Stigmatization and Health Communication:
The Regulatory Fit of Health Messages for Stigmatized Groups
Opsteller
Naam: Inge Ruitenberg
Collegekaartnummer: 10506152
Begeleider(s)
Binnen programmagroep: Alvin Westmaas Tweede beoordelaar: Michael Vliek
Onderzoeksinstelling: Universiteit van Amsterdam, Sociale Psychologie
Datum: 05-06-2016
Totaal aantal woorden: 6230 Aantal woorden abstract: 138
2 Contents Abstract ... 3 Introduction ... 4 Method ... 8 Participants ... 8
Procedure and materials ... 8
Survey 1 ... 9 Manipulation ... 9 Health messages ... 10 Survey 2 ... 11 Manipulation check ... 12 Debriefing letter ... 12 Results ... 12 Participants ... 12 Manipulation check ... 13 Stroop task ... 13 Health messages ... 14 Main analysis ... 14 Explorative analysis ... 16 Discussion ... 19 Future directions ... 23 Conclusions ... 24 Acknowledgements ... 25 References ... 25
3 Abstract
This study examined the regulatory fit of health messages for stigmatized groups by looking at the effect of depletion on the effectiveness of, and preference for, loss or gain
framing. Using a within-subject design depletion was manipulated in an experimental setting by means of a Stroop task. Subsequently, participants read a loss and a gain framed health message after which its effectiveness and their preference was measured. The results show that loss and gain framed health messages are equally effective in stigmatized groups, but stigmatized
individuals prefer gain framing. An explanation for these results could be that depletion leads to an indistinct regulatory focus. With the advice therefore to combine loss and gain framing, this study provides a first insight in the implications of stigmatization for health communication.
4 Introduction
There is a large amount of research that shows a relationship between stigma and health (e.g. Pascoe & Richman, 2009; Quinn & Chaudoir, 2009). It seems that when people experience stigmatization, they are more likely to behave in ways that are problematic to health such as overeating or smoking (Richman & Lattanner, 2014). Most health interventions don't pay attention to stigmatization (Phelan, Lucas, Ridgeway & Taylor, 2014; Viruell-Fuentes, Miranda & Abdulrahim, 2012). Yet, the concept of stigmatization in its dynamical context might have important implications for the health communication of interventions (Richman & Lattanner, 2014), which can be seen as the development of effective messages to communicate with the target group (Bernhardt, 2004). The aim of this study was to identify these implications.
In its simplest form, stigmatization can be defined as “describing or regarding
(something, such as a characteristic or group of people) in a way that shows strong disapproval" (www.merriam-webster.com). Link and Phelan (2001) offer a more complex definition of stigmatization which the present study adopts. In Conceptualizing Stigma they explain that stigma contains five interrelated components: labelling, stereotyping, separation, status loss and discrimination. Labelling means that people differentiate between each other and label these differences. A negative stereotype develops when people associate the labels with undesirable characteristics. People usually separate themselves from stereotyped persons. As a result these persons experience status loss and discrimination, for example through disapproval, rejection and exclusion, which can lead to unequal outcomes. Additionally, social, economic and political power plays an essential role in the development of stigma. The degree of stigmatization people experience, for instance, depends on the power distance between the stigmatized and the one who stigmatizes (Link & Phelan, 2001). People can experience structural and interpersonal stigmatization; the first meaning stigmatization at an environmental level (for example by laws and institutions) and the latter meaning stigmatization in social interactions (Phelan et al., 2014).
5 In research stigmatization is often interpreted as a low status, which means individuals or groups have a low social value (Anderson & Kilduff, 2009). There are several studies that show that status threats can lead to attention problems, troubles performing difficult tasks and troubles with resisting temptations. In other words an experienced low status can lead to a decreased capacity to control or alter one's behavior, also known as self-regulation (for an overview see Richman & Lattanner, 2014). There is also a whole body of research about the effects of stereotype threat. Just like an experienced low status, stereotype threat impairs self-regulation (for an overview see Aronson, Burgess, Phelan & Juarez, 2013). Studies on the relation between stereotype threat and cognitive performance show that stereotype threat induces a prevention focus, for it makes the risk of failure salient and leads to a goal of avoiding this failure (Seibt & Förster, 2004). This prevention focus can be beneficial on the short term, because a person with a prevention focus is motivated to avoid failure (Ståhl, Van Laar & Ellemers, 2012). On the long term, however, such motivation can be counterproductive as the extra effort a person shows often impairs the performance on difficult tasks (Aronson et al., 2013). A promotion focus can moderate for this effect (Trawalter & Richeson, 2006). For example, reframing the threat as a challenge helps improve the performance of the target of the stereotype threat (Alter, Aronson, Darley, Rodriguez & Ruble, 2010). Next to this, numerous studies on the effect of power found that a feeling of low power leads to the activation of inhibitory processes which cause a
heightened vigilance for threats and avoidance as well (Keltner, Gruenfeld & Anderson, 2003). Consistent with the research on stereotype threat and cognitive performance chronic inhibition can cause impaired functioning on self-regulation tasks (Guinote, 2007; Richman & Lattanner, 2014). In summary, stigmatization (which can be experienced through a low status, stereotype threat and/or low power) seems to lead to failures in self-regulation.
