Using the Extended Parallel Process Model to create ...

Using the Extended Parallel Process Model to create and evaluate the effectiveness of brochures to reduce the risk for noise-induced hearing loss in college students Michael R. Kotowski, Sandi W. Smith1, Patti M. Johnstone2, Erin Pritt2 School of Communication Studies, University of Tennessee, Knoxville, 293 Communications Building, Knoxville, TN 37996, 1 Department of Communication, Michigan State University, 573A Communication Arts and Sciences Building, East Lansing, MI 48824, 2Department of Audiology and Speech Pathology, University of Tennessee, Health Science Center, 444 S. Stadium Hall, Knoxville, TN 37996

Abstract Brochures containing messages developed according to the Extended Parallel Process Model were deployed to increase intentions to use hearing protection for college students. These brochures were presented to one-half of a college student sample, after which a questionnaire was administered to assess perceptions of threat, efficacy, and behavioral intentions. The other half of the sample completed the questionnaire and then received brochures. Results indicated that people receiving the brochure before the questionnaire reported greater perceptions of hearing loss threat and efficacy to use ear plugs when in loud environments, however, intentions to use ear plugs were unchanged. Distribution of the brochure also resulted in greater perceptions of hearing loss threat and efficacy to use over-the-ear headphones when using devices such as MP3 players. In this case, however, intentions to use over-the-ear headphones increased. Results are discussed in terms of future research and practical applications. Keywords: Extended Parallel Process Model, noise-induced hearing loss, health communication

Introduction Using the Extended Parallel Process Model (EPPM) to create and evaluate the effectiveness of brochures to reduce the risk for noise-induced hearing loss (NIHL) in college students who are exposed to excessive exposure of loud noise is a well-recognized hazard. According to the National Institute on Deafness and Other Communication Disorders,[1] nearly 15% of American Adults (26 million people) between the ages of 20 and 69 years may have suffered permanent hearing damage as a result of hazardous noise exposure. NIHL may be caused by one-time or prolonged exposure to hazardous noise.[2] The intensity and duration of noise determines its level of harm. Depending on the intensity of the sound and the duration of the exposure, permanent sensorineural hearing loss and tinnitus (ringing in the ears) can ensue.[3] The intensity of sound is measured in decibels Access this article online Quick Response Code:

Website: www.noiseandhealth.org DOI: 10.4103/1463-1741.82958 PubMed ID: ***

261

(dB) ranging from 0-180 dB(A). Reference points often cited include: Whisper (30 dB[A]); normal conversation (60 dB[A]); lawnmower (90 dB[A]); pneumatic drill (100 dB[A]); auto horn (115 dB[A]); jet engine (140 dB[A]); and rocket pad during launch (180 dB[A]). Consistent exposure to noise above 85 dB(A) causes an excess risk of hearing impairment[4] and every 3 dB increase in intensity halves the allowable exposure time. NIHL is 100% preventable and the most effective ways to avert NIHL are by avoiding or reducing hazardous noise. Although, NIHL is well recognized as an occupational hazard, less recognized perhaps, are non-occupational leisure activities or environments that can put people of all ages at risk for hazardous noise exposure.[5,6,7,8] In fact, college students are exposed regularly to noisy situations that exceed NIOSH recommended sound intensity limits.[9] For example: Music concerts, 92–110 dB(A);[10] nightclubs, 96–110 dB(A);[11] sporting events, 100–140 dB(A);[12,13] and use of portable listening devices, 90 -120 dB(A).[14,15] In addition, various studies have shown that young adult survey respondents (age 19–27 years) report frequent symptoms of NIHL including temporary hearing threshold shift (61%), tinnitus (43%), and/or ear pain (28%) after participating in leisure activities associated with hazardous noise levels.[16,17] Noise & Health, Noise July-August & Health, July-August 2011, Volume 2011, 13:53,261-71 Volume 13

