Heartfelt communication effectively communicating personal CVD risk through visual aids

(1)

Heartfelt Communication

Effectively communicating personal CVD risk through visual aids

Name: Malu Susanna Pasman Student ID: 6043569 Subject: Master’s Thesis University: University of Amsterdam Institution: Graduate School of Communication Master’s Programme: Persuasive Communication

Supervisor: Annemiek Linn Date: 27th June 2014

(2)

1

tailoring these aids to the receivers’ preferred mode of presentation yields larger effects and whether education, health literacy and numeracy had moderating effects. It turned out that numerical aids led to higher risk perceptions and thus were more effective in communicating risk. Tailoring the aids to the receivers’ mode of preference did influence risk perception however. Moderating effects were not found for numeracy, health literacy and education in regarding both the type of visual aid and tailoring the aid to the receivers’ preference. Based on these results, it can be said that the type of aid indeed influences risk perception regarding and that numerical information is useful when communicating risks. Future research should look into creative ways of adapting this lesson in presentation formats. Furthermore, it should be assessed whether other delivery modes can be used when tailoring a message to individual preferences.

Introduction

The renowned American journalist Sydney J. Harris once famously stated that “the two words information and communication are often used interchangeably, but they signify quite

different things. Information is giving out, communication is getting through”. In other words, people are often confronted with a substantial amount of detailed and complex health

information which is often difficult to understand (Kessels, 2003) Especially health risks often are misunderstood and as a consequence, people have an incorrect risk perception, which makes desired behavioural changes unlikely to occur (Kreuter & Stretcher, 1995; Han & Dieckmann, 2009; Kreuter & Stretcher, 2009).

Health related behavioural theories state that communicating risk adequately is important when trying to change health-related behaviours (Paling, 2003). Risk understanding leads to recognition of personal risk, which encourages people to take precautions to reduce their risk (Kreuter & Stretcher, 1995; Weinstein & Nicolich, 1993). Perceived risk thus is an important motivator for behavioural change and people must therefore understand the benefits and risks involved with the options they face in order to make informed decisions. Research shows that understanding personal risk is difficult for most people (Weinstein & Nicolich, 1993; Kreuter & Stretcher, 1995; Reyna, Nelson, Han & Dieckmann, 2009).

(4)

3 In an attempt to improve risk communication, health communication practitioners have

adapted several creative commercial marketing strategies (Pauling, 2003). This has led to a transformation from communication based on text-only messages into communication through textual messages that are supported by visual aids (Garcia-Retamero & Cokely, 2013b). Research has shown that adding these visual aids to textual information yields positive effects (Reyna et al., 2009, Garcia-Retamero & Cokely, 2010, Garcia-Retamero & Cokely, 2011, Gaissmaier et al., 2012). Not all aids are equally effective however and communicating risk information is complicated further because not all people process

information the same way (Carpenter & Shah, 1998). It has been proven that numeracy, health literacy and education may make a difference, but it is still unclear what works best in

overcoming these differences and how exactly they influence risk perception (Reyna et al., 2009; Garcia-Retamero & Cokely, 2013a).

Adding aids can be fruitful, but often people prefer a certain type of aid over another (Ancker et al., 2006). Thus, they prefer a certain mode of presentation. The development of new technologies provides possibilities to tailor modes of presentation to individual characteristics and preferences. (Noir, Benac & Harris, 2007). Tailoring to individual characteristics has proven to be successful and scholars assume that this effect will also occur when tailoring to mode of preference (Kreuter & Rimer, 2006; Noir, Benac & Harris, 2007; Jensen, King, Cacioppo & Davis, 2012). Mode preference tailoring would lead to a higher appreciation of the message, which in turn leads to higher attention and more considerate message processing. These are prerequisites for larger effects (Stanczyk, Crutzen & De Vries, 2013). However, early studies in this field mostly show that this is not the case (Lewis et al., 2006;

Vandelanotte et al., 2012). This study therefore aims to investigate whether there is a difference between numerical and visual aids and whether adapting these aids to individual preference yields larger effects:

To what extent does the difference among visual aids influence CVD risk perception, does tailoring these visual aids to the receiver’s preference lead to larger effects and how are these relationships moderated by numeracy level, education, age and health literacy?

This research is relevant because it provides more insight into the demographic variables that might moderate the influence of visual aids on risk perception among a wide array of people. The outcome of this research can then have important consequences for tailored health interventions, as these provide many possibilities for interactivity and tailoring to individual

(5)

4 needs. Furthermore this is the first time research in this field has been conducted in The Netherlands. This is important, as previous cross-cultural comparison has proven that the effects of visual aids differ among countries (Garcia-Retamero & Galesic, 2010). Finally, this research is relevant because current health care is highly dependent on the choices made by consumers and patients (Peters, Hibbard, Slovic & Dieckmann, 2007). Informing people adequately therefore is important when they have to make decisions concerning their health.

Furthermore, research regarding the communication of risk about Cardiovascular Diseases (CVDs) is scarce. This is interesting, because CVDs are the leading cause of death in Western countries and in The Netherlands many people have a lifestyle that makes them susceptible for CVD development (Epping-Jordan et al., 2005) Guidelines on primary prevention of CVD emphasize identifying high-risk patients for intensive risk-reducing management (Waldron et al., 2011). People however tend to underestimate their risk for developing a CVD, which means that they are not very inclined to change their behaviour. Waldron et al. (2011) further notice that information regarding CVD development is complex and therefore not easily communicated. Therefore, more research in this field can be highly relevant.

Theoretical Framework

Communicating health risk

Health issues are often communicated through informing consumers about the benefits and risks concerned with the options they have (Rothman & Kieviemini, 1999). This strategy of risk communication in the field of public health emerged in the late 1970s and is defined as “any public or private communication that informs individuals about the existence, nature, form, severity or acceptability of risks” (Plough & Krimsy, 1987). Because health costs were increasing rapidly, policy makers decided that individual’s should gain more responsibility in health decisions. Health interventions therefore became more focused on taking prevention measures and people were informed about their personal risk when not adhering to these measures. Successful interventions are able to provide individuals with the ability to make an accurate risk perception (Rothman & Kieviemini, 1999, Plough & Krismy, 1987). As a result of this accurate perception, beneficial changes in health behaviours may occur (Weinstein & Nicolich, 1993; Reyna et al., 2009).

In order to be able to increase effectiveness of health interventions, scholars have looked into delivery methods that increase accuracy of risk perception. Adding visual aids to text-based and descriptive numerical information are found to positively affect risk perception

(6)

5 (Rothman & Kieviemini, 1999; Paling, 2003; Reyna et al., 2009; Garcia-Retamero & Cokely, 2013a). Visual aids play an important role in increasing understanding of risk information overcomes important problems regarding understanding (Kreuter & Stretcher, 2009). People often do not understand texts because their numeracy and literacy are to low and as a result, they find that there is no need to alter their behaviour (Reyna et al., 2009; Kreuter &

Stretcher, 2009; Epping-Jordan et al., 2005).

Regarding CVD development, an accurate assessment of risk is extremely important (Kreuter & Stretcher, 1995, Waldron et al., 2011). Many people in modern Western countries are very susceptible for developing a CVD, but awareness about this susceptibility generally is low and people tend to underestimate the severity of CVDs (Epping-Jordan et al., 2005; Waldron et al., 2010). This lack of awareness prohibits people from changing their behaviour desirably (Brendryen, 2013). CVD development risk often is conducted through textual information and descriptive numbers, which people often are not able to understand (Waldron et al., 2010). Policy makers and professionals in the field of health communication therefore often have not been able to change risk perception through interventions (Waldron et al., 2011). As visual aids are effective in communicating risk, Waldron et al. (2011) have suggested to look into possibilities to incorporate these in interventions regarding CVD risk. These interventions should be able to increase risk understanding, worry about heart disease, emotional affect and behavioural intention, which altogether compromise ‘perceived personal risk’ in the case of CVD development (Waldron et al., 2010).

