You Can Produce Effective PSAs
Research is the Key to Effective Communication
by William D. Novelli
Despite widespread criticism, public health PSAs can be effective if
you maximize your efforts by taking a
systematic, evaluative approach to
Public service television announcements (PSAs) have become
part of America's
mass media lore. Late-night TV watchers have seen hundreds of PSAs for
virtually every illness
imaginable, often with appeals based on guilt (e.g.,
the crippled-child visual) or fear ("20 million Americans
disease, and it usually strikes without warning.").
One of the more negative descriptions of PSAs was offered a few years
ago by the director of community
affairs for a major TV station in Boston.
"My image of PSAs," he said, "is seven dull doctors
kidneys at 6 A.M."
Aside from predictability and dullness, other criticisms of health PSAs
have been leveled at their
producers. Stephen J. Weber of Market Facts,
Inc., believes that many creators of PSAs do not understand their
In his book. "Government Marketing Through Public Service Advertising"
Mr. Weber describes some PSA producers as "naively
optimistic": They cite statistics about the millions
of TV watchers per
day, extrapolate from this the "awesome might of advertising,"
and assume this
power extends to PSAs. He characterizes other producers
as "naive pessimists" who consider unpaid
advertising a complete
waste of effort.
Public relations and advertising people largely involved in promoting
commercial goods and services, also
have been criticized for their handling
of PSAs. These campaigners for toothpaste, cosmetics, automobiles and
products sometimes take a simplistic view of health and social issues communication.
While they might
carefully research and plan each step of a commercial
program, they may toss off PSA creation and production
with little understanding
or attention to the complex issues involved.
Applying Commercial Methods
Despite the criticisms, mass media campaigns can be important sources
of information on public health. The
challenge is to maximize your communications
by taking a systematic evaluative approach to public service programs.
One common barrier to effective health PSA development has been inadequate
pre-campaign research and analysis.
Usually little effort is made to understand
audience needs, expectations and perceptions. Segmentation by meaningful
audience characteristics (e.g., demographics, benefits sought, risk factors,
readiness to adopt new health habits)
often is ignored. Also, PSAs and
other mass media messages frequently are not pretested for message clarity,
recall or appeal.
Such message pretesting (also known as copy testing) is a common practice
among public relations firms and
advertising agencies when promoting commercial
goods and services. Health agencies and organizations, recognizing the
need to measure effectiveness of their communication, have made great strides
in applying commercial copy testing methods
to health messages. In 1976,
the National Cancer Institute's Office of Cancer Communications and the
branch of the National Heart Lung, and Blood Institute
assembled a working group of health communication researchers,
planners, and marketing researchers to develop a standardized system or
health message pretesting.
At the initial meeting, the Health Message Testing Service (HMTS ) began
to take shape. The working group identified
two key objectives: to help
message producers improve the quality of health PSAs prior to final production;
isolate those characteristics of health PSAs that contribute to
more effective communication.
The Service determined that health message testing should be: open-ended
(to allow for respondent probing);
able to recruit from specific target
audiences; based on natural (at home) or "reasonably" forced
exposure to the messages; capable of ingredient analysis (e.g., a character
or the music can be isolated and
examined); valid and reliable; reproducible;
affordable; and timely.
In addition to these criteria, the
working group considered it important
for the testing approach to accommodate prefinished messages in line with
the objective to improve the message before the expense and effort of final
Members of the
working group debated test methods. Some wanted to develop
new methodology tailored to the specific needs of
others opted for a commercial testing service, with minor adaptations it
health PSAs. This argument finally was settled by budget. An
existing commercial system, modified for health spots,
proved to be far
more affordable than the creation of an entirely new service.
After an extensive search, a research supplier, McCollum/Spielman &
Company of Great Neck, NY, was chosen
to collect data on TV and radio
PSAs, and provide field services for the Health Message Testing Service.
used is forced exposure of l00 respondents with 100 from each
of four geographically representative cities.
Respondents are invited to attend a session to evaluate new TV and radio
programs. They are not told that they
will be asked to assess commercials
and public service announcements, Respondents are grouped around four TV
monitors. All interviewing is prerecorded to avoid interviewer bias.
Both commercials for goods and services
and health PSAs are tested,
along with control messages that are kept constant from test to test. The
shown between variety programs, and respondents are exposed
to each message twice. The first program consists of four
these are interrupted by a "station break" in which seven commercials
are played in sequence.
This "clutter" sequence lasts three-and-a-half
to seven minutes, depending on message length (30-second and
messages are tested), and simulates a normal on-air station break. Only
one health PSA is tested in a session.
It's felt that mixing the health
spot in with consumer-oriented commercials helps to create a reasonably
typical viewing climate.
To maintain the illusion that the programming is the area of interest,
respondents are asked a series of questions
about the variety programs.
Then recall of the commercials and PSAs is measured by two methods. The
first method is
to assess message attention. Respondents are asked to name,
aided, all the messages they can remember. Second, respondents
to write down the main idea each message was trying to communicate. These
two recall measures are taken about
30 minutes after respondents first
see the test messages.
Following this sequence, the test messages are "
shown again, this
time within a situation comedy. Unlike the earlier "clutter"
format of the station break,
the messages appear one at a time, separated
by about six minutes of program content. Then another battery of questions
is administered. This round of questions is designed to yield diagnostic
information about the respondents' perceptions
of the health PSA.
