks such as translat- the study, and with any relevant issues that arise.
ing languages, a process in which delicate shades We do this in the way that we conduct the study and
report on the findings.
Personal openness and integrity by the indi-
vidual researcher are central to a qualitative study.
Linear research path Research that proceeds in a
By contrast, in a quantitative study, we stress neu-
clear, logical, step-by-step straight line; often used in
quantitative research. trality and objectivity. In a quantitative study, we
rely on the principle of replication, adhere to stan-
Nonlinear research path Research that proceeds in
a cyclical, iterative, or back-and-forth pattern and is
dardized procedures, measure with numbers, and
often used in qualitative research. analyze the data with statistics.4 In a sense, we
try to minimize or eliminate the subjective human
Reconstructed Logic, Linear Path
Logic in Practice, Nonlinear Path
FIGURE 2 Graphic Representation of Linear and Nonlinear Paths
factor in a quantitative study. As Porter (1995:7, 74) relies more on a dependable, credible researcher and
has argued, her or his personal integrity, self-discipline, and trust-
worthiness.5 Four other forms of validation in quali-
Ideally, expertise should be mechanized and objec- tative research somewhat parallel the objective
tified . . . grounded in specific techniques. . . . This
procedures found in quantitative studies.6
ideal of objectivity is a political as well as scientific
one. Objectivity means rule of law, not of men. It
The first form indicates that the researcher has
implies the subordination of personal interests and carefully evaluated various forms of evidence and
prejudices to public standards. checked them for consistency. For example, a field
researcher listens to and records a student who
The issue of integrity in quantitative research says, “Professor Smith threw an eraser at Professor
mirrors the natural science approach. It relies on Jones.” The researcher must consider the evidence
using an explicit and objective technology, such as carefully. This includes considering what other
making statements in precise neutral terms, using people say about the event. The field researcher also
well-documented standard techniques, and making looks for confirming evidence and checks for inter-
replicable, objective numerical measures. nal consistency. The researcher asks whether the
Quantitative social research shares the hallmarks student has firsthand knowledge of the event, that is,
of natural science validation: explicit, standard pro- directly witnessed it, and asks whether the student’s
cedures; precise numerical measurement; and repli- feelings or self-interest might lead him or her to lie
cation. By contrast, validation in qualitative research (e.g., the student dislikes Professor Smith).
A second form of validation arises from the Preplanned and Emergent Research
great volume of detailed written notes in most qual- Questions
itative studies. In addition to verbatim description
Studies start in many ways, but the usual first step
of the evidence, other documentation includes ref-
is to select a topic.7 We have no formula for how
erences to sources, commentaries by the researcher,
to do this task. Whether we have experience or are
and quotes, photographs, videos, maps, diagrams,
just a beginning researcher, the best guide is to pick
paraphrasing, and counts. The huge volume of infor-
something that interests us. There are many ways to
mation, its great diversity, and its interlocking and
select topics (see Expansion Box 2, Sources of Top-
mutually reinforcing presentation help to validate
ics). We may begin with one topic, but it is too large
its authenticity.
and is only a starting point. We must narrow it into
A third kind of validation comes from other
a focused research question. How we do this varies
observers. Most qualitative researchers work alone,
by whether our study is primarily qualitative or
but many others know about the evidence. For
quantitative. Both kinds of studies work well with
example, we study people in a specific setting who
some topics; we can study poverty by examining
are alive today. Other researchers can visit the same
official statistics, conducting a survey, doing ethno-
setting and talk to the same people. The people
graphic field research, or completing a historical-
we studied can read study details and verify or raise
comparative analysis. Some topics are best suited
questions about it. Likewise, historical-comparative
for a qualitative study (e.g., how do people reshape
researchers cite historical documents, archival
their self-identity through participating in goth
sources, or visual material. By leaving a careful
youth subculture) and others for a quantitative study
“audit trail” with precise citations, others can check
(e.g., how has public opinion on the death penalty
the references and verify sources.
shifted over the past 50 years and whether one’s
A fourth type of truthfulness is created by the
opinion on this issue is influenced by views on
way we publicly disclose results. In a quantitative
related issues or by the amount of exposure the news
study, we adhere to a standard format for writing a
media gives to certain topics).
research report. We explain in detail how we fol-
Most qualitative studies start with a vague or
lowed accepted procedures. We describe each step
loosely defined topic. The specific topic emerges
of the study, display the quantitative data in charts,
slowly during the study, and it may change direc-
graphs, or tables, and make data files available to
tion based on new evidence. This was the case for
others to reanalyze. We offer to answer any ques-
Venkatesh’s study (2008). He began with an inter-
tions about the study. In a qualitative study, we can-
est in studying poverty in an inner-city housing
not publicly display or share the many mountains
project but shifted to studying a drug-selling gang.
of detailed notes, recorded interviews, photos, or
Focusing on a specific research question continues
original source materials in a research report. They
while we gather data. Venkatesh increasingly
might fill an entire room! Instead, we “spin a web”
focused his topic of gang activity into sharper ques-
of interlocking details and use tightly cross-refer-
tions: How and why did gangs in a low-income
enced material. Through our writing and presen-
housing project sustain an underground economy
tation, we provide sufficient texture and detail to
and provide housing project residents with protec-
build an “I-was-there” sense within readers. By pro-
tion and aid services?
viding rich specific descriptions supplemented with
Flexibility in qualitative research encourages
maps, photos, and verbatim quotations, we convey
us to continuously focus throughout a study. An
an intimate knowledge of a setting. We build a
emergent research question may become clear only
sense of shared familiarity in readers. A skilled
during the research process. We can focus and
qualitative researcher can recreate the visual
refine the research question after we gather some
images, voices, smells, sounds, tensions, and entire
data and begin a preliminary analysis. In many
atmosphere that existed by referring to the moun-
qualitative studies, the most important issues and
tains of empirical evidence.
EXPANSION BOX
2
Sources of Topics
1. Personal experience. You can choose a topic based support capital punishment, nor has anyone mapped
on something that happens to you or those you the geographic location of these religious groups.
know. For example, while you work a summer job at Your knowledge of the field suggests a topic for a
a factory, the local union calls a strike. You do not research project: beliefs about capital punishment
have strong feelings either way, but you are forced and religion in different regions.
to choose sides. You notice that tensions rise. Both 4. Solving a problem. Applied research topics often
management and labor become hostile toward each begin with a problem that needs a solution. For
other. This experience suggests unions or organized example, as part of your job as a dorm counselor,
labor as a topic. you want to help college freshmen establish friend-
2. Curiosity based on something in the media. Some- ships with each other. Your problem suggests friend-
times you read a newspaper or magazine article or ship formation among new college students as a
see a television program that leaves you with ques- topic.
tions. What you read raises questions or suggests 5. Social premiums. This is a term suggested by Sin-
replicating what others’ research found. For example, gleton and colleagues (1988:68). It means that some
you read a Newsweek article on people who are topics are “hot” or offer an opportunity. For example,
homeless, but you do not really know much about you read that a lot of money is available to conduct
who they are, why they are homeless, whether this research on nursing homes, but few people are inter-
has always been a problem, and so forth. This sug- ested in doing so. Your need of a job suggests nurs-
gests homeless people as a topic. ing homes as a topic.
3. The state of knowledge in a field. Basic research is 6. Personal values. Some people are highly committed
driven by new research findings and theories that to a set of religious, political, or social values. For
push at the frontiers of knowledge. As theoretical example, you are strongly committed to racial equal-
explanations are elaborated and expanded, certain ity and become morally outraged whenever you
issues or questions need to be answered for the field hear about racial discrimination. Your strong per-
to move forward. As such issues are identified and sonal belief suggests racial discrimination as a topic.
studied, knowledge advances. For example, you read 7. Everyday life. Potential topics can be found through-
about attitudes toward capital punishment and real- out everyday life in old sayings, novels, songs, sta-
ize that most research points to an underlying belief tistics, and what others say (especially those who
in the innate wickedness of criminals among capital disagree with you). For example, you hear that the
punishment supporters. You notice that no one has home court advantage is very important in basket-
yet examined whether people who belong to certain ball. This statement suggests home court advantage
religious groups that teach such a belief in wickedness as a topic for research.
most interesting questions become clear only after Typical qualitative research questions include
we become immersed in the data. We need to these: How did a certain condition or social situation
remain open to unanticipated ideas, data, and originate? How do people, events, and conditions
issues. We should periodically reevaluate our focus sustain a situation over time? By what processes
early in a study and be ready to change direction does the situation change, develop, or end? Another
and follow new lines of evidence. At the same time, type of question seeks to confirm existing beliefs
we must exercise self-restraint and discipline. If we or assumptions (e.g., do Southern and Northern
constantly change the focus of our research with- Whites act differently around people of other races
out end, we will never complete a study. As with as those in McDermott’s [2006] study of working
most things, a balance is required. class neighborhoods in Atlanta and Boston). A last
type of research question tries to discover new