One reason self-regulation can be difficult for stigmatized individuals is that stigma is ego-depleting (Inzlicht, McKay & Aronson, 2006). The term ego-depletion stems from the so-called strength model of Baumeister, Vohs and Tice (2007). This model compares
6 self-control with a muscle and states that people don't possess unlimited self-control. Instead, self-control requires strength and energy and declines with every demanding task one performs. Therefore, at the end of a challenging day, it's possible people don't have any self-control resources left (like muscles are worn-out after an intensive sport match) (Baumeister et al., 2007). There is some discussion about the strength model. Alternative explanations for the negative effects of ego-depletion found in studies are tiredness and a lack of motivation (Hagger, Wood, Stiff & Chatzisarantis, 2010). That being said, it is not the purpose of the present study to determine how ego-depletion works. Primarily relevant is the finding that because coping with stigma requires self-control, which is an exhaustible source, stigmatized individuals get depleted what causes self-regulation problems with unhealthy behaviors as a result (Inzlicht et al., 2006). Figure 1 combines the results of the studies of Richman and Lattanner (2014), Aronson et al. (2013) and Inzlicht et al. (2006) and shows the pathways by which stigmatization can lead to depletion, self-regulation problems and, finally, unhealthy behaviors. This figure will serve as a conceptual theoretical framework for the current research.
Figure 1: Pathways by which stigmatization can lead to unhealthy behaviors
Stigmatization Self-control Depletion
Low status Self-regulation
problems Unhealthy behaviors Stereotype threat Prevention focus
Low power Inhibitory
7 In the research on stigmatization there seems to be given little attention with regard to the implications for health communication. Particularly relevant in the field of health
communication is the concept of regulatory fit, which means that the framing of effective health messages is tailored at the regulatory focus of the individual (Ludolph & Schulz, 2015). In short, a promotion focus is related to eagerness and positive emotions and a prevention focus is related to vigilance and negative emotions (Lin & Johnson, 2015). Research on the effect of
stigmatization on regulatory fit is rare. There are only two contradictory studies in the field of behavioral economics that looked whether depleted individuals are more susceptible to framing effects (i.e. show different responses due to the framing of information (Mandel, 2013)). In an experimental setting Pocheptsova, Amir, Dhar and Baumeister (2009) found that depletion impairs deliberate decision making which leads to more sensitivity towards heuristics such as the use of a reference point. In a similar study, De Haan and Van Veldhuizen (2015) did not find this effect. Both studies, however, did not focus on framing in terms of loss and gain (Rothman & Salovey, 1997). Therefore the present study examined what works best on stigmatized groups: health information that focuses on the negative aspects of the unhealthy behavior or health information that focuses on the positive aspects of the healthy behavior by looking at the effect of depletion on the effectiveness of, and preference for, respectively loss or gain framing.
Using a within-subject design depletion was manipulated in an experimental setting by means of a Stroop task (Stroop, 1935). Subsequently, participants read a loss and gain framed health message about a health goal they chose after which its effectiveness and the preference of the participants was measured. For stigmatization induces a prevention focus Richman and Lattanner (2014) predict that loss framing may be more effective in stigmatized groups. In line with this, the expectation was that depletion leads to more effect of, and a preference for, loss framing.
8 Method
Participants
An a-priori power analysis determined that, in the within-subject design, a minimum number of 66 participants was required for a multivariate analysis of variance with an effect size of .25 and a power of .8. Participants were recruited in two ways. First, participants could enrol for the study via the website of the lab of the Psychology department of the University of Amsterdam. These participants read the invitational letter at the website, and earned participation points or five euro as a reward. Secondly, participants were recruited via the Applied University of Amsterdam. These participants received the invitational letter in a flyer (students) or in an e-mail (employees) and earned a small gift worth five euro as a reward. The invitational letters did not reveal that the study was about stigmatization. There were no criteria for any of the participants. Participants were randomly assigned to a condition by drawing a ticket out of a box upon arrival in the laboratory.
Procedure and materials
The study took place on two locations: 1) a cubicle in the laboratory of the Psychology department of the University of Amsterdam and 2) a room at the Applied University of
Amsterdam. First, participants chose a health goal that was important to them from the national most popular health goals, which are: quit smoking, drink less alcohol, exercise more and eat healthier (Bregman, 2013; Buckley, 2015; RTL Nieuws, 2014). The health messages
participants read later on in the study were about this health goal in order to prevent little effects due to non-applicable information (for example, when a participant who doesn’t drink alcohol reads a message about drinking less alcohol). The participants then filled in questions about personal data, the current state of depletion to serve as a covariate, and the health goal they chose as a baseline measure (survey 1). After this, the participants performed a five minute Stroop task. Next, the participants read a loss framed health message and a gain framed health
9 message in random order. After each message, they answered questions about the effectiveness and their preference (survey 2). Then followed a manipulation check. Finally, the participants received a hardcopy letter as debriefing. In total the study lasted approximately 20 minutes. The used materials will be described below.
