Teen perceptions of good drivers and safe drivers: implications for reaching adolescents
- 1Center for Injury Research & Prevention, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, USA
- 2Department of Family Medicine and Community Health, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, USA
- 3Department of Anthropology, University of Pennsylvania School of Arts and Sciences; Philadelphia, Pennsylvania, USA
- 4Penn Center for the Integration of Genetic Healthcare Technologies, Division of Medical Genetics, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, USA
- 5Craig- Dalsimer Division of Adolescent Medicine, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, USA
- 6Division of General Pediatrics, Department of Pediatrics, Leonard Davis Institute for Health Economics, Philadelphia, PA, USA
- 7Center for Public Health Initiatives, University of Pennsylvania, Philadelphia, Pennsylvania, USA
- Dr F K Barg, Family Medicine and Community Health, 2 Gates, Hospital of the University of Pennsylvania, 3400 Spruce Street, Philadelphia, PA 19104, USA;
- Accepted 14 August 2008
Objective: To understand definitions of the phrases “good driver” and “safe driver” among teen pre-drivers and early drivers in order to appropriately tailor messages about driving safety.
Design: Qualitative study using freelisting, an anthropological research technique, to explore nuances in the ways that teens define a good driver and a safe driver
Setting: Classes in six high schools each in a different state in the USA.
Subjects: 193 adolescent pre-drivers and early drivers, aged 15–17.
Main outcome measures: Meaning of the phrase good driver and safe driver was identified for subgroups of adolescents.
Results: Teen pre-drivers and early drivers define a good driver and a safe driver as one who is cautious, alert, responsible, does not speed, obeys the law, uses seatbelts, and concentrates. There are subtle and potentially important differences in the way that subgroups define a good driver and a safe driver.
Conclusions: Injury prevention experts need to attend closely to the implicit meanings that teens attach to everyday terms. Freelisting is a method that identifies perceptions about the meaning of health communication messages and suggests differences in meaning among subgroups.
Teen crash prevention strategies make frequent use of the phrases “good driver” and “safe driver.” The Teen Safe Driver Program1 and the Partners for Safe Teen Driver Program2 offer interventions targeted at reducing crashes among teen drivers. Many insurance companies offer good-driver discounts. California Governor Arnold Schwarzenegger recently signed a teen driver safety law designed to “eliminate any extra distractions so (teens) can focus on paying attention to the road and being good drivers.”3
Injury prevention specialists acknowledge the importance of word choice in public health communications about teen driving,4–67 but little attention has been paid to the use of the phrases good driver and safe driver. In this paper, we explore teen perceptions of these phrases using freelisting, a technique used by anthropologists that generates terms that define the domain of interest and delimits the boundaries of that domain for specific subgroups. We have identified meanings that teens ascribe to the phrases good driver and safe driver, and we demonstrate how freelisting can be used to clarify nuances in the meaning of words used in injury prevention communication.
In an attempt to understand cultural phenomena, anthropologists make the distinction between etic and emic explanations.8 Expert ideas (etic explanations) about injury prevention are derived from knowledge of best practices and solid reviews of the literature. By contrast, the emic, or insider, perspective represents what members of the target group understand about the phenomenon. Emic and etic notions do not always overlap. A core task in anthropology as well as in public health is to find ways to elicit the emic perspective to understand the meaning of concepts from the point of view of the people with whom one wants to communicate.
Using words that mean one thing to experts and another to the lay public can lead to unintended (negative) consequences.6 For example, the injury prevention literature notes that use of the word “accident” infers unpredictability in the minds of the lay public.46 Arguing that the word accident should be eliminated from the injury prevention lexicon, Evans5 emphasizes that terminology alters public perception. The Evans study also shows that lay perceptions about the word accident differ from those of injury prevention specialists. Word choice also affects risk perception, which is often a function of the emotions that the words evoke.9–11 Similarly, positive versus negative word choices (message framing) in health communications, and whether the message refers to the targeted individual or their social group, has been shown to affect the uptake of the message.12
Although experts use the phrases good driver and safe driver frequently, little attention has been paid to the meanings of these phrases in the minds of teen drivers and pre-drivers. In a study examining teen and adult drivers’ perceptions of a good driver, McCormick and colleagues13 used semantic differential scales to ascertain how drivers viewed themselves on eight characteristics of good drivers. A “very good driver” was defined (in order) as more safe, reliable, predictable, considerate, responsible, and wise than an average driver. The concept of a safe driver was not explored in the McCormick study. We used freelisting to understand from the point of view of teens who are at the age when they are learning to drive how they understand the concepts of a good driver and a safe driver.
