66.4, November 2019

Expository Warnings in Public Recreation and Tourism Spaces

By Emil B. Towner


Purpose: Every day, people are exposed to risks they either do not know exist or do not understand how to avoid. This is especially true in locations where the public may not expect or understand the danger around them—such as amusement parks, national parks, or unfamiliar vacation spots. Although a number of factors influence the effectiveness of a warning, this study focuses specifically on the use of expository versus abridged language in warning signs.

Method: This research was conducted in two phases. First, I distributed an online survey that was completed by 303 participants who were randomly assigned to view one of two warning sign options and then were asked to describe the sign’s meaning. Second, I conducted a content analysis of the survey responses using coding terms (related to Bloom’s Taxonomy) to assess the levels of understanding that resulted from the two warning sign versions.

Results: My findings indicate that people who viewed a longer, more detailed expository warning from a national park were more likely to understand the risks and consequences as well as how to apply the information to new situations or specific actions. In other words, they demonstrated a higher level of understanding risks and risk avoidance.

Conclusions: In public recreation and tourism spaces, detailed expository warning signs may be more effective at helping the public understand unfamiliar risks and how to avoid them. However, reading times (max. 30 seconds) and contexts (near resting/standing areas) should be considered.

Keywords: safety signage, risk communication, public safety, expository warnings

Practitioner’s Takeaway:

  • Longer, more detailed expository warnings (especially in recreation spaces where people frequently gather or expect signs to be posted) may support better understanding of risks and risk avoidance than abridged warnings (such as “No Swimming”).
  • People responsible for public safety in outdoor recreation and tourism spaces should use the findings to identify inadequate warnings that are currently in place (before an injury occurs) and develop more effective signs.
  • People responsible for safety in non-recreation locations, such as dangerous workplaces, should also consider testing expository warnings in breakrooms or changing areas where employees rest or have time to read.


Every day, people are exposed to risks they either do not know exist or do not understand how to avoid. In some cases, the public may anticipate (though not completely understand) that risks exist when entering dangerous locations (such as factories, construction sites, etc.) or engaging in risky behavior (such as smoking, drinking and driving, texting and driving, etc.). However, in other locations—such as amusement parks, national parks, or unfamiliar vacation spots—the public either may not expect or is not focused on the potential risks that surround them. In fact, the opposite may often be the case. Instead of expecting hazards and considering risks, the public may think of these locations as controlled environments containing fun, safe activities. As Espiner (1999) explained, tourists have a “sense of security based on the controlled, predictable and urban communities in which most people now live”; however, the natural public recreation areas they visit “are often not entirely controlled or predictable—or, at least, they have the potential to become inhospitable or dangerous” (p. 12). In some instances, people may simply be unaware of the risks that exist in such places (Whittlesey, 2014). In other instances, people may be aware that some level of risk exists, but they may not believe they have any control over the risks in vacation or recreation locations, may assume the responsibility for risk mitigation belongs to someone else, or may be unaware of specific actions they should take to avoid the risks (Espiner, 1999; Jackson et al., 1996; Njome et al., 2010; Wachinger et al., 2013).

Take, for example, the fatal attack of a two-year-old child by an alligator at Disney’s Florida resort in 2016. That tragedy highlighted the inadequacy of the warning sign used by Disney, which merely told visitors not to swim in the water. That warning sign was problematic in two ways. First, the sign used the term swimming but did not define swimming as wading or even walking a few inches into the water. The official report of the tragedy indicated that the child’s parents were near the two-year-old victim while he was standing on the shoreline with his feet “ankle deep or less in the water,” as he bent over to fill his bucket with water to make a sandcastle (Florida Fish and Wildlife, 2016, p. 4). It is unlikely that the child’s parents (or an average person for that matter) would have characterized that act as swimming. Second, the sign failed to inform the parents (and other park visitors) about either the presence of dangerous wildlife or about the severity of the threat posed by such wildlife. Yet, the official report of the tragedy explicitly stated that “Alligators routinely hunt and kill prey along the shore and drag it into the water” (Florida Fish and Wildlife, 2016, p. 12). Unfortunately, that information was not communicated to the parents via the warning sign. Instead, the sign merely commanded park visitors not to perform an action (i.e., “no swimming”) without explaining why or what potential harm may result. Legal experts summed up why the missing information would have been useful to park visitors: “Although Floridians know that alligators live in nearly every freshwater body in the state…it would be reasonable to argue that vacationers coming from somewhere like Nebraska wouldn’t share the same knowledge” (Mettler & Sharma, 2016, para. 3).

As the Disney resort tragedy highlights, not all warning signs effectively “alert and inform” the public about “potential hazards, so people can act to avoid the consequences” (Ayres, 2013, p. 1698). Obviously, a number of factors can influence the effectiveness of a warning, including the placement, color, shape, appearance of a border around the warning, size and font of the text, interactivity, and so on (Rogers, Lamson, & Rousseau, 2000). This study, however, is focused on the language or verbiage used in warning signs. In contrast to many of the previous studies on warning language that are discussed in the literature review below (which studied the warning language used on product packaging), I explore warnings used in tourism and recreation spaces that are unfamiliar to the average visitor and, therefore, may pose risks people could not be expected to anticipate.

Like the Disney resort, Yellowstone National Park provides another example of a location where the public is exposed to risks they may not be accustomed to or fully understand. Each year, people are exposed to (and, at times, disregard) safety warnings regarding hazards, including thermal water pools, geysers, and wildlife. For example, nearly every year, park visitors are attacked by bison. In fact, bison injure more park visitors than any other animal in Yellowstone. According to one study, bison charged at park visitors 81 times over a 22-year period, and, back in 1987, more than 40 people were injured by bison in Yellowstone (Miller, 2015). In many cases, reports of the attacks indicate that the victims were told not to approach wildlife; however, having been to Yellowstone and having witnessed the abundance of wildlife in close proximity to park visitors, I question whether the public understands the warnings and risks well enough to choose safe actions. I do not mean to imply that the National Park Service or its employees are misleading visitors or overlooking safety obligations. After all, safety information from flyers to signs to verbal warnings from park rangers are present throughout Yellowstone on a wide range of hazards including wildlife and thermal attractions. In some cases, people visiting recreation spaces simply choose to ignore the warnings or deny they’re at risk (Albers, 2012) and, in doing so, knowingly place themselves in dangerous situations. In other cases, people may see a warning but not completely understand the risk or how to mitigate it. For instance, one of Yellowstone’s warning signs features the silhouette of a bison with text below it that reads “DANGER” followed by “DO NOT APPROACH WILDLIFE” (in all caps), but the sign does not provide information about a bison’s speed or the severity of injury that is likely (see Figure 1). This warning sign exemplifies what I call an abridged warning sign. I use the term abridged because such warnings present shortened versions of risk and risk avoidance information while maintaining the general sentiment (e.g., what not to do); however, they are typically vague and lack explicit details regarding the exact nature of the danger or steps to avoid it.