Survey 1. The questions about personal data concerned gender, age, country of origin and educational level. For the measurement of the current state of depletion the short version of the State Self-Control Capacity Scale was translated in Dutch (Ciarocco, Twenge, Muraven & Tice, 2004). This version consists of ten questions about tiredness (e.g. “I feel drained”) and willpower (e.g. “I feel like my willpower is gone”) at the moment which have to be answered on a 7-point Likert scale. In the current research the State Self-Control Capacity Scale had a
reliability of α = .84. For the baseline measure the attitude and intention towards the chosen health goal were measured by six attitude items on a 7-point semantic-differential scale (α = .67) and four intention items on a 7-point Likert scale (α = .75). The same items were used to
measure the effect of the health messages. More information regarding these items can be found under ‘Survey 2’.
Manipulation. For the manipulation of depletion two computerized Stroop tasks (Stroop, 1935) were programmed. Previous research used this method to manipulate depletion effectively (for an overview see De Haan & Van Veldhuizen, 2015). Next to the proved
effectiveness, the Stroop task was chosen for the manipulation of depletion because it’s possible to check whether participants actually performed the task in comparison to the “white bear” paradigm that can also be used as a manipulation of depletion. Also, the Stroop task is relatively neutral and objective in contrast to watching a funny or scary video without expressing any emotions (Fischer, Kastenmüller & Asal, 2012).
The Stroop task for the experimental condition, from now on called the depleted condition, showed words of four different colors (blue, green, red and orange) in colors that matched the word (e.g. blue printed in blue) and didn’t matched the word (e.g. blue printed in
10 red). Just like the study of De Haan and Van Veldhuizen (2015, p. 50) in the depleted condition “the font color and color name [were] randomly matched, so that they [were] identical in only approximately 33 percent of all cases”. The Stroop task for the control condition only showed words of colors in colors that matched the word. In both conditions participants were instructed to indicate the font color of each word as soon as possible by pressing on the corresponding key on the keyboard. After pressing a key, the task showed whether the answer was right or wrong and the next word was automatically shown. All participants were able to practice with ten words before the actual Stroop task started. A Stroop task manipulates depletion because people have an urge to immediately read the word the see in front of them which makes it hard to focus on the font color. Focusing on the font color therefore requires energy and self-control strength (Burger, Charness & Lynham, 2011). Both Stroop tasks were programmed to stop after five minutes. According to previous research this should be enough time to manipulate depletion (De Haan & Van Veldhuizen, 2015).
Health messages. Eight health messages were designed based on a pre-test of concept messages that were evaluated by 18 persons. For each health goal two health messages were designed: a loss framed message and a gain framed message. Every loss framed message stated five short negative consequences of the unhealthy behavior of which one possible deadly consequence (cancer or heart diseases), two less severe long term consequences and two short term consequences. Likewise, every gain framed message stated five short positive
consequences of the healthy behavior. Every consequence existed of two sentences, that were build up as followed: “[Unhealthy behavior/health goal] [disadvantage/advantage]: [unhealthy behavior/health goal] [exemplification disadvantage/advantage]”, e.g.: “Drinking alcohol increases the chance of developing different types of cancer: for example, if you drink alcohol, you have a higher risk of colon cancer”. The advantages and disadvantages differed from each other within each health goal (so if an increased chance of cancer was a disadvantage,
11 and disadvantages were as equal as possible. The lay-out of all the messages was the same, with the exception of a picture that suited the goal (a cigarette, an alcoholic beverage, training shoes and an apple).
Survey 2. The questions regarding the effectiveness of the health messages measured attitudes about the health goal and behavioral intentions. For these questions were based upon the Theory of Planned Behavior, guidelines for constructing such questionnaires were strictly followed (Francis et al., 2004). In a similar questionnaire Lee, Cappella, Lerman and Strasser (2012) measured the attitude about smoking abstinence by averaging six attitude items on a 7-point semantic-differential scale: “1 = bad/good, 2 = unenjoyable/enjoyable, 3 = unpleasant/ pleasant, 4 =foolish/wise, 5 = difficult/easy, 6 = harmful/beneficial” (Lee et al., 2012, p. 529) which is a commonly used valid and reliable method (Fishbein & Ajzen, 2010). They measured the intention to abstain from smoking by averaging four intention items on a 4-point Likert scale: “1 = definitely will not, 4 = definitely will” (Lee et al., 2012, p. 529). As an extra check they weighed the intention items by corresponding confidence items. This study followed the example of Lee et al. (2012) by constructing six attitude items and four intention items for each health goal, all in Dutch. In contrast to Lee et al. (2012) intention items were not weighed by corresponding confidence items, for such a measure is not standard and necessary for the purpose of the present study, and were answered on a 7-point Likert scale. The reliabilities of these scales were α = .75 for loss framed attitude, α = .79 for loss framed intention, α = .69 for gain framed attitude, α = .83 for gain framed intention.