Kotowski, et al.: College student hearing loss

Wearing hearing protection, turning down the volume, and use of headphones with personal listening devices have all been found to help reduce the risk of NIHL, but young adults typically do not choose to heed warnings to protect their hearing.[3,18] Some possible reasons for this include discomfort, design, lack of knowledge about NIHL, and peer pressure.[10,18] It has been proposed, however, that young adults would be more likely to take precautions if they understood that they were personally vulnerable to permanent NIHL.[9] It is imperative, therefore, to determine if prevention programs targeting young people can effectively convey the risk of NIHL and change their behavior to help reduce risk of NIHL.[8] The purpose of this study is to determine the effectiveness of a specific message model, the Extended Parallel Process Model (EPPM)[19] in informing college students of the risks of NIHL associated with MP3 player usage and how use of headphones instead of earbud transducers can lower the risk. The focus of this research endeavor on headphone use with MP3 players (instead of lowering the volume) was used for the following reasons. First, students entering college in the USA are very likely to have been exposed to NIHL prevention messages that focus on reducing volume. An example of two recent national campaigns targeting young people to lower the volume while using portable listening devices include: “Turn It to the Left,” a campaign promoted by the American Academy of Audiology;[20] and “Listen to Your Buds,” a campaign promoted by the American Speech Language Hearing Association.[21] Second, incoming college students are unlikely to have been exposed to information detailing the difference in level between headphones and earbuds (~3 – 4 dB SPL less for headphones),[15] which, according to NIOSH standards, translates into a doubling of exposure time without increasing risk. In other words, by switching to headphones a student could listen to music for twice as long as when using earbuds. Finally, by selecting a message that teens are unlikely to have been exposed to while in primary school, the effect of a specific communication model or theory on health risk knowledge and intentions could be more accurately measured in this study.

Using an approach based in communication theory to achieve the goal of effectively informing young people of the threat they face for NIHL and then outlining effective ways to avert that threat seems particularly useful. This study used the EPPM[19] to design and evaluate the creation of brochures targeting college students that highlight threats associated with exposure to extreme noise and the efficacy with which hearing protection can be used. The current study represents the first step towards the goal of reducing risk of NIHL in undergraduate college students (typically 18 to 21 years of age) and simultaneously tests the ability of the EPPM to guide effective message development in this population. This goal is particularly important given that frequent or prolonged exposure to hazardous noise levels leads to permanent NIHL (and often constant tinnitus) the consequence of which can include decreased enjoyment and quality of life, depression, lower academic performance, and reduced productivity.[3,22,23] Furthermore, NIHL has implications for the communication abilities of those suffering from the ailment. Specifically, the impact of NIHL on communication includes: Difficulty hearing speech from a distance, in noisy and reverberant environments, and when it is softly spoken.[24,25] These symptoms are particularly problematic to college students who are expected to listen and learn in complex acoustic environments.

The Extended Parallel Process Model The EPPM is a model of persuasive communication that posits a mechanism responsible for the effectiveness of a particular class of persuasive messages referred to as fear [Figure 1].[26] Fear appeals should present a relevant threat and then show a relatively simple response that will avert the threat, when enacted. To the extent that this combined presentation of threat and action instigates protective motivation, recipients reduce the threat in the way advocated by the appeal.[19,27] The EPPM has been successfully applied in a variety of public health contexts such as: Health promotion;[28] breast self exams;[29] alcohol warnings;[30] firearms;[31] drinking water quality;[32] cardiovascular disease;[33] tobacco use;[34] burn prevention;[35] and road safety.[36]

External stimuli

Message processing (1st and 2nd appeals)

Message components

Perceived efficacy (Self-efficacy, Response efficacy)

Self-efficacy response efficacy susceptibility severity

Perceived threat (Susceptibility, severity)

Outcomes Protection motivation

Process

Message Acceptance

Danger control process

Fear Feedback loop

No threat perceived (No response)

Defensive motivation

Message rejection

Fear control process

Individual differences Adapted from: Witte, K. (1994). Fear control and danger control: A test of the extended parallel proces model (EPPM). Communication Monographs,61,113-134.