ideas.8
EXPANSION BOX
3
Techniques for Narrowing a Topic
In a quantitative study, we narrow a topic into
into a Research Question
a focused question as a discrete planning step before
we finalize the study design. Focusing the question
1. Examine the literature. Published articles are excel-
is a step in the process of developing a testable lent sources of ideas for research questions. They are
hypothesis (to be discussed later). It guides the usually at an appropriate level of specificity and sug-
study design before you collect any data.9 gest research questions that focus on the following:
In a qualitative study, we can use the data to a. Replicating a previous research project exactly or
help narrow the focus. In a quantitative study, we with slight variations.
must focus without the benefit of data and use other b. Exploring unexpected findings discovered in pre-
techniques. After picking a topic, we ask ourselves: vious research.
What is it about the topic that is of greatest interest? c. Following suggestions an author gives for future
For a topic about which we know little, we must research at the end of an article.
first acquire background knowledge by reading d. Extending an existing explanation or theory to a
new topic or setting.
studies about the topic. Reading the research liter-
e. Challenging the findings or attempting to refute
ature can stimulate many ideas for how to focus a a relationship.
research question. f. Specifying the intervening process and consider-
In most quantitative studies, research ques- ing any linking relations.
tions refer to relationships among a small number 2. Talk over ideas with others.
of variables. This means that we should list vari- a. Ask people who are knowledgeable about the
ables as we try to focus the topic into a research topic for questions about it that they have
question (see Expansion Box 3, Techniques thought of.
for Narrowing a Topic into a Research Question). b. Seek out those who hold opinions that differ from
For example, the question what causes divorce? is yours on the topic and discuss possible research
not a good research question. A better one is, is age questions with them.
3. Apply to a specific context.
at marriage associated with divorce? The second
a. Focus the topic onto a specific historical period or
question has two variables: age of marriage and time period.
whether or not a divorce occurred (also see Example b. Narrow the topic to a specific society or geo-
Box 2, Examples of Bad and Good Research Ques- graphic unit.
tions). c. Consider which subgroups or categories of people/
Personal experience can suggest topics. Per- units are involved and whether there are differ-
haps personal experience suggests people released ences among them.
from prison as a topic as it did for Pager (2007). We 4. Define the aim or desired outcome of the study.
can read about former inmates and their reentry and a. Will the research question be for an exploratory,
about probation in dozens of books and hundreds explanatory, or descriptive study?
of articles. A focused research question might be b. Will the study involve applied or basic research?
whether it is more difficult for someone who has a
nonviolent criminal record to get a job offer than
someone without a criminal record. This question is
more specific in terms of type of criminal record how does racial category (Black versus White) and
and the specific outcome for a former prisoner. It whether a person had a criminal record affect the
focuses on two variables, whether a person has a chances of getting a job? Did race make a differ-
criminal record and whether the person gets a job ence, did being a former prisoner make a difference,
offer. A common type of research question asks did the two factors operate separately, cancel out
which factor among several had the most significant one another, or intensify one another in their impact
impact on an outcome. We might ask, as Pager did, on getting a job offer?
EXAMPLE BOX
2
Examples of Bad and Good Research Questions
BAD RESEARCH QUESTIONS GOOD RESEARCH QUESTIONS

Not Empirically Testable, Nonscientific Questions Exploratory Questions
Should abortion be legal? Has the incidence of new forms of child abuse
Is it right to have capital punishment? appeared in Wisconsin in the past 10 years?
General Topics, Not Research Questions Descriptive Questions

Treatment of alcohol and drug abuse Is child abuse, violent or sexual, more common in
Sexuality and aging families that have experienced a divorce than in
intact, never-divorced families?
Set of Variables, Not Questions
Are the children raised in impoverished households
Capital punishment and racial discrimination
more likely to have medical, learning, and social-
Urban decay and gangs
emotional adjustment difficulties than children who
Too Vague, Ambiguous are not living in poverty?
Do police affect delinquency?
Explanatory Questions
What can be done to prevent child abuse?
Does the emotional instability created by experienc-
Need to Be Still More Specific ing a divorce increase the chances that divorced
Has the incidence of child abuse risen? parents will physically abuse their children?
How does poverty affect children? Is a lack of sufficent funds for preventive treatment
What problems do children who grow up in poverty a major cause of more serious medical problems
experience that others do not? among children raised in families in poverty?
We also want to specify the universe to which we must narrow the question to fit the practical
we generalize answers to a research question. All limitations.
research questions and studies apply to some cate- Time is always a consideration. However, it is
gory of people, organizations, or other units. The very difficult to estimate the time required for a
universe is the set of all units that the research ques- study. A specific research question, the research
tion covers or to which we can generalize. For techniques used, the complexity of the study, and
example, in Pager’s (2007) study, his units were the amount and types of data we plan to collect all
individuals, specifically young White and Black affect the amount of time required. Experienced
men. The universe to which we might generalize researchers are the best source for getting good esti-
his findings includes all U.S. males in their twen- mates of time requirements.
ties of these two racial categories. Cost is another limitation, and we cannot
As we refine a topic into a research question and answer some research questions because of the great
design a study, we also need to consider practical expense involved. For example, our research
limitations. Designing the perfect research project question asks whether sports fans develop strong
is an interesting academic exercise, but if we expect positive feelings toward team mascots if the team
to carry out a study, practical limitations must shape has a winning season but negative feelings if it has
its design. Major limitations include time, costs,
access to resources, approval from authorities, ethi-
Universe The entire category or class of units
cal concerns, and expertise. If we have 10 hours a that is covered or explained by a relationship or
week for 5 weeks to conduct a research project but hypothesis.
answering the research question will require 2 years,
TA B L E 1 Quantitative Research versus Qualitative Research
QUANTITATIVE RESEARCH QUALITATIVE RESEARCH
Researchers test hypotheses that are stated Researchers capture and discover meaning once
at the beginning. they become immersed in the data.
Concepts are in the form of distinct variables. Concepts are in the form of themes, motifs,
generalizations, and taxonomies.
Measures are systematically created before data Measures are created in an ad hoc manner and are
collection and are standardized. often specific to the individual setting or researcher.
Data are in the form of numbers from precise Data are in the form of words and images from
measurement. documents, observations, and transcripts.
Theory is largely causal and is deductive. Theory can be causal or noncausal and is often inductive.
Procedures are standard, and replication is Research procedures are particular, and replication is
frequent. very rare.
Analysis proceeds by using statistics, tables, or Analysis proceeds by extracting themes or generalizations
charts and discussing how what they show relates from evidence and organizing data to present a coherent,
to hypotheses. consistent picture.
a losing season. To examine the question for all pay for another person’s services, the research
sports teams across a nation across a decade would question may not be practical.
require a great investment of time and money. The In sum, qualitative and quantitative studies
focus could be narrowed to one sport (football), to share a great deal, but they differ on several design
sports played in college, and to student fans at just issues: logic, research path, mode of verification, and
four colleges across three seasons. As with time, way to arrive at a research question (see Table 1). In
experienced researchers can help provide estimates addition, the research approaches speak different
of the cost to conduct a study. “languages” and emphasize distinct study design
Access to resources is a common limitation. features, issues that we consider in the next section.
Resources include expertise, special equipment,
and information. For example, a research question QUALITATIVE DES IGN ISS U ES
about burglary rates and family income in many
different nations is nearly impossible to answer. The Language of Cases and Contexts
Data on burglary and income are not collected or Most qualitative studies involve a language of cases
available for many countries. Other questions and contexts, employ bricolage (discussed later in
require the approval of authorities (e.g., to see this chapter), examine social processes and cases in
medical records) or involve violating basic ethical their social context, and study interpretations or
principles (e.g., lying to a person and endangering meanings in specific socio-cultural settings. We
her or him). Our expertise or background as examine social life from multiple points of view and
researchers is also a limitation. Answering some explain how people construct identities. Only rarely
research questions involves the use of data collec- do we use variables, test hypotheses, or create pre-
tion techniques, statistical methods, knowledge of cise measures in the form of numbers.
a foreign language, or skills we may not have. Most qualitative studies build on the assump-
Unless we acquire the necessary training or can tion that certain areas of social life are intrinsically
qualitative. For this reason, qualitative data are not theoretical questions may arise that suggest future
imprecise or deficient but are very meaningful. observations. In this way, we tailor new data to
Instead of trying to convert fluid, active social life answer theoretical questions that arose only from
into variables or numbers, we borrow ideas and thinking about previous data.
viewpoints from the people we study and situate In grounded theory, we build from specific
them in a fluid natural setting. Instead of variables, observations to broader concepts that organize
we examine motifs, themes, distinctions, and per- observational data and then continue to build prin-
spectives. Most often, our approach is inductive ciples or themes that connect the concepts. Com-
and relies on a form of grounded theory. pared to other ways of theorizing, grounded theory
Qualitative data may appear to be soft, intan- tends to be less abstract and closer to concrete obser-
gible, and elusive. This does not mean that we vations or specific events. Building inductively from
cannot capture them. We gather qualitative data the data to theory creates strong data-theory link-
by documenting real events, recording what actual ages. However, this can be a weakness as well. It
people say (with words, gestures, and tone), may make connecting concepts and principles
observing specific behaviors, examining written across many diverse settings difficult, and it may
documents, and studying visual images. These are slow the development of concepts that build toward
specific, concrete aspects of the social world. As creating general, abstract knowledge. To counter-
we closely scrutinize photos or videotapes of act this weakness, we become familiar with the con-
people or social events, we are looking at “hard” cepts and theories developed in other studies to
physical evidence.10 The evidence is just as “hard” apply shared concepts when appropriate and to note
and physical as the numeric measures of attitudes, any similarities and differences. In this way, we can
social pressure, intelligence, and the like found in establish cross-study interconnections and move
a quantitative study. toward generalized knowledge.
Grounded Theory The Context Is Critical