Due to the within-subjects design participants evaluated a loss framed health message as well as a gain framed health message. This created the opportunity to examine the preference of the participants which could deviate from the results on effectiveness (e.g. when the loss framed health message is more effective, but participants prefer a gain framed message). Therefore, survey 2 ended with a dichotomous question about which message the participants preferred.
12 Manipulation check. As a manipulation check, all participants answered four questions on a 7-point Likert scale based upon Webb and Sheeran (2003, p. 281-282), that were: “‘How difficult did you find performing the task?’ (very easy–very difficult), ‘How much effort did it take to perform the task?’ (very little effort–a great deal of effort), ‘How strong a desire did you feel to stop working on the task?’ (not very strong–very strong), and ‘How much did you feel you were forcing yourself to work on the task’ (not at all–a great deal)”, all translated in Dutch. This scale had a reliability of α = .76. Lastly, participants were asked if they could guess the purpose of the task and the study.
Debriefing letter. The debriefing letter explained the purpose of the task and the study and provided participants with contact details of the researcher.
Results
Participants
A number of 104 participants completed the study, of which 96 were recruited via the University of Amsterdam and 8 were recruited via the Applied University of Amsterdam. An independent samples t-test showed there were no significant differences between these
participants on the test-variables, thus they were considered as one group. When examining the dataset two outliers were detected in the control condition. These participants scored relatively high on the manipulation check scale, which means they found the easy Stroop task rather difficult and felt depleted afterwards. There was one participant who didn’t answer one question of the manipulation check scale, but – with a high score on all the other questions and most mistakes of all participants in the control condition – would have probably been an outlier if he/she did. Next to this, there was one outlier with relatively negative attitudes and low intentions towards the health goal. For these four participants could bias the results, they were excluded from the study.
13 Seventy-six percent of the 100 remaining participants were women. The average age was 24.69 (SD = 10.13) with a minimum age of 18 and a maximum age of 66. The educational level of most participants was medium (63%) and the majority of the participants was born in the Netherlands (91%) (table 1).
Table 1: Characteristics of participants in the study (N = 100)
Sex % Country of origin % Health goal %
Women 76 Netherlands 91 Exercise more 47
Men 24 Germany 5 Eat more healthy 39
Educational level* % Canada 1 Drinking less alcohol 10
Low 6 China 1 Quit smoking 2
Medium 63 Curacao 1
High 31 Suriname 1
* Based upon the International Standard Classification of Education 2011 (UNESCO, 2012).
Manipulation check
Stroop task. An independent samples t-test showed that the average number of correctly answered items was significantly higher in the control condition (M = 292.04, SD = 26.81) than in the depleted condition (M = 235.78, SD = 37.20), t(98) = 8.68, p < .01. There wasn’t a significant difference between the two conditions in incorrect answers on the Stroop task. Participants in the control condition made an almost equal amount of mistakes as the depleted condition, but answered more questions in total. The average percentage of incorrect answers in the control condition (M = 4.75, SD = 2.82) did differ significantly from the depleted condition (M = 6.76, SD = 3.85), t(89.87) = -2.98, p < .01. Moreover, the average score on the four questions that served as a manipulation check was significantly higher in the depleted condition (M = 3.16, SD = 1.22) than in the control condition (M = 2.51, SD = .83), t(86.12) = -3.07,
14
p < .01. This led to the conclusion that the depletion manipulation was successful. None of the
participants could guess the purpose of the task and the study.
Health messages. The health messages had a significant effect on the attitudes and intentions of the participants. A paired samples t-test showed that the attitude scores after a loss framed message were significantly higher than the baseline attitude scores, t(99) = -4.88,
p < .01. The same effect showed up for the intention scores after a loss framed health message, t(99) = -5.71, p < .01, for the attitude scores after a gain framed health message, t(99) = -6.82, p < .01 and for the intention scores after a gain framed health message, t(99) = -6.57, p < .01
(table 2).
Table 2: Average attitude and intention scores (M) and standard deviation (SD) before and after reading loss and gain framed health messages (N = 100)
Baseline Loss framed Gain framed
M SD M SD M SD
Attitude 5.47 .72 5.70* .79 5.74* .70
Intention 5.49 .94 5.76* .85 5.81* .96
* Significantly higher than baseline, p < .01.
Main analysis
The assumption of (multivariate) normality was checked by examining the univariate normality for the baseline measures and the outcome variables. All measures were significantly skewed (D(100) = 0.10, p < .01 for baseline attitude, D(100) = 0.09, p < .05 for baseline
intention, D(100) = .10, p < .01 for loss framed attitude, D(100) = .13, p < .01 for gain framed attitude, D(100) = .11, p < .01 for loss framed intention and D(100) = .11, p < .01 for gain framed intention). Since the participants chose a health goal that was important to them, most of them had a positive attitude about the health goal and a high intention to obtain it. Besides this,
15 all assumptions for parametric testing (independence of the covariate, homogeneity of regression slopes, homogeneity of covariance matrices (Field, 2009)) were met.
