Improving Implementation Intentions
Examining the Effectiveness of Reasoning Implementation and Implementation Intentions on Exercising Behaviour and Outcomes of the Protection Motivation Theory
A master thesis by Clio Carlotta Rosebrock
Health Communication
Supervisor: Dr. Prof. Bas van den Putte Student ID: 11817399
Amsterdam, 29.06.2018
Abstract
The current study had two objectives. First, direct effects of reasoning implementation intentions compared to implementation intentions and the formulation of goal intentions on exercising behaviour were explored. Second, it was examined whether reasoning
implementation intentions were more effective to increase goal-intention related cognitions: Perceived severity, perceived susceptibility, perceived self-efficacy and perceived response-efficacy. Data was generated by a longitudinal online survey over the course of two weeks. Even if there was a small albeit significant growth in exercising behaviour over time, there were no significant differences between groups. In this way, reasoning implementation intentions were not more likely to increase outcome variables than II’s or the formulation of goal intentions were. Neither were II’s more effective than the formulation of goal intentions.
It is likely that the results were influenced by sample characteristics as most participants who indicated to have a positive goal intention to exercise already exercised.
Table of Contents
Introduction ... 3
Theory ... 6
General Effects of II’s ... 6
Extending II’s by combining motivational and volitional elements ... 7
Protection Motivation Theory and RII’s ... 10
Hypotheses ... 11
Outcome variable: Exercising behaviour ... 12
Outcome variables: PMT elements... 12
Outcome variable: Goal intentions ... 12
Method ... 13
Sample ... 13
Design and Procedures ... 14
Measures ... 15 Independent Variables ... 15 Dependent Variables ... 16 Other Measures ... 18 Data-Analysis ... 19 Results ... 20 Discussion ... 22 Limitations ... 25 Future Research ... 27 Conclusion ... 29 Literature ... 30 Appendix ... 34
A1. Survey Time 1 English and German ... 34
A1A. Time 1 English... 34
A1B. Time 1 German ... 45
A2. Time 2 English and German ... 58
A2A. Time 2 English... 58
A2B. Time 2 German ... 64
B. Tables ... 72
B1. Factor and Reliability Analyses ... 72
B2. Randomization Checks ... 72
Introduction
Behavioural risks have emerged to be the most frequent sources of death and disabilities in the Netherlands (for Health Metrics and Evaluation IHME, 2016). Dietary risks including low physical activity are one of the major behavioural factors contributing to those negative outcomes. According to the World Health Organization (2018), there is a strong correlation between overweight and an inactive lifestyle. By means of example, overweight individuals appear to be twice as likely to develop coronary heart disease as non-overweight individuals (Milne, Orbell & Sheeran, 2002). However, large parts of western society do not meet exercising criteria. Recent numbers by Renew Bariatrics (2017) showed that 19.4% of Dutch citizens were obese1, not including the number of overweight individuals which is assumed to be much higher. This assumption is supported by the IHME (2016), which stated that 49.4% of Dutch citizens above 18 years were overweight or obese in 2014. Accordingly, enhancing exercising behaviour could significantly help to reduce diseases and disabilities.
However, one problem that comes with many behavioural changes is that often, individuals cannot act on their goal intentions. The literature refers to this problem as intention-behaviour gap (Gollwitzer & Sheeran, 2006), which has gotten plenty of scientific attentions so far (e.g., De Bruijn, Nguyen, Rhodes & van Osch, 2017; Papies & Hamstra, 2010). A seemingly promising method to close this gap is the use of implementation intentions (II’s) which are commonly used to change great varieties of health-behaviours
(Hagger & Luszczynska, 2014). II’s are if-then plans that enhance individuals’ self-regulative capacities by specifying goal intentions (e.g., If I come home from work on Thursday
afternoon, then I will go to the gym for half an hour). There is scientific evidence for II’s including a meta-analysis (Gollwitzer & Sheeran, 2006) indicating a moderate to large effect on health promoting behaviours containing exercising. However, research results are mixed. There are several studies which found no effect (e.g., deVet, Oenema, Sheeran and Brug
1
(2009), thus not supporting II’s as method for behaviour change. As a positive goal intention to change is a crucial condition for II’s to work, research tries to enhance effects through a combination of II’s with goal intention enhancing methods (Milne, Orbell & Sheeran., 2002).