In qualitative research, we may develop theory dur- In qualitative research, we usually emphasize the
ing the data collection process. This largely induc- social context because the meaning of a social action,
tive method means that we are building theory from event, or statement greatly depends on the context in
data or ground the theory in the data. Grounded the- which it appears. If we strip social context from an
ory adds flexibility and allows the data and theory event, social action, or conversation, it is easy to dis-
to interact. This process also helps us remain open tort its meaning and alter its social significance.
to the unexpected. We can change direction of study Social context includes time context (when
and even abandon the original research question in something occurs), spatial context (where something
the middle of a project if we discover something occurs), emotional context (the feelings regarding
new and exciting.11 how something occurs), and socio-cultural context
We build theory by making comparisons. For (the social situation and cultural milieu in which
example, we observe an event (e.g., a police officer something occurs). For example, a social activity (a
confronting a speeding motorist who has stopped). card game, sexual act, or disagreement) occurs late
We may ponder questions and look for similarities at night on the street in a low-income area of a large
and differences. When watching a police officer, we city, a setting for drug use, fear and anger, violent
ask: Does the police officer always radio in the car’s crime, and prostitution within a cultural milieu of
license number before proceeding? After radioing extreme racial-economic inequality. The same
the car’s location, does the officer ask the motorist activity occurs midday in the backyard of a large
to get out of the car or some times casually walk up house in an affluent suburban neighborhood in a
to the car and talk to the seated driver? When we social setting of relaxation and leisure, surrounded
intersperse data collection and theorizing, new by trust and emotional closeness, and within a
cultural milieu of established affluence and privi- in the context. He or she may treat the same behav-
lege. The context will significantly color the activ- ior (e.g., voting for a presidential candidate) dif-
ity’s meaning. With different contextual meanings, ferently depending on the social context in which
the same activity or behavior may have different it occurs (see Example Box 3, Example of Impor-
consequences. tance of Context for Meaning).
In a quantitative study, we rarely treat context Context goes beyond social events, behaviors,
as important. We often strip it away as being and statements to include physical objects. One
“messy” or just “noise” and instead concentrate on handgun could be an art object, part of a recre-
precise counts or numerical measures. Thus, what ational hobby, a key element in committing a vio-
a qualitative study might treat as essential may lent crime, evidence of an irresponsible parent, a
be seen as irrelevant noise in a quantitative study. suicide facilitator, or a means of social peace and
For example, if a quantitative study counts the community protection, each depending on the con-
number of votes across time or cultures, a qualita- text. Without including the surrounding context, we
tive researcher might consider what voting means cannot assign meaning to an object.
EXAMPLE BOX
3
Example of the Importance of Context for Meaning
“Voting in a national election” has different meanings instead, they are tied to ethnic or religious groups
in different contexts: and are part of a person’s ethnic-religious identity.
Ethnic and religious group identities are very strong.
1. A one-party dictatorship with unopposed candi- They affect where one lives, where one works, whom
dates, where people are required by law to vote. one marries, and the like. Voting follows massive
The names of nonvoters are recorded by the police. parades and week-long community events organized
Nonvoters are suspected of being antigovernment by ethnic and religious groups.
subversives. They face fines and possible job loss for 5. A context in which one political party is very pow-
not voting. erful and is challenged by one or two very small,
2. A country in the midst of violent conflict between weak alternatives. The one party has held power for
rebels and those in power. Voting is dangerous the past 60 years through corruption, bribery, and
because the armed soldiers on either side may shoot intimidation. It has the support of leaders through-
voters they suspect of opposing their side. The out- out society (in religious organizations, educational
come of the vote will give power to one or the other institutions, businesses, unions, and the mass media).
group and dramatically restructure the society. Any- The jobs of anyone working in any government job
one over the age of 16 can vote. (e.g., every police officer, post office clerk, school-
3. A context in which people choose between a dozen teacher, and garbage collector) depend on the polit-
political parties of roughly equal power that repre- ical party staying in power.
sent very different values and policies. Each party has 6. A context in which the choice is between two parties
a sizable organization with its own newspapers, social with little difference between them. People select
clubs, and neighborhood organizers. Election days are candidates primarily on the basis of television adver-
national holidays when no one has to work. A person tising. Candidates pay for advertising with donations
votes by showing up with an identification card at any by wealthy people or powerful organizations. Voting
of many local voting locations. Voting itself is by secret is a vague civic obligation that few people take seri-
ballot, and everyone over age 18 can vote. ously. Elections are held on a workday. In order to
4. A context in which voting is conducted in public by vote, a person must meet many requirements and
White males over age 21 who have regular jobs. register to vote several weeks in advance. Recent
Family, friends, and neighbors see how one another immigrants and anyone arrested for a crime are pro-
vote. Political parties do not offer distinct policies; hibited from voting.
Bricolage sis becomes the sequence of events: what occurred

