Media Effects Research Lab - Research Archive
Exploring the Significance of Modality and Topic on Political Messages
Student Researcher(s)
Richard Wirth (Ph.D Candidate);
Sara Erlichman (Ph.D Candidate);
Ashley Jiang (Masters Candidate);
Faculty Supervisor
This project was done based on COMM 517 course.
INTRODUCTION
The current investigation aims to shed light on how message modality and topic affect attitude towards content and the ability to retain pertinent information. The innumerable forms in which political information is delivered raises concerns over the efficacy of the various modalities being employed. It is well acknowledged by many research industries such as Pew that Americans are more politically polarized than ever before. However, beyond just measures of general ideology, a well-established trait in those with strong party identity is that of attitude preservation. Attitude preservation is characterized as the opposition to counter-attitudinal messages through selective processing and lowered motivated reasoning (Redlawsk, 2002). Prior research has shown that two of the most critical factors contributing to significantly reduced preservation behavior are useful information (Hart et al., 2009) and an abundance of information (Redlawsk et al., 2010). This research suggests that the ability to retain information from political messages is critical to promoting cross-cutting behavior.
In addition to this severe political polarization, it is important to consider the ways in which media perceptions are affected. According to a Pew study, liberals and conservatives perceive the news with stark differences in trust levels (Mitchell et al., 2014). Due to these differences between the two parties, trust and related constructs such as perceived credibility may play a significant role in preferred news consumption.
Another concern in need of addressing in the complicated relationship between political news consumption, modality, political ideology, perceived credibility, and attitude, is the increasing amount of political information to sift through, is the possibility of information overload. The Limited Capacity Model of Mediated Message Processing (LC4MP) assumes that information processors have a fixed amount of cognitive resources, and that media richness and higher levels of multimodality exacerbate this issue. However, with the ubiquitous nature of information technology, power users have emerged for whom media is highly familiar. Research has shown that power users are notably more efficient and innovative in their usage of media devices (Zhong, 2013) and significantly prefer higher levels of interactivity and customization (Sundar & Marathe, 2010). Despite the established relevance of power usage as a factor in media usage studies, it remains underexplored within the context of political messages.
In light of the contemporary surge in political information and the concerns surrounding its consumption, the present study reviews research on topics surrounding political messages in media, before reporting the results of empirical tests of hypotheses derived from this review.
RESEARCH QUESTIONS/HYPOTHESES
For the general population, what is the relationship between modality, topic, and power usage, on cued and free recall, and attitude toward political news stories?
H1: As the complexity of modality increases (text to audio to video), there will be a significantly positive relationship with perceived source and message credibility.
H2: Cognitive elaboration will significantly mediate the relationship between modality and attitude.
H3a: Higher levels of message modality will be positively related to Cued Recall.
H3b: Higher levels of message modality will be negatively related to Free Recall.
H4a: Power Usage will positively moderate the effects of video modality on memory, and negatively moderate the effects of text modality on memory.
H4b: Power usage will significantly moderate the relationship between topic and attitude.
METHOD
A 3 (Modality: text versus audio versus video) x 2 (Topic: Immigration versus Healthcare) factorial, between-subjects experimental questionnaire was conducted. Participants were randomly categorized into one of the three modalities (text, audio, and video) and one of the two topics (Immigration and Healthcare). Participants answered pre-stimulus questions that measured the dependent variables, and then were asked to read, listen, or watch a news story about either Immigration or Healthcare. After the stimulus, participants answered questions investigating their memory of the content and other independent, mediating, and moderating variables.
RESULTS
H1. Our findings showed that as complexity of modality increased (text to audio to video), participants perceived the stimuli with more message and source credibility.
H2.
H3a. Modality was found to have a significant relationship with cued recall, particularly in the video condition were there significantly higher memory scores compared to audio; the text condition was not significantly different than either conditions.
H3b. Two three-way interactions demonstrated the interaction between Topic, Modality (text and audio), Involvement on Free Recall. However this interaction differed by Topic. For the Immigration condition, video and audio held a positive relationship, and text a negative. The inverse was true for healthcare, text held a positive relationship and video and audio was a negative relationship. Overall, Topic had a significant relationship with Free Recall, a positive one with Immigration and negative relationship with Healthcare.
H4a. This hypothesis was not supported.
H4b. Power Usage had a significant relationship with Value Judgment, Message Credibility, and Source Credibility.
CONCLUSIONS/DISCUSSION
Of particular note is the three-way interaction identified between Topic, Modality, and Power Usage on Cued Recall. The collected data revealed a significant relationship between power usage and cued recall, such that for the immigration topic, the video condition resulted in significantly higher recall, while the text condition lowered recall. More interesting yet is that the second topic -- healthcare reform -- showed an inverse relationship with modality on cued recall. In comparing the two video conditions (the base form of the stimuli), we can see that there were previously unnoticed differences between the two. The video for the immigration topic was notably more visually active, including a series of cuts to different speakers, visual representations of the legislative process, maps of the U.S., a change in speaker, and more to read on screen. This suggests that as power usage increases, individuals benefit from more visually rich information and a higher rate of encoding, whereas a text-only stimuli does not adequately engage the four dimensions of power usage, likely that of expertise and behavior. Conversely, the healthcare video is much less visually rich, and is narrated by a single individual with the only supplementary content being a bulleted list of policy reforms. This supports our earlier interpretation that power users benefit from information presented in appropriate forms, as the simpler and more direct text-based delivery of such straightforward information would satisfy the power user’s need for efficiency.
Based on existing literature, we hypothesized that as the complexity of modality increased, there would be a significant increase in perceived message and source credibility. Our results suggest that the most valued modality is video, and that audio and text are less likely to gain traction in political contexts, where credibility and trust are critical elements of information seeking behavior (Rieh & Hilligoss, 2008). Similarly, we investigated the role of trust in the processing of political messages, and found that it significantly moderated both factors of our attitude measure (evaluative trust and value judgments) as well as message and source credibility. The relationship to credibility and attitude can be summarized as the increasingly important role of trust in the media in the face of a constantly shifting landscape of news content and punditry. The interaction effect is an unexpected but impactful finding, which suggests that due to the impersonal nature of text compared to an individual’s voice, there is a significant difference in the evaluative trust attitude factor.
Another important finding regarding the factors contributing to information-seeking behavior and news content selection is that of power usage. Power usage was found to be significantly related to both the evaluative trust factor of attitude, as well as on message and source credibility. This is likely due to the fact that power users are logically more evaluative and expertise-driven in their information technology usage.
The relationships identified between topic and modality on free recall were largely in line with theory drawn from prior literature, which suggests that when message content is incongruous with modality, there will be lower levels of attitude but higher levels of free recall.
For more details regarding the study contact
Dr. S. Shyam Sundar by e-mail at sss12@psu.edu or by telephone at (814) 865-2173