However, such information is often based on what researchers consider as important and not the participants themselves (Canning & Harackiewicz, 2015) and demands more resources of participant and researcher than II’s alone. Those limitations could be diminished by using reasoning II’s (RII’s) a novel form of II’s based on the concept of self-persuasion (Prestwich, Ayres and Lawton, 2008). RII’s connect the specific plan (II’s) directly with an individual’s
major motivation behind performing it. In this way, individuals are not passively receiving persuasive information but are rather actively generating it by mentally making available their major motivation to exercise. In this way, individuals’ estimates of negative consequences of not exercising as well as coping beliefs are supposed to be enhanced by RII’s. According to the Protection Motivation Theory (PMT; Rogers,1983), threatening individuals by presenting negative consequences of not performing a particular healthy behaviour, followed by coping advices enhance positive goal intentions to engage in that behaviour. As II’s seem to only work in combination with goal intention enhancing methods (Prestwich, Lawton & Connor., 2003), RII’s might function as such a combined method. Further, RII’s could help overcome limitations of previous research as they are not depending on huge resources and profit from the benefits of self-persuasion. RII’s as optimized form of II’s could enhance exercising behaviour and significantly reduce diseases and disabilities due to behavioural risks. Accordingly, the current study aims to help health professionals to optimize exercising strategies. As scientific evidence for RII’s is scarce with only two studies experimentally examining their effects (Prestwich et al., 2008; Koka, 2016) the current study adds to existing research by investigating RII’s relative strength compared to II’s. The current study aims to extend existing research as it will be explicitly determined in which way RII’s might
influence PMT variables, which in turn are supposed to enhance individuals’ goal intentions
to exercise. Based on this argumentation, following research questions (R1-R3) were created.
RQ1. Compared to II’s and a control group, to what extent can RII’s enhance exercising behaviour?
RQ2. Compared to II’s and a control group, to what extent can RII’s enhance PMT variables? RQ3. Compared to II’s and a control group, to what extent can RII’s enhance goal intentions to exercise?
In the following, the intention-behaviour gap will be presented, followed by research about general positive effects of II’s. Subsequently, limitations of II’s will be explained in detail and the combination of volitional (II’s) and motivational (enhance goal intentions)
methods will be presented as possible solution (e.g., Milne et al., 2002). However, as research based on this approach contains several limitations, RII’s as novel method will be discussed. Finally, the PMT will be presented and hypotheses are formulated.
Theory
General Effects of II’s
According to many health models including the Theory of Planned Behavior (TPB; Ajzen, 1991) and the Protection Motivation Theory (PMT; Rogers, 1983) a positive intention is regarded as main predictor for behaviour change. In health communication, to such an intention is often referred to as goal intention (Gollwitzer & Sheeran, 2006). Goal intention describes a motivational state in which the individual is willing to achieve a specific goal. An example of a goal intention is: I intend to exercise. However, based on recent theoretical and experimental evidence, there appears to be an intention-behaviour gap. This gap describes that motivating individuals to change does not equal actual behaviour change (e.g., de Bruijn et al., 2017; Sheeran, 2002). It is clearly presented by Sheeran (2002) who conducted a meta-analysis of meta-analyses, showing that significant changes in goal intentions at the most resulted in small to medium behaviour changes. The intention-behaviour gap seems to apply to the context of exercising. Results of a recent meta-analysis showed, that only 54% of individuals who wanted to exercise managed to act accordingly (Rhodes & de Bruijn, 2013). According to Sheeran (2002), the intention-behaviour gap is influenced by several factors. In this context, Sheeran particularly (2002) stressed the type of goal intention. More specifically, a particular form of goal intentions, called implementation intentions (II’s) seems to be much more effective in changing behaviours. II’s specify goal intentions and add a context
specifying the behaviour that should be changed (e.g., going to the gym), the time (e.g., five PM on Tuesday afternoon) and the duration of the desired behaviour (e.g., for half an hour). These elements are then combined into one simple plan (e.g., If I come home at five PM on Tuesday afternoon after work, then I will go to the gym for twenty minutes). According to deBruijn et al. (2017), two psychological mechanisms are responsible for the success of II’s. The first mechanism refers to critical cues (if I come home at five on Tuesday after work)
being activated by the specificity of the context. The second mechanism describes a growth in strength of the response towards that cue by directly linking it to the desired behaviour (then I will go to the gym for twenty minutes). II’s are cheap and simple to utilize in health
interventions. More cost-effective than other methods, e.g. counselling, II’s might be used to reach large groups (Prestwich, Perugini & Hurling., 2009). The added value of II’s compared to goal intentions is supported by meta-analytical evidence from 94 empirical studies showing a moderate to large effect of II’s for a diversity of health-related behaviours (d = .52, k = 94; Gollwitzer & Sheeran, 2006).