Figure 1. An abridged warning sign conveys the general sentiment of a warning in a shortened version that lacks details or specificity. The original version of this sign was posted at Yellowstone National Park by the National Park Service.

Contrast that sign to another warning sign that is also displayed at Yellowstone, which I refer to as an expository warning sign (see Figure 2). I use the term expository because such warning signs use a longer form to explain risk and risk mitigation with facts and figures (rather than opinions) written in a logical order. For example, the warning sign in Figure 2 explicitly visualizes the danger of bison attacks by showing an illustration of a person being tossed into the air by a bison. This sign also provides specific information about bison size, speed, and unpredictability.

Figure 2. A longer expository warning provides facts and details in a logical sequence. The original version of this sign was posted at Yellowstone National Park by the National Park Service.

These two signs provide different approaches to warning the public about the hazards of bison—including visual differences (colors, shapes, types of illustrations) and textual differences (length, tone, type of information). The signs are not posted in the same location but, instead, are distributed separately throughout the park in areas where people commonly congregate and where bison are known to frequently wander. The abridged warning is often seen on sign posts along boardwalks or near open fields, while the expository warning can be seen attached to sign posts along boardwalks and tacked to bulletin boards. Due to the vast size of the park and innumerable directions that visitors may undertake as they navigate the roads, trails, and tourist spots within the park, the signs are not posted in a specific sequence. That said, the park does provide other safety materials throughout the park in the form of handouts, displays, and so on.

Those safety materials (including warnings signs) deliver risk communication messages to people who visit Yellowstone. Risk communication is a multidisciplinary field with technical communication scholars and practitioners focusing on “predisaster, postdisaster, and prevention and participation discourse related to accidents and crises” (Youngblood, 2012, p. 40). At its most basic definition, risk communication can be defined as “the potential to lose something of value” (Espiner, 1999; Priest & Baillie, 1987). More nuanced definitions include aspects such as personal choice and uncertainty (Roehl & Fesenmaier, 1992) as well as cultural practices and social norms (Douglas & Wildavsky, 1982). Often, risk communication follows a “technocratic” approach that views “risk as determined by experts” and delivered to an audience in “linear (one-way)” communication that attempts to “educate/influence the public to think about risk the way the experts do” (Grabill & Simmons, 1998, pp. 421–422). More recently, however, technical communication scholars have argued against the linear, expert-driven (even male-dominated assuming) approach, opting instead for approaches that acknowledge different perspectives, experience-based knowledge, and different types of uncertainty in risk assessment and communication (Grabill & Simmons, 1998; Sauer, 1992; Sauer, 2003). Uncertainty is a critical factor in risk communication and is often thought of as “a lack or distortion of crucial technical information” (Walsh & Walker, 2016, p. 71). However, a broader view of uncertainty (based on spheres of arguments) distinguishes between technical uncertainties, personal uncertainties, and political or public uncertainties (Walsh & Walker, 2016). Such distinctions enable technical communicators to better understand and predict how uncertainties “travel from their home sphere to another” (Walsh & Walker, 2016, p. 83), which ultimately shapes risk assessment and communication.

In addition to the aspects of risk assessment and uncertainty, previous research on warnings has uncovered a number of factors that influence the effectiveness of a warning, including the use of images, colors, borders around warnings, sizes and fonts of the text, placement/location, interactivity, and so on. This study, however, focused specifically on the language (length, specificity, etc.) used in warning signs. Therefore, before discussing the methods, findings, and implications of this study in more detail, I begin with a brief literature review of previous research related to warning-sign language.

Literature Review of Warning-Sign Language

For the purposes of this study, I have grouped the major findings from previous research on warning-sign language into five key characteristics that influence effectiveness: signal words, length, complexity, explicitness, and hazard information.

Signal Words

One way to increase the chance that a warning is noticed and taken seriously is to include a signal word—such as “danger,” “warning,” or “caution.” In fact, Young et al. (1995) conducted two experiments with nearly 200 participants to test the importance of four warning sign elements: the signal word, a statement about the hazard, consequences of non-compliance, and instructions for mitigating the hazard. Results from both experiments indicated that participants viewed the signal word as the most important aspect included in a warning. That said, past research conflicts on which word is most effective. The American National Standards Institute (ANSI) provides some guidance on the use of signal words (and how they relate to both the likelihood and severity of risk, as indicated in italics below):

  • DANGER: The hazard will result in death or serious injury
  • WARNING: The hazard could result in death or serious injury
  • CAUTION: The hazard could result in minor or moderate injury

Unfortunately, people (especially people unaccustomed to working in hazardous environments) may not understand that each term relates to a specific risk level. In that sense, the perceptions of people are more relevant to this study than the actual ANSI definitions. In terms of public perceptions, some studies have found that people perceive the word “danger” to indicate a more serious threat than other words, such as “warning,” “caution,” or “notice” (Braun & Silver, 1995; Wogalter & Silver, 1990). Other studies, however, found the words “danger,” “warning,” and “caution” to be similar in people’s perceptions of severity (Leonard, Matthews, & Karnes, 1986; Wogalter et al., 1994). Despite those different findings, the general consensus is warnings that include a signal word are perceived as describing more of a threat than warnings without a signal word (Rogers et al., 2000). For example, Wogalter et al. (1994) found that when signal words were included in warnings, people perceived the level of risk or danger as being higher than when the same warning was shown without a signal word. However, Wogalter et al. (1994) did not find any statistical difference between the words “danger,” “warning,” or “caution,” which indicates that all three terms are at least more effective at conveying risk than no signal word at all.