Figure 1: The Extended Parallel Process Model[26, 27] Noise & Health, July-August 2011, Volume 13

262

Kotowski, et al.: College student hearing loss

The first step in applying the EPPM is to present a threat of a hazard (a message) to the target population, in this case college students. According to the EPPM, perceptions of threat require that the young person receiving the message believe that the health hazard is severe and that they are susceptible to injury or harm. When applying the EPPM to leisure NIHL, the college student, for example, must perceive that hearing loss is a real threat to their well-being and that they are susceptible to hearing loss. If a threat is considered to be trivial or irrelevant, the model predicts that the student will not be motivated to process and accept the message and no action will be taken to prevent injury.[19]

consequences, and fail to use hearing protection while engaged in noisy leisure activities.

Once a threat is established, the next step is to induce perceived efficacy, i.e. the perception that the college studnet can effectively avert the hazard by performing a simple task. According to the EPPM, efficacy is composed of two elements: Perceived self-efficacy and response efficacy. Perceived self-efficacy refers to a student’s perception of his/her ability to perform the recommended response to avoid harm.[27] For college students, selfefficacy involves the perception that hearing protection devices (HPDs) are easy to use while engaging in leisure activities or headphone-style transducers are easy to use while listening to an MP3 player. Response efficacy refers to the perception that the recommended action to avert the threat (or injury) really will protect a college student from harm. Response efficacy refers to the perception that HPDs will adequately protect against future hearing loss due to hazardous leisure noise exposure and will protect against the negative personal and financial consequences that hearing loss can bring.

Hypothesis 2: College students receiving messages describing the efficacy of the recommended response and their selfefficacy of enacting the response will rate the efficacy of using hearing protection higher than those who do not receive the messages.

According to the EPPM, an effective prevention message is created with a balance between threat and efficacy. If both perceived threat and efficacy are high (i.e., real danger is recognized but perceived to be easily and effectively avoided), the EPPM predicts that college students will engage in a danger-control process. The danger-control process is desired because it motivates a young person to employ protection behaviors, which will prevent the negative outcomes associated with health hazard. In the application of leisure NIHL, if a danger-control process is engaged the college student would use hearing protection because he/she perceives it to be an effective way to avoid a real hazard with undesirable, negative consequences. When the perceived threat far outweighs the perceived efficacy, the EPPM predicts that people will engage in a fear-control process. In the fear-control process, college students engage in denial, defensive avoidance, reactance, and ultimately reject the message in order to escape their fear Witte.[27] In relation to leisure NIHL, if the risk of NIHL is exaggerated or the steps to avoid injury too onerous, a fear-control process could be engaged and the young person would deny the severity of the noise hazard, minimize the 263

Therefore, because the brochures developed for this study were designed to include the EPPM components of severity, susceptibility, response efficacy, and self-efficacy messages the following hypotheses are offered: Hypothesis 1: College students receiving messages describing the severity of NIHL and their susceptibility to NIHL will rate the threat of NIHL higher than those who do not receive the messages.

Hypothesis 3: College students who receive messages describing the threat of NIHL and the efficacy of the recommended response will have higher intent to use hearing protection than will those who do not receive the messages.

Methods Subjects Undergraduates at a large southeastern university were convenience sampled at several different campus events, at several different points in time over the course of a semester until a sample of 200 subjects was obtained. No more than 40 subjects were grouped at any one event, and subjects were only allowed to participate in the study once. This sample was reduced to 176 after removing subjects with a self-reported history of hearing loss. This plan resulted in a sample that comprised 45% males and 55% females, where subjects were on average, Mean (M)=21.02 (Standard Deviation [SD]=3.62) years old and in their M=2.88 (SD=1.60) year in school. Design and procedures This study employed a randomized two group post-test only design where after selection for inclusion in the study, subjects were assigned randomly to either a control condition or treatment condition. In the control condition, subjects responded to a questionnaire designed to measure the EPPM variables of severity, susceptibility, response efficacy, self-efficacy, and behavioral intentions, as well as several demographic variables. After completing the questionnaire, these subjects received the brochure so as not to withhold any benefits of the brochure from the control condition. Subjects assigned to the treatment condition were presented with the same brochure first. After being given a few minutes to review the brochure, subjects responded to the same questionnaire that was administered to subjects in the control Noise & Health, July-August 2011, Volume 13

