Professional Documents
Culture Documents
Research
Ted Gournelos
Joshua R. Hammonds
Maridath A. Wilson
First published 2019
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
52 Vanderbilt Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2019 Ted Gournelos, Joshua R. Hammonds, and Maridath A. Wilson
The right of Ted Gournelos, Joshua R. Hammonds, and Maridath A.
Wilson to be identified as authors of this work has been asserted by them
in accordance with sections 77 and 78 of the Copyright, Designs, and
Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced
or utilized in any form or by any electronic, mechanical, or other means,
now known or hereafter invented, including photocopying and recording,
or in any information storage or retrieval system, without permission in
writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or
registered trademarks, and are used only for identification and explanation
without intent to infringe.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
A catalog record for this book has been requested
ISBN: 978-0-367-20791-5 (hbk)
ISBN: 978-0-367-20793-9 (pbk)
ISBN: 978-0-429-26355-2 (ebk)
Typeset in Galliard
by Apex CoVantage, LLC
Visit the eResources: https://www.routledge.com/9780367207939
Contents
List of figuresvii
Preface: two professors and a librarian
walk into a bar . . . viii
Part I
Get the party started right 1
2 Where do I start? 14
4 Choosing a method 42
Part II
Methodology overviews 71
8 Surveys 119
Part III
Data analysis and the final push 145
Glossary181
Index185
Figures
prepared for it, and they always end up taking that at the last
possible moment because it terrifies them.”
“OK, but that’s a problem, right? They’ll be doing this for the
rest of their lives, or at least they will if they want to get a decent
job. That’s what being a good manager means: research.”
Maridath, the librarian that works with Ted and Josh to intro-
duce students to research methods and how to find sources,
walked in predictably late and ordered something from Brook-
lyn to match her artfully distressed leather jacket. “They always
come to me for help for exactly that reason, but I don’t know
what to tell them. It’s one thing to learn about research. It’s
another thing to learn how to actually do research.”
They chatted for a while and realized that it wasn’t just their stu-
dents. It was them, too. Even graduate students have often never
learned how to conduct various methods and why you’d choose
one over the other for any given question or context. Students,
graduate students, and even faculty often just use the same meth-
ods over and over again; sometimes they even avoid reading work
applying different methods. It was a reality check, as Maridath
reminded them. “We can’t blame our students, right? It’s not like
there’s a book out there that does this, either. They’re all huge,
dense and often boring, and still don’t give you a practical guide.”
They all sat for a little while, and then Ted looked up and
said, “Um, well, why don’t we just write it?”
*****
That all actually happened, more or less. We probably sounded a
lot dumber, and probably added in a lot more complaining. But
that’s why memory is so great; you can avoid thinking about
unpleasant truths. This book is obviously that book, born in a
bar and written in offices, coffee shops, libraries, at home, and,
again of course, the occasional bar.
What we’re going to do here is take you through a few steps.
Part 1 (Chapters 1 through 4) is about understanding what
x
Preface
research is, how to choose a topic, and how to get the back-
ground so that you actually know what you’re talking about and
can write a “literature review.” Part 2 (Chapters 5 through 9)
takes you through a few of the main methods that you’ll encoun-
ter and shows you what they do and why you’d choose them.
Part 3 (Chapters 10 through 12) shows you how to actually
analyze the data you gathered through your method, and then
how to put all of that together into a finished product. While
this book is suitable for anyone doing academic research, it’s
intended mainly for undergraduate and graduate students. Basi-
cally, it’s what we wish we’d read when we were starting out.
We’ve added in a couple of things to make this even more
useful as a practical guide. We offer some tips and tricks that we
use to make things manageable, and we hope those will help you
along the process. We also offer “tips from your friendly neigh-
borhood librarian,” in which Maridath takes you through how
librarians can help you along each stage of your research project.
Chapters also feature “IRL” (in real life) sections that show you
how each part of the process might appear in the workplace, in
fields ranging from marketing and restaurant management to
human resources and customer experience monitoring. Finally,
we supply a set of online resources for each chapter. We’ve
included articles from many different disciplines that use the
methods described in the book, videos that guide you on how
to use technology that will make your life a lot easier, and links
to citation guides, tutorials, and even paid services that can do
everything from transcribing your data to keeping your sources
organized to arranging for people to take your survey.
We hope you enjoy the book and that it makes the whole
process a lot less intimidating. If you’re not interested in your
research project, and if it’s not at least a little fun, then you’re
doing it wrong. This isn’t something you should dread; it’s a
rare opportunity to figure things out, to find out what makes
you you, and what makes the world what it is. Run with it.
PART
Get the party
started right I
Learn to love,
CHAPTER
1 not dread,
research
It’s Thursday afternoon and your boss has just come back
from a meeting with her boss. She asks to chat with you
for a second, which is always scary, and tells you she needs
a comprehensive report on how your company is doing
around the world. And she needs it tomorrow. At this
point you’re more than nervous; you’re worried about
keeping your job if you can’t complete an impossible task.
So here’s what you do (or what one of us did, since this is a
real story): First, ask what your boss really needs to know.
In other words, if she just wants numbers that you can
find on a company database, that’s not so bad. If she wants
those numbers along with detailed information about cus-
tomer satisfaction, you’ve got a big problem. Second, ask
who she needs the research for, and why they want it. If
the CEO just wants to be able to show the Board of Direc-
tors that you’ve got stores in 20 countries as one tiny part
of a presentation, no problem. If they’re evaluating where
to invest in new marketing campaigns and where to hire
staff to support an international branch, it’s going to be a
long, long night and you still won’t finish. Remember that
research is a process, and there is always room for more
depth or more reach. You have to know what questions
you need to answer, and why, before you can do it well. Be
careful, because you can get in over your head really fast.
a little peephole in the wall that you can look through. Maybe
eventually you’ll be able to see everything, but it’s going to take
you a while, and you’ll have to shift your perspective, depth of
focus, and sight angle to see much. And that’s if nothing in the
room changes. Good luck with that. People, media, minds, cul-
ture, politics – everything is changing all the time. The trick is to
get a perspective on what’s happening in a way that meets your
goals as well as your timeline. Sometimes that’s a matter of how
much you study, and sometimes it’s a matter of how deeply you
study it. But time, like money, is never infinite.
Once you’ve chosen a general topic, you need to figure out
how you want to do the study (see Figure 1.1). Researchers
often split our work into three categories or perspectives, or
what we call “paradigms”: the Empirical, the Interpretive, and
the Critical. Empirical scholars tend to believe that the topic
they’re studying is discoverable, and often through measurable
data with numbers. They want certainty, and so they tend to
focus on large amounts of data that they can extract and analyze
in a very methodical way that leaves little room for variation. In
other words, they’re looking for a snapshot through that keyhole
without drawing conclusions about anything they can’t see. For
example, in a project Josh conducted on healthcare, he wanted
to know how often patients did what their doctors told them
to do. He decided to study whether or not the amount of time
doctors spent with patients (the independent variable, or what
we use to explain changes) predicted how satisfied the patient
was with the experience (the dependent variable, or what actu-
ally changes). There are several ways to tackle how time affects
patient satisfaction, but for his study, he decided that the length
of time and degree of satisfaction told the best story.