The length of text also impacts the public’s perception of risk. According to past research, the longer the text, the more likely the public is to read it (Leonard et al., 1989; Silver et al., 1991). Rogers et al. (2000) noted that this may be due to the public’s belief that “the greater the length of text, the more hazardous the product must be, and therefore it should be read” (p. 118). At first glance, the notion that longer text may be more effective than shorter text seems to conflict with previous research. For example, Wogalter et al. (1987) found that when people’s time or the amount of space available for text is limited, “people may be more likely to read a short, concise message . . . than a long, wordy one” (p. 611). Similarly, Velotta (1987) suggested that technical writers “say as much as possible in as few words as possible” when writing safety information (p. 123). More important than the number of words or length of text, however, is the amount of time required to read a warning sign. Previous research indicates that visitors to natural recreation areas (such as wilderness trails) spend approximately 25 seconds (Cole et al., 1997) to 30 seconds (Hughes & Morrison-Saunders, 2002) reading signs that are posted near visitor centers or on bulletin boards. When reading signs at trail heads or along the actual wilderness trail, people spend approximately 10 to 15 seconds (Hughes & Morrison-Saunders, 2002). In other words, people responsible for public safety and risk communication should consider the context when determining the appropriate length of warning-sign text. In areas where visitors are more likely to rest or to expect signs to be posted, warnings should include only enough text to be read within the 25 to 30 seconds (Cole et al., 1997; Hughes & Morrison-Saunders, 2002). Conversely, in areas where people are less likely to read signs (i.e., areas in which they are constantly moving or are focusing their attention on other aspects, such as the scenery or their footing on a steep trail), warnings should include only enough text to be read within 10 to 15 seconds (Hughes & Morrison-Saunders, 2002).


Effective risk communication must convey “unambiguous, definitive and easily interpreted” information (Breakwell, 2000, p. 116). According to past research, the more explicit (i.e., clear and detailed) the warning, the more likely people are to comprehend risk and risk mitigation (Laughery et al., 1991; Laughery et al., 1993; Laughery & Stanush, 1989; Rhoades et al., 1990; Trommelen, 1997; Velotta, 1987). For example, an explicit warning for sleep medication would state that the patient must get eight hours of sleep (rather than “adequate sleep”) between taking the medication and operating a car or equipment. Clear, easily interpreted information may be connected to the use of concrete (instead of abstract) language. In simple terms, concrete language refers to objects that can be experienced by the senses—that is, seen, touched, heard, etc. Abstract language, on the other hand, refers to intangible ideas, actions, and concepts. From a psychological perspective, concrete words “have single, bounded, identifiable referents,” while abstract words “lack bounded and clearly perceivable referents, even if they might evoke situations, scenes, introspection and emotional experiences” (Borghi et al., 2017, p. 263). As a result, the meanings of abstract words can change over time or can be influenced by personal life experiences (Barsalou, 1987). In terms of communication, researchers have known for decades that concrete words are more memorable than abstract words (Brener, 1940; Epstein, 1962; Paivio, 1963). That tendency also extends to complete sentences, with people comprehending concrete sentences faster and storing them more completely in their memory than abstract sentences (Holmes & Langford, 1976). According to Paivio, Yuille, and Madigan (1968), concrete words also tend to correlate with higher ratings in imageability. In fact, imageability—which “is a psycholinguistic variable that is used to indicate how well a word gives rise to a mental image or sensory experience”—is so closely aligned with concreteness that numerous scholars have mistakenly used the two terms interchangeably (Rofes et al., 2017, p. 1). In reality, imageability is more like a scale, with some concrete words rating higher in imageability than others. Overall, however, concrete words are higher in imageability, and abstract words are lower in imageability (Borghi et al., 2017). This concept is relevant to warning signs because words with higher imageability ratings are believed to be processed faster and more accurately (Rofes et al., 2017).

Another way to make warnings more explicit—and, as a result, to at least marginally increase comprehension and compliance—is to use similes or metaphors in warnings (Bowles et al., 2002). When a warning includes information or details that are unfamiliar to the public, the information can be compared to details that the public already understands, which can increase both comprehension and recall (Harris & Mosier, 1999; Johnson, 1996). For example, a warning that states a “cell phone can initiate the combustion of some flammable gasses” or a warning that body parts “might be forcefully drawn into” a machine may be more effectively written as similes or metaphors, such as “This cell phone is like a match and causes flammable vapors to explode” or “the shredding device is a vacuum that might suck in body parts during operation” (Bowles et al., 2002, p. 1704).


Similar to explicitness, the complexity of the text influences how easily a warning can be read and understood. Complexity includes aspects such as voice (passive or active), syntax, and vocabulary (Leonard, Creel, & Karnes, 1991; Main et al., 1993), as well as the positive or negative tone of the warning-sign language (Harris & Wiklund, 1989). For example, Velotta (1987) argued that warning messages should be written in an active voice with “strong auxiliary verbs” to “ensure that the message comes across forcefully” (p. 124). Moreover, threatening or even morbid terms (e.g., death or dismember) tend to attract more attention and to be more memorable (Harris & Wiklund, 1989). That said, threatening or morbid signs may be less likely to be posted by business owners who may fear their customers will be scared away (Harris & Wiklund, 1989). Based on that, it may not be realistic to expect private businesses to display signs with morbid language; however, such signs could still be useful in some tourism and public recreation spaces where serious risks exist. I argue that the number of separate or different messages that are communicated should also be factored into the complexity of a warning sign. For example, Cole et al. (1997) found that people can experience information overload when reading a sign with more than two messages. As a result, “attention per message and retention of message content both decline” (Cole et al., 1997, p. 69). Therefore, practitioners who are responsible for safety and risk communication in recreational spaces should select one or two important messages that need to be conveyed for the given context or environment.

Hazard Information

A final characteristic of an effective warning is hazard information. This information can include multiple components. First, the warning should communicate the “severity” of consequences that may result if people do not comply with the warning (Wogalter et al., 1987; Wogalter & Barlow, 1990). For example, warnings related to viewing a solar eclipse should state whether staring directly at the sun would result in “temporary vision problems” or in “permanent eye damage and even blindness.” Second, the warning should indicate the likelihood that a person will experience the hazard. The likelihood may be communicated explicitly (such as stating the number or percentage of victims) or implicitly (such as writing “will be injured” or “may experience injury”). Third, the hazard information should provide specific actions (as opposed to vague behavioral commands) that would help visitors avoid or mitigate the hazard. For example, Frantz (1994) found that the phrase “wear rubber gloves and protective glasses” was more effective than a warning that merely told people to “avoid contact with eyes and skin.” That type of statement can help overcome situations in which people are unaware of specific actions they can take to help ensure their own safety. As Wachinger et al. (2013) argued, “it is not just a matter of raising risk perception but also of providing individuals with the physical and mental capacity to affect their own situation” (p. 1059). Fourth, I argue that “motivating” information should be included in a warning (Cornelissen, van Hoof, & van Vuuren, 2014). As Geller (2003, as cited in Saleh, 2011, p. 5) stated, when people understand why they should follow rules, they’re more likely to be personally motivated. Including such information can also help ensure warnings are not overly technical (Herrero et al., 2002) or too focused on the organization’s interests (Elling, 1997).