first, second, third, and so on. This focus on process
A bricoleur is someone who has learned to be adept
helps to reveal how an issue evolves, a conflict
in diverse areas, can draw on a variety of sources,
emerges, or a social relationship develops.
and makes do with whatever is at hand.12 The
bricolage technique involves working with one’s
hands and combining odds and ends in a practical,
Interpretation
skilled, and inventive way to accomplish a task.
A successful bricoleur possesses a deep knowledge To interpret means to assign significance or coher-
of materials, a set of esoteric skills, and a capacity ent meaning. In quantitative research, meaning
to combine or create flexibly. The typical bricoleur comes from using numbers (e.g., percentages or sta-
is often a highly inventive and skilled craftsperson, tistical coefficients), and we explain how the numer-
repairperson, or jack-of-all-trades. ical data relate to the hypotheses. Qualitative studies
A qualitative study draws on a variety of skills, rarely include tables with numbers. The only visual
materials, and approaches as needed. This usually presentations of data may be maps, photographs, or
happens when we are unable to anticipate the need diagrams showing how ideas are related. We instead
for them. The process of mixing diverse source weave the data into discussions of the ideas’ signif-
materials, applying disparate approaches, and assem- icance. The data are in the form of words, including
bling bits and pieces into a whole is analogous to quotes or descriptions of particular events. Any
the bricolage of a skilled craftsperson who is able numerical information is supplementary to the tex-
to create or repair many things by using whatever tual evidence.
is available at the time. Qualitative studies give data meaning, trans-
late them, or make them understandable. We begin
with the point of view of the people we study and
The Case and Process then find out how they see the world and define
situations. We learn what events, behaviors, and
We can divide all empirical social research into two
activities mean for them. To begin qualitative inter-
groups: case study (with one or a few cases) or
pretation, we first must learn the meanings of things
cross-case (comprising many cases).13 Most qual-
for the people we are studying.14
itative studies use a “case-oriented approach [that]
People who create social activities and behav-
places cases, not variables, center stage” (Ragin,
ior have personal reasons or motives for what they
1992a:5). Thus, we examine many aspects of a few
do. This is first-order interpretation. As we dis-
cases. The intensive, in-depth study a handful of
cover and reconstruct this first-order interpretation,
cases replaces the extensive, surface-level study of
it becomes a second-order interpretation because
numerous cases as is typical in quantitative
we come from the outside to discover what has
research. Often a case-oriented analysis emphasizes
occurred. In a second-order interpretation, we elicit
contingencies in “messy” natural settings (i.e., the
an underlying coherence or sense of meaning in the
co-occurrence of many specific factors and events
in one place and at one time). Rather than precise
measures of a huge number of cases, as is typical
Bricolage Improvisation by drawing on diverse
of quantitative research, we acquire in-depth of materials that are lying about and using them in
knowledge and an astute insight into a small num- creative ways to accomplish a pragmatic task.
ber of cases.
First-order interpretation Interpretations from
The study of cases tends to produce complex the point of view of the people being studied.
explanations or interpretations in the form of an
Second-order interpretation Qualitative inter-
unfolding plot or a narrative story about particular pretations from the point of view of the researcher
people or specific events. This makes the passage who conducted a study.
of time integral to the explanation. Often the empha-
data. Meaning develops only in relation to a large set a variable; it can take the value of never married
of other meanings, not in a vacuum. In a second- single, married, divorced, or widowed. Type of
order interpretation, we place the human action crime committed is a variable; it can take values of
being studied into a “stream of behavior” or events robbery, burglary, theft, murder, and so forth. Fam-
to which it is related: its context. ily income is a variable; it can take values from zero
If we were to adopt a very strict interpretive to billions of dollars. A person’s attitude toward
approach, we might stop at a second-order inter- abortion is a variable; as a woman’s basic right can
pretation, that is, once we understand the signifi- range from strongly favoring legal abortion to
cance of the action for the people we study. Most strongly believing in the sanctity of fetal life.
qualitative researchers go further. They want to A variable’s values or categories are its attri-
generalize or link the second-order interpretation butes. It is easy to confuse variables with attributes.
to a theory or general knowledge. They move to The confusion arises because one variable’s attri-
a broad level of interpretation, or third-order bute can itself be a separate variable in its own right
interpretation by which they assign general theo- with only a slight change in definition. This rests on
retical significance to the data. a distinction between concepts that vary and the
Because interpreting social meaning in context conditions within concepts that vary. For example,
is often a major purpose and outcome of qualitative “male” is not a variable; it describes a category of
studies, keep in mind that the three steps or orders gender. Male is an attribute of the variable gender,
of interpretation help provide a way to organize the yet a related idea, degree of masculinity, is a
research process. variable. It describes the intensity or strength of
attachment to a set of beliefs, orientations, and
behaviors that are associated with the concept of
QUANTITATIVE DES IGN ISS U ES masculine within a culture. Likewise, “married” is
not a variable; it is an attribute of the variable
The Language of Variables and Hypotheses
marital status. Related ideas such as number of
Variation and Variables. Simply defined, a variable years married or depth of commitment to a mar-
is a concept that varies. In quantitative research, we riage are variables. In a third example, “robbery” is
use a language of variables and relationships among not a variable; but an attribute of the variable type
variables. of crime. Number of robberies, robbery rate,
Previously, we discussed two types of con- amount taken during a robbery, and type of robbery
cepts: those that refer to a fixed phenomenon (e.g., are all variables because they vary or take on a
the ideal type of bureaucracy) and those that vary in range of values.
quantity, intensity, or amount (e.g., amount of edu- In quantitative research, we redefine all con-
cation). Variables are this second type of concept cepts into the language of variables. As the examples
and measures of the concepts. of variables and attributes illustrate, the redefini-
A variable must have two or more values. Once tion often requires only a slight change in defi-
we become aware of them, we see variables every- nition. Concepts are the building blocks of theory;
where. For example, gender is a variable; it can take they organize thinking about the social world. Clear
one of two values: male or female. Marital status is concepts with careful definitions are essential in
theory.
Third-order interpretation Qualitative interpreta- Types of Variables. As we focus on causal rela-

tions made by the readers of a research report. tions among variables, we usually begin with an
Variable A concept or its empirical measure that effect and then search for its cause(s). We can clas-
can take on multiple values. sify variables depending on their location in a causal
Attributes The categories or levels of a variable. relationship or chain of causality. The cause variable,
or the force or condition that acts on something else,
is the independent variable. The variable that is affects suicide (dependent variable). Specifying
the effect, result, or outcome of another variable the chain of causality makes the linkages in a
is the dependent variable. The independent vari- theory clearer and helps a researcher test complex
able is “independent of” prior causes that have explanations.15
acted on it whereas the dependent variable depends Simple theories have one dependent and
on the cause. one independent variable whereas complex ones
It is not always easy to determine whether a can contain dozens of variables with multiple
variable is independent or dependent. Two ques- independent, intervening, and dependent vari-
tions can help to identify the independent variable. ables. For example, a theory of criminal behavior
First, does it come before other variables in time? (dependent variable) identifies four independent
Independent variables must come before any other variables: an individual’s economic hardship,
type. Second, if two variables occur at the same opportunities to commit crime easily, membership
time, does one variable have an impact on another in a deviant subgroup that does not disapprove of
variable? Independent variables affect or have an crime, and lack of punishment for criminal acts.
impact on other variables. We often phrase research A multicause explanation usually specifies which
topics and questions in terms of the dependent independent variable has the most significant
variable because dependent variables are the phe- causal effect.
nomena we want to explain. For example, an exam- A complex theoretical explanation has a string
ination of the reasons for an increase in the crime of multiple intervening variables. For example, fam-
rate in Dallas, Texas would have the dependent ily disruption causes lower self-esteem among chil-
variable as the crime rate in Dallas. dren, which causes depression, which causes poor
A simple causal relationship requires only an grades in school, which causes reduced prospects
independent and a dependent variable. A third vari- for a good job, which causes a lower adult income.
able type, the intervening variable, appears in The chain of variables is family disruption (indepen-
more complex causal relations. Coming between dent), childhood self-esteem (intervening), depres-
the independent and dependent variables, this vari- sion (intervening), grades in school (intervening), job
able helps to show the link or mechanism between prospects (intervening), adult income (dependent).
them. Advances in knowledge depend not only on Two theories on the same topic can differ as to
documenting cause-and-effect relationships but the number of independent variables. In addition,
also on specifying the mechanisms that account for theories might agree about the independent and
the causal relation. In a sense, the intervening vari- dependent variables but differ on the intervening
able acts as a dependent variable with respect to the variable or causal mechanism. For example, two
independent variable and acts as an independent theories say that family disruption causes lower
variable toward the dependent variable. adult income, each for different reasons. One theory
For example, French sociologist Émile Durk-
heim developed a theory of suicide that specified a
causal relationship between marital status and sui-
cide rate. Durkheim found evidence that married Independent variable A type of variable that pro-
people are less likely to commit suicide than single duces an effect or results on a dependent variable in a
causal hypothesis.
people. He believed that married people have more
social integration (i.e., feelings of belonging to a Dependent variable The effect or result variable
that is caused by an independent variable in a
group or family). He thought that a major cause of
causal hypothesis.
one type of suicide was that people lacked a sense
of belonging to a group. Thus, his theory can be Intervening variable A variable that comes logi-
cally or temporally after the independent variable
restated as a three-variable relationship: marital sta- and before the dependent variable and through
tus (independent variable) causes the degree of which their causal relation operates.
social integration (intervening variable), which
holds that disruption encourages children to join

deviant peer groups, which are not socialized to the
EXPANSION BOX 4
Five Characteristics of Casual Hypotheses
norms of work and thrift. Another theory empha-
sizes the impact of the disruption on childhood
1. They have at least two variables.
depression and poor academic performance. In the 2. They express a causal or cause–effect relationship
second theory, depression and limited school learn- between the variables.
ing directly cause poor job performance. 3. They can be expressed as a prediction or an expected
In one study, we usually test only one or a few future outcome.
parts of a causal chain. For example, a research 4. They are logically linked to a research question and
project examining six variables may take the six a theory.
from a large, complex theory with two dozen vari- 5. They are falsifiable; that is, they are capable of being
ables. Explicit links to a larger theory strengthen tested against empirical evidence and shown to be
and clarify a research project. true or false.
Causal Theory and Hypotheses