Interpretive researchers, on the other hand, look at their
research slightly differently. They’re not looking for a snapshot,
they’re looking for what’s moving. More specifically, they’re
looking at how things might be portrayed, received, understood,
Topic 1: Topic 2: Quality Topic 3: Fighting
Cognitive healthcare hunger in the USA
disabilities in
the media
Focus: Focus: Doctor/ Focus: Nonprofit
Portrayal of patient organizations’ use
autism and interaction of social media
dementia in the (interpretive to communicate
media (critical paradigm) (critical/empirical
paradigm) paradigm)
RQ1: How do RQ1: How RQ1: Who do
plays portray/ do doctors nonprofits tend to
include people communicate target in their social
with different with patients media, and how?
ways of about health How often? With
thinking? issues? what response from
viewers?
RQ2: How do RQ2: How do RQ2: What stories
portrayals patients connect do nonprofits tell,
of autism in their interaction and what visuals do
children and with doctors to they use to tell those
dementia in the their perception stories? How often?
elderly differ? of the quality of With what response
care? from viewers?
RQ3: What RQ3: What might RQ3: How might
kinds of be improved approaches by
problems in setting various nonprofits
and solutions expectations differ, and how
do the plays and establishing do they align with
focus on, and effective ways of industry “best
who causes/ communicating practices” for
solves those during effective website
problems? interactions? development?
2 I start?
Now that you have a topic that you don’t hate and that you can
feasibly accomplish, when do you get to the actual really real
research stuff? To be honest, you’ve already started. Choosing
the topic, asking the questions, and finding people to help you
are part of the research. In fact, they are the most important
parts. Everything else is comparatively easy, even if it’s hard work
(yes, we know that sounds weird). Your librarian in particular
can help you here, because librarians are excellent at helping you
find stuff out, no matter what “stuff” means. They don’t do that
through a research method the way we’re describing it; instead,
they help you by giving you a step up. To use a common saying
in English, they help you “stand on the shoulders of giants.”
So what does that mean, realistically? It means librarians help
you find other people who have tried to answer similar questions,
or find ways in which other people have studied similar topics,
so that you don’t try to do everything (or that you don’t waste
time doing something that someone else has already done). The
trick is that you need to be able to keep this process methodical.
If you just start randomly choosing articles or books that have
to do with your topic, you can run into some major problems.
The main reason you might find “bad” sources is if they don’t
pass the “CRAP” test for a source’s “currency” (how recent it
is), “relevance” (how related to your topic and its particular dis-
cipline), “authority” (how credible the publisher or writer is,
and if they have authority in that particular field), and “purpose”
(are they just trying to persuade you of their opinion, or do
they want to inform you?). In general, for secondary sources
you’ll want to stick with scholarly, peer-reviewed journals. In
some cases, like with the project on nonprofit organizations, it
can also be important to get sources from established “trade” or
16 Part I: Get the party started right
So who are these “peers?” And what exactly are they reviewing?
Librarians love to talk about “scholarship as conversation” –
seriously, get one going on this for a minute and see how
excited they get. Articles that appear in peer-reviewed aca-
demic journals are written by experts for experts. After per-
forming their research, scholars submit their work to these
journals, which then undergo a review process. Their work
is checked for style, substance, and quality by other schol-
ars in the field, usually in a double blind method where no
one knows the identity of anyone else. Articles and books
are then either accepted for publication in the journal,
accepted with revisions (where the scholar must make
often substantial changes and then re-submit), or outright
rejected. Usually, from start to finish, this process can take
a long time – sometimes years. It’s important to note the
length of this process, because this notion of “currency”
(the “C” in the CRAP test above) really differs between
2: Where do I start? 17
While this depends on what your project is and who has assigned
it (your boss, your professor, yourself), usually you want to make
sure your sources are current and useful. While not necessarily
“bad,” you might find that out-of-date sources are easier to find
because they are “kind of” about your topic, but not really useful
18 Part I: Get the party started right
in the end. Sources that are “kind of” about your topic probably
don’t pass the “R” (relevance) part of the CRAP test. In some
cases, older sources may be appropriate (e.g., if you need to trace
the origin of a problem or track its historical trajectory). That
said, it’s perfectly fine to consult sources that inform you about
the topic more generally. Just be careful to keep them to a mini-
mum. For Ted’s project on disability and theater, for instance, it
was useful to look at sources on mental disabilities in the media in
general, and ageing in the media in general. However, looking at
medical studies on autism or on Alzheimer’s probably wouldn’t
be useful for a paper on how it’s portrayed in the theater, or how
audiences react to those portrayals. Similarly, be sensitive to how
old a source is. Old doesn’t mean bad, but it does mean that the
author(s) weren’t able to react to conversations that happened
since. Remember, scholarship is a constantly evolving discussion.
material, make sure you look for citations and give those
the CRAP test!
Problem 3: Cherry-picking
Citation exporters
There are many ways to cite sources, but they’re all system-
atic. You’ll most likely encounter the styles from the Asso-
ciated Press, the American Psychological Association, the
2: Where do I start? 23
Annotated bibliographies
You also want to make sure that you don’t choose some sources,
read them, and then forget what you read and have to read it all
over again in a week (or a month, or a year). The best way to
prevent that is to create an “annotated bibliography,” which is
a bibliography to which you’ve added notes about each source
(including whether it was useful or a waste of time to read). What
we recommend is keeping this organized for yourself, either with
bullets under each source or in a short paragraph. While there are
various way to do this, there are a few things you want to make
sure to hit, some of which we discussed in Figures 1.1 and 1.2:
3 I do? What’s a
lit review?
Identity and The Last Jedi Re-thinking James Bond Identity in Orange is the New Politics and identity in Zootopia (2016)
(2017) I. White masculinity and the Black and Moana (2016)
I. Sci-fi, politics, identity media I. Portrayal of criminal justice in I. Race, authenticity, and politics in
II. Star Wars, politics, II. James bond and identity the media (vs. reality) animated film
identity III. Evolution of a franchise and II. Portrayal of women in the II. Disney and identity
III. Evolution of a franchise identity shifts CJS A. Princess vs. anti-princess films
and identity shifts III. Orange is the New Black III. Zootopia and Moana
The Office and Adaptations, power, and Images of plastic and pollution Fast fashion and environmentalism
mockumentary TV The Blind Side I. Activist photojournalism I. Advertising and fashion
I. Portrayals of work and I. Football films II. Activism and social media II. Fashion and unsustainability
office life in the media II. “White savior” films III. Environmental activism III. Fashion and environmentalism
II. Mockumentary genre The Blind Side
III. The Office series (US III. Adaptations of “true” stories
and UK)
Figure 3.1 From topic to literature review (samples from popular culture and society)
Gender and cyberpunk Social media addiction Representations of women in the Revenge and identity in Insatiable
2020 I. Addictive behaviors and Harry Potter series (2018)
I. Sci-fi, identity, and treatments I. Gender representations in I. Teen-targeted media, narrative,
gender II. Social media addiction children’s or teen literature and identity
II. Role-playing games (psychology focus on II. Harry Potter series II. Revenge in the media
III. Cyberpunk individuals) III. Gender in Harry Potter series III. Body image and the media
III. Social media use (trends and (obesity-focus)
addiction) (communication
focus on media)
Superman and religious Anti-Photoshop discourse and (Mis)representation of Africa in Female celebrity identity on Instagram
iconography legislation US media I. Celebrity branding and gender/
I. Portrayals of superman I. Advertising, gender, and I. Representations of Africa in sexuality
II. Superman and other body image the media II. Celebrity and social media
comic book icons II. Photoshop and/or image A. Positive over time III. Parasocial relationships and
as cultural/political alteration B. Negative over time aspirant identity
allegory III. Anti-alteration legislation II. Black Panther or other Africa- IV. Social media and self-portrayal
III. Superman and religion and “natural” body portrayal focused films from Black V. Social media and self-perception
movements African perspectives
Ted’s method
one so I could find it again. This created three very, very long
documents. Suddenly it wasn’t so much about what to write,
but what to cut; I had five to eight pages that I could use for this
at most, and I had fifty pages or so of quotations alone. So what
did I do? I highlighted them again, in different colors depend-
ing on what part of the literature review (what theme) they
fit. Then I organized them by color/theme and read through
them all together, putting quotations that were similar together
and reducing quotations by paraphrasing where it made sense.