One caveat to hazard information (and warnings in general), however, is that people with alternative prior experiences (that is, people who have experienced or witnessed the hazard but not the negative consequences) may not believe the likelihood or severity of the risk—or even notice the warning in the first place—regardless of the warning’s language (Godfrey & Laughery, 1984; Goldhaber & deTurck, 1988).

Based on the characteristics above, I argue that expository warnings are distinct from abridged warnings because they include most (but not necessarily all) of the aspects listed in Figure 3.

Figure 3. A checklist of common characteristics in expository warnings

Although those characteristics provide a baseline for creating and even critiquing warning-sign language, this study extends previous research by focusing on the effectiveness of differing approaches to warning-sign language and how they impact the public’s actual understanding of risks and application of the information, especially in recreational areas where people may unknowingly be exposed to risks.


Specifically, this study tested the hypothesis that expository-warning language would result in the public being more informed about the risks and risk mitigation than abridged-warning language. Based on this hypothesis, I chose to conduct a large-scale, online survey using an open-ended question. Although other user experience research methods, such as usability or observation, might have been used to test the public’s ability to perform tasks or study their interaction with warning signs, the goal of this study was to gauge people’s perceptions and understanding of information conveyed by different versions of warning signs. Therefore, a survey was chosen because it provided one of the best methods for investigating user perceptions and for statistically describing segments of the population (Kuniavsky, 2012). In addition, aside from the demographic data questions at the beginning of the survey, this study focused on a single open-ended question: “In the space below, please describe in your own words what the sign above means to you.” This open-ended question was chosen because it allowed “respondents to express their opinions in their own words,” had the potential to “reveal unanticipated responses,” and helped reduce researcher bias that could have influenced people to “answer toward an expected response or outcome” (Hughes & Hayhoe, 2008, p. 97). In that sense, this study is aligned with Grabill and Simmons’ (1998) call for more qualitative research, such as surveys and interviews that include users in the construction of knowledge, as well as with Espiner and Weiss’ (2010) use of surveys to study the public’s interpretation of messages contained in signs posted in a national park. That said, the use of an online survey resulted in removing the participants from the original context of use—i.e., the national park setting with wide-open landscapes and potential distractions. That limitation will be discussed in the closing section of this paper.

To test this study’s hypothesis, two current Yellowstone National Park warning signs were tested. Yellowstone National Park’s warning signs about bison provide a real example of two different approaches (one being an abridged warning and the other being an expository warning) that are currently being used at the same time and in the same national park. They also provide an opportunity to research warnings aimed at helping people understand and avoid hazards in tourism and outdoor recreation spaces where danger and risk may not be at the forefront of the experience. Because numerous factors influence a warning’s effectiveness, aspects such as different images, colors, and so on were removed from the signs that participants were shown in order to test only the language. Additionally, the warning signs included in this study were intentionally designed using a similar, generic design—which helped focus the participant’s perceptions and responses on the actual warning language (rather than font differences, etc.). Moreover, the warning-sign language that was tested was pulled verbatim from the actual warnings that are currently posted throughout Yellowstone National Park. While that enabled me to test actual warning language currently used in a tourism and outdoor recreation space, it did limit the types of signs tested to two extremes (a purely abridged versus a purely expository). That limitation will be discussed in the closing section of this paper. In the paragraphs below, I explain the two phases of the research—an online survey to collect data, followed by content analysis of the survey responses—in more detail.

Online Survey

I administered an IRB-approved survey online from March 2016 through November 2016. Participation in the survey was completely voluntary and no identifiable information was collected. I recruited participants in two stages. First, I recruited students from college courses at a Midwestern university. Second, I recruited non-university participants via snowball effect by distributing the survey link to online connections (such as co-workers, family members, friends, etc.) and encouraging them to distribute the link to their contacts.

The online survey resulted in 303 completed responses. Approximately 55% were female and 45% were male. Although respondents ranged from 18 to 60+ years old, 64% were 18 to 29 years old at the time the survey was conducted. Finally, respondent education levels ranged from “some high school” to “graduate degree” (see Figure 4).

Figure 4. Demographic profile of survey respondents

Multiple-choice questions were used to generate demographic information. After completing the demographic questions, participants were randomly assigned to view one of two warning signs. In order to make sure other aspects (such as images, font size, color, etc.) did not influence the results, I tested two generic-looking signs featuring only the language that appears on two currently posted warning signs in Yellowstone National Park. In other words, the signs that were tested were intentionally shown with as little design or differences as possible to focus exclusively on the information conveyed through the wording, rather than color, images, etc. Although the two signs looked nearly identical in their lack of design, they did feature different signal words (one used “DANGER” and the other used “WARNING”) because the language that was tested was taken verbatim from the actual signs. Simply put, the test signs featured actual language, rather than contrived or manipulated wording.

For the purposes of this paper, one of the warnings signs is referred to as the “expository warning” because it featured longer text with facts and details describing the size and speed of bison as well as loaded language, such as “gored” (see Figure 5).

Figure 5. An expository warning describing bison size, speed, and unpredictability

The other warning sign is referred to as the “abridged warning,” because it featured only an authoritative (yet vague) statement about not approaching bison (see Figure 6).

Figure 6. An abridged warning featuring short text that lacked specific details

The survey software randomly assigned participants—49.5% of participants viewed the abridged warning, while 50.5% viewed the expository warning. After viewing their assigned warning sign, participants were asked a single open-ended question: “In the space below, please describe in your own words what the sign above means to you.” The answers were recorded by the online software and then analyzed (as described below) to determine how the different warning-sign language influenced respondents’ comprehension of the risks visitors face when they encounter bison at Yellowstone National Park.

Content Analysis

After the survey was closed, I conducted an analysis of survey responses to determine the public’s comprehension of the warning sign messages. The results were analyzed in two phases. First, I analyzed all participant responses using software that identified common themes and word frequency. Second, based on the frequent themes identified and an initial review of responses, I developed, tested, and then implemented a coding sheet based on Bloom’s Taxonomy (Anderson et al., 2001; Bloom et al., 1956). Bloom’s Taxonomy provides a hierarchy of cognitive processes “from simple to complex” (Krathwohl, 2002, p. 215). Specifically, the revised taxonomy (Anderson et al., 2001) uses verbs to describe cognitive processes in terms of remembering, understanding, applying, analyzing, evaluating, and creating—each with its own subcategories. Often, that hierarchy is used in education to “describe intended learning outcomes as the result of instruction” (Krathwohl, 2002, p. 213). Because the goal of this study was to analyze the public’s comprehension of warning sign messages, the category levels used in the content analysis were based on the taxonomy of cognitive processes—in particular, the three lower levels were used because no evidence was found of the higher-level processes. However, the exact terms were changed for this study to better align with the context of this survey and to better categorize the types of responses. For example, even though participants did not actually “remember” information (because it was still visible to them during the test), many participants repeated key words. That repetition would not accurately be termed “remembering,” but it would still align with the verbs often associated with that level, such as duplicate, list, and repeat. In Figure 7, I summarize the categories and characteristics that were used for this study’s coding.