Communicating risk through visual aids

Studies have shown that adding visual and graphical representations of personal risk improves the comprehension of risk (Garcia-Retamero & Cokely, 2013a; Reyna et al., 2009). Adding visual aids leads to a higher understanding of part-to-whole relations, which means that people are more able to understand how high their relative risk actually is. Furthermore, properly designed visual aids help people in selecting the most important information that is presented in a text , which increases risk understanding, worries and the emotional

affectedness (Reyna et al., 2009; Waldron et al., 2010).They therefore are more likely to correctly estimate their personal risk and when people perceive themselves at risk, they are more likely to alter their behaviour (Kreuter & Stretcher, 1995).

Adding visual aids to textual information thus helps in overcoming several barriers that usually prohibit risk understanding. Studies have shown that adding visual aids to text-based

(7)

6 health communication messages have got a positive effect on risk perception regarding a wide array of health issues (Waldron et al., 2011). Through experimental and qualitative research, Ancker et al. (2006) found that people find health information easier to understand when it is presented to them through a visual aid. Meta-analysis of quantitative studies by Reyna et al. (2009) furthermore shows that adding visual aids reduces susceptibility to how information is presented and overcomes framing effects. Garcia-Retamero and Galsesic (2010) found that visual aids increase comprehension of part-to-whole relationships and provide possibilities for people to grasp the concept of relativity. Altogether, it can be stated that the adding of visual aids lead to higher comprehension, reduction of presentation bias and discarding of framing effects (Ancker et al., 2006; Reyna et al, 2009; Garcia-Retamero & Galesic, 2010; Waldron et al., 2011; Gaissmaier et al., 2011). These factors contribute to higher effectiveness of

messages and as a result, the accuracy of personal risk perception is increased (Garcia-Retamero & Cokely, 2013a).

Through meta-analysis, Reyna et al. (2009) found that many studies show that visual aids indeed are effective, but that the strength of these effects are highly dependent on the type of aid that is used. Reyna et al. (2009) have identified the level of abstraction of visual aids as the main determinant for this difference in effectiveness. This distinction in abstraction is determined by the level of iconicity (Gaissmaier et al, 2012). Iconicity refers to how much a representation resembles what it is supposed to represent versus the extent to which it is an abstraction. Aids that still use numerical information are low in iconicity, because they are the strongest abstractions of information . Gaissmaier et al.’s (2012) research furthermore shows that presentation formats that used numerical information indeed led to a more accurate risk perception. This verifies earlier research by Feldman-Stewart et al. (2000), who showed that people were more able to make accurate risk perceptions when they were able to read the numbers from charts. This might be due to the fact that numerical information strips risk percentages to their essential elements (Gaissmaier et al., 2012). Furthermore, other distraction factors (such as icons) are removed. Also, people might be more familiar with numerical aids than with other visual representations, which makes the numerical aids easier to understand (Ancker et al., 2006).

Despite vast evidence that numerical information works better, there is also theoretical

foundation and empirical evidence that states otherwise (Reyna et al., 2009). Often descriptive information is inadequate in communicating risk, because it is too abstract. People’s

(8)

7 they lack the skills to process the numerical information, even when it is presented through a visual delivery method. Numerical aids therefore might be less effective than visual

representations (Garcia-Retamero & Cokeley, 2013b).

The meta-analysis by Reyna et al. (2009) shows that visual representation through higher iconic aids in some cases led to a better estimation of risk. These findings may be explained by the fact that visual aids that are high in iconicity are appreciated better and because they provide people with the possibility to more easily recall information (Ancker et al., 2006; Gaissmaier et al., 2012). Furthermore, they are helpful in grasping difficult concepts such as chance and uncertainty (Ancker et al., 2006). However, the fact that intermediate steps –such as the counting of icons- need to be taken might prohibit effects from occurring (Gaissmaier et al., 2012). The answer to the question whether a numerical or a visual aid leads to better risk estimation still is unclear and therefore the following supporting Research Question has been formulated:

RQ1: To what extent do types of visual aids (numerical vs. visual) differ in their effects on risk perception?

Moderators

Numeracy

Previous studies show that the effects of visual aids can be moderated by numeracy level (Reyna et al., 2009; Garcia-Retamero & Galesic, 2010; Garcia-Retamero & Cokely, 2011). Numeracy refers to the basic understanding of numbers and mathematical knowledge. People with higher levels of numeracy have more mathematical insight and may have a higher understanding of relative risk (Reyna et al., 2009). These differences in numeracy lead to a difference in how people process the information as it is presented in visual aids (Carpenter & Shah, 1998). High numerate people have more mathematical insight and may have a higher understanding of relative risk (Reyna et al., 2009). As a result, they may benefit more from visual aids with lower iconicity, as they believe these provide them with a more transparent and complete assessment of risk (Lipkus & Hollands, 1999; Lipkus, 2007). Low numerate people might not possess the skills to process these numerical aids sufficiently however and often they also find it difficult to filter relevant information presented through numbers (Reyna et al., 2009). Therefore, they may benefit more from visual aids that are high in iconicity (Garcia-Retamero & Galesic, 2010; Garcia-Retamero & Cokely, 2013a):

(9)

8

H1a: The effect of visual aids is moderated by numeracy, in which numerical aids are more effective for people with high numeracy, whereas visual aids are more effective for people with low numeracy.

Education

Although studies have identified numeracy level as the most important moderator of the effects of visual aids on risk perception, recent research has looked into and suggested other variables as well (Garcia-Retamero & Galesic, 2010). Most importantly, the general education level of individuals is thought to be a significant moderator of this relationship (Lipkus, Samsa & Rimer, 2001). Naturally, high-educated people are able to more easily and swiftly understand health information that low- and middle-educated people (Lipkus, 2007). They still benefit from adding visual aids to textual information, but the effects probably are smaller for them than for lower-educated people (Lipkus, Samsa & Rimer, 2001; Garcia-Retamero & Cokely, 2011). It is thought that high-educated people benefit more from less iconic visual aids than lower-educated people however, because these aids contain more information and provide people with more possibilities to see the bigger picture. These characteristics align with the need for information higher educated people often have (Lipkus & Hollands, 1999). Studies also indicate that high-educated people still can have low numeracy however, which again might influence the moderating effect of education (Lipkus, Samsa & Rimer, 2001; Lipkus, 2007):

H1b: The effect of visual aids is moderated by education, in which numerical aids are more effective for higher educated people, whereas visual aids are more effective for lower-educated people.

Health literacy

Health literacy refers to the capability to understand health information and the precautions they could take, along with basic knowledge about health issue (Peters et al., 2007; Williams, 2002; Nutbaum, 2000). It is often defined as ‘the ability to read, understand, and act on health information’ (Osborne, 2006). Low health literacy can be problematic, because modern healthcare requires some level of independence of patients and consumers (Peters et al., 2007). However, this independence is limited when people are not able to process health information. Previous research has shown that many people have inadequate or moderate health literacy, which complicates information processing (Giute et al., 2012). For these

(10)

9 people, adding visual aids can be very effective, because they make information more

comprehensible (Osborne, 2006; Giute et al., 2012). Studies have shown that this group of people especially profit from pictorial visual aids that are easy to understand as these take less cognitive effort to process (Ancker, 2006; Kripalani et al., 2007; Garcia-Retamero & Galesic, 2010). Thus, the following hypothesis has been formulated:

H1c: Numerical aids are more effective for people with high health literacy whereas visual aids are more effective for people with low health literacy.