I outlined the literature review with those sub-themes as head-
ings, and then started to write, cutting/citing/quoting from my
sources as I went. Fifty pages turned into ten, as the literature
seemed to just write itself, logically and in an order that made
sense. I was telling a story, in other words, drawing from the
stories of other people. Then I just went through and edited,
usually removing quotations or pieces that I thought were going
to be relevant but really weren’t, or needed to be one sentence
instead of an entire paragraph. It helps to remember that you’re
trying to answer a question, and sometimes there’s some great,
interesting research that gives you context but doesn’t help you
with your question. You can usually reduce or cut all that stuff.
That was it! End of literature review. I’ve used this for every-
thing from a proposal for a book to a journal article to a full,
three-hundred-page book. It works, and it keeps you sane.
Josh’s method
Maridath’s method
4 method
Once your literature review is done, you should have a clear idea
of what others before you have argued about your topic and
through what methods. At that point you have four choices for
your own method:
Textual Close examination of two scenes Conversational analysis of ten Analysis of three highest “liked”
analysis from two plays about disability interactions between staff and Facebook posts from three highest
(one about autism and one about patients in the waiting room, budget hunger nonprofits in the USA,
dementia), dwelling on issues of dwelling on tone, information dwelling on use of imagery, tone,
narrative style, imagery, and tone, as provided, and two-way feasible call to action, and “shareability.”
well as place within the story. communication.
Content Analysis of trailers or advertisements Discussion of forty observed Review of all social media posts
analysis of twenty films and television shows patient/staff interactions in from ten anti-hunger nonprofits for
featuring a character with a disability, a waiting room, coding for the months of December 2018 and
particularly autism and dementia, best practices in information January 2019, coding for aesthetics
coding for issues of representation conveyance. choices, use of SMART guidelines, and
and focus/erasure of disability in type of persuasive appeal.
society.
Discourse Discussion of the concept “disability” Discussion of guides for patient/ Review of user comments on all social
analysis as it appears or is erased/concealed provider interactions in medical media posts from 2017 and 2018 on the
in critic reviews of and newspaper textbooks, hospital brochures, and Facebook and Instagram channels for
articles on the Netflix Original training videos, focusing on two five national and five local/regional anti-
Atypical and the CBS series The Big hospital chains and two medical hunger nonprofits, looking for elements
Bang Theory, as well as on specific schools in the state of Florida. of engagement with the issue or the
media relating to dementia in organization.
contemporary mainstream media.
In-depth Interviews with five spectators of the Interviews with ten sets of Interviews with marketing managers
interviews play The Curious Incident of the Dog doctors and patients after their from ten different hunger-focused
in the Nighttime and five spectators consultation to determine their nonprofit organizations discussing their
of the play Forget Me Not within a individual perceptions regarding strategies for social media posts.
week of seeing each play. Participants the interaction’s emotional content,
should include parents and children informational content, and next
who saw the film together. steps.
Focus groups Four groups of five, one group with Three groups of five, one of Four groups of five, one of social
people who know someone with doctors and nurses, one of staff media managers of local nonprofits
autism, and one group people who members, and one of patients. fighting hunger, one with their social
do not. Discussions about their Discussions about what makes media producers, and one with the
perceptions of autism after seeing for a good experience, what are target audiences for the organizations.
The Curious Incident of the Dog in challenges to that experience, and Discussions about intended and actual
the Nighttime and how it might alter how they think things could be uses and gratifications from the posts,
their understandings or behaviors improved. as well as perceptions of the posts and
towards autistic individuals. Identical their strategy overall.
framework for Forget Me Not.
Surveys Although surveys don’t necessarily Providing pre-consultation and A survey here could focus on potential
suit studying the plays themselves, post-consultation surveys to donors, asking them their perceptions
pre- and post-viewing surveys can be patients and doctors is a very of the organization, the issue as a
an effective way to understand how effective way to measure whether whole, their donation practices, and
theater impacts audiences. In this there is a disconnect in how each even specific questions about what they
case, asking audiences about their feels about the experience, and would like to see in a nonprofit website.
understanding of, experiences with, then those surveys could actually Ranking various possibilities rather than
or personal connections to autism be discussed with the doctors to listing individual ones as important can
and Alzheimer’s disease could reveal try to solve any issues that occur. be very useful here, since for nonprofits
whether the plays expand or change Asking questions about what there is the risk that respondents
Case studies In many ways the textual analyses are Looking at patient-provider The case studies used in this project
case studies. However, you could also interaction in a case study would would focus on specific efforts from one
study cases in which the plays were most likely study a specific organization, with information about the
performed in certain contexts (Forget hospital, private practice, or challenges, goals, and changes made in
Me Not) is performed in elder care hospital ward. The tighter the focus, the process by the managers and other
facilities, for instance. The key to a the better. This is telling the story of staff.
case study is specificity. that ward, rather than giving more
general data.
Archival Studying medical archives, media Doctors have changed quite a bit Looking at the archives of hunger-
research archives, or other sources could in the public imagination over focused nonprofit organizations
provide valuable information about time. Studying archives of medical could reveal how those organizations
how these medical issues have been materials, from advertising to (individually or collectively) view their
understood or portrayed over time. American Medical Association place in society, view their donors and
These could be notes, for instance, documents for doctors to clients, and structure programming.
from production, casting, and direction educational documents, could It can also reveal how they have
meetings for shows like The Big Bang provide valuable perspective into the approached communicating about that
Theory, but they could also be from construction of the profession, but programming or those stakeholders.
the stage adaptation of The Curious also how that profession has viewed
Incident of the Dog in the Night-time. patients differently over time.
Ethnography An ethnography of a new production In this case ethnography is a good Studying the lives of clients impacted
of the plays might be a way to fit and could be used to study by hunger is one way to approach the
approach this, but you could also residencies by accompanying new topic through ethnography (living with a
conduct an ethnography of autistic doctors throughout their workdays client of the organization, for instance),
children and their families (perhaps in to see how they treat and learn to but you could also work alongside the
a school) and compare it to another interact with patients and families. volunteers or organization staff for
of people with Alzheimer’s disease a set amount of time to understand
and their families and professional their culture, their struggles, and their
caregivers. various rewards/triumphs.