Adolescents in grades 9–11 at six high schools, each in a different state in the USA, were recruited to participate in research to investigate ideas about teen driving. Schools (and the states in which they were located) were stratified first by whether they had low, medium, or high mortality of teen drivers and then by the presence or absence of graduated driving laws and type of seatbelt laws. Schools were also chosen on the basis of whether they were located in urban, suburban, or rural areas, by their ethnic make-up, and by poverty indices. Details about the sample and the larger study can be found in Ginsburg et al.14
Written individual freelists were collected from 18 classes of high school students. Freelisting is a standard semi-structured interviewing technique15 that is commonly used to identify the elements and boundaries of a particular domain (in this case, good drivers and safe drivers) and determines the relative salience or importance of words defining the domain.16–18 A salient word in a domain is one that is mentioned frequently and spontaneously by members of the group.18 Respondents are asked to list all the words that describe a specific concept. Amalgamating lists across participants shows what words members of that group typically use to describe the concept. Central to the method is the idea that shared experiences (such as the experience of learning to drive) or shared cultural values will tend to yield a common notion about a domain among members of that group.
Students in each class were asked by a facilitator to generate two written lists: “What words do you think of when you think of a good driver?” and “What words do you think of when you think of a safe driver?” Classes were randomly assigned as to whether they would first generate the list for “good” or “safe.” The study was approved by the Institutional Review Board of The Children’s Hospital of Philadelphia. Students provided written assent and their parents provided written consent before study participation.
The research team met to review freelists, to standardize word forms, and to combine synonyms. For example, terms such as “seatbelts,” “buckles up,” and “wears a seat belt” were all coded as “seatbelts.” The cleaned lists were entered into Anthropac (version 4.0, 1996; Analytic Technologies, Natick, MA, USA), a software program designed specifically to analyze freelists. Firstly, the output was examined to determine which words were common to both lists and which words were unique. For each word, Anthropac calculates a salience index (Smith’s S) using the formula
S = ((Σ (L−Rj +1))/L)/N
where L is the length of each list, Rj is the rank of item J in the list, and N is the number of lists in the sample.18 Salience describes the words that prototypically19 define the domain of interest among members of a group by taking into account both the frequency and rank of the word on the list.
Salience scores were sorted from high to low and then plotted as scree plots (see fig 1 for an example) using the salience scores as values on the y axis. The scree plots were inspected to select a natural breaking point, which demarcated a flattening of the slope for the less-salient words. For each group’s list, all words with salience scores above that breaking point were retained as the list of salient terms. Because the range and distribution of salience scores differed for each word list, the breaking point also differed. Salience scores were generated for all adolescents, and by gender and ethnicity.
A total of 193 15–17- year olds completed two freelists enumerating characteristics of a good driver and a safe driver. Table 1 presents the demographic characteristics and license status of the study participants.
Freelisting results were analyzed by looking at salient terms as well as words that were either common or unique to the good driver and safe driver lists. Table 2 contains the frequency of words listed by at least three respondents, the average position of each word on the list, and salience scores for a good driver and a safe driver. Salient words are in bold. Overall, participants used many of the same words to describe a good driver and a safe driver: “cautious,” “concentrates,” “alert,” “responsible,” “doesn’t speed,” “obeys the law,” and “seatbelts.” (Salient words common and unique to the good driver and safe driver lists can be found in table 3.)
Although the word “seatbelts” was mentioned on both lists, it was mentioned 78 times on the safe driver list compared with 38 times on the good driver list. Words that were unique to the good driver list included: “not angry,” “knows the road,” “makes good turns,” “skillful,” “taken driving classes,” “listens,” “good eyesight,” “determined,” “nervous,” and “nice car.” Words that were unique to the safe driver list included: “(checks) mirrors,” “well-rested,” “has insurance,” “me,” “checks condition of car,” “confident,” “disciplined,” “hard to influence,” “drug free,” “male,” and “female.” When the data were divided by gender and ethnicity, different words became salient.
Males and females used many similar words to describe a good driver and a safe driver (table 4). However, only females said that a good driver “concentrates,” is “patient,” and “smart,” and a safe driver “obeys signs.” Only male respondents used the words “female,” “courteous,” and “calm” and behaviors such as “takes their time,” “(checks) mirrors,” and “uses signals” to illustrate a safe driver.
Ethnic group comparisons
African–American, white, and Hispanic teens used the words “cautious,” “alert,” “obeys the law,” and “responsible” to describe a good driver and a safe driver. However, there were differences in both the number of unique terms and the types of terms used to describe a safe driver and a good driver (table 5). African–American respondents identified many more unique terms (“experienced,” “smart,” “stays in lane,” “knowledgeable,” “female,” “takes their time,” “in control”) for a good driver than the white (“drive safely”) and Hispanic (“truthful”) respondents. Hispanics named “seatbelts” on the safe driver list, but not the good driver list.
This study was the first application of freelisting to injury prevention research. We identified definitions of a good driver and a safe driver among various groups of teens. Teens described good drivers and safe drivers using the words “cautious,” “alert,” “responsible,” “doesn’t speed,” “obeys the law,” “seatbelts,” and “concentrates.” Injury prevention messages that use the phrases good driver or safe driver would likely invoke these positive traits in the minds of teens.