Kotowski, et al.: College student hearing loss

condition. In both conditions, subjects were instructed to ask the researcher if they had questions or needed clarification while completing the measure. None of the subjects needed any clarification. Brochure and measurement Brochure The brochure was designed specifically for a target population of college students focussing on two different types of behavior. Specifically, the two qualitatively different activities of portable audio device use and student leisure activity engagement were considered. While the former is an individual behavior, the latter is a collective behavior. Consequently, people’s motivations to engage in behaviors to protect hearing may be very different across the two behaviors resulting in the possibility of observing a differential pattern of results. Thus, each page of the brochure contained messages developed to make salient the threat college students face for NIHL and increase the perceived efficacy of behaviors that can avert the threat. To increase perceptions of threat, the first through third pages of the brochure presented students with messages hypothesized to increase perceptions of NIHL severity and susceptibility (e.g., “Hearing loss is permanent and irreversible!” and “College students are at high risk for hearing loss!”). These first three pages also summarize NIOSH[4] and Occupational Safety and Health Administration[37] standards for noise level exposure in addition to images of college student leisure activities (e.g., football games and nightclubs) and portable listening devices (e.g., CD and MP3 players) along with corresponding decibel levels. The rear two pages of the brochure presented additional severity messages such as a listing of the side effects of NIHL (e.g., “Difficulty hearing conversations in groups.” and “The need to use hearing aids.”). These latter pages of the brochure, however, were dedicated primarily to messages designed to increase perceptions of response efficacy and self-efficacy by emphasizing the ease of preventing NIHL (e.g., “Hearing loss is one of the easiest hazards to prevent against.”) and recommending behaviors that can avert the threat of NIHL (e.g., “You can begin to save your hearing loss today even if you have experienced some hearing loss by using formable foam plugs when in loud environments or using over-theear headphones when listening to your MP3 player”). The messages contained in the brochure were collaboratively designed by an EPPM expert and an audiologist with NIHL expertise. The NIHL data contained in the brochure was provided by an audiologist with expertise in the area. A copy of the brochure can be found in Figure 2. Measurement The questionnaire was a combination of several demographic items and items designed to measure the EPPM variables of threat (i.e., severity and susceptibility), efficacy (i.e., response Noise & Health, July-August 2011, Volume 13

efficacy, and self-efficacy), and behavioral intentions adapted from the risk diagnosis behavior scale. [38] The demographic variables consisted of the subject’s age, sex, year in school, whether they have a history of hearing loss, and how frequently they are exposed to environmental risk factors for NIHL. The EPPM variables were measured using five-point Likert-type scales ranging from “disagree strongly” to “agree strongly” where higher scores corresponded to greater item endorsement. Using this scaling format, the EPPM portion of the questionnaire was designed to measure the severity of NIHL, the susceptibility to NIHL, the response efficacy and self-efficacy of using formable ear plugs when in loud environments, the response efficacy and self-efficacy of using over-the-ear headphones when using MP3 players, and behavioral intentions to enact the recommended behaviors (i.e., using formable ear plugs when in loud environments and using over-the-ear headphones when using MP3 players). The Flesch-Kincaid grade level of the measure was 7.5. A copy of the measure can be found in Appendix A.