Experiments Experiments would be extremely As with the disability project, Experiments for this project could use
difficult to do here, because the studying how patients interact with multiple websites or social media posts
populations are so vulnerable. multiple forms of doctors could for various stakeholders (e.g., different
It would be easier to conduct reveal a lot about preconceptions categories of donors, volunteers,
an experiment with caregiver and differences connecting due to clients, etc.) to see their responses.
professionals in schools or nursing identity or attitude, and doing the Then those different media could be
homes, comparing how they treat same with doctors could reveal correlated with other data, like donation
patients who they have been told similar biases and inconsistencies activity or volunteer registrations, to
have autism or Alzheimer’s disease in approach. determine which types of posts work
with those they have been told have best with which stakeholder.
another condition, using actors to
simulate the experiences.
Perception/reception
In-depth interviews
• Structured, semi-structured, or unstructured, interviews
often conducted with fewer participants but in great depth,
usually from 1–2 hours in length; in-depth interviews (IDIs)
try to get participants to tell stories and explore concepts in
detail, often with follow-up questions or examples.
• Example: So say we didn’t really care about what 13 Rea-
sons Why was doing, but instead we wanted to see how
54 Part I: Get the party started right
Focus groups
• Small groups of individuals that are targeted to gain specific
information about a topic; usually random groups or groups
that are intended to get a range of opinions, focus groups
are widely used in industry studies of media, like whether a
new TV show will be popular, or in social sciences to gain
perspectives on things like body image, and so forth; focus
groups are great because you can use a semi-structured
interview guide but encourage conversation among individ-
uals, which you record and then analyze; however, they can
also result in “groupthink” or silenced voices.
• Example: Focus groups are often used in media analy-
sis, including by industry professionals who are testing a
new show/film/advertisement or deciding on whether to
renew/continue/change what they’re already doing. For
13 Reasons Why, we could do focus groups with counselors/
psychologists/psychiatrists about the benefits and problems
they can see with the show for their clients, or we could
do focus groups with students like in the interview exam-
ple (both again with IRB approval). In this case, we want
56 Part I: Get the party started right
Surveys
• A questionnaire of some sort given to a sample of individ-
uals to gain their perspective on a topic; samples are usu-
ally large, but depth of reaction is often not as intricate as
focus groups or interviews; common scales are Likert scales,
4: Choosing a method 57
Observational analysis
Using archives
ETHICS ALERT!
ETHICS ALERT!
Textual Look at how your competitors are While it’s possible to look at customer Do a textual analysis of a break-room
analysis advertising their products and if blogs, videos, etc. if you have an issue or an office space, connected to best
there are missing demographics (e.g., with United Airlines), usually practices in the industry for productive
or situations (e.g., selling SUVs to you wouldn’t choose this method workers. For instance, Clif Bar &
“soccer moms” instead of to dads). because it’s about perception, not Company has an office filled with
portrayal. athletic equipment, and an in-house
massage therapist and gym.
Content Compare your competitors’ social Analyze all of the complaints and Although this often wouldn’t be the
analysis media presence to your own, coding other communication that comes best method for this question, you
for types of messages, media, through email and on social media, could analyze suggestions made by
demographics, even tone (in addition like comments on posts, tweets about employees through an anonymous
to likes, shares, etc.). Target at least your products or customer service, email survey, or responses made
the past year. and hashtag use. Code for tone, issue, during exit interviews when
and positivity/negativity, as well employees leave the company. Code
as whether it was resolved and/or for tone, similar concepts that might
responded to by your staff. be positives and negatives, and
forcefulness of feeling.
Discourse Because discourse analysis is about One of the most useful things you As sites like Glassdoor start to provide
analysis perception and social construction, can research is how people are extensive, anonymous reviews of
it doesn’t really help much when speaking about your company and its companies, it’s increasingly important
figuring out what your competition competitors. You could look at social to monitor them to determine what
is doing. However, it could certainly media accounts of “influencers,” or employees are saying when they work
reveal opportunities and where you just look at what people are saying for you and when they leave. This
and your competitors are succeeding when they hashtag you or comment can head off some serious problems,
or failing. For instance, when you on your own social media. You could showing you issues with ethical
In-depth If you want the most rich, detailed By selecting a few representative Similarly to the customer interviews,
interviews information on why people are customers (for example by location, you might select employees from
choosing your competition, in-depth revenue amount, or company size), different departments or levels in the
interviews can really become and conducting IDIs with a set list of organization, and administer your
immersed into the perspective of questions. Be careful to avoid leading IDI. While you can expect richness
the customer. Rich descriptions and questions of course, and if possible of data with this method, you most
unique stories can help paint a clear get a third-party to administer the IDI. likely will spend time looking for
picture of why customers/clients are common themes that emerge from the
choosing your competition. qualitative data.
Focus To help flush out some of the major Focus groups can be extremely A focus group to determine employee
groups stories why people are choosing your helpful in determining how customers satisfaction might be useful, however,
competitor, you can conduct focus feel about your company. This method in order to protect anonymity and
groups to hear from several people can also be helpful when attempting avoid groupthink, you’d probably
at once. Focus group research can to make claims about which particular want it administered by an outside
help remind and confirm the attitudes groups like or do not like your facilitator. For this particular question,
associated with competitor allegiance. company, based upon the makeup a focus group might not be the best
of the focus groups you build. Focus method. Surveys and especially IDIs
groups can sometimes suffer from might be better in terms of providing
“group think,” so be careful of this rich feedback in a setting that avoids
pitfall. bias and makes your participants feel
as though they can speak candidly.
Surveys You want to find out what are the Surveys can be useful in getting a Surveys provide employees with the
most common reasons people wide array of responses fairly quickly, anonymity they need to speak freely
are flocking to your competitors, but you might limited on how much about the company, but you won’t
conducting surveys can help you information you get. It’s nice to have a get the same amount of qualitative
find trends and patterns. It can also summary of overall attitudes, but you responses as from an IDI or focus
identify how high/low customer’s may not get the whole story. group.
attitudes rank when they purchase
your competitors products
Ethnography An ethnography of a group that you Embedding yourself into campus It’s important sometimes to get a
want to sell your product to is a great tours, dealership test drives, and feeling for what work is like on the
way to get good information. You can other situations can help you gain ground. Are people overworked? Do
easily figure out what people buy, a close look at how people interact they feel undervalued? What do they
when, and how much they spend, in those strange situations. What worry about, and what might make
but ethnography helps you truly do families talk about when they’re their lives better and easier? Those
understand the many reasons why. buying a car or choosing a school, and answers are rarely as simple as “more
what are the things you didn’t expect? money,” because you still need to
know why.
Experiments Experiments are a great way to The challenge and benefit of Again, this would be far more involved
confirm whether specific variables experiment is that the variables you than some of the other methods, but
are causing a specific change in a are measuring need to be exceedingly you could develop interventions to
group of people. In examining a clear. When assessing attitude/affinity determine employee productivity and
competitor’s business strategies, you often experiment design researchers satisfaction and compare the overall
could determine if a specific strategy will assess hormone levels and/or scores across the conditions.
is influencing customer acquisition or brain activity to determine how much
sales. someone likes/dislikes a company//
person.
5 and thematic
analysis
it only mean “stop?” Who’s telling you that, and why? What are
the consequences if you don’t stop? A stop sign is actually say-
ing “stop . . . or else.” That last bit is important, because while
a stop sign is signifying “stop,” it is also signifying a system of
ethics in which cars should be careful to avoid people, other
cars, and other objects, as well as signifying a system of power
in which if you don’t stop, you might get a ticket from a police
officer. Either way, it’s going to be a bad day. So that’s textual
analysis. You look at as many details as you can to get an under-
standing of all the different levels that build the sign, or what
we call “signification.” We’ll take you through some examples
throughout this chapter that will make this clearer.