Using Pillai’s trace, a MANCOVA showed a borderline significant effect of depletion on attitude and intention whilst controlling for current depletion, F(4,94) = 2.34, p = 0.06.
However, separate univariate ANOVA’s on the outcome variables showed no significant effects of depletion on attitude and intention (F(1,97) = .98, p = 0.32 for loss framed attitude,
F(1,97) = .28, p = 0.60 for gain framed attitude, F(1,97) = 2.42, p = 0.12 for loss framed
intention and F(1,97) = .12, p = 0.73 for gain framed intention). Within the conditions paired samples t-tests revealed that there was a significant difference between the loss framed intention (M = 5.63, SD = .90) and the gain framed intention (M = 5.78, SD = .99) in the control
condition, t(49) = -3.13, p < .01. There were no significant differences in attitude in the control condition or in attitude or intention in the depleted condition.
Although participants from the depleted condition preferred loss framed health messages more often (graph 1), a Pearson’s chi-square analysis showed there wasn’t a significant
association between the condition and the preference of participants for a certain framing,
X2 (1) = 2.55, p = .11. In both conditions, gain framed health messages were preferred by most
participants. In conclusion, there’s no significant effect of depletion on the effectiveness of, and preference for, loss or gain framing.
16
Graph 1: Preference for loss and gain framed health messages in both conditions (N = 100)
Explorative analysis
To examine the absence of a significant effect some explorative analysis have been conducted. First, the two graphs below visualize the results of the main analysis. As can be seen in the depleted condition the attitude and intention towards the health goal increase after a loss framed health message, but increase less or stay the same after a gain framed health message. In the control condition the attitude and intention towards the health goal also increase after a loss framed health message, but increase more after a gain framed health message. These differences lean towards the hypothesis, but are too little to be significant.
17
Graph 2: Attitude and intention at baseline, after a loss framed health message and after a gain framed health message for both conditions (N = 100)
Secondly, the notion came up that the absence of significant results might be caused by the within-subject design. It’s possible that after reading the first health message, the second health message had little “extra” effect. To examine this the second measures were deleted from the dataset.
Between conditions, two independent samples t-tests (one for the participants who read a loss framed health message first and one for participants who read a gain framed health message first) in this new dataset showed no significant differences between the attitude and intention of depleted and non-depleted participants. So, the loss framed health messages as well as the gain framed health messages weren’t more effective in the depleted condition than in the control condition (table 3).
18
Table 3: Condition, number of participants (N), average score on attitude and intention (M) standard deviation (SD) and p-value of the independent samples t-test (p) for participants who read a loss framed health message and for participants who read a gain framed health message
Condition N M SD p
Loss framed health message
Attitude Control condition 33 5.65 .72
Depleted condition 20 5.71 1.00 .79
Intention Control condition 33 5.53 1.00
Depleted condition 20 5.74 .83 .44
Gain framed health message
Attitude Control condition 17 5.60 .60
Depleted condition 30 5.71 .73 .59
Intention Control condition 17 5.82 .75
Depleted condition 30 5.88 .83 .81
Between health messages, two independent samples t-tests (one for the depleted condition and one for the control condition) in this new dataset showed no significant
differences between the attitude and intention after a loss framed health message and after a gain framed health message. So, in both conditions the health messages were equally effective
(table 4). It seems that when a between-subjects design replaces the within-subject design, depletion still has no effect on the effectiveness of loss or gain framing.
19
Table 4: Health message, number of participants (N), average score on attitude and intention (M), standard deviation (SD) and p-value of the independent samples t-test (p) for the control condition and the depleted condition
Health message N M SD p
Control condition
Attitude Loss framed 33 5.65 .72
Gain framed 17 5.60 .60 .81
Intention Loss framed 33 5.53 1.00
Gain framed 17 5.82 .75 .29
Depleted condition
Attitude Loss framed 20 5.71 1.00
Gain framed 30 5.71 .73 .99
Intention Loss framed 20 5.74 .83
Gain framed 30 5.88 .83 .55
Discussion
This study examined the regulatory fit of health messages for stigmatized groups by looking at the effect of depletion on the effectiveness of, and preference for, loss or gain
framing. Assuming that depletion is a valid manipulation of stigmatization, the results show that there are no differences in effectiveness between stigmatized and non-stigmatized individuals. Loss and gain framed health messages are equally effective in stigmatized groups. In
non-stigmatized groups gain framing has more effect on the intention towards the health behavior than loss framing. Although not significant, explorative analysis demonstrated that a reversed effect might be true for stigmatized individuals: loss framed health messages seem to increase their intention to perform the health behavior slightly more than gain framed health
20 messages. Conversely, just like non-stigmatized individuals, stigmatized individuals prefer a gain framed health message.