Figure 7. Three categories were used to code the types of responses. Each category was related to a taxonomy term and level of learning.

Using those categories, I conducted a content analysis of all survey responses. After my coding was completed, I recruited and trained a second coder (who was not familiar with the study’s purpose) on using the categories to code a random sample of 10% of the responses. After the coder completed the analysis, a reliability test was conducted using Cohen’s kappa with quadratic weighting, which indicated an acceptable correlation (0.8148) between my results and the second coder’s results.


Of the 303 responses that were collected using the online survey, 12 were removed from the study because the responses did not actually relate to the survey’s question (e.g., when asked to describe what the warning sign meant, one participant wrote: “You do not have to try”). Of the 12 removed responses, six were from participants who saw the abridged warning, and the other six were from participants who saw the information sign. The removal of those responses left 291 responses for analysis (144 related to the abridged warning and 147 related to the expository warning). Those responses were analyzed using software (which focused on common themes and word frequency) and, as a result, two key findings regarding the content and language were uncovered. First, people who viewed the expository warning were more likely to focus their responses on maintaining a safe distance than people who viewed the abridged warning (p = .039). Second, people who viewed the abridged warning were more likely to use abstract or vague terms (such as “harm” or “hurt”) to describe the consequences of disobeying the warning (p = .029), while people who viewed the expository warning were more likely to use concrete, descriptive (even morbid) terms (such as “kill” or “gore”) (p = .002).

The responses were also analyzed using the content analysis categories (i.e., repeating, rephrasing, and relating) described above. Overall, approximately two-thirds of all responses consisted of rephrasing or summarizing key concepts from the signs (but without elaborating or applying the information to new or specific situations), such as: “Wildlife may be near, and could be hazardous to my being.” Those types of responses demonstrate a basic understanding of the information. After analyzing the rephrasing responses, I determined there was no significant statistical difference between the rephrasing responses by participants who viewed the abridged warning as compared to participants who viewed the expository warning. In other words, regardless of which sign was viewed, two-thirds of all participants demonstrated a basic level of comprehension that could be categorized as “understanding” in Bloom’s Taxonomy.

Although that may at first appear that each sign effectively communicated the risk to the public, it is important to remember that the goal of warning signs is to “alert and inform” the public about “potential hazards, so people can act to avoid the consequences” (Ayres, 2013, p. 1698). In that sense, an effective sign would result in participants demonstrating a thorough understanding of the risks as well as the ability to relate the information to specific situations or actions that were not originally described in the warning-sign language, so that people can actually avoid risks and consequences. Based on that emphasis, my findings uncover two significant differences between the warning signs tested. First, participants who viewed the abridged warning were more likely to merely repeat key terms, which correlated with the lowest level of learning and comprehension in Bloom’s Taxonomy (p = .001). In fact, 25% of people who viewed the abridged warning provided a “repeating response,” compared to only 9% of those who viewed the expository warning. For example, one participant wrote: “Do not approach wildlife.” Those types of responses demonstrate that the participant was able to recognize important words from the warning; however, the participant did not demonstrate enough understanding of the information to either rephrase it or to apply it to a different or more specific situation. Second, participants who viewed the expository warning were more likely than participants who viewed the abridged warning to provide a “relating response” (the highest level of learning and comprehension in this study), which demonstrated a thorough understanding of the sign’s message by applying the information to new situations or specific actions (p = .006). Specifically, 20% of participants who viewed the expository warning provided a “relating response,” compared to only 8% of participants who viewed the abridged warning. For example, one participant wrote: “Animals in the wild can be dangerous if they’re not used to having people around so you should not go up to them or try to pet them.” In that response, the participant demonstrated thorough understanding of the warning sign information by relating it to specific situations or actions—i.e., “you should not go up to them or try to pet them”—that were not originally described in the warning-sign language. In the discussion below, I relate those findings to the characteristics of effective warning signs.


The warning signs tested in this study provided contrasting tactics for warning the public about potential dangers: one abridged warning and one expository warning. Although they differed significantly in their language, they did share one similar aspect: the use of signal words. Both signs provided strong, attention-getting wording near the top of the sign. The abridged warning featured the word “danger,” while the expository warning featured the word “warning.” Although researchers disagree about which word may be more effective (Braun & Silver, 1995; Leonard et al., 1986; Wogalter, Jarrard, & Simpson, 1994; Wogalter & Silver, 1990), the fact that both signs include a signal word means they are likely to be more effective at highlighting a threat than they would if the signal words were absent (Wogalter et al., 1994). The use of signal words, however, is where the similarities end.

In terms of length, the expository warning featured longer text, which the public is more likely to read (Leonard et al., 1989; Silver et al., 1991), while still remaining short enough to be read in the 10 to 15 seconds that park visitors would be willing to spend on a sign posted along trails or boardwalks (Hughes & Morrison-Saunders, 2002). The expository warning was also more explicit than the abridged warning because it provided clear and detailed information, which increases the chances that people believe the situation is hazardous (Laughery et al., 1991; Laughery et al., 1993; Laughery & Stanush, 1989; Rhoades, Frantz, & Miller, 1990; Trommelen, 1997). Although the expository warning did not include an actual metaphor, it provided an explicit comparison of a bison’s speed to information the public already understands (i.e., the speed of humans), which can increase both comprehension and recall (Harris & Mosier, 1999; Johnson, 1996). That comparison also makes the warning “more specific to the individual,” which can increase the likelihood that readers will accept and comply with the warning (Williams & Noyes, 2007, p. 21). The expository warning also aligned with the characteristic of complexity because it used the active voice, understandable yet precise vocabulary, and even a negative (possibly morbid) tone—all of which can attract more attention and be more memorable (Harris & Wiklund, 1989; Leonard et al., 1991; Main et al., 1993). Finally, unlike the abridged warning, the expository warning featured hazard information, including information about the severity of non-compliance and likelihood of the hazard (Wogalter & Barlow, 1990; Wogalter et al., 1987). That said, the expository warning (and the abridged warning for that matter) could be more specific or concrete in stating actions to avoid or mitigate the hazard (Frantz, 1994).