Extending II’s by combining motivational and volitional elements
Although there is strong evidence in favour of II’s, some empirical studies failed to demonstrate significant effects. For example, Arbour and Martin-Ginis (2004) found significant II effects on self-efficacy of participants to exercise (two times a week) but no behaviour change. Another study by deVet et al., (2009) investigated the impact of II’s on physical activity without finding significant effects of II’s compared to a control condition.
Further, Jackson et al. (2005) conducted an experimental study of the effect of II’s on fruit intake but could not find significant differences between II- and control groups. As II’s specify goal intentions, a positive goal intention to engage in the desired behaviour seems to be a necessary pre-condition (Prestwich & Kellar, 2014). Accordingly, the stronger positive goal intentions are, the more likely II’s are to be successful. However, II’s are not a
motivational method (Milne et al., 2002). More specifically, they are not likely to enhance goal intentions. This argumentation is based on the Model of Action Phases by Heckhausen and Gollwitzer (1987) which differentiates between motivational and volitional elements. Motivation, which equals positive goal intentions, is created by an estimation of
dis/advantages of a specific behaviour (Prestwich, Lawton & Conner, 2003). By contrast, II’s are rather volitional by nature (Milne et al., 2002). Volitional means that the decision to
change behaviour is taken and the individual searches for strategies to implement its goal intention (Heckhausen & Gollwitzer, 1987). Through specifying goal intentions, II’s provide
individuals with self-regulative structures (Milne et al., 2002). Prestwich et al. (2003)
provided evidence for the synergic effects of combining motivational and volitional elements by showing that the combination of both methods had larger effects on exercising behaviour than each method in isolation even when participants already had strong positive goal intentions to exercise (M = 6, 7-point Likert scale). This assumption underlies a number of (experimental) studies showing positive effects of a combination of motivational and
volitional methods with regard to health behaviours (e.g., Caudwell, Mullan & Hagger, 2016; McGowan et al., 2013; Milne et al., 2002). However, within the quoted studies, there remain three relevant limitations.
The first limitation concerns conditions included and variables measured in previous research. Milne et al., (2002) did not include a group which only formed II’s. Subsequently,
there was no direct comparison possible between II effects and the combination of II’s with motivational methods. Prestwich et al., (2003) did include an II group but did not explicitly measure goal intention as outcome variable. Accordingly, Prestwich et al., (2003) could not conclude that their motivational intervention was successful in enhancing positive goal intentions. Further, none of the quoted studies used control groups that received intervention-related information. According to Jackson et al. (2005) control groups who are not reminded of the desired behaviour could bias results. Although the cited studies mostly included control groups, those groups did not receive any exercising-related information. For example, Milne et al., (2002) included a control group that needed to read pages of a novel instead of
intervention-related information. In this way, there is no certainty whether the mere mentioning of exercising could have influenced participants.