Looking at the conceptual theoretical framework in figure 1 (p. 6) an explanation for these results could be that stigmatization leads to inhibitory processes and a prevention focus at first (Aronson et al., 2013; Richman & Lattanner, 2014), but the depletion that follows
evaporates this effect, which is in line with the literature that states depletion impairs inhibition (Baumeister, 2014). One then might expect that depletion leads to a promotion focus, because it is followed by disinhibitory tendencies (Richman & Lattanner, 2014) that are commonly
associated with an approach state of mind (see for example Keltner et al., 2003). If this was the case, gain framing should have had more effect on stigmatized individuals in this study as well, but the opposite seems true. We therefore propose that stigmatization on the short term leads to inhibition and a prevention focus and on the long term leads to disinhibition and might lead a promotion focus, but the in-between-stage of depletion in the process of stigmatization leads to an indistinct regulatory focus. Maybe during this stage there’s a transition going on from inhibition to disinhibition which results in a temporary insensitivity for framing in terms of loss and gain. This could mean that depleted individuals indeed aren’t more susceptible to framing effects than non-depleted individuals, like De Haan and Van Veldhuizen (2015) found. Figure 2 shows what this might look like.
Figure 2: Pathways by which stigmatization can lead to a certain regulatory focus
Stigmatization Inhibition Depletion Disinhibition
Prevention focus Indistinct regulatory focus Promotion focus
21 The fact that a promotion focus seems to help stigmatized individuals (Alter et al., 2010; Trawalter & Richeson, 2006) could be an explanation for their preference. People who are stigmatized and/or depleted might be deliberately looking for positive information that motivates and encourages them and doesn't take too much energy to process.
The implications of stigmatization for health communication are that although the initial prevention focus of stigmatized individuals pleads for loss framing, the eventual ego-depleted state calls for a different approach. In this approach a combination of loss and gain framed messages seems the best decision so that the possible effects of loss framing as well as the preference for gain framing is being met. By doing this it’s feasible that the chance at a health message that fits the regulatory focus increases.
Limitations
An important limitation of the current research is that the sample is “WEIRD”, which means the participants are mostly western, educated and from industrialized, rich and
democratic countries (Henrich, Heine & Norenzayan, 2010). In addition, there’s no equal distribution of men and women in the sample. Therefore, the results can’t be generalized to groups that aren’t “WEIRD” and might not be applicable to men. Next to this, the distribution was skewed due to highly motivated participants. The reason for this is probably the earlier mentioned fact that participants got to choose a health goal that was important to them in order to prevent little effects due to irrelevant information. This logic, however, can also be reversed: maybe there weren’t any large effects because the participants were already motivated. Still, the health messages were effective which means the motivation of the participants had not reached its ceiling yet. Another reason for the skewed-distribution could be the fact that there was no random sampling, but self-selection. The participants that signed up for the study possibly were more interested in health behavior than average persons. The measures therefore resemble attitudes and intentions about self-chosen health behaviors of motivated individuals in real life,
22 but can’t be generalized to behaviors that are more forced than self-chosen or the less motivated (e.g. someone who doesn’t want to quit smoking).
Regarding the Stroop task two limitations have to be addressed. First, unexpectedly, depleted participants only made more mistakes than non-depleted participants relatively speaking. An explanation for this could be that the participants in the control condition got bored and made mistakes due to a lack of concentration, whilst participants in the depleted condition were more focused. The finding that depleted participants found the Stroop task harder than the non-depleted participants shows the manipulation still was successful. Secondly, there was no check on color blindness. With the colors being blue, green, red and orange this could be a problem. However, after they completed the study, the researcher asked all participants how the Stroop task went. One may expect that color blind participants would have pointed out their disability. Even if they didn’t, approximately 8% of all western men and 0.4% of all women are color blind (www.kleurenblindheid.nl). With 25% percent male and 75% female participants the chance of bias in this study is very small.
The use of self-designed health messages also has its limitations. In the pre-test of the health messages a lot of effort was put in to designing them as similar as possible. The great similarities between all the health messages, increased the chance that the chosen health goal had any effect on the outcomes. Nevertheless, such messages are always subject to the
interpretation of the participant. What’s interpreted as a severe consequence for one participant, can be interpreted as mild by another and vice versa. Yet, in real life these differences are also present, therefore this shouldn’t be problematic.
As mentioned in the results, the design of the study might be a limitation as well, for the second measures are possibly based on the information in both health messages instead of the information in the second health message. Although this problem is minimized by randomizing the order participants read the health messages in, one could argue that a between-subjects design would have been better. On the other hand, if the hypothesis was true, depleted people
23 would pay extra attention to loss framed health messages regardless of what they read before this and one may expect an effect of a second health message nonetheless. The explorative analysis that shows there also weren’t any effects in a between-subjects design confirms this. The study was, however, underpowered for such a design. Besides, by deleting the second measures, the distribution of participants amongst conditions wasn’t equal anymore and the number of participants per cell decreased, which makes these results less reliable. Although the sample size should be sufficient for the used design, it’s possible a replication of this study with a larger sample and higher power results in significant outcomes.