Details in a textual analysis come from many different places.
In literature they might be the way words work together
(meter, rhyme, assonance, etc.), in film they might be light-
ing, placement in a scene (mise en scène), and color, and both
might examine setting, narrative, and dozens of other elements.
It’s helpful to categorize textual analysis into some general
concepts: formal/aesthetic (how it sounds, looks, etc.), allusions
(references to other texts), narrative (the story), setting (place
and mise en scène), etc. All of this depends on what exactly
you’re analyzing. So let’s start with a couple of possibilities.
6 discourse
analysis
OK, let’s take a step back. Textual analysis is awesome, but it’s
also limited. When you look at that many signifiers, how do you
know what’s important? How do you know that what you’re
reading into the text has broader significance beyond the text
(e.g., to other advertisements by the same company) or con-
nects to audiences the same way (e.g., what is it from the viral
video that made it viral, and was that a good thing or a bad
thing)? Those kinds of questions really need a higher sample size
(N), and while this will sacrifice some of the depth, mini-textual
analyses of specific examples can often be a useful way to incor-
porate depth into other, broader methods like content analysis
or discourse analysis.
Content analysis
After you’ve rated them, you can then calculate what we call
“inter-rater reliability,” which basically means you’re figuring
out how close you were in the analysis. Using Excel or SPSS, you
simply tabulate each of the coders scores in adjacent columns
and run a Cohen’s Kappa (two raters) or Fleiss’ Kappa (3+) anal-
ysis to determine the consensus (or lack thereof) between the
coders. If you score 82%–100% reliability (k =.90) your coders
are probably sharing a brain. Strong reliability is within 64%–
83% (k =.80 –.90). Much less than that, and you will want to
consider having another norming session. This type of analysis
allows you to see how subjective your material is and whether or
not the best way to describe it is through quantitative data, or if
it only allows you to see general trends that you’ll then have to
spend more time explaining in deeper, qualitative descriptions.
See our online resources for some step-by-step guides on how
to do this in Excel.
those customers want (to see, hear, etc.) drawn from other
studies and from interviews or focus groups he has con-
ducted. As a health management consultant, Josh com-
pares the various interactions patients have with doctors,
nurses, and other staff across not just the one hospital he’s
working with, but also other hospitals. He codes for best
practices to increase satisfaction and outcomes found in
the literature on patient-provider interaction. Maridath
works with multiple library databases to determine what
they offer for their prices, coding for not just what they
have and how they work, but also what the university fac-
ulty and students need to access, and how much they over-
lap with the existing content the library offers. These are
all content analyses, even if they aren’t always done in the
same systematic way that we have advocated here.
Graphic? (presence/absence)
Of what? (category: map, infographic, drawing/ani-
mation, event marketing, other)
Tone? (category: informational, funny/cute, inspira-
tional, other)
Aesthetic quality (emphasis: 1 = low, 2 = medium,
3 = high)
Video? (presence/absence)
Of what? (category: donation process, sourcing pro-
cess, client profile, inspiration, program informa-
tion, impact information, volunteer profile, donor
profile, employee profile, other)
Tone? (category: sad, happy, inspirational/hopeful,
angry, detached, other)
Aesthetic quality (emphasis: 1 = low, 2 = medium,
3 = high)
Length? (minutes)
Clear call to action? (presence/absence)
To do what? (volunteer, donate, attend an event, other)
sources, etc. You will most likely code for themes just like you
would in a content analysis, but probably not at the same level
of detail because you are looking at longer and/or a larger
number of texts. The key here is that you’re not studying the
texts themselves, you’re studying what they’re saying about
something else. You’re still looking for patterns, though, or
what discourse analysts often call “interpretive repertoires.”
There are a few things you can focus on: outliers, or what
have been called “deviant” cases; consistency and coherence
with previous work on the topic (e.g., your literature review);
understandings of an event or issue (e.g., letters to the editor,
movie reviews, advice columns). You can also simply present
your data and ask readers to come to their own conclusions.
Here are some examples of discourse that might help you get
a feel for it:
across the nation (or globe) could reveal quite a bit about
how various regions treat (or conceal, erase, and misrep-
resent) the Civil War, slavery, and indigenous populations.
Connecting that to records of school board meetings in
which boards, parents, and teachers meet, as well as the
presence or absence of discussions about this issue on the
news and social media, can provide a snapshot of how
the culture is actively constructing and silencing certain
histories.
3. Arts: Art isn’t just made of the artifacts themselves; in fact,
that is only a small part of what we know of as “art.” The
well-known, award winning, controversial Broadway musi-
cal Hamilton, for instance, could be studied in terms of
how news sources, professional and amateur reviewers, and
actors spoke about the play. This was particularly interest-
ing as actors came out actively to contest popular under-
standings of history, noting that their versions in the play
weren’t accurate, but that neither were the versions in his-
tory books. In other words, Hamilton and the discourse
around it was in many ways a struggle over who gets to
write history and how.
4. Brand perception: Airlines are an oligopoly of compa-
nies that everyone hates. No one likes to fly, but many
of us have to put ourselves through the misery of air-
ports, security lines, immigration, and tiny seats just to
save some time. They’re all around the same price, with
mostly the same service. But perceptions of the airline
can make the difference, as United Airlines has found
out repeatedly. Internet sensations such as the “United
Breaks Guitars” song from Dave Carroll or the infamous
video of security personnel beating a passenger as they
dragged him off a plane in order to give his seat to an
airline staff member, have caused incredible damage to
the brand. To fix it, any analyst would need to look
96 Part II: Methodology overviews
7 interviews, and
focus groups
Ethnography
that group in some way, and you observe them, taking “field
notes” along the way that include detailed information about
events, verbal and nonverbal communication, and other ele-
ments. These notes can then be coded in a similar way to textual
analysis or discourse analysis, which we discuss in depth in Chap-
ter 10. It’s important to remember that what you’re coding for
and how you’re coding for it can change, depending on your
questions, background and discipline, and what your literature
review has taught you. There are also three primary ways to
conduct ethnography: undercover ethnography, non-participant
observation, and participant observation.
Undercover ethnography
This method was often used to study groups from within as
a completely embedded participant. That means that research-
ers would claim to be a part of that group (like an undercover
police officer) in order to observe participants as organically as
possible. There are many reasons to do this. For instance, what if
you want to study prison culture? Gangs? Human trafficking or
organ trafficking? All of those have been the subjects of under-
cover ethnography. However, at the same time we have to be
very careful here. Because the participants can’t give consent if
you’re undercover, you can end up exploiting and misrepresent-
ing them. That means that undercover ethnography is rarely if
ever ethical as research. Leave it to the police or journalists, who
have a different job to do.