The Hypothesis and Causality. A causal hypoth- connect the hypothesis to a research question and
esis is a proposition to be tested or a tentative state- to a broader theory; after all, we test hypotheses to
ment of a relationship between two variables. answer the research question or to find empirical
Hypotheses are guesses about how the social world support for a theory. Statements that are logically
works; they are stated in a value-neutral form. or necessarily true, or questions that are impossible
Kerlinger (1979:35) noted that, to answer through empirical observation (e.g., What
is the “good life”? Is there a God?) are not scientific
Hypotheses are much more important in scientific
hypotheses.
research than they would appear to be just by know-
ing what they are and how they are constructed. We can state causal hypotheses in several ways.
They have a deep and highly significant purpose of Sometimes we use the word cause, but it is not nec-
taking man out of himself. . . . Hypotheses are essary. For example, we can state a causal hypoth-
powerful tools for the advancement of knowledge, esis between religious attendance and a reduced
because, although formulated by man, they can be likelihood of divorce in ten different ways (see
tested and shown to be correct or incorrect apart Example Box 4, Ways to State Causal Relations).
from man’s values and beliefs. In scientific research, we avoid using the term
proved when talking about testing hypotheses. Jour-
A causal hypothesis has five characteristics
nalism, courts of law, and advertisements use the
(see Expansion Box 4, Five Characteristics of
word proof, but a research scientist almost never
Causal Hypotheses). For example, we can restate
uses it. A jury says that the evidence “proves” some-
the hypothesis that attending religious services
one guilty, or a television commercial will state,
reduces the probability of divorce as a prediction:
“Studies prove that our aspirin cures headaches
Couples who attend religious services frequently
the fastest.” This is not the language of scientific
have a lower divorce rate than do couples who rarely
research. In science, we recognize that knowledge
attend religious services. We can test the prediction
is tentative and that creating knowledge is an ongo-
against the empirical evidence. We should logically
ing process that avoids premature closure. The word
proof implies finality, absolute certainty, or some-
thing that does not need further investigation. It is
Causal hypothesis A statement of a causal expla- too strong a term for the cautious world of science.
nation or proposition that has at least one indepen-
dent and one dependent variable and has yet to be
We might say that the evidence supports or con-
empirically tested. firms, but does not prove, the hypothesis. Even after
hundreds of studies show the same results, such as
contention. Theorists and researchers constantly cre-