Also vital is that this study makes the assumption that depletion is a valid manipulation of stigmatization, whereas depletion due to stigmatization could be different than the depletion that is manipulated by a Stroop task. After someone experienced (one of) the different concepts of stigmatization and then gets depleted, he or she might be in a more negative state of mind then some who never experienced any status loss, stereotype threat or low power. Moreover, several studies suggest that depletion might play a smaller role in self-regulation problems than assumed (see Carter & McCullough, 2014; De Haan and Van Veldhuizen, 2015; Xu et al., 2014 ). If this is true, the conceptual theoretical framework of this study needs to be revised.
Lastly, this study makes assumptions about the regulatory focus based on the measurement of effectiveness of loss and gain framing. It might have been better if the regulatory focus was actually measured, but the indirect approach was chosen in order to prevent such a direct measurement having an effect on the effectiveness of, and preference for, the health messages.
Future directions
Based on the limitations of this study, future research should use a larger, more representative sample that consists of motivated as well as unmotivated persons. Researchers have to let go of the idea of a self-chosen health goal and just rule-out health behaviors that
24 really aren’t applicable instead. Ideally, they use a 2x2 between-subjects design and measure the regulatory focus after the effects of loss and gain framing. They are also ought to manipulate stigmatization in multiple ways. Future research could, for example, manipulate stigmatization via status loss, stereotype threat and/or low power and measure or manipulate depletion next. More insight into the short and long term effects of stigmatization and the interaction between inhibition and disinhibition in stigmatized individuals is necessary. Next to this, the idea of different stages of stigmatization as visualized in figure 2 (p. 20) could be a subject of future research. Finally, it’s very important to examine the implications of stigmatization within actual stigmatized groups, for a lab-study could never truly simulate stigmatization in real life.
Conclusions
With the finding that stigmatized individuals don’t seem to have a distinct regulatory focus in the stage of depletion and the advice therefore to combine loss and gain framing, this study provides a first insight in the implications of stigmatization for health communication, which are yet to be confirmed by future research. This study also offers more information about the mechanisms underlying the unhealthy behavior of stigmatized individuals. The finding that depletion seems to change the regulatory focus of stigmatized individuals at least partially, might explain the switch a stigmatized person makes from a prevention focus to behaviors that are a risk for the health. The study is not only of value to the research on stigmatization. Research on ego-depletion, regulatory focus and framing could use the results as well for a broader understanding of the used theories and concepts. Although this study took place in the lab, it's also relevant for the practice of health communication towards stigmatized group. Interventions can adjust their framing based upon the results of this study and communicate more effective health messages.
25 Acknowledgements
I want to thank dr. Alvin Westmaas for his valuable feedback and ir. Dennis Pullens for his assistance in programming the Stroop Task.
References
Alter, A. L., Aronson, J., Darley, J. M., Rodriguez, C., & Ruble, D. N. (2010). Rising to the threat: Reducing stereotype threat by reframing the threat as a challenge. Journal of
Experimental Social Psychology, 46, 166-171.
Anderson, C., & Kilduff, G. J. (2009). The pursuit of status in social groups. Current
Directions in Psychological Science, 18(5), 295-298.
Aronson, J., Burgess, D., Phelan, S. M., & Juarez, L. (2013). Unhealthy interactions: The role of stereotype threat in health disparities. American Journal of Public Health, 103(1), 50-56.
Baumeister, R. F. (2014). Self-regulation, ego depletion, and inhibition. Neuropsychologia,
65, 313-319.
Baumeister, R. F., Vohs, K. D., & Tice, D. M. (2007). The strength model of self-control.
Current Directions in Psychological Science, 16(6), 351-355.
Bernhardt, J. M. (2004). Communication at the core of effective public health. American
Journal of Public Health, 94(12), 2051-2053.
Bregman, R. (2013, januari 5). Goede voornemens duren vaak nog geen week. De Volkskrant. Opgehaald 25 januari 2016, van
http://www.volkskrant.nl/wetenschap/goede-voornemens-duren-vaak-nog-geen-week~a3372352/
Buckley, P. (2015, december 29). De populairste goede voornemens voor 2016. Metro
Lifestyle. Opgehaald 25 januari 2016, van
http://www.metronieuws.nl/lifestyle/life/2015/12/de-populairste-goede-voornemens-voor-2016
26 Burger, N., Charness, G., & Lynham, J. (2011). Field and online experiments on self-control.
Journal of Economic Behavior & Organization, 77, 393-404.
Carter, E. C., & McCullough, M.E. (2014). Publication bias and the limited strength model of self-control: Has the evidence for ego depletion been overestimated? Frontiers in
Psycholy, 5, 1-11.
Ciarocco, N., Twenge, J., Muraven, M., & Tice, D. (2004). The state self-control capacity scale:
Reliability, validity, and correlations with physical and psychological stress.
Unpublished manuscript, San Diego State University.
De Haan, T., & Van Veldhuizen, R. (2015). Willpower depletion and framing effects. Journal
of Economic Behavior & Organization, 117, 47–61.
Field, A. (2009). Discovering Statistics Using SPSS. London: Sage.