Undercover ethnography is also misleading, because it’s very
difficult to be completely organic as a part of a group if you’re not
actually a part of that group. That means that if you are separated
a bit through the act of observation and covert note-taking, you
might introduce elements of bias and expectation that can lead to
changes in the behavior of the people around you. For instance, if
you’re studying gang violence and are undercover, should you be
violent? And if you feel you need to be violent for “authenticity,”
7: Ethnography, interviews, focus groups 101
Non-participant observation
In this method, you’ve received IRB approval and informed
consent from all participants (see the online resources for some
guides on how to do that successfully). However, you are
attempting as much as possible to not participate in the group’s
interactions. This could happen through video recording, but
there are challenges in terms of the depth and detail that can be
captured. Of course, being there and taking notes might also
reduce depth, since you won’t catch everything and might be
immersed in some detail or emotional moment (e.g., a riveting
story from a recovering alcoholic in an Alcoholics Anonymous
meeting) and miss something important (e.g., that half of the
room isn’t paying attention to that meeting, or that someone in
the back is shaking from alcohol withdrawal). The biggest chal-
lenge here, however, is that simply the act of observing some-
thing can change that thing (according to particle physics, it
always does). Certainly having an observer present during an AA
meeting could change the dynamic of that meeting, especially if
people are communicating trauma. Video recording the inter-
actions might even make the problem worse. Since researchers
are also given a kind of cultural power and authority, this impact
could be massive. In other words, it’s almost impossible to have
non-participant observation; the best you can get in most cases
is a limited form of participant observation. Imagine yourself
as a child playing with friends. Even if your parents aren’t in
the room, you know they’re there and often act accordingly. If
102 Part II: Methodology overviews
Participant observation
Participant observation basically means that you’ve gotten IRB
approval and informed consent, and you consider yourself an active
participant in the group you are observing. That means that you’ll
have to observe yourself as well, and of course in some cases like
auto-ethnography you’re focusing primarily on yourself. However,
in most cases you’re simply acknowledging the biases, changes, and
limitations to observing interactions in any organic or “authentic”
way. That means that your work will need to be deep enough to
acknowledge what is going on in what you’re observing, but also
engage how you might be impacting the events and interactions.
It’s a great way to try to understand group interactions, but chal-
lenging to do both ethically and in a way that provides truly useful
data. Combining ethnography with other methods can often be
a useful way to add elements that might be missed, especially if
it happens before and after your observations. Surveys are great
here, but so are the other two methods we’ll discuss in this chap-
ter: in-depth interviews and focus groups.
In-depth interviews
Bias
You have bias. Admit it, then try your best to avoid it. While
you can conduct an experiment behind a one-way mirror and
make every effort to remove your presence from the study,
as long as you’re studying it, the bias is still there. Qualita-
tive interviewers make no apologies about their involvement
and participation in co-creating and reporting on the truth
of people’s experiences. Remember, you have just developed
an extensive literature review, a research design, and a set of
research questions about your project. At this point, you are
knowledgeable and have a good idea about what people think/
feel about your topic. That is a good thing, but it can also be
a problem, since you might end up accidentally guiding the
research towards what you’re expecting to find. There are some
ways to reduce this effect though, and creating a good inter-
view guide is a good start.
Interview guides
One of the first actions you can take to help curb your own bias
is to construct an interview guide, which is a prepared script with
a list of questions that you’ll use as you do the interview. There’s
a balancing act here between very scripted and very unscripted,
though. Do I write out every question and deliver it verbatim?
Is it okay to ask questions that aren’t on my script? Do I need
to stick with the same order each time? The best interviews will
often feel like a natural conversation, and at the same time, it’s
important to understand that structure is necessary to compare
responses and prevent leading questions.
In a “non-structured interview,” there is no interview guide
at all. Often corporate executives who have been conducting
employment interviews for several years will claim that they
need no prepared list of questions. They allow the interview to
flow like an organic conversation and make decisions based on
intuition. There are a few downsides to this interview approach:
it says here on your resume that you lived in Boston for a while.
I’m a huge Red Sox fan – are you?” Well, if you weren’t before
the question was asked, our bet is that you are now (and will go
home and burn your Yankees jersey).
On the other side of the continuum is the “highly-structured”
interview guide. In this case, interview questions are written and
delivered verbatim. We call this format “Stick to the Script,” and
it’s not unusual when there are many interviewers speaking with
many candidates (say, for entrance into college). You state them
exactly as they are written, in the exact order, with no deviation.
Comparing multiple responses is much easier to do in this case,
but there are some problems:
Questions
Like interview guides in general, the questions you ask also range
from open to closed. On the one hand, highly open questions give
the participant complete freedom to respond in any direction they
would like. Questions such as, “Can you tell me about yourself?”
and “What does it all mean?” can be useful sometimes, especially
if the participant starts going in a direction that you didn’t antic-
ipate; however, questions that are too open can often lead your
participant to go off on a tangent and take up way too much of the
time you’ve set for the interview. Remember that at some point
you’ve got to get to the questions you need answered!
On the other hand, highly closed questions are great for short
interviews, and like highly structured interview guides they make
responses easier to compare. Unfortunately this also means that
participants will often just try to answer your question, which
reduces the chances of unexpected directions and responses, or
even the richness of detail that interviews can provide. They can
also make the interview feel cold and distant, reducing the likeli-
hood that your participants will tell the truth or disclose informa-
tion that might be more controversial, personal, or uncomfortable.
For instance, a closed question might ask something like “did that
make you happy?” or “did you have a close relationship with your
parents?” but it might not encourage the participant to explain
why they felt a certain way, how they felt if they didn’t feel a certain
way, or examples of situations that shaped their feelings.
We suggest that the key for a useful in-depth interview is to
focus on having enough structure to facilitate getting to the
point, but that you also focus on having your participants tell
stories and go deeper (see Figure 7.1). The deeper you get into
their examples, the more data they show you rather than tell
108 Part II: Methodology overviews
It’s not just problems you should watch out for. You should also
remember that there are a few things that can make things a lot
easier on you, saving you time or even saving an interview from
being unusable. Here are a few:
begin the interview, you need to let them know all of this
through a prepared written statement about your project.
Let them know that:
1. You will be recording the interview, transcribing it, and
removing all personal names and places, even though
you may use excerpts in your project.
2. Some of the questions may be personal in nature – and
at any time they can choose not to respond.
3. At any time, you will stop the recording if they ask you to.
4. At any time, they can leave with no hard feelings.
5. If the interview stirs up any unanticipated emotions, you
have contact information for counselors if necessary.
Put all of these statements in your written consent form,
read it aloud to them, and then ask them to sign it if they
are still willing to participate. If they say consent, then you
can begin recording. Small digital recorders are fairly easy
to find, and your phone can do this too (just make sure you
turn it to “silent”). Please record through two devices, and
test them first to make sure the audio is usable. Trust us. Let
them know when you begin recording, place the recorder(s)
on the table, and begin your questions.
5. Take field notes: Focus on the questions, creating a natural
dialogue, and record any notes you think might be neces-
sary. We like to write down the setting, time of day, length
of the interview, and any nonverbal elements that you think
might be significant. For instance, if there was an emotional
response to one of the questions or a noticeable pause, write
it down. You might find that over half of your participants
became emotional over one of the questions, and if so it’s a
good thing you noted it from the start. In the end, you’re
not sure what will and won’t have a significant impact on
your findings. To be safe, write down more than you think
you might need.
7: Ethnography, interviews, focus groups 115
Focus groups
The process for creating focus groups is fairly similar to that for
creating interviews, but sometimes it gets a little more challeng-
ing because you’re trying to work around many people’s sched-
ules. We like to use online signups like WeJoinIn or Doodle for
these, depending on who we’re including in the focus group.