Tailoring and mode of presentation

With the rise of new technologies, it has become possible to tailor messages and delivery methods to individual beliefs, characteristics, abilities and preferences (Noir, Benac & Harris, 2007; Jensen, King, Cacioppo & Davis, 2012). Tailoring plays an important role in realising customized and personalised communication and is regarded as a means to facilitate

behavioural change by influencing some key intermediate steps between exposure and behavioural intention (Rimer & Kreuter, 2006). It has been theorised that the main reason why tailoring works can be derived from the Elaboration Likelihood Model (ELM), which explains conditions under which communication is more effective. The ELM states that people can take a ‘central’ and ‘peripheral’ route when processing information (Hawkins et al., 2007). Tailoring the message to individual characteristics increases personal relevance, which then elicits ‘central route processing’. Taking the central route means that people more carefully consider the persuasive message, which can then yield larger effects on a wide array of outcomes (Stanczyk et al., 2013). This theory has found to be applicable for wide array of health related issues, such as smoking (Civljak, Sheikh, Stead & Car, 2010), weight loss (Kreuter, Bull, Clark & Oswald, 1999) and cancer screening (Emmons et al., 2010) . However, taking this route also makes the message more susceptible for counter arguing, which can decrease the effects of tailoring (Hawkins et al., 2007).

Recently, scholars have suggested that messages should not only be tailored to individual characteristics, but also to individual preferences for mode of delivery (Jensen, King,

Cacioppo & Davis, 2012). Often, people prefer one presentation format over another, because they can more easily relate to it or because it aligns with their preferred learning method (Reyna et al., 2009; Giuse et al., 2012; Stanczyk, Crutzen & De Vries, 2013). When the mode of delivery meets personal preference, appreciation of the message increases yield higher

(11)

10 effects (Kreuter et al., 1999). As a result, attention for the message is higher, which influences message comprehension, self-reference, appreciation and message processing (Stanczyk, Crutzen & De Vries, 2013). Thus, tailoring a message to the receivers’ mode of preference can motivate people to take the central route of message processing (Jensen et al., 2012).

Research in the field of tailoring to mode preference is limited and the results of empirical research are inconsistent. Mode preference tailoring mainly has been researched regarding the issue of physical activity. These studies have shown that tailoring to mode of preference predominantly does not yield higher effects than when the delivery mode is not tailored. Lewis, Napolitano, Whiteley and Marcus (2006) compared telephonic interventions to print interventions and found that despite the fact many people had a preference for telephonic interventions, matching the intervention to this preference did not influence their physical activity and compliance with a proposed sports protocol. Vandelanotte et al (2012) also focused on physical activity and created an experimental intervention in which they informed people through textual brochures or video messages about their personal risk regarding chronic illnesses -such as CVDs. This study also found that there was no difference between matched and mismatched intervention methods. However, lower educated people did improve their risk understanding when the mode of delivery was matched to their personal preferences.

Furthermore, Guise et al. (2012) conducted an experiment in which they matched messages about hypertension risk with learning style preferences (visual vs. textual) and found that understanding of hypertension risk was higher when messages were matched with

participants’ preferred learning style. Guise et al. (2012) also found that adapting materials to participants’ health literacy yielded positive effects and that tailoring could provide better possibilities for processing for people whose health literacy is inadequate. They recommended that the influence of health literacy on the relationship between tailoring and risk perception therefore should be researched more thoroughly.

Both Lewis et al. (2006) and Vandelanotte et al. (2012) acknowledge that their lack of results may have been caused by the fact their samples were homogeneous and because participants were forced to choose between two modes of presentation they were not familiar with. Thus, other formats need to be researched. It is interesting to assess this issue regarding visual and numerical aids, because these represent two distinct learning methods and it is very likely that people have a strong preference for either of these (Reyna et al., 2009; Lipkus, 2007).

(12)

11 preferred mode of presentation indeed effects risk perception positively. Furthermore, the moderating influences of numeracy, health literacy and education will also be assessed regarding the relationship between tailoring and risk perception:

RQ2: Does adapting the visual aid to the receiver’s mode of preference have a positive effect on patients’ risk perception?

RQ3: To what extent is the influence of mode of preference on risk perception moderated by numeracy, education and health literacy?

The research is summarised in the following conceptual model:

Risk perception Tailoring type of visual

Education Numeracy Health literacy

(13)

12 Method

Design

A 2 condition (numerical vs. visual)between-subjects factoral design was used to examine the effect of the type of visual aid on risk perception. Another 2 condition (tailoring vs. non-tailoring) between-subjects factoral design was used to examine the effect of tailoring on risk perception Secondly, it was assessed whether there was a difference in risk perception

between participants who got to choose their aid of preference and participants who did not (tailored vs. non-tailored). The participants (N=115) were randomly assigned to one of the following conditions: numerical (N = 57) or visual (N = 58) and tailored (N = 39) or non-tailored (N = 79). In the non-tailored condition, participants were provided with the possibility to choose for a numerical (N = 17) or a visual aid (N = 19). This means that there was not a significant preference for either of the aids.

For this experiment, an educational message was developed, through which people were presented with a standardised chance of developing a CVD. Two different versions of this information brochure were created, containing either textual information supported by a numerical aid or textual information supported by a visual aid. The numerical aid provided participants with information regarding their CVD risk through a bar chart and the visual aid provided participants with this information through an icon array.

Procedure

Participants (N = 115) were recruited via social media, e-mail and snowballing in order to obtain a heterogeneous sample. Participation was completely voluntary and no incentive was used in order to recruit more participants. Participants were eligible to participate when they (1) were 18 years or older and (2) were able to read and write Dutch. When participants entered the online experiment, an introductory text was shown, in which they were informed that in this experiment they would receive information about their chance of developing a CVD. According to ethical guidelines, people were able to stop participation throughout the experiment.

After participants gave informed consent, they had to fill out a questionnaire in order to assess their chance of developing a CVD. This questionnaire measured several demographic variables and other background characteristics, such as age, education, whether people smoked and whether several factors contributing to CVD development were common in their families. Because the time span of this research was too limited to develop a professional

(14)

13 measurement, all participants were given the hypothetical chance of 29.7%, as this is close to the average chance of developing a CVD in the Netherlands (Factsheet Hart-en Vaatziekten, 2012) . Furthermore, this percentage indicates that people are at high risk and should

undertake action in order to defer the development of a CVD.

Next, participants were presented with a text which informed them about their chance of developing a CVD. This text furthermore contained information about risk factors that

contribute to CVD development and actions they could undertake in order to reduce their risk. Participants were randomly assigned to a text supported by a numerical aid, by a visual aid or to the tailored condition. In the tailored condition, participants were presented with the two options (numerical vs. visual) and got to select their aid of preference. Participants were then asked to carefully read the text and to pay specific attention to the aid supporting the text. They were able to spend as much time on the page as they felt was necessary. Immediately after exposure to the text, risk comprehension, health literacy and involvement were assessed.

At the end of the experiment, a disclaimer was shown. Through this disclaimer, participants were told that their chance of CVD development was manipulated and that all participants were presented with an equal chance of developing a CVD. It was explained that this was done in order to be able to objectively analyse risk comprehension. Furthermore, participants were told that their actual chance might be higher or lower than the manipulated chance they just had been presented with. If participants felt in need of more information, they were refered to the website of the Nederlandse Hartstichting (www.hartstichting.nl), on which more information regarding CVDs can be found.