Based on that comparison, it is not surprising this study’s findings confirmed the hypothesis that more expository-warning language may result in people more thoroughly understanding the risks in a public recreation or tourism space as well as how to avoid or mitigate such risks. When the survey responses were analyzed using terms related to the levels of comprehension and critical thinking in the revised version of Bloom’s Taxonomy (Anderson et al., 2001; Bloom et al., 1956), this study’s findings indicated that people who viewed the expository warning sign were more likely to be able to “relate” the information, which demonstrated that they understood the severity of the risk and could apply the information to new situations they may face or specific actions they should or should not take when visiting the park.

Implications and Future Research

Broadly speaking, this study’s results lead to an important conclusion regarding warning-sign language: People who view longer, more detailed expository warnings may be more likely to understand a situation’s risks and consequences as well as how to avoid or mitigate them. That conclusion is based on data indicating that the people who viewed expository warnings were more likely to focus on maintaining a safe distance from potential hazards, to use vivid (or even morbid) terms to describe the consequences of disobeying the warning, and to obtain a higher level of understanding—akin to the third level of Bloom’s Taxonomy (i.e., apply). In contrast, the people who view abridged warnings may be more likely to have a limited understanding of the risk—akin to the lowest level of Bloom’s Taxonomy.

Because the ultimate goal of a warning sign is to actually “alert and inform” the public about “potential hazards, so people can act to avoid the consequences” (Ayres, 2013, p. 1698), warning signs in certain contexts (such as locations where people have time to read and where detailed information is necessary to understand unfamiliar risks) should feature expository warning-sign language as opposed to vague abridged warnings. For example, rather than merely command the public with a sign that reads: “Do Not Swim” or “Do Not Approach,” effective warning signs should include explicit details, such as the source of the hazard, safe distance required, severity of consequences, likelihood of injury, ways to mitigate harm, and even metaphors or concepts that help the public process the information in terms of ideas they already understand. After all, if the Disney resort sign had included wording similar to the tragedy’s report’s language—that is, “Alligators routinely hunt and kill prey along the shore and drag it into the water” (emphasis added, Florida Fish and Wildlife, 2016, p. 12)—the parents of the two-year-old victim would likely have had a much better idea about the danger that lurked just a few feet away from their child.

This study’s findings are relevant to a number of people working in risk communication. Teachers of business and professional communication can apply the findings to help students critique and create more effective warning signs—especially in tourism and outdoor recreation spaces where people may not be expecting danger, such as parks, community events, and even public tours of workplaces. Suggestions include field trips to public recreation spaces, document design assignments related to warnings and risk communication, and even partnering with a local park to create and test warning signs as part of a service-learning project.

Practitioners who are responsible for visitor safety in public recreation and tourism spaces can use this study’s findings—in collaboration with Wogalter, Conzola, and Smith-Jackson’s (2002) summary of evaluation techniques—to identify inadequate warning signs that are currently in place (before an injury occurs) as well as to develop and test more effective warnings signs based on specific aspects (such as the common characteristics provided in Figure 3 above). In doing so, care should be taken to consider the context and time required to read the warning. For example, in an area where visitors frequently rest or where they expect to read information (such as visitor centers), longer expository warnings may be used as long as they can be read within 25 to 30 seconds (Cole et al., 1997; Hughes & Morrison-Saunders, 2002). Although this study focused only on public recreation risk warnings, practitioners who are responsible for safety in dangerous workplaces may consider testing similar expository warnings in areas where employees rest or have more time to read warnings, such as breakrooms or in changing areas.

Finally, researchers can use this study’s findings to examine additional nuances of warning-sign language. First, this study focused on testing the content of warning signs rather than the context. As Albers (2012) explained, the content focuses on the “actual words . . . which are presented” to the public, whereas the context focuses on the “real-world situation” in which “information gets presented” including “mitigating factors within the information presentation, social interactions, and overall environment” (p. 7). In other words, this study provided data regarding effective warning-sign language in isolation. Future studies should test these findings further in actual public recreation spaces where risks as well as distractions are present. Suggestions include field studies to observe visitors and track differences in noticeability, reading time, and compliance as well as follow-up surveys or interviews to confirm if this study’s findings hold true in a natural setting. Such research would be in line with Grabill and Simmons’ (1998) call for more qualitative research methods.

Second, because this study focused specifically on language, other aspects (such as images, color, font size, etc.) were removed from the test. Doing so enabled me to test the effectiveness of the wording by itself; however, it limited the results to language-only signs with limited design. Future studies could test the usage of images, color, etc. combined with expository and abridged warning-sign language to determine if this study’s findings regarding effective wording are supported when the text is not the main or only way in which visitors obtain risk information.

Third, this study tested actual warning-sign language that is currently used in a national park where people gather near potential risks. While that enabled me to test realistic (rather than contrived) warning language, it limited the warning sign scenarios to two extremes (a purely abridged versus a purely expository). Future studies might consider abridged warnings and expository warnings less as a rigid dichotomy and more as a spectrum. Researchers could test signs that feature varying degrees of middle-ground language (such as short yet concrete text that conveys one concise fact or shorter text that includes more morbid terms or metaphors). For example, a warning sign at a public pool might use language that reads: “WARNING! No Diving! Shallow Water. Crippling injuries may result from diving.” That type of language would include aspects of the expository warning (i.e., a signal word, slightly longer text with facts/details, two or fewer different messages, “why” information, threatening or morbid terms, appropriate length based on reading time and contextual expectations, etc.) in a slightly abridged format.

Finally, future research could focus on different populations, such as gender differences, age differences, and alternative prior experiences. Such studies could be conducted in combination with the field research suggestion above to collect data on the influence that mitigating factors and that different warning-sign language options have on various populations. The results of such studies could help scholars and practitioners better understand how to communicate to different target populations in recreational spaces that are designated for specific segments of the population.

Note: This study was approved by the Institutional Review Board of St. Cloud State University.


Albers, M. J. (Ed.). (2012). Human-Information Interaction and Technical Communication: Concepts and Frameworks: Concepts and Frameworks. IGI Global.

American National Standards Institute (ANSI). 2011. ANSI Z535.2: American National Standard for Environmental and Facility Safety Signs.