Secondly, the methods used in prior research (Mc Gowan et al., 2013; Milne et al., 2002; Prestwich et al., 2003) were cognitive resources and time demanding for participants as
well as researchers. For instance, McGowan et al. (2013) arranged several face-to-face meetings, all lasting around one hour. This could be a relevant threat to the external validity since participants might not be motivated to invest high amounts of cognitive resources in a real-life scenario. This notion is based on a view of individuals as economic minded who generally invest the least possible amount of (cognitive) effort (Lang, 2000). Further, such extensive methods diminish a major advantage of II’s: Combining II’s with personal
counselling would not result in the cheap and easy-to-apply method they could be.
Thirdly, in the quoted studies motivation- enhancing information was based on what the authors deemed important. However, research has shown that it seems to be more effective letting individuals decide by themselves why they want to change (Canning & Harackiewicz, 2015). This process is called self-persuasion and places the individual in an active role rather than being a mere receiver of convincing information (Tiro et al., 2016). The power of self-persuasion strategies received empirical evidence in health communication (Canning & Harackiewicz, 2015; Ryan & Deci, 2000; Tiro et al., 2016). Two mechanisms seem to underlie the concept of self-persuasion: Choice and depth of processing (Tiro et al., 2016). Firstly, individuals have the opportunity to choose the most convincing arguments. This is in line with the self-determination theory which states that providing options predicts motivation for behaviour change (Ryan & Deci, 2000). The second mechanism is that
individuals are more likely to invest cognitive resources for self-generated content (Tiro et al., 2016). This leads to a greater probability of the argument to be in top-of –mind awareness which in turn leads to stronger motivational power.
The current study will directly address those limitations of a) conditions and variables measured, b) time and resource intensive methods and c) motivations based on what authors deemed as important. Regarding the first limitation, a control group will be included which motivated to exercise by formulating goal intentions (e.g., I want to enhance my exercising). Subsequently, exercising will be also made mentally accessible for this group. Further, goal
intentions will be explicitly measured at all measurement times. The second and third limitations will be addressed through the use of a particular form of II’s, RII’s which is a novel but poorly investigated method, introduced by Prestwich et al. (2008). Ain of RII’s is to combine motivational and volitional aspects in form of a simple, specific plan to help
individuals to act on their goal intentions. RII’s have the additional advantage of adding the
main motivation of an individual to engage in the desired behaviour to II’s (Prestwich et al., 2008).Hence, they might overcome the second limitation of earlier studies as they do not require a large amount of resources, neither by the sender nor the receiver. To deal with the third limitation, individuals will have to formulate their main motivation to exercise by themselves, based on the advantages of self-persuasion (choice and depth of processing; Tiro et al., 2016). A possible formulation of a RII is “If I come home on Friday afternoon at four O’clock from work, than I will say to myself that I need to exercise for at least 20 minutes because I do not want to die of coronary heart disease”.
Protection Motivation Theory and RII’s
To explore the potential effect of RII’s on individuals’ goal intentions to exercise; they will be embedded into the framework of Protection Motivation Theory (PMT), which focuses on positive goal intentions to change (Rogers, 1983). Specifically, high scores on PMT variables are expected to enhance positive goal intentions to exercise (Prestwich et al., 2008). Therefore, PMT focuses on threat and coping elements as important predictors (Milne, Sheeran & Orbell, 2000). The threat element includes perceived vulnerability and perceived severity. Perceived vulnerability is the degree to which the individual thinks to be affected by the threat whereas perceived severity determines the strength of that threat. Only if both factors are high, the threat is thought to be sufficient to motivate an individual to change (Rogers, 1983). According to the PMT, a goal intention enhancing message further needs to include coping responses to the threat which can be divided into perceived self-efficacy and
perceived response-efficacy. Regarding exercising, perceived response-efficacy describes the degree to which individuals feel that the intended exercising works in preventing the
perceived threat, whereas perceived self-efficacy refers to the perceived ability to exercise (Gaston & Prapavessis, 2014). Compared to an II- intervention, RII’s are expected to enhance all PMT variables which will be discussed in the following. Perceived self-efficacy is the only PMT variable that is expected to be equally influenced by II’s and RII’s.