Fischer, P., Kastenmüller, A., & Asal, K. (2012). Ego depletion increases risk-taking. The
Journal of Social Psychology, 152(5), 623-638.
Fishbein, M., & Ajzen, I. (2010). Predicting and changing behavior: The reasoned action
approach. New York: Psychology Press.
Francis, J. J., Eccles, M. P., Johnston, M., Walker, A., Grimshaw, J., Foy, R.,...Bonetti, D. (2004). Constructing questionnaires based on the theory of planned behavior: A manual
for health services researchers. Newcastle upon Tyne: Centre for Health Services
Research.
Guinote, A. (2007). Power affects basic cognition: Increased attentional inhibition and flexibility. Journal of Experimental Social Psychology, 43, 685-697.
Hagger, M. S., Wood, C., Stiff, C., & Chatzisarantis, L. N. D. (2010). Ego depletion and the strength model of self-control: A meta-analysis. Psychological Bulletin, 136(4), 495-525.
Henrich, J., Heine, S. J., & Norenzayan, A. (2010). Most people are not WEIRD. Nature,
27 Inzlicht, M., McKay, L., & Aronson, J. (2008). Stigma as ego depletion: How being the target
of prejudice affects self-control. Psychological Science, 17(3), 262-269.
Keltner, D., Gruenfeld, D. H., & Anderson, C. (2003). Power, approach, and inhibition.
Psychological Review, 110, 265-284.
Lee, S., Cappella, J. N., Lerman, C., & Strasser, A. A. (2012). Effects of smoking cues and argument strength of antismoking advertisements on former smokers’ self-efficacy, attitude, and intention to refrain from smoking. Nicotine & Tobacco Research, 15(2), 527-533.
Lin, S., & Johnson, R. E. (2015). A suggestion to improve a day keeps your depletion away: Examining promotive and prohibitive voice behaviors within a regulatory focus and ego depletion framework. Journal of Applied Psychology, 100(5), 1381-1397.
Link, B. G., & Phelan, J. C. (2001). Conceptualizing stigma. Annual Review of Sociology, 27, 363-85.
Ludolph, R., & Schulz, P. J. (2015). Does regulatory fit lead to more effective health communication? A systematic review. Social Science & Medicine, 128, 142-150. Mandel, D. R. (2013). Do framing effects reveal irrational choice? Journal of Experimental
Psychology: General, 143(3), 1185-1198.
Pascoe, E. A., & Richman, L. S. (2009). Perceived discrimination and health: A meta-analytic review. Psychological Bulletin, 135(4), 531-554.
Phelan, J. C., Lucas, J. W., Ridgeway, C. L., & Taylor, C. J. (2014). Stigma, status, and population health. Social Science & Medicine, 103, 15-23.
Pocheptsova, A., Amir, O., Dhar, R., & Baumeister, R. F. (2009). Deciding without resources: Resource depletion and choice in context. Journal of Marketing Research, 46(3), 344-355.
Quinn, D. M., & Chaudoir, S. R. (2009). Living with a concealable stigmatized identity: The impact of anticipated stigma, centrality, salience, and cultural stigma on psychological
28 distress and health. Journal of Personality and Social Psychology, 97(4), 634-651. Richman, L. S., &Lattanner, M. R. (2014). Self-regulatory processes underlying structural
stigma and health. Social Science & Medicine, 103, 94-100.
Rothman, A. J., & Salovey, P. (1997). Shaping perceptions to motivate healthy behavior: The role of message framing. Psychological Bulletin, 121, 3-19.
RTL Nieuws. (2014). Opgehaald 25 januari 2016, van
http://www.rtlnieuws.nl/nieuws/binnenland/goede-voornemens-dit-zijn-de-meest-gemaakte-voor-2015
Seibt, B., & Förster, J. (2004). Stereotype threat and performance: How self-stereotypes influence processing by inducing regulatory foci. Journal of Personality and Social
Psychology, 87(1), 38-56.
Ståhl, T., Van Laar, C., & Ellemers, N. (2012). The role of prevention focus under stereotype threat: Initial cognitive mobilization is followed by depletion. Journal of Personality
and Social Psychology, 102(6), 1239-1251.
Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental
Social Psychology, 18(6), 643-662.
Trawalter, S., & Richeson, J. A. (2006). Regulatory focus and executive function after interracial interactions. Journal of Experimental Social Psychology, 42, 406-412. UNESCO (2012). International Standard Classification of Education: ISCED 2011. Montreal:
UNESCO Institute for Statistics.
Viruell-Fuentes, E. A., Miranda, P. Y., & Abdulrahim, S. (2012). More than culture: Structural racism, intersectionality theory, and immigrant health. Social Science &
Medicine, 75, 2099-2106.
Webb, T. L., & Sheeran, P. (2003). Can implementation intentions help to overcome ego-depletion? Journal of Experimental Social Psychology, 39, 279-286.
29 Xu, X., Demos, K. E., Leahey, T. M., Hart, C. N., Trautvetter, J., Coward, P., Middleton, K. R.,