Anderson, L. W., Krathwohl, D. R., Airasian, P. W., Cruikshank, K. A., Mayer, R. E., Pintrich, P. R., . . . Wittrock, M. C. (2001). A taxonomy for learning, teaching, and assessing: A revision of Bloom’s Taxonomy of educational objectives (Complete edition). New York, NY: Longman.

Ayres, T. (2013). Warnings, anti-warnings, and pacifiers. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 1698–1701. Los Angeles, CA: Sage Publications.

Barsalou, L. W. (1987). The instability of graded structure: Implications for the nature of concepts. In U. Neisser (Ed.), Concepts and conceptual development: Ecological and intellectual factors in categorization (pp. 101–140). Cambridge, UK: Cambridge University Press.

Bloom, B. S., Engelhart, M. D., Furst, E. J., Hill, W. H., & Krathwohl, D. R. (1956). Taxonomy of educational objectives: The classification of educational goals. New York, NY: David McKay.

Borghi, A. M., Binkofski, F., Castelfranchi, C., Cimatti, F., Scorolli, C., & Tummolini, L. (2017). The challenge of abstract concepts. Psychological Bulletin, 143(3), 263–292.

Bowles, C. T., Fisk, A. D., & Rogers, W. A. (2002). Inference and the use of similes and metaphors in warnings. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 1703–1707. Los Angeles, CA: Sage Publications.

Braun, C. C., & Silver, N. C. (1995). Interaction of signal word and colour on warning labels: Differences in perceived hazard and behavioural compliance. Ergonomics, 38, 2207–2220.

Breakwell, G. M. (2000). Risk communication: Factors affecting impact. British Medical Bulletin, 56(1), 110–120.

Brener, R. (1940). An experimental investigation of memory span. Journal of Experimental Psychology, 26, 467.

Cole, D. N., Hammond, T. P., & McCool, S. F. (1997). Information quantity and communication effectiveness: Low-impact messages on wilderness trailside bulletin boards. Leisure Sciences, 19, 59–72.

Cornelissen, P. A., van Hoof, J. J., & van Vuuren, M. (2014). Enabling employees to work safely: the influence of motivation and ability in the design of safety instructions. Technical Communication, 61, 232–244.

Douglas, M., & Wildavsky, A. (1982). Risk and culture: An essay on the selection of technical and environmental dangers. Berkeley. Cal.: University of California Press.

Elling, R. (1997). Revising safety instructions with focus groups. Journal of Business and Technical Communication, 11, 451–468.

Epstein, W. (1962). Backward Association as a Function as Meaningfulness. The Journal of General Psychology, 67, 11–20.

Espiner, S. R. (1999). The use and effect of hazard warning signs. Science for Conservation, 108, 1–40.

Espiner, S., & Weiss, R. (2010). Evaluating aspects of hazard awareness, safety message effectiveness and behavioural compliance among visitors to the Glaciers, Westland National Park, New Zealand. South Westland: Department of Conservation.

Florida Fish and Wildlife Conservation Commission. (2016, August 22). Case summary: Fatal alligator attack. (Publication No. FWNE16OFF7368).

Frantz, J. P. (1994). Effect of location and procedural explicitness on user processing of and compliance with product warnings. Human Factors, 36, 532–546.

Godfrey, S. S., & Laughery, K. R. (1984). The biasing effects of product familiarity on consumers’ awareness of hazard. Proceedings of the Human Factors Society 28th Annual Meeting, 483–486. Santa Monica, CA: Human Factors and Ergonomics Society.

Goldhaber, G. M, & deTurck, M. A. (1988). Effects of consumers’ familiarity with a product on attention to and compliance with warnings. Journal of Products Liability, 11, 29–37.

Harris, R. J., & Mosier, N. J. (1999). Memory for metaphors and similes in discourse. Discourse Processes, 28(32), 257–270.

Harris, J. E., & Wiklund, M. E. (1989). Consumer acceptance of threatening warnings in the residential environment. Proceedings of the Human Factors Society 33rd Annual Meeting, 989–993. Santa Monica, CA: Human Factors and Ergonomics Society.

Herrerro, S. G., Saldaña, M. A. M., Campo, M. A. M. del, & Ritzel, D. O. (2002). From the traditional concept of safety management to safety integrated with quality. Journal of Safety Research, 33, 1–20.

Holmes, V. T., & Langford, J. (1976). Comprehension and recall of abstract and concrete sentences. Journal of Verbal Learning and Verbal Behavior, 15, 559–566.

Hughes, M. A., & Hayhoe, G. F. (2008). A research primer for technical communication: Methods, exemplars, and analyses. New York, NY: Routledge.

Hughes, M., & Morrison-Saunders, A. (2002). Impact of trail-side interpretive signs on visitor knowledge. Journal of Ecotourism, 1(2–3), 122–132.

Jackson, M. S., White, G. N., & Schmierer, C. L. (1996). Tourism experiences within an attributional framework. Annals of Tourism Research, 23(4), 798–810.

Johnson, A. T. (1996). Comprehension of metaphors and similes: A reaction time study. Metaphor & Symbolic Activity, 11(2), 145–159.

Krathwohl, D. R. (2002). A revision of Bloom’s taxonomy: An overview. Theory into Practice, 41(4), 212–218.

Kuniavsky, M. (2012). Observing the user experience: A practitioner’s guide to user research (2nd ed.). San Francisco, CA: Morgan Kaufman.

Laughery, K. R., Sr., Rowe-Hallbert, A. L., Young, S. L., Vaubel, K. P., & Laux, L. F. (1991). Effects of explicitness in conveying severity information in product warnings. Proceedings of the Human Factors Society 35th Annual Meeting, 481–485. Santa Monica, CA: Human Factors and Ergonomics Society.

Laughery, K. R., Sr., & Stanush, J. A. (1989). Effects of warning explicitness on product perceptions. Proceedings of the Human Factors Society 33rd Annual Meeting, 431–435. Santa Monica, CA: Human Factors and Ergonomics Society.

Laughery, K. R., Sr., Vaubel, K. P., Young, S. L., Brelsford, J. W., Jr., & Rowe, A. L. (1993). Explicitness of consequence information in warnings. Safety Science, 16, 597–613.

Leonard, S. D., Matthews, D., & Karnes, E. W. (1986). How does the population interpret warning signals? Proceedings of the Human Factors Society 30th Annual Meeting, 116–120. Santa Monica, CA: Human Factors and Ergonomics Society.

Leonard, S. D., Creel, E., & Karnes, E. W. (1991). Adequacy of responses to warning terms. Proceedings of the Human Factors Society 35th Annual Meeting, 1024–1028. Santa Monica, CA: Human Factors and Ergonomics Society.