Measures

Risk perception

There have been several studies that have assessed the impact of communicating (risk) information about CVD development on the perceived personal risk of the receiver. The meta-analysis by Waldron et al. (2010) concludes that risk perception is … by several components: behavioural intention, understanding, emotional affect and worries. Altogether, these components form the construct ‘risk perception’. The scale for risk perception was constructed by using three studies that Waldron et al. (2010) identified as measuring these aspects. The items measuring understanding and worries was based on the items of the scale used by French et al. (2004). The items used in research by Scott and Curbow (2006)

(15)

14 participants. Finally, items that assess risk understanding and behavioural intention were derived from the study by Fair et al. (2008). All items were measure on 1 (totally disagree) to 7 (totally agree) point Likert-scales. These items together form the scale ‘risk perception’. Furthermore, people were asked to assess their risk likelihood on a scale of 1 to 10, as this assesses risk accuracy (Asimakopoulou et al., 2008).

Finally, a Factoranalysis was conducted in order to validate the scale. This shows that risk perception consists of three components that had an Eigenvalue above 1 (respectively 2.79, 1.32 and 1.05). The first component is ‘perceived risk susceptibility’ and consists of three items such as “how likely is it for someone in your situation to develop a CVD”. Reliability of the scale is good (α = 0.73) and cannot be improved. The scale was recoded to a scale that runs from 1 (low risk susceptibility) to 7 (high risk susceptibility). The second component is ‘emotional affect’ and includes items such as “how worried are you after viewing these

results” (3 items, α = 0.69) and thus the scale is reliable. This scale also was recoded to a scale that runs from 1 (low emotional affect) to 7 (high emotional affect). The third component finally was not transformed into a scale, because reliability was low (α = 0.03).

Numeracy

Numeracy will be assessed with the Scale of Numeracy scale developed by Lipkus, Isaac, Samsa and Rimer (2001). This scale consists of eight basic mathematical questions, such as “If Person A’s risk of getting a disease is 1% in ten years, and Person B’s risk is double that of A’s, what is B’s risk?” and “If the chance of getting a disease is 20 out of 100, this would be the same as having a ____% chance of getting the disease.”. Numeracy then was assessed through the number of answers the participants answered correctly (Peters et al., 2005). The more questions participants answered correctly, the more numerate they were.

The mean numeracy score was 6.65 (median = 7) out of 8 possible (range = 4-8, alpha = ..). The distribution was highly skewed (skewness = 1.16).

Health Literacy

In order to measure health literacy, the Set of Brief Screening Questions (SBSQ) developed by Chew, Bradley and Boyko (2004) was used. This scale has been used in previous studies, which show that it provides an objective mean of measuring health literacy (Jordon, Osborne & Buchbiner, 2011). The SBSQ includes six items, such as “How often are medical forms difficult to understand and fill out?” and “How often do you have difficulty understanding written information your health care provider gives you?” (six items, α = 0,82), all to be

(16)

15 scored on a 7-point Likert scale ranging from 1 (totally disagree) to 7 (totally agree). The scale then was recoded to a scale that runs from 1 to 7.

The mean health literacy score was 3.45 (median = 3,33) out of 7. The distribution was highly skewed.

Education

Education was measured among the Dutch educational groups. Because distribution was highly skewed (skewness = -0.80), a median split was done in order to determine two educational groups (median = 6). Because HBO is regarded as higher education in The Netherlands, this educational level was appointed to the higher educated group. Primary school up to MBO-level was appointed to ‘lower educational level’.

Involvement

Involvement with the educational message was measured, because the ELM proposes that people who feel more involved with the message, might process the message more

considerately. For this study, a validated scale developed by was used. This scale consisted of 6 items such as “I have read the message with great interest” and “I paid much attention to the supportive picture” and were measured on 1 to 7 point Likert Scales. The construct

‘involvement’ was computed and then recoded to a scale that runs from 1 to 7. On average, participants felt rather involved with the information (M = 4.30, SD = 1.07).

Statistical Analysis

Descriptive statistics were used in order to describe the sample characteristics. For validating the scales, a Principal Component Analysis (PCA) was done and reliability was measured through Cronbach’s Alpha. The distribution of the health literacy scale and the numeracy score were highly skewed. Therefore, a median split was used on the measures for both numeracy (median = 6) and health literacy (median = 3.33). Thus, the analysis compared the participants who were most numerate (7 or 8 correct) with those who were less numerate (4-6 items correct) and the participants who were most literate (average score of 3,33 or higher) with those who were less numerate (average score of 3,33 or lower). The fact a dichotomous split was used is justified by the skewness of the results (Peters et al., 2005). Finally, a median split was used to split the participants in ‘high involvement’ (average score of 4.33 or higher) and ‘low involvement’ (average score of 4.32 or lower) groups and it was assessed whether the difference in involvement moderated the results.

(17)

16 For testing the effects of the type of aid on risk perception (RQ1) and of tailoring on risk perception (RQ2), two separate analyses of variance (ANOVA’s) were conducted.

Furthermore, moderating effects of numeracy (H1) education (H2a) and health literacy (H2b) on the relationship between type of aid and risk perception were assessed through ANOVA’s, with risk perception as a dependent variable and type of aid and moderating variables as fixed factors. In order to analyse the moderating influence of numeracy, education and health literacy (RQ3) on the relationship between tailoring the aid to the preferred mode of preference and risk perception, an ANOVA was conducted with the tailoring condition (tailored vs. non-tailored) as an independent variable, risk perception as a dependent variable and the moderating variables as fixed factors.

In order to assess whether participants had been equally divided among the numerical and visual aids regarding education, gender and age, Chi-statistics and F-statistics were conducted where appropriate. This analysis shows that people were equally divided among groups regarding education (χ2

= 0.55, p = 0.30), gender (χ2 = 0.07, p = 0.47) and age F (1,113) = 1.41, p = 0.24. The conditions regarding tailoring did not differ significantly on education (χ2 = 0.01, p = 0.54), gender (χ2

= 0.01, p = 0.53) or age F (1,113) = 0.3, p = 0.57.

Results

Sample characteristics

One-hundred and fifteen participants participated in the whole experiment (N = 113). On average, the group was fairly young (M = 25.53, SD = 8.47) and the majority was HBO-educated or higher (N = 79). Slightly more women (N = 68) than men (N = 47) participated in this survey. Further important demographic characteristics are shown in the following table:

Demographic variables Number (N) Percentage (%)

Gender Male 47 40,9%

Female 68 50,1%

Education level Low 36 31,3%

High 79 68,7%

Nummeracy Low 49 39,1%

High 70 60,9%

Health Literacy Low 50 43,5%

High 65 56,5%

(18)

17

Effect of type of aid on risk perception

First, a one-factor analysis of variance (ANOVA) was conducted in order to answer RQ1, which was concerned with the influence of the type of aid (numerical vs. visual) on risk perception. Through this ANOVA, the influence of type of visual aid on perceived risk susceptibility was assessed A very small, but significant effect was found F (1,113) = 4,92; p < 0.05; η2

= 0.05. Participants who were exposed to a numerical aid had a higher perceived risk susceptibility (M = 3.68, SD = 1,20) than subjects who were confronted with a visual aid (M = 3.18, SD = 1.22). Thus, participants who were exposed to a text supported by a

numerical aid, were more able to make an accurate perception of their risk susceptibility.