EXAMPLE BOX
4 ate new hypotheses to challenge those that have
Ways to State Casual Relations
received support (see Figure 3). From Figure 3 we
see that in 2010, three hypotheses are in contention,
Religious attendance causes reduced divorce.
Religious attendance leads to reduced divorce.
but from 1970 to 2010, eleven hypotheses were con-
Religious attendance is related to reduced divorce. sidered, and over time, eight of them were rejected
Religious attendance influences the reduction of in one or more tests.
divorce. Scientists are a skeptical group. Supporting a
Religious attendance is associated with reduced hypothesis in one study is not sufficient for them to
divorce. accept it. The principle of replication says that a
Religious attendance produces reduced divorce. hypothesis needs several tests with consistent and
Religious attendance results in reduced divorce. repeated support before it can gain broad accept-
If people attend religious services, then the likelihood ance. Another way to strengthen confidence in a
of divorce will be reduced. hypothesis is to test related causal linkages in the
The higher religious attendance, the lower the like-
theory from which it comes.
lihood of divorce.
Religious attendance reduces the likelihood of
As scientists, we accept the strongest contender
divorce. with the greatest empirical support as the best expla-
nation at the time. The more alternatives we test a
hypothesis against, the more confidence we have
in it. Some tests are called crucial experiments or
the link between cigarette smoking and lung cancer, crucial studies. This is a type of study whereby
scientists do not say that we have absolute proof.
Instead we can say that overwhelming evidence, or two or more alternative explanations for some phe-
all studies to date, support or are consistent with nomenon are available, each being compatible with
the empirically given data; the crucial experiment
the hypothesis. Scientists never want to close off
is designed to yield results that can be accounted
the possibility of discovering new evidence that for by only one of the alternatives, which is thereby
might contradict past findings. They do not want to shown to be “the correct explanation.” (Kaplan,
cut off future inquiry or stop exploring intervening 1964:151–152)
mechanisms. History contains many examples of
relationships that people once thought to be proved Thus, the infrequent crucial experiment is an impor-
but were later found to be in error. We can use proof tant test of theory. Hypotheses from two different
when referring to logical or mathematical relations, theories confront each other in crucial experiments,
as in a mathematical proof, but not for empirical and one is knocked out of the competition. It is rare,
research. but significant, when it occurs.
Testing and Refining a Hypothesis. Knowledge Types of Hypotheses. Hypotheses are links in
rarely advances on the basis of one test of a single a theoretical causal chain and are used to test the
hypothesis. In fact, researchers can get a distorted direction and strength of a relationship between vari-
picture of the research process by focusing on a single ables. When a hypothesis defeats its competitors,
study that tests one hypothesis. Knowledge develops it supports the researcher’s explanation. A curious
over time as many researchers across the scientific aspect of hypothesis testing is that researchers treat
community test many hypotheses. It slowly grows
from shifting and winnowing through many hypothe-
ses. Each hypothesis represents an explanation of a Crucial experiment A direct comparison and eval-
uation of competing explanations of the same
dependent variable. If the evidence fails to support phenomenon designed to show that one is supe-
some hypotheses, they are gradually eliminated from rior to the other.
consideration. Those that receive support remain in
1970
1 2 3 4 5
There are five possible hypotheses.
1980
Test
Reject Remain under
Two of the original five hypotheses
Consideration
are rejected.
A new one is developed.
1 3 2 4 5 6
1990
Test
Reject Remain under
Two hypotheses are rejected.
Consideration
Two new ones are developed.
2 5 4 6 7 8
2000
Test
Reject Remain under
Three hypotheses are rejected.
Consideration
A new one is developed.
4 7 6 8 9
2010
Test
Reject Remain under
One hypothesis is rejected.
Consideration
Two new ones are developed.
9 8 10 11
FIGURE 3 How the Process of Hypotheses Testing Operates over Time
evidence that supports a hypothesis differently from and with the use of null hypotheses (see later in this
evidence that opposes it: They give negative evi- section).
dence more importance. The idea that negative evi- Recall the preceding discussion of proof.
dence is critical when evaluating a hypothesis comes We never prove a hypothesis; however, we can dis-
from the logic of disconfirming hypotheses.16 It is prove it. With supporting evidence, we can say only
associated with Karl Popper’s idea of falsification that the hypothesis remains a possibility or that
it is still being considered. Negative evidence is
Logic of disconfirming hypothesis The logic for the more significant. With it, the hypothesis becomes
null hypothesis based on the idea that confirming “tarnished” or “soiled” because a hypothesis makes
empirical evidence makes a weak case for the exis- predictions. Negative and disconfirming evidence
tence of a relationship; instead of gathering support- shows that the predictions are wrong. Positive or
ing evidence, testing that no relationship exists
provides more cautious, indirect support for its pos-
confirming evidence for a hypothesis is less criti-
sible existence. cal because various alternative hypotheses may
make the same prediction. When we find confirm-
ing evidence for a prediction, we may elevate one the tested relationship does not exist. This implies
explanation over its alternatives that could also that the alternative hypothesis is false. On the
have confirming evidence. other hand, if we find evidence to reject the null
For example, a man stands on a street corner hypothesis, the alternative hypotheses remain a
with an umbrella and claims that his umbrella pro- possibility. We cannot prove the alternative; rather,
tects him from falling elephants. He has supporting by testing the null hypotheses, we keep the alter-
evidence for his hypothesis that the umbrella pro- native hypotheses in contention. When we add null
vides protection. He has not had a single elephant fall hypothesis testing to confirming evidence, the
on him in all of the time he has had his umbrella open, argument for alterative hypotheses can become
yet such supportive evidence is weak; it also is con- stronger over time.
sistent with an alternative hypothesis: elephants do If all this discussion of null hypothesis is con-
not fall from the sky. Both hypotheses predict that the fusing to you, remember that the scientific com-
man will be safe from falling elephants. Negative evi- munity is extremely cautious. After all, it is in the
dence for the hypothesis—the one elephant that falls business of creating genuine, verified truth. It
on him and his umbrella, crushing both—would would prefer to consider a causal relationship as
destroy the hypothesis for good! false until mountains of evidence show it to be true.
We can test hypotheses in two ways: in a This is similar to the Anglo-American legal idea of
straightforward way and in a null hypothesis way. innocent until proved guilty. We assume, or act
Many quantitative researchers, especially experi- as though, the null hypothesis is correct until
menters, frame hypotheses in terms of a null reasonable doubt suggests otherwise. When we use
hypothesis based on the logic of the disconfirming null hypotheses, we can also use specific statisti-
hypotheses. These researchers look for evidence cal tests (e.g., t-test or F-test) designed for this way
that will allow them to accept or reject the null of thinking. Thus, we say there is reasonable doubt
hypothesis. Most people talk about a hypothesis as in a null hypothesis if a statistical test suggests that
a way to predict a relationship. The null hypothesis the odds of it being false are 99 in 100. This is what
does the opposite. It predicts no relationship. For we mean when we say that statistical tests allow us
example, Sarah believes that students who live on to “reject the null hypothesis at the .01 level of
campus in dormitories get higher grades than stu- significance.”
dents who live off campus and commute to college. Another type of hypothesis is the double-
Her null hypothesis is that there is no relationship barreled hypothesis.17 It shows unclear thinking
between residence and grades. Researchers use the and creates unnecessary confusion and should be
null hypothesis with a corresponding alternative avoided. A double-barreled hypothesis puts two
hypothesis or experimental hypothesis. The alter-
native hypothesis says that a relationship exists.
Sarah’s alternative hypothesis is that students’ on- Null hypothesis A hypothesis stating that there is no
campus residence has a positive effect on grades. significant effect of an independent variable on a
For most people, the null hypothesis approach dependent variable.
seems like a backward way to think about hypoth- Alternative hypothesis A hypothesis paired with
esis testing. Using a null hypothesis rests on the the null hypothesis that says an independent
assumption that we want to discover a relationship. variable has a significant effect on a dependent
Because of our inner desire to find relationships, variable.
we need to design hypothesis testing to make find- Double-barreled hypothesis A confusing and
ing relationships very demanding. When we use the poorly designed hypothesis with two independent
variables in which it is unclear whether one or
null hypothesis approach, we directly test only the
the other variable or both in combination produce
null hypothesis. If evidence supports or leads us an effect.
to accept the null hypothesis, we conclude that
separate relationships into one hypothesis. For Tautology. A tautology is a form of circular rea-
example, we say that poverty and a high concentra- soning. We appear to say something new but are
tion of teenagers in an area cause property crime to really talking in circles and making a statement that
increase. This is double barreled. We might mean is true by definition. We cannot test tautologies with
either of two things: that poverty or a high concen- empirical data. For example, I heard a news report
tration of teenagers causes property crime or that about a representative in the U.S. Congress who
only the combination of poverty with a high con- argued for a new crime law that would send many
centration of teenagers causes property crime. If more 14- and 15-year-olds to adult courts. When
“either one” is intended and only one independent asked why he was interested only in harsh punish-
variable has an effect, the results of hypothesis test- ment, not prevention, the representative said that
ing are unclear. For example, if the evidence shows offenders would learn that crime does not pay and
that poverty causes crime but a concentration of that would prevent crime. He believed that the only
teenagers does not, is the hypothesis supported? If prevention that worked was harsh punishment. This
we intend the combination hypothesis, then we sounded a bit odd when I heard it. So, I reexamined
really mean that the joint occurrence of poverty with the argument and realized it was tautological (i.e.,
a high concentration of teenagers only, but neither it contained a logic error). The representative essen-
alone, causes property crime. If we intend the com- tially said punishment resulted in prevention
bination meaning, it is not double barreled. We need because he had redefined prevention as being the
to be very clear and state the combination hypothe- same as punishment. Logically, he said punishment
sis explicitly. The term for a combination hypothesis caused prevention because harsh punishment was
is the interaction effect (interaction effects are dis- prevention. Politicians may confuse the public with
cussed later; also see Figure 4). circular reasoning, but social researchers need to
learn how to see through and avoid such garble.
Potential Errors in Causal Explanation
Example. A conservative is a person with certain
Developing a good explanation for any theory (i.e., attitudes, beliefs, and values (desires less govern-
causal, interpretive, or network) requires avoiding ment regulation, no taxes on upper income people,
some common logical errors. These errors can enter a strong military, religion taught in public schools,
while starting a study, while interpreting and analyz- an end to antidiscrimination laws, etc.). It is a tau-
ing quantitative data, or while collecting and ana- tology to say that wanting less regulation, a strong
lyzing qualitative data. Such errors can be referred to military, and so on causes conservatism. In sloppy
as fallacies or false explanations that may deceptively everyday usage, we can say, “Sally is conservative
appear to be legitimate on the surface but have seri- because she believes that there should be less reg-
ous problems once they are more deeply investigated. ulation.” This appears to be a causal statement, but
it is not. The set of attitudes is a reason to label
Sally as a conservative, but those attitudes cannot
be the cause of Sally’s conservatism. Her attitudes
Tautology An error in explanation in which the are conservatism, so the statement is true by defi-
causal factor (independent variable) and the result nition. It would be impossible ever to obtain evi-
(dependent variable) are actually the same or restate- dence showing that those attitudes were not
ments of one another, making an apparent causal rela- associated with conservatism.
tionship true by definition.
Teleology An error in explanation in which the Teleology. A teleology is something directed
causal relationship is empirically untestable because by an ultimate purpose or goal. It can take two forms.
the causal factor does not come earlier in time than
the result or because the causal factor is a vague, gen-
First, it is associated with an event that occurs
eral force that cannot be empirically measured. because it is in “God’s plan” or in some overarching,
mysterious unseen and unknowable force. In other
HYPOTHESIS: Poverty and a high concentration of teenagers in an area cause property crime to increase.
DOUBLE-BARRELED HYPOTHESIS: This can mean one of three things:
Poverty
CRIME
OR
High
Concentration
of Teens
CRIME
OR
Poverty
and High
Concentration
of Teens
CRIME
INTERACTION EFFECT: This means a combination of things:
Poverty
and High
Concentration
of Teens
Together
CRIME
FIGURE 4 Double-Barreled Hypothesis versus Interaction Effect