Leonard, D. C., Ponsi, K. A., Silver, N. C., & Wogalter, M. S. (1989). Pest-control products: Reading warnings and purchasing intentions. Proceedings of the Human Factors Society 33rd Annual Meeting, 436–440. Santa Monica, CA: Human Factors and Ergonomics Society.

Main, B. W., Frantz, J. P., & Rhoades, T. P. (1993, July). Do consumers understand the difference between “flammable” and “combustible?” Ergonomics in Design, 14–17, 32.

Mettler, K., & Sharma, S. (2016, June 17). Disney installing signs warning of alligator’s after boy’s death. The Washington Post. Retrieved from https://www.washingtonpost.com/news/morning-mix/wp/2016/06/17/disney-will-begin posting-alligator-warning-signs-after-boys-death/

Miller, M. E. (2015, July 23). Bison selfies are a bad idea: Tourist gored in Yellowstone as another photo goes awry. The Washington Post. Retrieved from https://www.washingtonpost.com/news/morning-mix/wp/2015/07/23/bison-selfies-are-a-bad-idea-tourist-gored-in-yellowstone-as-another-photo-goes-awry/?utm_term=.2fc32bdf7e36.

Njome, M. S., Suh, C. E., Chuyong, G., & de Wit, M. J. (2010). Volcanic risk perception in rural communities along the slopes of mount Cameroon, West-Central Africa. Journal of African Earth Sciences, 58(4), 608–622.

Paivio, A. (1963). Learning of adjective-noun paired associates as a function of adjective-noun word order and noun abstractness. Canadian Journal of Psychology/Revue Canadienne de Psychologie, 17(4), 370-379.

Paivio, A., Yuille, J. C., & Madigan, S. A. (1968). Concreteness, imagery, and meaningfulness values for 925 nouns. Journal of Experimental Psychology, 76(1, Pt. 2), 1–25.

Priest, S., & Baillie, R. (1987). Justifying the risk to others: The real razor’s edge. Journal of Experiential Education, 10(1), 16–22.

Rhoades, T. P., Frantz, J. P., & Miller, J. M. (1990). Emerging methodologies for the assessment of safety-related product communications. Proceedings of the Human Factors Society 34th Annual Meeting, 998–1002. Santa Monica, CA: Human Factors and Ergonomics Society.

Roehl, W. S., & Fesenmaier, D. R. (1992). Risk perceptions and pleasure travel: An exploratory analysis. Journal of Travel Research, 30(4), 17–26.

Rofes, A., Zakariás, L., Ceder, K., Lind, M., Blom Johansson, M. B., de Aguiar, V., . . . Howard, D. (2017). Imageability ratings across languages, Behavior Research Methods, Springer Verlag, 1–11. doi: 10.3758/s13428-017-0936-0.

Rogers, W. A., Lamson, N., & Rousseau, G. K. (2000). Warning research: An integrative perspective. Human Factors, 42(1), 102–139.

Saleh, T. A. (2011). Testing the effectiveness of visual aids in chemical safety training. Journal of Chemical Health and Safety, 18(2), 3–8.

Sauer, B. A. (1992). The engineer as rational man: The problem of imminent danger in a non-rational environment. IEEE Transactions on Professional Communication, 35, 242–249.

Sauer, B. A. (2003). The rhetoric of risk: Technical documentation in hazardous environments. New York, NY: Routledge.

Silver, N. C., Leonard, D. C., Ponsi, K. A., & Wogalter, M. S. (1991). Warnings and purchase intentions for pest-control products. Forensic Reports, 4, 17–33.

Trommelen, M. (1997). Effectiveness of explicit warnings. Safety Science, 25, 79–88.

Velotta, C. (1987). Safety labels: What to put in them, how to write them, and where to place them. IEEE Transactions on Professional Communication, (3), 121–126.

Wachinger, G., Renn, O., Begg, C., & Kuhlicke, C. (2013). The risk perception paradox—implications for governance and communication of natural hazards. Risk Analysis, 33(6), 1049–1065.

Walsh, L., & Walker, K. C. (2016). Perspectives on uncertainty for technical communication scholars. Technical Communication Quarterly, 25, 71–86.

Whittlesey, L. H. (2014). Death in Yellowstone: Accidents and foolhardiness in the first national park. Lanham, MD: Roberts Rinehart.

Williams, D. J., & Noyes, J. M. (2007). How does our perception of risk influence decision-making? Implications for the design of risk information. Theoretical Issues in Ergonomics Science, 8(1), 1–35.

Wogalter, M. S., & Barlow, T. (1990). Injury severity and likelihood in warnings. Proceedings of the Human Factors Society 34th Annual Meeting, 580–583. Santa Monica, CA: Human Factors and Ergonomics Society.

Wogalter, M. S., Conzola, V. C., & Smith-Jackson, T. L. (2002). Research-based guidelines for warning design and evaluation. Applied Ergonomics, 33(3), 219–230.

Wogalter, M. S., Godfrey, S. S., Fontenelle, G. A., Desaulniers, D. R., Rothstein, P. R., & Laughery, K. R., Sr. (1987). Effectiveness of warnings. Human Factors, 29, 599–612.

Wogalter, M. S., & Silver, N. C. (1990). Arousal strength of signal words. Forensic Reports, 3, 407–420.

Wogalter, M. S., Jarrard, S. W., & Simpson, S. N. (1994). Influence of warning label signal words on perceived hazard level. Human Factors, 36, 547–556.

Young, S. L., Wogalter, M. S., Laughery, K. R., Sr., Magurno, A., & Lovvoll, D. (1995). Relative order and space allocation of message components in hazard warning signs. Proceedings of the Human Factors Society and Ergonomics Society 39th Annual Meeting, 969–973. Santa Monica, CA: Human Factors and Ergonomics Society.

Youngblood, S. A. (2012). Balancing the rhetorical tension between right to know and security in risk communication: Ambiguity and avoidance. Journal of Business and Technical Communication, 26, 35–64.

About the Author

Emil B. Towner is an associate professor of business communication and the internship director for the Herberger Business School at St. Cloud State University in Minnesota. His research interests focus on business and technical communication, including pedagogy, risk communication, tactical technical communication, and public apologia/apologies. His work has been published in various journals, including Business and Professional Communication Quarterly, Technical Communication Quarterly, and Rhetoric Review. He can be reached at ebtowner@stcloudstate.edu.

Manuscript received 24 January 2018, revised 7 April 2018; accepted 14 May 2018.