There are slight nuances in meaning between the two phrases as well. A good driver (but not a safe driver) was described by teens with words connoting positive personal characteristics such as “experienced,” “patient,” and “smart.” A safe driver (but not a good driver) was described as “compliant” (obeys signs). Words that were unique to the good driver list infer a driver who is in control and skilled at navigating the road (eg, “not angry,” “knows the road,” “makes good turns,” “skillful”). Some of the words unique to the safe driver list (eg, “checks mirrors,” “well-rested,” “has insurance,” “checks car,” “disciplined,” “hard to influence,” “drug free”) describe a driver who is careful. These may not be trivial distinctions.
Injury prevention specialists argue for the importance of word choice in the development of messages about teen driver safety.
Although the phrases are used frequently, little is known about how teens define a good driver and a safe driver.
In this study, teen pre-drivers and early drivers define a good driver and a safe driver as one who is cautious, alert, responsible, doesn’t speed, obeys the law, uses seatbelts, and concentrates.
There are subtle differences in the ways that subgroups define a good driver and a safe driver.
Freelisting is a method that can be used by injury prevention specialists to identify “local” meanings of words or phrases that are used in health communication messages.
Subtle messages in health promotion communication can have paradoxical effects. In response to the mandates from the Master Settlement Agreement to conduct non-smoking campaigns among adolescents, the tobacco company Philip Morris produced a program called “Think. Don’t Smoke,” which presented words and images of smoking as an “adult choice”, “a forbidden fruit,” and an “act of rebellion”.2021 Although on the surface this subtle messaging conveyed disapproval of youth smoking, these advertisements may have appealed to developmental imperatives that encourage teens to challenge restrictions and appear competent enough to assume adult roles,22 and there is evidence that these ads promoted pro-tobacco outcomes.23 In evaluating the “boomerang” effect of anti-smoking messages in the “Think. Don’t Smoke” campaign, Hendriksen and colleagues24 discuss how psychological reactance theory helps to explain adolescents’ reactions. Psychological reactance theory states that objects that are forbidden or restricted for a particular group will tend to be more attractive to members of that group.
Although there was considerable agreement in the definition of a good driver and a safe driver among African–American, white, and Hispanic respondents, there were also substantial differences in these lists as well. The use of different words by race or ethnicity suggests that more research needs to be carried out to elucidate ways in which culture plays a role in the interpretation and integration of health promotion messages. Insight into these mechanisms may lead to a better understanding of some of the observed national differences in safety belt use, crash rates, and fatality rates among African–Americans, Hispanics, and American Indians.25
The same is true about gender. Males and females use many similar words to describe a good driver and a safe driver. Yet there are numerous examples in the lists of words that only males or only females use to describe a good driver or a safe driver. Males described safe driving with words that denote a slower pace (“calm,” “courteous,” “takes their time”), and some included “female” in their list of characteristics of safe drivers, thus reinforcing the gendered notion of safe driving. Because crash rates do indeed differ for male and female adolescents,26 further research needs to explore whether “female” in this context, when used by male adolescents, has positive, neutral, or negative connotations.
Freelisting allows researchers a glimpse into how a target audience might interpret a word. In our case, it demonstrated that, while the meanings of “good” and “safe” overlap, there are nuanced differences that hold the possibility of influencing a teenager’s response to motor vehicle injury prevention messages. Some of these differences exist on a population level, whereas others vary by subgroups. Just as injury prevention experts have eschewed the word accident because of its unintended interpretation, we need to learn to test concepts such as good drivers and safe drivers with teenagers before incorporating words into safety campaigns. Freelisting would be a first step toward developing messages that could be tested through other techniques including focus groups and conjoint analysis before the implementation of targeted social marketing campaigns.
There are several limitations to this study. Freelists are only strings of words that participants generate in response to a specific command. They were not recorded or observed in a natural driving environment. Secondly, freelists were collected in discussions in which teens were aware that the topic was teen driver safety. This awareness is likely to affect responses in unknown ways. Finally, although these data provide insights into how teens perceive language related to driving, they do not reflect actual behavior of teens.
IMPLICATIONS FOR PREVENTION
As health communications about teen driver safety are developed, it is important to ensure that the messages have the intended meaning for the target group. Experts need to attend closely to the implicit meanings that teens attach to everyday terms. Freelisting is a useful tool to help uncover those meanings. This method would also be useful for evaluating whether injury prevention messages are heard by the target group in the way that they are intended.
This paper was written as part of the Youthful Driver Research Initiative, a collaborative research program between the Center for Injury Research and Prevention at The Children’s Hospital of Philadelphia (CHOP) and State Farm Insurance Companies TM (State Farm). The views presented are those of the authors and not necessarily the views of CHOP or State Farm. We would like to acknowledge the important contributions made to this paper by Danielle Erkoboni, Dr J Felipe Garcia-España, and Dr Lela Jacobsehn.
Contributorship: FKB, KRG, and FKW conceptualized the study. FKW and KRG conceptualized and implemented the parent study upon which this paper is based. Data collection was by KRG, FKW, and SK. Data management was by SK and FKB. Data analysis was performed by FKB, KRG, FKW, SK. FKB wrote the paper, and KRG, FKW, and SK added substantially to subsequent drafts of the paper.
Funding: External funding was provided by the State Farm Insurance Companies TM (State Farm).
Competing interests: None.