Next, an ANOVA was conducted to assess the influence of the type of visual aid on emotional affect. A small, marginal significant effect was found F (1,113) = 3.57; p = 0.06, η2 = 0.03. for the influence of the type of visual aid on emotional affect, indicating that a numerical aid (M = 3.90, SD = 0.15) leads to higher emotional affect than a visual aid (M = 3.49, SD = 0.15). Hence, the answer to RQ1 is that the type of aid used indeed makes a difference in perceived personal risk and that numerical aids work better than visual aids.

Next, it was assessed whether demographic factors moderate the relationship between the type of aid and risk perception. It turned out that there are no interaction effects for education F (1,111) = 0,97; p = 0.34, numeracy F (1,111) = 0.93, p = 0.34 and health literacy F (1,111) = 0; p = 0.98 regarding the influence of the type of visual aid and risk susceptibility.

Involvement did not act as a moderator regarding this relationship either, F (1,111) = 1.47, p = 0.23. No interaction effects were found either for education F (1,111) = 0.66; p = 0,42,

numeracy F (1,111) = 0.06; p = 0.80 or health literacy F (1,111) = 1.37; p = 0.24 regarding the relation between the type of visual aid and emotional affect. Also, involvement did not

moderate this relationship, F (1,111) = 1.80, p = 0.18. This absence of interaction effects means that these variables to not influence the relationship between the type of visual aid and risk perception. This means that hypotheses H1a, H1b and H1c are rejected.

These findings show that numerical aids generally are slightly more effective when

communicating risk than visual aids on both a rational and emotional level. No moderating effects for education, numeracy and health literacy were found. This indicates that

demographic variables do not influence the relationship between these type of aids and risk perception.

(19)

18

Effect of tailoring on risk perception

The second part of this study assessed whether tailoring an aid to the receiver’s mode of preference yielded higher effect than when it was not tailored. Furthermore, the moderating effects of numeracy, health literacy and education were assessed. An ANOVA was conducted to assess the influence of tailoring on perceived risk susceptibility. . The results of this

analysis show that participants that were in the tailored condition did not score significantly higher (M = 3.43, SD = 1.23) than participants that were in the non-tailored condition (M = 3.40 , SD = 1.24), F (1,113) = 0; p = 0.94. This means that tailoring the aid to the receiver’s mode of preference makes no difference in the perceived risk susceptibility.

A second ANOVA was conducted in order to assess the relationship between tailoring and risk perception. Through this second ANOVA, the relationship between the type of visual aid and emotional affect was assessed. This analysis shows that participants in the tailored condition score lower (M = 3.56, SD = 1.19) than participants in the non-tailored condition (M = 3.75, SD = 1.16), but this effect was not significant, F (1,113) = 0.66; p = 0.42.

Furthermore, no interaction effects for education, F (1,111) = 0.24; p = 0.63, health literacy, F (1,111) = 0.06; p = 0.81 or numeracy, F (1,111) = 6.40; p = 0.18 on the relationship between tailoring and risk susceptibility were found. Interaction effects also were not found for

numeracy, F (1,111) = 1.26; p = 0.27 and health literacy, F (1,111) = 0.69; p = 0.41 regarding the relationship between the tailoring condition and emotional affect. However, a small, a small and marginal significant moderating effect for education was found, F (1,111) = 2.77; p = 0.09; η2

= 0.02 regarding the tailored (Mlower = 3.18, SD = 1.20 vs. Mhigher = 3.73) versus the non-tailored (Mlower = 3.93, SD = 1.09 vs. Mhigher = 3.67, SD = 1.18) condition. This means that education interacts with the relationship between tailoring and emotional affect. Thus, education moderates this relationship. Finally, involvement did not moderate the relationship between both tailoring the type of aid and risk susceptibility, F (1,111) = 1.11, p = 0.29 and between tailoring the type of aid and emotional affect, F (1,111) = 2.19, p = 0.14.

These findings show that there were no significant differences between the tailoring and non-tailoring conditions regarding risk perception after exposure to the message. Thus, the answer to RQ2 is that tailoring the message to the preferred mode of delivery does not significantly improve risk perception. Furthermore, this relationship was not moderated by numeracy and health literacy. Education level however did make a small difference, as emotional affect was slightly higher for low-educated people that were in the non-tailored condition. The answer to

(20)

19

RQ3 therefore is that numeracy and health literacy do not moderate the relation between tailoring and risk perception, but that education seems to influence this relationship marginally.

Discussion

This study examined the influence of the type of aid (numerical vs. visual) on the perceived personal risk of individuals who (hypothetically) are at high risk and showed that using numerical aids lead to a better risk perception. Furthermore, this study assessed whether adapting the aid to an individual’s delivery mode of preference, i.e. tailoring, yields larger effects on risk perception and found that tailoring the aid to the receiver’s preferred mode of delivery does not lead to a significantly different risk perception compared to when the aid is not tailored. The moderating influence of numeracy, health literacy and education were assessed for both the relationship between type of aid and risk perception and for the relationship between tailoring to mode of preference and risk perception, but they were generally found to not influence the results. In this section, the theoretical contribution of this study is highlighted along with recommendations for future research, followed by limitations. Finally, the practical implementations are discussed.

The data shows some interesting findings regarding the difference between numerical aids and visual aids. Participants’ perceived risk was higher when they were presented with a numerical aid. This effect occurred for both risk susceptibility and emotional affect. This means that participants in the numerical condition were more aware of the fact they were at high risk. This is important, as this realization is an important prerequisite for actual

behavioural change (Kreuter & Stretcher, 1999). Thus, numerical aids are more effective when trying to communicate personal risk regarding CVDs.

These results align with previous studies that found that conveying numerical information through visual aids yields the largest positive effects (Gaissmaier et al., 2012). An explanation for this is that this presentation format strips down the communication of risk to its essential elements, because people often are more familiar with numerical formats of presentation than with visual formats of presentation and because they provide people with more information (Ancker et al., 2006). Based on the existing theoretical framework, the option that the visual representation of risk worked better was not ruled out, because these representations often are liked better, are easier to understand and more easily recalled (Gaissmaier et al., 2012, Garcia-Retamero & Cokely, 2013a). It might be that these factors are not relevant when assessing

(21)

20 CVD risk. Furthermore, CVD development is such a complex issue that numerical

information might still be needed in order to adequately communicate about this illness (Ancker et al., 2006; Gaissmaier et al., 2012).

Contrary to what was expected, the relationship between the type of visual aid and risk perception was not moderated by numeracy, health literacy and education. This contrasts pervious research, which shows that adding visual aids are specifically useful for people that are low numerate (Reyna et al., 2009), low literate (Kripalani et al., 2007) and low educated (Lipkus, Samsa & Rimer, 2001). Previous research does show however that adding aids can be especially effective for people whose numeracy, literacy and education levels are

extremely low (Garcia-Retamero & Cokely, 2013a). Despite the fact that there was a vast difference between these levels in this study, these might not have been large enough to define differences

This study also sheds some light on the influence of tailoring the message’s delivery mode to the receiver’s preference. Previous tailoring studies had shown that tailoring the message’s mode of delivery to individual beliefs, characteristics and abilities can yield larger positive effects on desirable outcomes among a wide array of health topics (Kreuter et al., 1999; Rimer & Kreuter, 2006; Jensen et al., 2012). Based on these results, scholars have recently

recommended to look into possibilities for tailoring messages to the receiver’s preferred mode of delivery. Because tailoring to individual characteristics increases an intervention

effectiveness, it was assumed to be very likely that tailoring to mode of preference also would increase effectiveness (Jensen, King, Cacioppo & Davis, 2012). This assumption is based on the Elaboration Likelihood Model, which states that there are several factors that influence whether people elaborate on a message, such as appreciation of the message and the fact that it becomes more personally relevant (Kreuter et al., 1999). Tailoring a message to mode of delivery would therefore elicit more elaborate processing.