words, an event occurs because God, or an unseen, grade in a course. The mental condition can be a
unknowable master force has predetermined that it motivation that causes me to engage in certain
must occur. It is a teleology to say that something behaviors, such as studying (an intervening vari-
occurs because it is part of the “natural unfolding” able). The studying behaviors could increase the
of some all-powerful inner spirit or Geist (German chances that a future event (a course grade) will
for spirit). Thus, it is a teleology to say that a soci- occur. Conscious human goals differ from the will
ety develops in a certain direction because of the of God, a society’s Geist, or system needs, which
“spirit of the nation” or a “manifest destiny.” Sim- we cannot empirically measure, have no fixed exis-
ilar teleogical arguments rely on human nature as tence in time, and always match what occurs.
a cause, such as “Crime occurs because it is just
human nature.” Teleology has appeared in theories Example. The statement The nuclear family is the
of history when someone says we are moving dominant family form in Western industrial soci-
toward an “ideal society” or a utopia, and this move- eties because it is functional for the survival of the
ment explains events that are occurring today. Tele- society is an untestable teleological statement from
ology has also been found in functional arguments. structural functional theory. It is saying “society’s
It is a teleology to say the family takes a certain survival” causes “development of family form,” yet
form (e.g., nuclear) because the nuclear family ful- the only way we can observe whether a society sur-
fills social system “needs” for societal continua- vives is after the fact, or as a consequence of its hav-
tion. Logically, this says that the functional needs ing had a form of the family. Here is another
of the social system’s survival into the distant future example of a teleological statement: Because it was
are the cause of the family form we see today. It is the destiny of the United States to become a major
impossible to measure the cause and empirically world power, we find thousands of immigrants
test teleologies. entering the Western frontier during the early nine-
Teleology violates the temporal order require- teenth century. This says that “becoming a major
ment of causality. There is no true independent vari- world power,” which occurred from 1920 to 1945,
able because the “causal factor” is extremely vague, caused “westward migration,” which took place
distant, and unseen. Many people confuse goal between 1850 and 1890. It uses the obscure term
motivation (i.e., a desire for something yet to occur) destiny, which, like other similar terms (e.g., “in
with teleology. I might say a goal causes an action. God’s plan”), cannot be observed in causal rela-
For example, my goal to get an A in a class caused tionships.
me to get a good grade. My conscious goal or desire
could be a legitimate cause and not be teleological. Ecological Fallacy. The ecological fallacy arises
To show this, I need to outline the causal chain. from a mismatch of units of analysis. It refers to a
First, we can empirically measure my mental con- poor fit between the units for which we have empir-
dition (e.g., goals, desires, or aspirations) at some ical evidence and the units for which we want
time point. This clarifies both the empirical evi- to make general statements. Ultimately, it comes
dence and temporal order issue. Second, we can down to imprecise reasoning and generalizing well
compare my mental condition to future events that beyond what the evidence warrants. Ecological fal-
may or may not occur, such as getting a specific lacy occurs when we gather data at a higher or an
aggregated unit of analysis but want to say some-
thing about a lower or disaggregated unit. It is a fal-
lacy because what happens in one unit of analysis
Ecological fallacy An error in explanation in which does not always hold for a different unit of analy-
empirical data about associations found among large- sis.18 Thus, when we gather data for large aggre-
scale units of analysis are greatly overgeneralized and
treated as evidence for statements about relationships
gates (e.g., organizations, entire countries) and draw
among much smaller units. conclusions about the behavior of individuals from
those data, we are creating an ecological fallacy. To
avoid this error, we must ensure that the unit of one. Or perhaps one rich family and five poor ones
analysis we use in an explanation is the same as or in Joansville own motorcycles. To make a state-
very close to the unit on which we collect data (see ment about the relationship between family own-
Example Box 5, The Ecological Fallacy). ership of motorcycles and family income, we have
to collect information on families, not on towns as
Example. About 45,000 people live in Tomsville
a whole.
and in Joansville. Tomsville has a high percentage
of upper income people. More than half of the
Reductionism. Another problem that involves a
households in the town have family incomes of
mismatch of units of analysis and imprecise
over $160,000. The town also has more motorcy-
reasoning about evidence is reductionism, also
cles registered in it than any other town of its size.
called the fallacy of nonequivalence (see Example
The town of Joansville has many poor people. Half
Box 6, Error of Reductionism). This error occurs in
of its households live below the poverty line. The
an explanation of macro-level events using evidence
town also has fewer motorcycles registered in it
about specific individuals. It occurs when a person
than any other town of its size. But it is a fallacy to
observes a lower or disaggregated unit of analysis
say, on the basis of this information alone, that rich
but makes statements about the operations of higher
people are more likely to own motorcycles or that
or aggregated units. In a way, it is a mirror image
the evidence shows a relationship between family
of the mismatch error in the ecological fallacy. A
income and motorcycle ownership. The reason is
that we do not know which families in Tomsville or
Joansville own motorcycles. We know about only
the two variables—average income and number of
motorcycles—for the towns as a whole. The unit Reductionism An error in explanation in which
of analysis for observing variables is each town as empirical data about associations found among small-
scale units of analysis are greatly overgeneralized and
a whole. Perhaps all of the low- and middle-income treated as evidence for statements about relationships
families in Tomsville belong to a motorcycle club, among much larger units.
but not a single upper income family belongs to
EXAMPLE BOX
5
The Ecological Fallacy
Researchers have criticized the famous study Sui- Later researchers (van Poppel and Day, 1996)
cide ([1897] 1957) by Émile Durkheim for the ecolog- reexamined nineteenth century suicide rates with
ical fallacy of treating group data as though they were only individual-level data that they discovered for
individual-level data. In the study, Durkheim com- some areas. They compared the death records and
pared the suicide rates of Protestant and Catholic dis- looked at the official reason of death and religion, but
tricts in nineteenth-century western Europe and their results differed from Durkheim’s. Apparently,
explained observed differences as due to dissimilarity local officials at that time recorded deaths differ-
between people’s beliefs and practices in the two reli- ently for people of different religions. They recorded
gions. He said that Protestants had a higher suicide “unspecified” as a reason for death far more often
rate than Catholics because the Protestants were for Catholics because of the religion’s strong moral
more individualistic and had lower social integration. prohibition against suicide. Durkheim’s larger theory
Durkheim and early researchers had data only by dis- may be correct, yet the evidence he had to test it
trict. Because people tended to reside with others of was weak because he used data aggregated at
the same religion, Durkheim used group-level data the group level while trying to explain the actions of
(i.e., region) for individuals. individuals.
EXAMPLE BOX
6
Error of Reductionism
Suppose you pick up a book and read the following: ment’s ideology, popular mobilization, politics, orga-
nization, and strategy are absent. Related macro-level
American race relations changed dramatically during historical events and trends that may have influenced
the Civil Rights Era of the 1960s. Attitudes among the the movement (e.g., Vietnam War protest, mood shift
majority, White population shifted to greater tolerance with the killing of John F. Kennedy, African American
as laws and court rulings changed across the nation. separatist politics, African American migration to
Opportunities that had been legally and officially urban North) are also ignored.
closed to all but the White population—in the areas of This error is not unique to historical explanations.
housing, jobs, schooling, voting rights, and so on—were Many people think in terms of only individual actions
opened to people of all races. From the Brown vs. Board and have an individualist bias, sometimes called
of Education decision in 1955, to the Civil Rights Act of methodological individualism. This is especially true
1964, to the War on Poverty from 1966 to 1968, a new, in the extremely individualistic U.S. culture. The error
dramatic outlook swept the country. This was the result is that it disregards units of analysis or forces beyond
of the vision, dedication, and actions of America’s fore- the individual. The error of reductionism shifts expla-
most civil rights leader, Dr. Martin Luther King, Jr. nation to a much lower unit of analysis. One could
continue to reduce from an individual’s behavior to
This says: dependent variable ! major change in biological processes in a person, to micro-level neu-
U.S. race relations over a 10- to 13-year period; rochemical activities, to the subatomic level.
independent variable ! King’s vision and actions. Most people live in “social worlds” focused on
If you know much about the civil rights era, you local, immediate settings and their interactions with
see a problem. The entire civil rights movement and a small set of others, so their everyday sense of real-
its successes are attributed to a single individual. Yes, ity encourages seeing social trends or events as indi-
one individual does make a difference and helps build vidual actions or psychological processes. Often, they
and guide a movement, but the movement is miss- become blind to more abstract, macro-level entities—
ing. The idea of a social-political movement as a causal social forces, processes, organizations, institutions,
force is reduced to its major leader. The distinct social movements, or structures. The idea that all social
phenomenon—a movement—is obscured. Lost are the actions cannot be reduced to individuals alone is the
actions of hundreds of thousands of people (marches, core of sociology. In his classic work Suicide, Émile
court cases, speeches, prayer meetings, sit-ins, riot- Durkheim fought methodological individualism and
ing, petitions, beatings, etc.) involved in advancing a demonstrated that larger, unrecognized social forces
shared goal and the responses to them. The move- explain even highly individual, private actions.
person makes this error when he or she has data on ter of fact, no useful understanding of the higher-level
how individuals behave but wants to talk about the structure can be obtained from lower-level analysis.
dynamics of macro-level units. It occurs because it . . . If we are interested in the higher-level processes
is often easier to obtain data on individuals. Also, and events, it is because we operate with the under-
the operation of macro-level units is more abstract standing that they have distinct qualities that are not
simply derived by summing up the subunits.
and nebulous. Lieberson argued that this error pro-
duces inconsistencies, contradictions, and confu-
As with the ecological fallacy, to avoid the
sion. He (1985:108, 113–114) forcefully stated:
error of reductionism, we must make certain that
Associations on the lower level are irrelevant the unit of analyses in our explanation and for
for determining the validity of a proposition about which we have empirical evidence are very close.
processes operating on the higher level. As a mat- When we fail to think precisely about the units of
analysis and fail to couple the data closely with the is likely that a war would have broken out at about
theory, we might commit the ecological fallacy or that time even if the assassination had not occurred.
error of reductionism. These are mistakes about
having data that are appropriate for a research Spuriousness. To call a relationship between vari-
question and seriously overgeneralizing from the ables spurious means that it is false, a mirage. We
data. often get excited if we think we have found a spu-
It is possible to make assumptions about units rious relationship because we can show the world
of analysis other than the ones we study empiri- to be more complex than it appears on the surface.
cally. Thus, research on individuals rests on Because any association between two variables
assumptions that individuals act within a set of might be spurious, we must be cautious when we
social institutions. We base research on social insti- discover that two variables are associated; upon
tutions on assumptions about individual behavior. further investigation, it may not be the basis for a
We know that many micro-level units join to form causal relationship. It may be an illusion, just like
macro-level units. The danger is that it is easy to the mirage that resembles a pool of water on a road
slide into using the behavior of micro units, such as during a hot day.
individuals, to explain the actions of macro units, Spuriousness occurs when two variables are
such as social institutions. What happens among associated but are not causally related because an
units at one level does not necessarily hold for dif- unseen third factor is the real cause (see Example
ferent units of analysis. Sociology as a field rests on Box 7, Spuriousness and Example Box 8, Night-
the belief that a distinct level of social reality exists Lights and Spuriousness). The third variable is the
beyond the individual. Explanations of this level cause of both the apparent independent and the
require data and theory that go beyond the individ- dependent variable. It accounts for the observed
ual alone. We cannot reduce the causes, forces, association. In terms of conditions for causality, the
structures, or processes that exist among macro unseen third factor represents a more powerful alter-
units to individual behavior. native explanation.
How can you tell whether a relationship is spu-
Example. Why did World War I occur? You may rious? How do you find out what the mysterious
have heard that it was because a Serbian shot an third factor might be? You will need to use statisti-
archduke in the Austro-Hungarian Empire in 1914. cal techniques (discussed later in this book) to test
This is reductionism. Yes, the assassination was whether an association is spurious. To use them,
a factor, but the macro-political event between you need a theory or at least a guess about possible
nations—war—cannot be reduced to a specific act third factors. Actually, spuriousness is based on
of one individual. If it could, we could also say that some commonsense logic that you already use.
the war occurred because the assassin’s alarm clock For example, you know that an association exists
worked and woke him up that morning. If it had not between the use of air conditioners and ice cream
worked, there would have been no assassination, cone consumption. If you measured the number
so the alarm clock caused the war! The cause of the of air conditioners in use and the number of ice
event, World War I, was much more complex and cream cones sold each day, you would find a strong
was due to many social, political, and economic
forces that came together at a point in history.
The actions of specific individuals had a role, Spuriousness An apparent causal relationship that
but only a minor one compared to these macro is illusionary due to the effect of an unseen or initially
forces. Individuals affect events, which eventually, hidden causal factor; the unseen factor has a causal
impact on both an independent and dependent vari-
in combination with large-scale social forces and able, and produces the false impression that a rela-
organizations, affect others and move nations, but tionship between them exists.
individual actions alone are not the cause. Thus, it
EXAMPLE BOX 7 EXAMPLE BOX