Hypotheses
By specifying an individuals’ major motivation to exercise in form of RII’s, threat
should be evoked by making accessible the potential harm that could result of not-exercising (e.g., becoming sick). In the current study, it is expected that perceived susceptibility can be affected by RII’s, as individuals choose their own motivation to exercise, meaning that
meeting their goal is highly relevant to them (Tiro et al., 2016). Subsequently, the motivation behind the change is completely individualized leading to the assumption that RII’s positively influence susceptibility, rather than II’s or a formulating positive goal intention alone.
Perceived severity could also be affected by RII’s by making available the potential threat.
Other variables proposed by the PMT are coping with the threat by perceived self- and
response-efficacy. Perceived response-efficacy is expected to be affected by RII’s because the main motivation to exercise is mentally connected to the performance of the behaviour itself. Creating a mental link between exercising and the reason behind it might strengthen
individuals’ beliefs about its effectiveness. Perceived self-efficacy has been examined in the context of II’s. For example, Arbour and Martin Ginis (2004) did conduct an experimental study testing II’s that resulted in higher of perceived self-efficacy because the desired
end-goal seems to be easier to accomplish. However, according to Milne et al., (2002) II’s do not help to change scores in perceived self-efficacy. Therefore, the effect of II’s on perceived self-efficacy remains unclear. As RII’s are expected to enhance PMT variables, it is expected
that RII’s do positively enhance individuals’ goal intentions to exercise but also specify goal
intentions. In this way, it is expected that RII’s provide the advantages of a combination of motivational and volitional methods. The current study will test following hypotheses.
Outcome variable: Exercising behaviour
H1A. Participants formulating II’s are more likely to enhance their exercising (for at least 20 minutes per week) than participants formulating goal intentions.
H1B. Participants formulating RII’s are more likely to enhance their exercising (for at least 20 minutes per week) than participants formulating II’s or goal intentions.
Outcome variables: PMT elements
H2A: Perceived severity, perceived susceptibility and perceived response-efficacy will be enhanced by formulating RII’s rather than II’s or formulating positive goal intentions. H2B: Perceived self-efficacy will be enhanced by formulating RII’s and II’s rather than by formulating positive goal intentions.
Outcome variable: Goal intentions
H3: Goal intentions to exercise will be enhanced for participants who formulated RII’s rather than for the individuals who formulated II’s or positive goal intentions.
Method
Sample
Participants had to be at least 18 years to avoid ethical concerns. Further, they needed to be able to read and understand either German or English as the study was distributed in those languages. Since II’s seem to be only effective if the individual is already motivated to change (Adriaanse et al., 2011), the final sample exclusively consisted of individuals who had a positive goal intention to exercise. Participants were recruited through the use of personal contacts of the researcher, as well as via social media posts. Additionally, several German fitness bloggers were approached to distribute the surveys on their promotional pages. The chance to win an Amazon voucher of 20 Euro was offered as incentive.
A total of296 participants completed the first survey. Participants who indicated to have a low to neutral goal intention at Time1 (n = 79) and drop outs (Time2; n = 89) were excluded from main analyses. Further, participants who did not form II’s or RII’s as intended by researcher were excluded (n = 6). This was either because they provided motivations for their exercising in the II condition, did not indicate their motivations in the RII condition or indicated to only act on their plan under hard to reach conditions (e.g., If I get to find a bike, just moved to new city, and am free after 7pm, then…). The final sample consisted of 122 participants originating from 18 nationalities2 (age range 20-60 years, M = 25.8, SD = 6.6). Most participants were native German, female and students. As Table 1 shows, exercising behaviour of the sample at Time1 was significantly above average, including 18 participants who already exercised more than four times a week. Table1 presents all baseline
characteristics at Time1.
2
During the analyses, participants’ nationality was treated as a three-dimensional factor, differentiating between German, Dutch and other.