The results of the current study align with earlier empirical research that predominantly had not found any evidence for positive effects of mode preference tailoring. (Lewis et al., 2006; Vandelanotte et al., 2012). Results furthermore show that the relationship between tailoring and risk perception was not moderated by involvement. Based on the ELM, it could have been expected that people who were in the tailoring condition felt more involved with the message (Jensen et al., 2012). The fact that the results show that there was no difference in involvement among conditions, makes it unlikely that participants in the tailored condition

(22)

21 elaborated more on the message than people who were in the non-tailored condition. On average, involvement with the message already was quite high and it could be that increasing it was challenging. Furthermore, it might be that the ELM does not apply to these formats, because they are not very dynamic and at first sight are quite similar.

These results do not rule out the possibility that tailoring to mode of preference may work however. Research in this field still is rather explorative and there are a lot of factors that have not been assessed yet. Furthermore, the fact that Giuse et al. (2012) did find significant results proves that in some situations mode preference tailoring can work and that differences were found when the characteristics of the delivery mode differed significantly. Future research should therefore look into other presentation formats that can be tailored to individual’s preferences, such as visual narratives, interactive videos and games (Vandelanotte et al., 2012, Garcia-Retamero & Cokely, 2013a). This study shows that there is no difference in tailoring versus non-tailoring when comparing numerical and visual aids, but it may be that results can be found for other pairs of delivery modes.

Although this study gives some insight into how the type of visual aid influences risk perception and whether tailoring to mode of preference influences the effectiveness of a message, some other factors might explain the results as well. First, the absence of significant results can be explained by the relatively small sample size. Due to time and budget

limitations, only 115 participants participated in this experiment. Future research should assess whether effects do occur with a larger sample.

Furthermore, average risk perception in both groups was rather high. Previous research has shown that no matter what visual aid is used, adding one to a textual message always

increases the effectiveness of a message (Garcia-Retamero & Galesic, 2010). An explanation therefore may be that adding these specific aids already increased the effectiveness of the textual message sufficiently. This may mean that perceived risk could not much further be increased.

However, the results do show that there was a small difference between the conditions

regarding risk perception. This supports previous studies that indicate that visual aids differ in their effectiveness (Reyna et al., 2009). Thus, different kinds of aids do yield different effects. Future research should explore other supporting visual aids as well, such as visual narratives and other aids that are high in iconicity .When researching tailoring effects, people should be

(23)

22 presented with more options to choose from, in order to minimize the chance that they do not prefer a certain format. Future research thus should also take into account that people might not have a preference for a certain presentation format and therefore should not force people to choose between formats, as was done in this study.

Implications

This study shows that numerical aids are more effective than visual aids in communicating risk regarding CVD development. CVD development is a complex issue to communicate and it turns out that people are better informed about their risk perception when presented with an aid that contains some numerical information. Numbers thus still are required to inform people adequately, which supports the theory that numbers are needed to support complex information in order to provide a comprehensive assessment of risk (Gaissmaier et al., 2012). Practitioners therefore should not discard numbers completely. However, they should always take in regard the fact that they should be presented comprehensibly (Ancker et al., 2006, Garcia-Retamero & Cokely, 2013b). This aligns with the notion that “that problems with understanding numerical information often do not reside in people's minds, but in the representation of the problem” (Garcia-Retamero & Galesic, 2010).

Practitioners can also look into manners to adopt this lesson in creative information formats. Previous tailoring studies have provided a theoretical foundation for the assumption that tailoring to mode of preference works. This study shows that tailoring does not influence results when comparing numerical and visual formats, but modern technology provides possibilities for a wide array of formats and these should be explored (Noir, Benac & Harris, 2007). Previous studies have shown that especially for people with low levels of numeracy, health literacy and education, presentation formats such as visual pictorial narratives and videos can be effective, because they are easier to understand (Garcia-Retamero & Cokely, 2013a). Therefore, appreciation for these formats is very likely to be higher among these groups of people (Ancker et al., 2006). For future research, it would be interesting to bring academics and practitioners together in order to develop new creative and interactive means of communicating risk. Effectiveness of these presentation formats should be assessed through experimental research. While developing these new formats, they should however always keep in mind that they cannot completely discard the use of numbers.

(24)

23 References

Ancker, J. S., Senathirajah, Y., Kukafka, R., & Starren, J. B. (2006). Design features of graphs in health risk communication: a systematic review. Journal of the American Medical Informatics Association, 13(6), 608-618.

Asimakopoulou, K. G., Skinner, T. C., Spimpolo, J., Marsh, S., & Fox, C. (2008). Unrealistic pessimism about risk of coronary heart disease and stroke in patients with type 2

diabetes. Patient education and counseling, 71(1), 95-101.

Brendryen, H., Johansen, A., Nesvåg, S., Kok, G., & Duckert, F. (2013). Constructing a theory-and evidence-based treatment rationale for complex eHealth interventions:

development of an online alcohol intervention using an intervention mapping approach. JMIR

research protocols, 2(1).

Brewer, N. T., Chapman, G. B., Gibbons, F. X., Gerrard, M., McCaul, K. D., & Weinstein, N. D. (2007). Meta-analysis of the relationship between risk perception and health behavior: the example of vaccination. Health Psychology,26(2), 136

Carpenter, P. A., & Shah, P. (1998). A model of the perceptual and conceptual processes in graph comprehension. Journal of Experimental Psychology: Applied, 4(2), 75.

Chew, L. D., Bradley, K. A., & Boyko, E. J. (2004). Brief questions to identify patients with inadequate health literacy. Health, 11, 12.

Civljak, M., Sheikh, A., Stead, L. F., & Car, J. (2010). Internet-based interventions for smoking cessation. Cochrane Database Syst Rev, 9

Emmons, K. M., Wong, M., Puleo, E., Weinstein, N., Fletcher, R., & Colditz, G. (2004). Tailored computer-based cancer risk communication: correcting colorectal cancer risk perception. Journal of health communication,9(2), 127-141.

(25)

24

Epping-Jordan, J. E., Galea, G., Tukuitonga, C., & Beaglehole, R. (2005). Preventing chronic diseases: taking stepwise action. The Lancet, 366(9497), 1667-1671.

Fair, A. K., Murray, P. G., Thomas, A., & Cobain, M. R. (2008). Using hypothetical data to assess the effect of numerical format and context on the perception of coronary heart disease risk. American Journal of Health Promotion, 22(4), 291-296.

Feldman-Stewart, D., Kocovski, N., McConnell, B. A., Brundage, M. D., & Mackillop, W. J. (2000). Perception of quantitative information for treatment decisions. Medical Decision Making, 20, 228–238.

French, D. P., Sutton, S. R., Marteau, T. M., & Kinmonth, A. L. (2004). The impact of personal and social comparison information about health risk. British Journal of Health Psychology, 9(2), 187-200.

Gaissmaier, W., Wegwarth, O., Skopec, D., Müller, A. S., Broschinski, S., & Politi, M. C. (2012). Numbers can be worth a thousand pictures: individual differences in understanding graphical and numerical representations of health-related information. Health

Psychology, 31(3), 286.

Garcia-Retamero, R., & Galesic, M. (2010). Who proficts from visual aids: Overcoming challenges in people's understanding of risks. Social science & medicine, 70(7), 1019-1025.

