Center of Excellence in Cancer Communication Research

CECCR I Message Core


The Message Core will provide assistance in message design and testing to major and pilot CECCR projects, evaluate new methods for assessment of message characteristics, and advance theories of message effects relevant to cancer control. It will cooperate with the Annenberg National Health Communication Survey (ANHCS) to maintain an online laboratory for message testing and collaborate with cancer-related research projects elsewhere at Penn and around the U.S.

Back to top


Much of the research in cancer communication requires the use of messages that are pre-tested for their effectiveness in the context of application. The clearest test of a message’s effectiveness is in the study employing the messages themselves where the psychosocial and behavioral effects can be directly observed. However, such a test is too risky without careful pre-testing. The purpose of the "Message Core" is to assist the CECCR’s major projects and pilots studies in message development and to provide efficient, reliable, and valid pre-testing of cancer control messages using relevant and representative populations. Recent development of efficient tools and procedures for message evaluation makes a message testing and evaluation core possible. These tools include measures of video PSAs executional features, extraction and efficient evaluation of a video ad’s arguments independent of the format of their presentation, and ratings of redundancy between visual and verbal streams in video materials. Other researchers have built a strong case that perceived effectiveness of a persuasive message is causally prior to actual effectiveness, the effects are robust and resistance to a message’s claims are readily assessed.

Primary Aims

Aim One: To provide to the CECCR II major projects and pilot research projects help in the selection, design, testing, and evaluation of messages employed in the research projects.
Aim Two: To provide an online laboratory for rapid and efficient message testing of cancer control messages in response to new developments with appropriate and representative target populations.
The Annenberg National Health Communication Survey (ANHCS) is a monthly survey of approximately 300 participants. The survey consists of a set of core questions with available space each month for embedding brief experiments or small survey modules. The Message Core will take advantage of ANHCS for rapid testing of messages with representative populations.
Aim Three: To collaborate with cancer-related research projects at the University of Pennsylvania and around the country in selecting, designing, testing, and evaluating messages for cancer communication research.

Secondary Aims

Aim Four: To develop and evaluate new methods and tools for the efficient and valid assessment of theoretically important message characteristics pertinent to cancer communication campaigns and interventions.
Aim Five: To advance theories of message effects relevant to cancer control that integrate with established theories of information processing and behavior and attitude change.

Back to top


Message design must be theory driven.

Messages can be designed uniquely for each context of application or theoretically using principles of effective message design. Unique, context-specific design not driven by theory is inefficient. Messages designed and tested in the absence of theory produce little knowledge that can be transferred to other contexts.

Most research in health communication that deals with individual behavior must address questions about how a behavior can change [10], how information pertinent to the behavior is processed [11-13], and how a message can be designed to maximize processing and change [14]. Integrating these theories and harnessing their insights into message design and testing to advance cancer control is one goal of the Message Core. Extant theories of message processing and behavior change direct attention to message content (specifically the arguments made within the message) and to how the content is presented (format, specifically executional features). Research on these broad domains has advanced steadily.

Core argumentative content in persuasive messages.

Argument strength appears to be the most commonly manipulated message feature in persuasion research [15]. When audiences process information centrally [11], strong arguments produce more attitude and belief change than do weak ones [16,17]. The most widely used measure of argument strength is thought-listing, which assesses audience’s cognitive responses to persuasive messages [11,18,19]. Messages that elicit predominantly favorable thoughts are likely to change attitudes in the advocated direction; messages eliciting predominantly unfavorable thoughts typically do not.

Message sensation value (MSV)

MSV describes a set of message features that can function to attract attention. The features are derived from theory about the sensation seeking personality [20,21] and are presumed to elicit sensory, affective and arousal responses [22-24]. These structural features include formal video features (e.g., cuts, edits, special visual effects, etc.), formal audio features (e.g., sound effects, music, voiceover, etc.), and content format features (e.g., act out vs talking head, surprise ending) [2].