Both services allow you to make meetings easily at a time that’s
convenient for everyone.
In many cases, you’ll estimate around four to five people per
focus group, although many use much larger numbers. The rea-
son you want to keep it small is to reduce issues of groupthink
(when people start to fall in line with each other in order to
preserve a sense of community) and power (when one person
dominates a conversation). On the other hand, sometimes those
interactions give you valuable insight as to how discourse plays
out, and you can get a feeling for people’s responses in post-focus
group surveys or interviews if you’re worried about people that
didn’t speak as much. The goal, however, is to have a conversa-
tion. Interviews are often about deep individual storytelling. In
focus groups you’re trying to learn more about how participants
relate to one another, and how they can learn from and bounce
off of each other, than you are about individual responses.
Focus groups use interview guides in a similar way to inter-
views, but you’re more a moderator than an interviewer in this
case. Often researchers will bring in media such as magazines
for participants to flip through or show short videos and ask
for responses/discussion. This allows you to see how the vari-
ous groups relate to specific content, and how they figure out
7: Ethnography, interviews, focus groups 117
Final thoughts
it’s a good idea to get IRB approval. Not only are your
librarians familiar with this part of the process, but they are
often even members of their institution’s IRB. They can
also help with finding quality human participants for your
research. Because they are engaged in their own research
and have an idea of other research that is happening on
campus, they might even be able to put you in touch with
groups on campus or in your wider community that could
serve as resources or participants.
Surveys are one of the most widely used (and misused) methods
that you see “in the real world.” In some ways that’s because
they’re so common, and seemingly so easy. We’ve all responded
to a survey at some point, and we’ve all given one. After all,
“How many of you like pizza?” is a survey. Simply put, surveys
120 Part II: Methodology overviews
are just a tool for collecting consistent data from a large sample.
Unfortunately, like anything that’s overused, many surveys are
poorly designed and don’t gather meaningful data. Part of the
reason is that there are just so many types of surveys and so
many types of questions you can ask. They can range from small
questionnaires about your customer experience to more open-
ended inquiries about why you quit your job. Before developing
a survey, you should ask yourself: What do I want to know? Who
do I want to ask? How do I want to send the survey out, and will
that impact responses? In the end, you’ll need to consider your
audience, the format, the questions, and the overall layout to get
the most out of your results.
Probability sampling
Probability sampling is a method for randomly selecting partic-
ipants, usually using computer programs (e.g., Microsoft Excel
or SPSS) that number and randomize an entire population.
8: Surveys 121
Non-probability sampling
If you have a specific population in mind, you might need
to turn to another method that doesn’t rely on the general
population. There are a few ways you can do this. In con-
venience sampling, you select people close by or who you
know to take your survey (e.g., team members, classmates,
122 Part II: Methodology overviews
1. Demographic questions
2. Closed questions
Categorical question
1. What are the main reasons you haven’t left your
current job? (check all that apply)
Strong coworker relationships
High job autonomy
High salary
Ethical management
Open questions
Survey types
9 methods in
practice
(participatory-action research,
experimental research, theory,
and metasynthesis)
Participatory-action research
Experimental research
Experiments seem like they’re easy to do, since the idea isn’t
that complex. Take some of one thing and don’t do anything
to it. Then take some of the same thing and do something to
it. Compare. Then go eat ice cream and celebrate a successful
experiment. In reality, experiments can be one of the more dif-
ficult research methods to design and pull off. We’ve already
9: PAR, experiments, theory, and metasynthesis 139
Metasynthesis
Theory
not connected at all to what the company did or stood for, who
its employees were, or who their customers were. Some theorists
then suggested that CSR wasn’t inconsistent in impact because
it wasn’t strategic. In other words, the problem wasn’t CSR
itself. Theorists like Coombs and Holladay argued for “strate-
gic corporate social responsibility,” and others like Porter and
Kramer argued that we think of CSR as “creating shared value”
between a company and a community. These arguments weren’t
focused on data; they were focused on ideas, and in many ways
they reframed both what was studied and how.
10 qualitative data
and making
sense of detail
This chapter goes into more detail about how you ana-
lyze qualitative data once you’ve gathered it. Focusing on
textual analysis and interviews/focus groups, we discuss
how you determine what to look for, and how to choose
between a tightly controlled and structured approach and
a loose, open one. It’s probably not a shock that we sug-
gest something in between, a “managed” approach that
draws from existing research and your research questions
rather than hoping that the data will inherently guide you
(or that it will do so without bias). In each case we sug-
gest that you look at what might be “silent” or “erased”
material rather than only what you see. Finally, we give
you some hints about how to make this process easier for
yourself as well in the way you go about it and by getting
some help from a computer program.
10: Analyzing qualitative data 147
can guide you here, as can any hints for themes that you might
find in the literature to help you out. For instance, as we dis-
cussed above, in Owen’s (1984) article on thematic analysis he
suggests looking for repetition (a participant that says the same
thing multiple times), forcefulness (the strength of a partici-
pant’s statement, which might or might not appear with other
participants), and recurrence (when several participants say the
same or similar things). This framework is really useful, because
it allows you to eliminate even more issues of bias. In many ways
you’re looking for both similarities and dissimilarities, and that
gives you a lot of power as a researcher to make sure you’re
doing the data (and your participants) justice.
The process of pulling out themes from the data, rather than
choosing themes and then laying them on top of the data or
forcing the data into them, is called “grounded theory.” A more
structured approach, like we saw above, is called “managed
grounded theory,” in which you have an idea of themes before
you start but you remain open to them evolving or even being
discarded as you look at your data. Rather than a spreadsheet,
however, we recommend that you take all the transcriptions
(and we can’t emphasize enough how important it is to tran-
scribe! See our online resources for some suggestions on how to
make that faster) and highlight them according to theme. This
looks a lot like the way we talked about doing literature reviews
in Chapter 3, right? That’s no accident. It really is the same
process. Find themes as you go. Anything that stands out, that
you see repeated within or across transcriptions, or moments
when you ask yourself “wait, but what about . . .” are things
to code for, even though in the last bit you’re just pointing out
the absence of something (an emotion, a logical connection,
etc.). Then you label each highlighted section with the partici-
pant number or pseudonym and place all the highlights of each
color together. Those groups of quotes then become the basis
for your analysis.
10: Analyzing qualitative data 153
The fun thing about this kind of data analysis is that you can
allow the participants to speak, just like your paradigmatic exam-
ples “speak” about your data in text-based analyses. This kind
of research tends to use a lot of paraphrases and a lot of quotes,
sometimes even extremely long quotes, to make their points.
You’re not necessarily trying to say what these mean, you’re sim-
ply showing in a structured way how participants make sense
of their reality, how they take in and then express various stim-
uli like a TV show, conversation with a boss, or relationship
counseling appointment. That doesn’t mean you aren’t being
persuasive and making an argument; you are. It’s just that the
data speak for themselves, and if you’re being ethical, they speak
accurately. Remember, these have a relatively small N; you’re
not interviewing thousands of people in most cases, so you have
very little ability to generalize. But you can certainly say that
these people felt these ways about the topic, which can be just as
powerful and accurate as a survey with thousands of responses.