Garcia-Retamero, R., & Cokely, E. T. (2011). Effective communication of risks to young adults: Using message framing and visual aids to increase condom use and STD

screening. Journal of Experimental Psychology: Applied, 17(3), 270.

Garcia-Retamero, R., & Cokely, E. T. (2013a). Communicating health risks with visual aids. Current Directions in Psychological Science, 22(5), 392-399.

(26)

25 Garcia‐Retamero, R., & Cokely, E. T. (2013b). The influence of skills, message frame, and

visual aids on prevention of sexually transmitted diseases. Journal of Behavioral Decision Making.

Giuse, N. B., Koonce, T. Y., Storrow, A. B., Kusnoor, S. V., & Ye, F. (2012). Using health literacy and learning style preferences to optimize the delivery of health information. Journal

of health communication, 17(sup3), 122-140.

Hawkins, R. P., Kreuter, M., Resnicow, K., Fishbein, M., & Dijkstra, A. (2008).

Understanding tailoring in communicating about health. Health Education Research, 23(3), 454-466.

Jensen, J. D., King, A. J., Carcioppolo, N., & Davis, L. (2012). Why are tailored messages more effective? A multiple mediation analysis of a breast cancer screening

intervention. Journal of Communication, 62(5), 851-868.

Jordan, J. E., Osborne, R. H., & Buchbinder, R. (2011). Critical appraisal of health literacy indices revealed variable underlying constructs, narrow content and psychometric

weaknesses. Journal of clinical epidemiology, 64(4), 366-379.

Kessels, R. P. (2003). Patients' memory for medical information. Journal of the Royal Society

of Medicine, 96(5), 219-222.

Kreuter, M. W., & Strecher, V. J. (1995). Changing inaccurate perceptions of health risk: results from a randomized trial. Health Psychology, 14(1), 56.

Kreuter, M. W., Bull, F. C., Clark, E. M., & Oswald, D. L. (1999). Understanding how people process health information: a comparison of tailored and nontailored weight-loss

(27)

26 Kreuter, M. W., & Wray, R. J. (2003). Tailored and targeted health communication: strategies for enhancing information relevance. American Journal of Health Behavior, 27, 227-232.

Kripalani, S., Robertson, R., Love-Ghaffari, M. H., Henderson, L. E., Praska, J., Strawder, A., & Jacobson, T. A. (2007). Development of an illustrated medication schedule as a

low-literacy patient education tool. Patient education and counseling, 66(3), 368-377.

Lewis, B. A., Napolitano, M. A., Whiteley, J. A., & Marcus, B. H. (2006). The effect of preferences for print versus telephone interventions on compliance and attrition in a randomized controlled physical activity trial. Psychology of Sport and Exercise, 7(5), 453-462.

Lipkus, I. M., & Hollands, J. G. (1999). The visual communication of risk. JNCI Monographs, 1999(25), 149-163.

Lipkus, I. M., Samsa, G., & Rimer, B. K. (2001). General performance on a numeracy scale among highly educated samples. Medical Decision Making,21(1), 37-44.

Lipkus, I. M. (2007). Numeric, verbal, and visual formats of conveying health risks:

suggested best practices and future recommendations. Medical Decision Making, 27(5), 696-713.

Nederlandse Hartstichting (2012). Factsheet cholesterol en hart- en vaatziekten. Retrieved 20th May, 2014, from http://www.nationaalkompas.nl/object_binary/o15424_Factsheet-Cholesterol-en-HVZ-nov-2012.pdf .

Noar, S. M., Benac, C. N., & Harris, M. S. (2007). Does tailoring matter? Meta-analytic review of tailored print health behavior change interventions.Psychological bulletin, 133(4), 673.

(28)

27 Nutbeam, D. (2000). Health literacy as a public health goal: a challenge for contemporary health education and communication strategies into the 21st century. Health promotion international, 15(3), 259-267.

Peters, E., Hibbard, J., Slovic, P., & Dieckmann, N. (2007). Numeracy skill and the

communication, comprehension, and use of risk-benefit information. Health Affairs, 26(3), 741-748.

Reyna, V. F., Nelson, W. L., Han, P. K., & Dieckmann, N. F. (2009). How numeracy

influences risk comprehension and medical decision making. Psychological bulletin, 135(6), 943.

Rimer, B. K., & Kreuter, M. W. (2006). Advancing tailored health communication: A persuasion and message effects perspective. Journal of Communication, 56(s1), 184-201.

Rothman, A. J., & Kiviniemi, M. T. (1999). Treating people with information: an analysis and review of approaches to communicating health risk information. JNCI Monographs, 1999 (25), 44-51.

Scott, L. B., & Curbow, B. (2006). The effect of message frames and CVD risk factors on behavioral outcomes. American journal of health behavior, 30(6), 582-597.

Stanczyk, N. E., Crutzen, R., Bolman, C., Muris, J., & de Vries, H. (2013). Influence of delivery strategy on message-processing mechanisms and future adherence to a Dutch computer-tailored smoking cessation intervention. Journal of medical Internet

research, 15(2).

Vandelanotte, C., Duncan, M. J., Plotnikoff, R. C., & Mummery, W. K. (2012). Do

participants’ preferences for mode of delivery (text, video, or both) influence the effectiveness of a web-based physical activity intervention? Journal of Medical Internet research, 14(1).

(29)

28 Waldron, C. A., Gallacher, J., van der Weijden, T., Newcombe, R., & Elwyn, G. (2010). The effect of different cardiovascular risk presentation formats on intentions, understanding and emotional affect: a randomised controlled trial using a web-based risk formatter

(protocol). BMC medical informatics and decision making, 10(1), 41.

Waldron, C. A., van der Weijden, T., Ludt, S., Gallacher, J., & Elwyn, G. (2011). What are effective strategies to communicate cardiovascular risk information to patients? A systematic review. Patient education and counseling,82(2), 169-181.

Weinstein, N. D., & Nicolich, M. (1993). Correct and incorrect interpretations of correlations between risk perceptions and risk behaviors. Health Psychology,12(3), 235.

(30)

29 Appendix A – Stimuli

Numerical condition

In de afbeelding aan de rechterkant, ziet u uw risico om een hart- en vaatziekte te

ontwikkelen. Indien u jonger bent dan 55, wordt de kans weergeven dat u deze op die leeftijd hebt ontwikkeld.

Idealiter is het risico om een hart- en

vaatziekte op te lopen rond de leeftijd van 55 jaar maximaal 10%. Er zijn echter

maatregelen om dit te voorkomen. Hieronder worden enkele mogelijkheden genoemd.

Gezond dieet en voldoende beweging

Een ongezonde levensstijl en overgewicht leiden tot een verhoogd risico op het ontwikkelen van een hart- en vaatziekte en andere kwalen. Een BMI van onder de 25 draagt bij aan het verkleinen van uw kans op een hart- en vaatziekte.

Roken

Roken verhoogt de kans op een hart- en vaatziekte aanzienlijk en zorgt voor een slechter hart. Een jaar na het stoppen met roken, is het hart terug in normale staat en is de kans om een aandoening te ontwikkelen verkleind met 50%.

Cholesterol

Cholesterol is een soort vet dat van nature in het lichaam voorkomt. Het heeft een vitale rol in de werking van cellen, maar een te hoog cholesterol leidt tot een grotere kans op hart- en vaatziekten. Idealiter is het cholesterol (200mg/dL).Het regelmatig laten controleren van uw cholesterol kan ertoe leiden dat een te hoog gehalte vroeg gedetecteerd wordt.