Table 1
Sociodemographic characteristics of the sample (N = 122) at Time 1
n (%) Gender Male Female 42 (34.4) 80 (65.6) Nationality German Dutch Other 77 (63.1) 20 (16.4) 25 (20.5) Highest level of education Primary school
High-school University - 43 (35.2) 79 (64.8) Occupation Student Working
Home duties/ not employed Other 84 (68.9) 33 (27.0) 4 (3.3) 1 ( .8) Workload 0-20 hours 21-36 hours 37-42 hours 43-50 hours More than 50 hours
25 (20.5) 44 (36.1) 35 (28.7) 16 (13.1) 2 (1.6) Exercising behaviour I currently don’t exercise
20 minutes once a week 20 minutes twice a week 20 minutes three times a week 20 minutes four times a week 20 minutes > 4 times a week
21 (17.2) 19 (15.6) 20 (16.4) 28 (23.0) 16 (13.1) 18 (14.8)
Design and Procedures
The current experiment was approved by the ethical commission of the University of Amsterdam. Participants were asked to complete two surveys, two weeks apart (Time1, Time2). Informed consent was provided for each survey separately. Participants were
informed that their data would be processed anonymously and that they could withdraw every time they wanted. Data was collected between 22.04.2018 and 14.05.2018. At 09.05.2018 an email containing a link to the second survey was send to all participants who finished the first
survey. Reminders were sent to participants who did not respond, yet. Time1 and Time 2 data of participants was combined for analyses via participants’ their email addresses. The current
study had a longitudinal 3(condition) by 2(time) mixed experimental design, including within- and between subjects measures. Changes in exercising behavior goal intentions and PMT scores were measured within subjects, at Time1 and Time2, as well as between conditions. Data was collected via online surveys created with an online tool to maximize reach and feasibility of the study.
After agreeing informed consent, the first survey (Appendix A1a+A1b) started with some baseline questions (gender, nationality, education, occupation, workload, age, goal intentions to exercise, and exercising behaviour). To avoid bias, it was important to measure goal intentions prior to the intervention. Subsequently, subjects were randomly assigned to two experimental (II; RII) and one control condition (formulation of goal intentions). This was followed by several questions assessing their PMT scores and reasons to exercise.
At the beginning of the second survey (Appendix A2a+b), PMT variables, reasons for exercise, exercising behaviour in the past two weeks and goal intention scores were assessed by the same scales used at Time1. Thereafter, participants were asked if they still remembered the structure of their plan and needed to indicate to what extent they had tried to act on their plan within the last two weeks.
Measures
Independent Variables
Participants in the II condition had to create their II’s step by step. They needed to answer 3 questions specifying a) their exercising behaviour (e.g., swimming), followed by b) the time (e.g., Friday evening after work), and c) the location (e.g., a pool in the
neighbourhood). Finally, participants had to form an II by embedding all elements into one sentence “Now, try to combine the above elements into one statement. One example could be
as follows: If I come home on Friday afternoon after work, then I will go swimming for at least 20 minutes in the Aquadrome swimming pool in my neighborhood. * Please embed your plan into following structure: If.... then...”. Participants were reminded that their exercise should at least last for 20 minutes. This procedure was mainly based upon McGowan et al. (2013) who used this II schema to create healthy feeding habits for parents. The only difference was that in the current study “habits” was replaced by “a specific exercise”. The McGowan et al.’s (2013) method was chosen as this method required participants to write
down their II’s in single steps. This procedure required cognitive resources from participants and hence increased the likelihood that participants would remember their plan at Time2.
Participants in the RII condition received exactly the same information as participants in the II condition, but additionally they had to add their main motivation to perform the chosen exercise. “Now, try to combine the above elements into one statement. One example could be as follows: If I come home on Friday afternoon after work, then I will go swimming for at least 20 minutes in the Aquadrome swimming pool in my neighborhood, because I want to lose weight. *Please embed your plan into following structure: If.... then...because…”.
Finally, pparticipants in the control condition (formulating a goal intention) were confronted with the following statement: “Now, please confirm that you want to exercise /
enhance your exercising for at least 20 minutes per week and try to act on this within the following two weeks. Therefore, key following statement into the textbox: I want to enhance my exercising”. They were then directed to the following question block.
Dependent Variables Goal Intention
Goal intention at Time1 and Time2 was assessed with a four-item scale developed and validated by Plotnikoff and Higginbotham (2002). The items ranged from 1(very applicable)
to 5(not applicable at all) and included items like “How likely or unlikely is that you will get adequate exercise during the next six months?” (Time1 and Time2, α =.71).