Results Measurement models – Validity and reliability Loud environments As specified by the EPPM, efficacy was hypothesized to be a second-order unidimensional construct composed of the constructs of response efficacy (RE) and selfefficacy (SE). Threat was also hypothesized to be secondorder unidimensional and composed of the constructs of susceptibility (SUS) and severity (SEV). RE, SE, SUS, and SEV were each measured by a set of five items. Only responses to the RE and SE items for wearing formable earplugs in loud environments were considered for these analyses. Analyses of responses to the RE and SE items for wearing over-the-ear headphones when listening to MP3 players were conducted separately. After examining the response distributions of each item for normality, the 20 item inter-item correlation matrix was examined for items lacking internal consistency and parallelism.[39] These initial analyses resulted in eliminating numerous items, either because they lacked either internal consistency, parallelism, or both. The results of these analyses yielded four potentially content valid indicators of RE, four of SE, four of SEV, and five of SUS. The AMOS 16.0 Confirmatory Factor Analysis (CFA) algorithm using a maximum likelihood parameter estimation method was employed to test the hypothesized measurement model. The analysis revealed that the model was consistent with the data. Examination of the obtained parameter estimates presented in Figure 3 reveal that they are all ample. Furthermore, examination of the modification indices suggested there were no localized areas of substantial model misfit. Global model fit was also acceptable: χ2 (114)=197.58, Goodness of Fit Index (GFI)=0.89, Root Mean Residual (RMR)=0.05. All of this is convincing evidence of this measure’s validity. 264

Kotowski, et al.: College student hearing loss

Figure 2: Two-sided tri-fold brochure for college students. The panels are in order of pages 5, 6, 1, 2, 3, and 4 in order to accommodate a tri-fold order

Given the CFA results and considering the face validity of the items, two indices were created. An efficacy index was created by averaging (arithmetic mean) across responses to the four RE items and four SE items, whereas a threat index was created by 265

averaging across the five SUS items and four SEV items. The distribution of the efficacy index was approximately normal: α=0.83, M=3.98, SD=0.61. The distribution of the threat index was also approximately normal: α=0.83, M=3.97, SD=0.58. Noise & Health, July-August 2011, Volume 13

Kotowski, et al.: College student hearing loss

MP3 players Only responses by subjects who self-reported as owning an MP3 player (N=137) and only responses to the RE and SE items for wearing over-the-ear headphones when listening to MP3 players were considered in these analyses. As was the case in the loud environment analysis, efficacy was hypothesized to be a second-order unidimensional construct composed of the constructs of RE and SE and threat was hypothesized to be second-order unidimensional and composed of the constructs of SUS and SEV for the MP3 player analysis. RE, SE, SUS, and SEV were each measured by a set of five items. Using the same procedures of initial analysis as applied to the loud environment data, the same four potentially content valid indicators of RE, SE, and SEV and five potentially content valid indicators of SUS emerged in the MP3 data. CFA also using the same procedures as applied earlier was performed to test the hypothesized measurement model with the MP3 data. The analysis revealed that the model was consistent with the data. The obtained parameter estimates presented in Figure 4 are all ample, and the modification indices suggested there were no localized areas of substantial model misfit. Again, global model fit was acceptable: χ2 (114)=225.21, GFI=0.85, RMR=0.06. Again, this evidence is indicative of the measure’s validity. Based on the evidence from the CFA, efficacy and threat indices were created for the MP3 data. The efficacy index was created by averaging across responses to the four RE items and four SE items, whereas a threat index was created by averaging across the five SUS items and four SEV items. The efficacy index was distributed approximately normal: α=0.92, M=3.42, SD=0.85. The distribution of the threat index was also approximately normal: α=0.86, M=3.91, SD=0.62.

intent to wear ear plugs in loud environments are presented in Table 1. Hypothesis one predicted that college students who received severity and susceptibility of hearing loss messages would rate the threat of hearing loss higher than students who did not receive the messages. The results of a multiple regression analysis showed that when threat was regressed onto experimental condition, current use of hearing protection, and weekly hours spent in loud environments, as well as all two- and three-way interaction terms, only the experimental condition emerged as a substantial and statistically significant predictor, β=0.26 (t (168)=3.40, P