Reactions are probed concerning message believability, personal relevance,
likes/dislikes, and elements of the message. Three measures are
standard in HMTS test's: attention, main idea
communication and believability.
These measures are then compared with the normative data for all health
that have been collected over the years.
Since its inception, the service has tested 63 TV and l9 radio PSAs
on health. Also, HMTS has developed a test for print PSAs used for health
pamphlets, flyers and posters. As with television,
the radio and print
tests use standard pretesting techniques to assess attention, main idea
communication, and believability.
HMTS has been particularly successful in meeting its objective of improving
PSA communication before final production
and distribution, For example,
test results of a PSA promoting sensible weight loss suggested that the
lacked credibility. About half of the respondents questioned
her as a source of weight-control information, since they
found it hard
to believe she had ever been overweight.
In another case, two prefinished PSAs on jogging for
fitness stemmed from
the same communication strategy and contained similar information. Test
findings indicated that
the message based on a testimonial approach (people
talking directly to the camera about personal exercise experience)
better than the other PSA, which employed a special-effects montage of
jogging images with narration.
Elements of The Effective Message
HMTS's goal is not to devise a rigid formula for message development,
since that could stifle creativity and lead to a dull sameness that might
actually reduce message impact. Rather, the point
is to create a general
understanding of what elements of the PSA have the most potential for gaining
recall and engendering believability.
To correlate message characteristics with high test score, the Department
Communication Arts and Theaters, University of Maryland, and Porter,
Novelli & Associates analyzed 58 PSAs judged
effective between l977
and 1980. Most (84 percent) of the PSAs analyzed were 30 seconds long. About
33 percent of the
messages used ''vignettes'' to tell their stories; 31
percent had one central presenter; 31 percent used a "slice
approach; and 3 percent used a "demonstration" approach.
Music was used on 40 percent of the spots. Forty percent used a voice-over
announcer, 24 percent had on-camera audio delivery, and 36 percent employed
a combination of these two techniques. About 30 percent of the characters
in the spots were adults only, while 38 percent
featured both adults and
children. Health professionals appeared in 10 percent of the PSAs, and
about the same percentage
In developing the coding instrument for the analysis, the various message
classified: character variables (e.g., number of characters,
number of main characters, sex and age of characters),
(e.g., number of words of audio, when the subject was introduced in the
message, presence of an audio
slogan); production variables (e.g., message
length, production method, whether message is clear from video alone):
and strategy variables (e.g., emphasis on detection, prevention, treatment,
message appeal on fear, humor, guilt,
social support, peer pressure, warm/touching
situations, straightforward presentation).
The dependent variables in the study were the three primary HMTS measures:
attention, main idea communication and
believability. The independent variables
were the character, content, production, and strategy variables cited
plus demographic characteristics of the test audiences (sex, education
Study findings disclosed a series of characteristics that resulted in
higher message attention: fewer characters
(fewer than the average of three
per PSA); audio repetition of the message subject (four times among the
scoring spots); repetition of an audio slogan; implied social benefits
of the health behavior; and a straightforward message appeal.
Regarding main idea communication, two PSA characteristics were associated
with higher recall: straightforward
presentation of facts and implied social
benefits of the health behavior.
Finally, the study results indicate
that an emotional appeal correlates
with higher message believability. The top five messages (in terms of
that were analyzed used either a high or moderate emotional
appeal. Examples include a New York State PSA on mental
urged the public to accept the retarded into their neighborhoods, and a
message on the importance
of breast feeding to babies' health.
Another variable found to be associated with believability is live-action
production (as compared with animated,
illustrations/sketches, still photos
and other techniques). HMTS is designed to test health PSAs prior to final
(live action) production, often using still photos with camera movement
or some other simulation of live action.
Producers of health PSAs should
realize that their messages may be more believable when finally produced
action, but they should not use expensive live-action production
techniques in pretesting simply to achieve higher scores.
Pretesting Pros and Cons
Copy testing is the subject of continuing and often heated debate in
advertising and public relations circles. Critics of copy research point
out that recall is only a partial measure
of persuasion, that prefinished
mood and image commercials are difficult to test and their effects hard
and that current test methods often lack the sensitivity to
discriminate between similar executions of the same strategy.
Underlying these and other criticisms is the fear that testing will
be used as a substitute for judgment and that
testing can lead to dull
This study is meant to extend the debate into the realm of health
PSAs. Clearly, test data and generalizations based on these data must not
be slavishly followed, but should be tempered
by good communication judgment,
based on experience and a thorough understanding of health communication
as well as the
However, it is also important to note that message pretesting can make
contribution to the health communication process by providing
creative guidance, encouragement and stimulation,
and by saving considerable
time and money.
Bill Novelli is a professor in the McDonough School of Business at Georgetown University and has created the
Global Social Enterprise Initiative at the School. He co-founded and was President of Porter Novelli, now one of
the world’s largest public relations agencies. He was also CEO of AARP, a membership organization of over 40 million
seniors. Prior to Prior to AARP, he was President of the Campaign for Tobacco-Free Kids, and Executive Vice President
of CARE, the world’s largest private relief and development organizations.