Silences
11 quantitative
data and
making sense
of numbers
Describing variables
Let’s say you want to tell your audience how satisfied a group
of people feel about a product. When you report this, you are
relaying your sample’s “bell curve” for that particular variable,
otherwise known as a “normal distribution” (see Figure 11.1).
A bell curve is often represented by the following figures:
11: Analyzing quantitative data 159
for the most part tell a more complete story. Do college seniors
know more than freshmen? About what? Are there significant
differences in how we pay men and women, and is that all men
and women or just those who have children? Does socioeco-
nomic status influence certain outcomes? Does race play a role?
Most statistical analyses focus on finding significant differences
between groups. We want to briefly walk you through three
common tests for identifying differences between groups and
variables. But before we do, let’s review a few definitions.
ANOVA
Similar to a t-test, the ANalysis Of VAriance Test, or as it’s more
commonly known the ANOVA (F) test, is useful when you
want to compare differences between groups. Unlike the t-test,
Dependent First independent Second Third independent
variable variable – year in independent variable – race
school variable –
gender
t-Test Communication (only two groups) X X
competence Freshman vs. Seniors
One-way ANOVA Communication (two or more groups) X X
competence Freshman vs.
Sophomores vs.
Juniors vs. Seniors
Two-way ANOVA Communication (two or more Male vs. X
competence groups) Freshman Female
vs. Sophomores vs.
Juniors vs. Seniors
Factorial ANOVA Communication (two or more Male vs. White (not Hispanic)
competence groups) Freshman Female vs. White vs. African
vs. Sophomores vs. American vs. Asian
Juniors vs. Seniors American vs. Other
Finding relationships
Correlation
The simplest inferential statistics test is the correlation. This type
of preliminary test is used if you want to quickly rule out or
confirm that two variables are related in some way. Say you’re
conducting research on nonprofit organizations. Is there a rela-
tionship between socioeconomic status and donor activity? Or
is there a connection between people who share the mission of
the nonprofit and those who volunteer? Relationships between
variables can happen in two ways: positive or negative. A posi-
tive correlation means that the two variables move in the same
direction. So, if you find that individuals who align closely to a
nonprofit’s mission also volunteer more often than those who
do not, then you have a positive correlation between “shared
mission” and “volunteerism.” In the same way, you could also
conclude that individuals reporting a low degree of affiliation
volunteer less. Correlations are reported using the (r) symbol.
After you’ve checked to makes sure the p-value is p < .05, you
can report how correlated the two variables are. Keep in mind
that a 1:1 correlation or (r) = 1.0 means that your two variables
168 Part III: Data analysis and the final push
Regression
The holy grail of inferential statistics is predicting future change,
and you often do this with what we call “regression analysis.”
While statisticians and theorists love to debate whether or not
you can ever truly show causation, regression analysis gets us
closer than the other tests. Linear regression examines how one
predictor, or independent variable influences one outcome, or
dependent variable. Similarly, multiple regression allows you
to include several different predictor variables (independent) to
determine the outcome of another variable (dependent). Keep
in mind that good statistical analysis is about telling the story
behind the numbers, and regression allows for that kind of com-
plex storytelling. It has four major benefits:
Descriptive
statistics
Mean (M) The average numeric value of all of the
data within a variable.
Median The numeric value in the middle when
the data is ordered from lowest to
highest value.
Mode The numeric value that appears the
most frequently.
Number of The number of participants or data
cases (N) points analyzed.
Range The lowest and highest number values
in the data set.
Standard The incremental distance each data
deviation (SD) point is from the mean and the other
data points.
Skewness When there is an uneven amount of
scores either higher or lower than the
average in a normal curve.
Independent The variable that is selected by the
variable research; it stands alone, and does not
depend on the value of the dependent
variable.
Dependent The variable that changes based on the
variable movement of the independent variable.
12 presentation,
and publication
We hate to say it, but no one really cares about your research. It
might be brilliant, but no one is excited to read a paper (often
not even if they know it’s good). It’s your job to make them
(ok, help them) care. There are a few ways to do this, but it all
174 Part III: Data analysis and the final push
starts with the beginning of the paper. Your title is really import-
ant, because it foregrounds what’s going to happen. You can try
to be cute about it, but at least try to be specific about what
you’re doing with a hint as to why. And keep it short. Sometimes
we start with a title and go from there; with this book, in fact,
we started with a title and that guided our approach. Then we
changed the title, because it was pretty bad. But it still inspired
us to structure everything in a specific way and for a specific set
of audiences. Hopefully you liked it.
After the title comes the abstract, which you again could write
before or after the paper is completed. However, you’ll always
want to change the abstract when you finish, since if you’re doing
your job right the data that you’ve analyzed should have guided
you to conclusions you didn’t necessarily expect when you wrote
the abstract. Basically an abstract is a very brief (fifty to one hun-
dred fifty words, usually) introduction to your paper that people
can read instead of (or hopefully as a prelude to) your full paper.
It should include your topic, your focus, your method, and, in a
very brief form, your conclusions. You might also talk a bit in the
beginning of your paper (in your abstract and your introduction)
about why the research matters. If this sounds a lot like an anno-
tated bibliography entry, that’s because it is. It’s exactly like that.
After the abstract comes the body of the paper, although you
still need an introduction. This is where you can either really get
someone interested in your project or make them stop reading.
You can’t just jump into the literature review; you might have
hated writing it, but they’ll hate reading it just as much. And
you definitely don’t want to start with general, vague statements.
There’s not much worse than reading someone writing about
“society” or “from the beginning of time” or “the beginning of
humankind” or some other such ridiculous statement. Yes, we
know that almost everyone has been trained to write that way.
Stop it. Please, for everyone’s sake, cut everything vague and
undefined like that out of your paper.
12: Persuasion, presentation, publication 175
1. Edit for tone. One of the biggest mistakes that even pro-
fessionals make is that they start ranting in their research.
They’re so passionate about their work, and so convinced
by their results, that they dive into what we call “polemic,”
which basically means a kind of verbal attack. This is not only
alienating for your reader, but it’s also a clear sign that you
haven’t given yourself the kind of distance necessary to eval-
uate your project for bias, errors, and unexpected results.
2. Edit for content. You’re probably exhausted by your proj-
ect at the end. So put it down, walk away, and come back
later to edit. You’ll find that you can probably cut a lot of
it out (maybe more than half of it), and tighten the rest of
it up. You also might find that there are some big chunks
missing that you thought you said because they were in your
head, but that you didn’t actually say. The best way to do
this is to have a clear word count for your paper, and write
double that amount in the first pass. If you do your literature
review and your data analysis right, that’s the very least you
should have. If you haven’t written too much, you’ve writ-
ten too little. And if you’ve written too little, you definitely
won’t have a good paper.
3. Peer review! After you edit it yourself, read it aloud to a friend.
That probably sounds horrible, and we understand that. But
if you want to catch all the places you didn’t make sense, read
it out loud to them rather than having them read it. That way
you’ll catch the problems with grammar and structure, but
you’ll also be able to have them stop you and say “hold on.
178 Part III: Data analysis and the final push
What? What are you talking about? I’m lost.” That’s the best
thing they can possibly do for you, so buy them pizza or ice
cream as a thank you. If you can then send the paper to some
friends that know something about your method (maybe
someone in your class, or another graduate student or col-
league), do that as well. They’ll give you some valuable insights
that will make the project much stronger, or at least make you
feel a little better about writing something strong.