Bloeddruk

Een hoge bloeddruk kan leiden tot een hart- en vaatziekte. Het is dus van belang deze laag te houden en deze regelmatig te laten detecteren.

(31)

30 Visual condition

In de afbeelding aan de rechterkant, ziet u uw risico om een hart- en vaatziekte te ontwikkelen. Indien u jonger bent dan 55, wordt de kans weergeven dat u deze op die leeftijd hebt ontwikkeld.

Idealiter is het risico om een hart- en vaatziekte op te lopen rond de leeftijd van 55 jaar maximaal 10%. Er zijn echter maatregelen om dit te voorkomen. Hieronder worden enkele mogelijkheden genoemd.

Gezond dieet en voldoende beweging

Een ongezonde levensstijl en overgewicht leiden tot een verhoogd risico op het ontwikkelen van een hart- en vaatziekte en andere kwalen. Een BMI van onder de 25 draagt bij aan het verkleinen van uw kans op een hart- en vaatziekte.

Roken

Roken verhoogt de kans op een hart- en vaatziekte aanzienlijk en zorgt voor een slechter hart. Een jaar na het stoppen met roken, is het hart terug in normale staat en is de kans om een aandoening te ontwikkelen verkleind met 50%.

Cholesterol

Cholesterol is een soort vet dat van nature in het lichaam voorkomt. Het heeft een vitale rol in de werking van cellen, maar een te hoog cholesterol leidt tot een grotere kans op hart- en vaatziekten. Idealiter is het cholesterol (200mg/dL).Het regelmatig laten controleren van uw cholesterol kan ertoe leiden dat een te hoog gehalte vroeg gedetecteerd wordt.

Bloeddruk

Een hoge bloeddruk kan leiden tot een hart- en vaatziekte. Het is dus van belang deze laag te houden en deze regelmatig te laten detecteren.

(32)

31 Appendix B – Questionnaire

1. ‘Meten’ van het risico op Hart- en Vaatziekten en verzamelen demografische gegevens

Hieronder worden u enkele vragen gesteld over uw levensstijl. Naar aanleiding van deze vragen, zal uw risico op het ontwikkelen van een hart- en vaatziekte gemeten worden.

1. Hoe oud bent u? 2. Wat is uw geslacht? o Man o Vrouw 2. Wat is uw opleidingsniveau? o Basisonderwijs o VMBO o Havo, VWO o MBO o HBO o WO Bachelor o WO Master o Doctoraal o Anders, namelijk… 3. Heeft u ooit gerookt?

o Ja, en ik rook nog steeds

o Ja, maar ik ben minder dan een jaar geleden gestopt o Ja, en ik ben meer dan een jaar geleden gestopt o Nee

4. Heeft iemand in uw familie op jonge leeftijd problemen gehad met hart en/of vaten? (NB.: Bijv. Hartaanval, Angina, Hartoperatie, beroerte of een TIA).

o Ja, ikzelf

o Ja, iemand anders uit mijn familie o Nee

o Weet niet

5. Hoe lang bent u in centimeters? ___ Centimeter

(33)

32 6. Wat is uw gewicht in kilogram?

___ Kilogram

7. Komen hartproblemen regulier in uw familie voor? o Ja

o Nee o Weet niet

8. Weet u hoe hoog uw cholesterol is? (Indien nee wordt het gemiddelde van Nederland genomen) 9. Hoe hoog is uw cholesterol in mg/dl?

10. Heeft een zorgverlener u ooit verteld dat u een hoog cholesterol-gehalte heeft? o Ja

o Nee

11. Weet u hoe hoog uw bloeddruk is? (Indien nee wordt het gemiddelde van Nederland genomen) o Ja

o Nee

12. Hoe hoog is uw bloeddruk?

13. Heeft een zorgverlener u ooit verteld dat u waarschijnlijk een hoge bloeddruk heeft? o Ja

o Nee

14. Heeft u ooit medicijnen genomen om uw bloeddruk te verlagen? o Ja, en deze neem ik nog steeds

o Ja, maar momenteel niet o Nee

15. Heeft u diabetes? o Ja, Type I o Ja, Type II o Nee

16. Welk format van informatie heeft uw voorkeur? (NB. Alleen in de Tailoring Conditie) o Numeriek

(34)

33 2. Risicoperceptie

17. Hoe aangetast voelt u zich door deze uitslag?

Totaal niet aangetast 1 2 3 4 5 6 7 Zeer aangetast 18. Hoe gerustgesteld voelt u zich door deze uitslag?

Totaal niet gerustgesteld 1 2 3 4 5 6 7 Zeer gerustgesteld 19. Hoe bezorgd bent u naar aanleiding van deze uitslag?

Totaal niet bezorgd 1 2 3 4 5 6 7 Zeer bezorgd 20. Hoe waarschijnlijk acht u het dat u een hart- en vaatziekte ontwikkelt, gebaseerd op deze uitslag? Totaal onwaarschijnlijk 1 2 3 4 5 6 7 Zeer waarschijnlijk 21. Hoe waarschijnlijk acht u het dat u een gemiddeld persoon van uw leeftijd hart- en vaatziekte ontwikkelt, gebaseerd op de informatie die u net heeft ontvangen?

Totaal onwaarschijnlijk 1 2 3 4 5 6 7 Zeer waarschijnlijk 22. Hoeveel vertrouwen heeft u erin dat u de informatie die u net is gepresenteerd heeft begrepen?

Helemaal geen 1 2 3 4 5 6 7 Zeer veel Vertrouwen 23. Hoeveel actie om het risico te reduceren moet iemand in uw situatie nemen?

Helemaal geen actie 1 2 3 4 5 6 7 Zeer veel actie 24. In hoeverre bent u gemotiveerd om actie te ondernemen?

Helemaal niet 1 2 3 4 5 6 7 Zeer gemotiveerd 25. Hoe waarschijnlijk acht u het dat u een hart- en vaatziekte ontwikkelt, op een schaal van één tot tien? ___

3. Betrokkenheid (als mogelijke moderator/covariabele)

26. Ik heb de informatie uit de boodschap goed gelezen (voor het format numeriek)

Helemaal niet mee eens 1 2 3 4 5 6 7 Helemaal mee eens

27. Ik heb de afbeelding goed bekeken (alleen voor het format tekst met afbeeldingen)

(35)

34 28. Ik heb lang nagedacht over de informatie

29. Ik was sterk betrokken bij het beoordelen van de informatie

30. Ik heb mijn best gedaan om de informatie goed te beoordelen

31. Ik heb veel moeite gestoken in het evalueren van de informatie

4. Numeracy (Moderator)

32. Welke van de volgende statements geeft de grootste kans aan om een ziekte op te lopen of te ontwikkelen?

o 1 in 100 o 1 in 1000 o 1 in 10

33. Welk van de volgende percentages geeft de grootste kans weer om een ziekte op te lopen of te ontwikkelen?

o 1% o 10% o 5%

34. Als het risico van persoon A om binnen 10 jaar ziek te worden 1% is en het risico van persoon B het dubbele is van dat van persoon A, wat is dan het risico van persoon B om binnen tien jaar ziek te worden?

____

35. Als de kans van persoon A om binnen tien jaar een ziekte op te lopen 1 op 100 is en het risico van persoon B het dubbele is van dat van persoon A, wat is dan het risico van persoon B om binnen tien jaar een ziekte op te lopen?

___

36. Als de kans om een ziekte op te lopen 10% is, hoeveel mensen krijgen de ziekte dan:

Heartfelt communication effectively communicating personal CVD risk through visual aids

Heartfelt Communication

Table of Contents