Empirical studies have generally supported the impact of certain MSV features (e.g., fast-paced editing) on arousal using both physiological measures such as heart rate and skin conductance [25-30] and self-reported measures [22,31]. Some individual features of MSV, such as edits, cuts [32-34], visual graphics [35], pace [36,37], and emotional messages [38,39] have shown effects on attention, arousal, memory and cognitive capacity [26,40]. Video messages can be readily and reliably coded for their sensation value [40,41], providing important information about an ad’s ability to attract attention and place processing demands on the audience.

Information Introduced.

Information Introduced (I2) is a measure of television message complexity [1]. The measure is designed to tap the cognitive resources required and employed in the processing of video messages. Coding rules are derived from structural changes in video messages known to produce cognitive and physiological responses. Camera changes (CCs) produce orienting responses [29,33] but the information introduced following a CC can alter demands on the resources viewers allocate. After each CC, I2 codes: Object change, new object, relatedness to the prior image, distance closer to focal object, perspective change, and emotion change. Lang et al. [1] detail the cognitive and physiological bases for each of these codes. The result is a measure that tracks the sometimes rapid changes in cognitive demand by a video message. Information introduced offers a more continuous measure of cognitive demand than does MSV even though the two are similar. The theoretical and empirical relationships between these measures require scrutiny but together then offer insight into the structural relations within complex video messages.

Redundancy between message modalities.

MSV and I2 refer primarily to the visual aspects of a video message while arguments are primarily delivered verbally or through written text (but not exclusively, [42]). These two modalities of presentation can work together or be at odds with one another. Visual cues that draw viewers’ attention could distract them from verbal segments carrying other information [43]. A message deploying strong arguments may still fail if the visual cues drawing cognitive processing resources distract from the central content. The degree of redundancy between the visual and verbal modality (VVR) can play an important role in understanding conditions under which persuasion is enhanced in audiovisual messages (through memory for consonant visuals and verbals) and when it might be undermined (through dissonance). VVR is a property of video messages that is rated by viewers rather than objectively coded [4].

Messages can be evaluated for their effectiveness efficiently.

Researchers in message effects are beginning to employ an efficient measure of actual message effectiveness based on perceptions of a message’s perceived effectiveness by the target audience [5,40,44]. The advantage of this shortcut is efficiency. Perceived effectiveness judgments preserve resources and provide valid - if imperfect - assessments of message effects. Dillard, Shen, & Vail [5] carried out five experimental studies to evaluate the causal direction between perceived effectiveness (PE) and actual effectiveness (AE) concluding that "PE may be viewed as a causal antecedent of AE" (p. 467). A meta-analysis of the PE - AE relationship from 40 studies reports a highly significant effect size of r=.41 (95% CI (.38, .43) (adjusted for attenuation) [6]. PE measures for messages can be valid and efficient tools during message pretesting. Messages can produce reactive responses as well in which audiences resist the source’s persuasive attempts in order to maintain their personal autonomy [45,46]. Building on research on message reactance [47,48], Quick and Stephenson [7,9] have offered consistent evidence that when negative thoughts and anger are elevated in response to a message, reactance increases as does resistance to the message. Attempts at freedom restoration includeing defensive processing and counter-arguing ensue [8].

Developments in perceived message effectiveness and in measures of message reactance have provided a basis for employing tools in message evaluation during pretesting that allows for efficient testing while also offering valid information about a message’s likely effects when actual effectiveness is observed.

ANHCS is an efficient platform for message testing.

The Annenberg National Health Communication Survey (ANHCS) is designed to capture national trends relating health behavior and intentions to media exposure, health knowledge and beliefs. It is the only survey that is continuously in the field with a nationally representative sample focusing on health communication. ANHCS has core questions about health and communication but also reserves limited space for online experiments and for temporary question modules. The opportunities for experimentation and additional question modules offer the ability to be responsive to immediate testing needs and to unexpected developments in the public arena.


The Message Core will allow studies carried out in CECCR II to employ message interventions that undergo careful pre-testing using a common set of reliable and valid message assessment tools. The Message Core can offer consultation about theory-based message design and assessment. The Message Core will also provide assistance in message design when messages are created anew and in message selection when pre-existing messages are deployed. In the process, message evaluation and theories of message effects in cancer control will be advanced.

Back to top