Content Analysis

1
Content Analysis
Andrew Austin 2009
2
Hypothesis PBS has more liberal bias than ABC
13 frequent guests 5 US officials (4
conservatives) and 4 "experts"- including
rightwing activists Robert Bork and Patrick
Buchanan. Rightwing activists (e.g. Bork and
Buchanan) were often presented as neutral
"experts." Despite dubious credentials on the
subject, Buchanan appeared on two Nightline
programs about terrorism. 89 of Nightline's US
guests were white and 82 were male. On programs
about international issues, 96 of US guests were
white and 90 were male. On domestic politics,
26 of guests were women. On Nightline, 34 of
its U.S. guests were current or former government
officials, 39 were professional and 5 were
corporate representatives. Compared to this 78
"elite opinion," 10 of ite guests represented
public interest, labor, or racial/ethnic groups.
19 frequent guests 9 US officials (6
conservatives) and 4 "experts" including Norman
Ornstein of the AEI and William Hyland of
Foreign Affairs. Guest list dominated by two
conservative Washington think tanks Center for
Strategic and International Studies (8
appearances) and AEI (6 appearances). Experts
from progressive think tanks never appeared.
90 US guests were white and 87 were male. On
programs about international issues, 94 of its
U.S. guests were white and 94 were male. On
domestic politics, 21 of its guests were women.
46 US guests were current or former government
officials, 38 were professional and 5 were
corporate representatives. Compared to this 89
"elite opinion" only 6 of its guests represented
public interest, labor or racial/ethnic groups.
FAIR study guest lists of the two shows for a
6-month period Feb 6-Aug 4,1989.
3
SPECIAL REPORT WITH BRIT HUME
WOLF BLITZER REPORTS
RepublicansAbigail Thernstrom (con) Andrew Card
(con) Angela Antonelli (con) Bill Bennett (con)
Bill Kristol (con) Bill Thomas (con) Bradley
Smith (con) C. Boyden Gray (con) Chester Finn
(con) Christine Todd Whitman Chuck Hagel (con)
Connie Mack (con) Dick Armey (con) Don Nickles
(con) Donald Rumsfeld (con) Doug Paal (con) Ed
Goeas (con) Gale Norton (con) George W. Bush
(con) Jack Kemp (con) Jeb Bush (con) Jim
Wilson (con) John Diulio (con) Ken Blackwell
(con) Ken Kies (con) Lionel Chetwynd (con)
Lynne Cheney (con) Mitch Daniels (con) Mitch
McConnell (con) Pat Toomey (con) Paul Gigot
(con) Pete Domenici (con) Randy Cunningham
(con) Christopher Shays Richard Allen (con)
Sally Satel Bill Frist (con) Stephen Moore
(con) Trent Lott (con) Bill Kristol (con)
Republicans Ann Veneman (con) Arlen Specter
Bob Dole (con) Charles Grassley (con)
Christopher Shays Chuck Hagel (con) Dan Burton
(con) David Dreier (con) Donald Rumsfeld (con)
Frank Keating (con) Frank Murkowski (con) Fred
Thompson (con) Henry Kissinger (con) James
Baker (con) James Brady Jeff Sessions (con)
John McCain (con) Karen Hughes (con) Kay
Bailey Hutchison (con) Laura Bush Linda Chavez
(con) Lynne Cheney (con) Mark Kirk (con) Mary
Matalin (con) Michael Deaver (con) Michael
Reagan (con) Pat Robertson (con) Richard
Thornburgh (con) Spencer Abraham (con) William
Bennett (con)
Only eight of Special Report's 92 guests during
the study period were women, and only six were
people of colormaking for a guest list that was
91 percent male and 93 percent white.
Democrats Bill Richardson Bob Kerrey Chaka
Fattah Charles Rangel Christopher Dodd Dianne
Feinstein Edward Kennedy Jack Quinn Jake
Tapper James Carville Joe Kennedy Joseph
Lieberman Kent Conrad Lanny Davis Mario Cuomo
Paul Wellstone Richard Durbin Richard
Holbrooke Robert Reich Robert Torricelli
Samuel Berger Sander Levin Thomas Daschle
Democrats Bob Graham Joe Lieberman John Breaux
John Spratt Jon Corzine Richard Holbrooke
?
?
Guests were 86 percent male and 93 percent white.
Source FAIR studied 19 weeks of Special Report
with Brit Hume (1/1/01-5/11/2001)
4
Types of Communication Media Newspapers,
magazines, and books TV programs, radio
broadcasts, commercials, and movies Official
documents and texts such as court records,
minutes of meetings, legal codes, and policy
statements Diaries, letters, and other personal
documents Photographs and art
5
Content analysisanalyzing and coding the content
of various communication media, i.e., existing
written, audio, or visual communication, in an
effort draw conclusions about the
characteristics, attitudes, and behaviors of the
individuals or groups that produced it and/or the
surrounding culture. Content analysis is used by
both qualitative and quantitative researchers.
Positivistic/quantitative Classifying and
counting occurrences Deploying preconceived
coding categories Emphasis on manifest
content. Focus on hypothesis testing
Interpretive/qualitative Classifying and
categorizing responses Developing categories
during data analysis that fit the data Emphasis
on latent content Focus on theory building
6
Doing Qualitative Content Analysis
Research questions General questions rather than
hypotheses Theory and sensitizing concepts Data
sources What type of data will you need to answer
questions? What types of data are available?
Sampling Unit of analysis and unit of
observation Sampling levels All methods may be
used Data analysis/coding Emphasis on latent
content Developing coding categories from the
data Reliability and validity
7
Doing Quantitative Content Analysis Identifying,
locating, and gathering data sources Choosing
media What is your research question and/or
hypotheses? What is the availability of the
desired media? Sampling and selection Stratified,
multistage, cluster sampling Sampling
levels Measuring and coding data Selecting
indicators and developing a coding guide Counting
occurrences of indicators Coding manifest versus
latent content Reliability and validity. Multiple
coders and inter-rater reliability
8
Advantages/Disadvantages of Content Analysis
Advantages Content analysis is unobtrusive and
non-reactive. It is excellent for addressing
broad cultural values and beliefs, especially as
they change over time. It is relatively easy,
cost efficient and safe, usually being conducted
in a library. Content analysis presents few
ethical problems. Disadvantages It can be very
time consuming. The quality of the research is
affected by the quality of the data. The
reliability and validity of coding can be
difficult to demonstrate. The unit of observation
is limited to previously "published" data,
limiting its application to specific types of
research questions.