8
Spuriousness Night-Lights and Spuriousness
In their study of the news media, Neuman and col- For many years, researchers observed a strong posi-
leagues (1992) found a correlation between type of tive association between the use of a night-light and
news source and knowledge. People who prefer to children who were nearsighted. Many thought that
get their news from television are less knowledgeable the night-light was somehow causing the children to
than those who get it from print sources. This corre- develop vision problems (illustrated below). Other
lation is often interpreted as the “dumbing down” of researchers could think of no reason for a causal link
information. In other words, television news causes between night-light use and developing nearsight-
people to know little. edness. A 1999 study provided the answer. It found
The authors found that the relationship was spu- that nearsighted parents are more likely to use night-
rious, however. “We were able to show that the entire lights; they also genetically pass on their vision defi-
relationship between television news preference and ciency to their children. The study found no link
lower knowledge scores is spurious” (p. 113). They between night-light use and nearsightedness once
found that a third variable, initially unseen, explained parental vision was added to the explanation (see b
both a preference for television news and a level of below). Thus the initial causal link was misleading or
knowledge about current events. They said, “We find spurious (from New York Times, May 22, 2001).
that what is really causing the television-is-the-prob-
lem effect is the preference for people with lower a. Initial relationship
cognitive skill to get their news from television” (p.
98). The missing or hidden variable was “cognitive
Night-Light Usage Nearsighted
skill.” The authors defined cognitive skill as a person’s
ability to use reason and manipulate abstract ideas.
In other words, people who find it difficult to process POSITIVE ASSOCIATION
abstract, complex information turn to television
news. Others may also use the high-impact, enter- b. Addition of the missing true causal factor
taining television news sources, but they use them
less and heavily supplement them with other more Parents Nearsighted
demanding, information-rich print sources. People
who have weak information skills also tend to be less Real Cause Real Cause
knowledgeable about current events and about other
topics that require abstract thought or deal with com-
plex information. Night-Light Usage Nearsighted
SPURIOUS ASSOCIATION
Spurious
Association Source: “Vital Signs: Update; New Study Vindicates Night
News Low Level of Lights” from The New York Times, Health Section, 5/22/2001
from TV Knowledge Issue, Page(s) 6.
Low Cognitive Real Cause

Level
consumption, and air conditioner use. In social
research, opposing theories help us figure out which
third factors are relevant for many topics (e.g., the
correlation with more cones being sold on the days
causes of crime or the reasons for war or child abuse).
when more air conditioners are in use. But you know
that eating ice cream cones does not cause people Example. Some people argue that taking illegal
to turn on air conditioners. Instead, a third variable, drugs causes suicide, school dropouts, and violent
hot days, causes both variables. You could verify acts. Advocates of “drugs-are-the-problem” posi-
this by measuring the daily temperature, ice cream tion point to the positive correlations between

Neuman Chapter 6

Uploaded by

Copyright:

Available Formats

Neuman Chapter 6

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Neuman Chapter 6

Uploaded by

Copyright:

Available Formats

Strategies of Research Design

Triangulation Qualitative Design Issues

Substantive problems must thus be translated into the vocabulary of social

TOPIC Theory: Compare how a feminist, a functional,

Reconstructed Logic, Linear Path

Logic in Practice, Nonlinear Path

FIGURE 2 Graphic Representation of Linear and Nonlinear Paths

type of research question tries to discover new

BAD RESEARCH QUESTIONS GOOD RESEARCH QUESTIONS

General Topics, Not Research Questions Descriptive Questions

TA B L E 1 Quantitative Research versus Qualitative Research

QUANTITATIVE RESEARCH QUALITATIVE RESEARCH

Grounded Theory The Context Is Critical

Bricolage sis becomes the sequence of events: what occurred

Third-order interpretation Qualitative interpreta- Types of Variables. As we focus on causal rela-

holds that disruption encourages children to join

Causal Theory and Hypotheses

contention. Theorists and researchers constantly cre-

FIGURE 3 How the Process of Hypotheses Testing Operates over Time

DOUBLE-BARRELED HYPOTHESIS: This can mean one of three things:

INTERACTION EFFECT: This means a combination of things:

FIGURE 4 Double-Barreled Hypothesis versus Interaction Effect

EXAMPLE BOX 7 EXAMPLE BOX

Low Cognitive Real Cause

You might also like