PMT Variables
To assess participants PMT scores at Time1 and Time2, the risk behaviour diagnosis (RBD) scale, created and validated by Witte, Cameron, McKeon and Berkowitz (1996) was used. The RBD scale was originally designed to measure another model, the extended parallel process model. This model, however completely overlapped with the PMT variables as they were used in the current study (severity, susceptibility, self-efficacy, response efficacy). A major advantage of the scale is that the items were formulated in a very general way and therefore applicable a wide range of (exercise) behaviors. The RBD is a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). In the context of the current study there was one shortcoming of the RBD as it only contained threat elements. Participants’ motivation to exercise might be achieving positive states rather than avoiding negative ones. Hence, for each item measuring perceived severity, perceived susceptibility and perceived response-efficacy a positive synonym was added.
The final scale of the RBD consisted of 21 items. In total, there were six severity items. Three items examined negative consequences of not exercising (perceived severity negative: Time1, α = .86; Time2, α = .77) e.g., “I believe that the negative consequences of not exercising are severe”. Three items investigated positive consequences of exercising (perceived severity positive: Time 1, α =.633; Time 2, α = .77) e.g., “I believe that the
positive consequences of exercising are desirable”. Perceived susceptibility was measured by six items. Those measuring negative consequences of not exercising (n=3) included items like “No exercising would result in me being likely to experience negative events” (perceived
susceptibility negative: Time 1, α = .86; Time 2, α = .85). Perceived susceptibility items measuring positive consequences of exercising (n = 3) included items like “Exercising would result in me being likely to experience positive events” (perceived susceptibility positive:
Time1, α =.75; Time2, α = .85). Perceived response-efficacy was measured by six items of which three items assessed negative consequences of not exercising e.g., “If I exercise, I am less likely to experience negative events” (perceived response-efficacy negative: Time1 α =
.91; Time2 = .96). Positive consequences of exercising were measured by three items like: “Exercising is effective in achieving positive events happening to me” (Time 1, α =.92; Time
2, α = .96). Perceived self-efficacy consisted of 3 items like: “Exercising is easy to do”
(Perceived self-efficacy: Time 1, α = .56; Time 2, α = .55).Results of the factor and reliability (Cronbach’s α) analyses are presented in Appendix B, Table B1.
Reasons for exercise
The current study added a modified version of the Reasons to Exercise Inventory, which reflects individuals’ motivations to exercise (Davis, Fox, Brewer & Ratusny, 1995). It served as addition to perceived response-efficacy items to check if scores on perceived response-efficacy may be dependent on explicit motivations. Motivations to exercise as proposed by the reasons for exercise inventory were: outward appearance, sexual
attractiveness, weight control, fitness and health, mood improvement and enjoyment. Each motivation was measured by one item like: “Exercising works in mood improvement”. The
items were embedded into 5-point Likert scales ranging from 1(totally agree) to 5 (totally disagree). The overall scale had a high reliability at both times (Time1, α = .97, Time2 α =.98).
Other Measures
At Time 2, participants were additionally asked whether they still remembered the structure of the plan they created two weeks ago: “Two weeks ago you were asked to form a plan to increase your amount of exercising by at least 20 minutes per week. Which form did that plan have (If..then..; If..then..because…; I want to exercise more)?”. Only 19 participants answered this question in a right way (n= 122).
Data-Analysis
Data-analysis was done by SPSS. To prepare the dataset, a chi square test and one way ANOVA’s revealed that gender, exercising behavior, workload, age, education, occupation,
nationality and goal intention were comparably distributed across the three conditions at Time1 (Appendix B, Table B2). Further, it was checked whether baseline characteristics (Time1) significantly influenced relevant outcome variables. In other words: if changes between Time1 and Time2 were significantly influenced by those characteristics. Therefore, RM ANOVA’s of survey language, workload, age, education, occupation, nationality and gender as independent variables and exercising behavior, goal intention, PMT variables or motivations at Time1 as dependent variables were conducted (Appendix B, Table B 3-11). The analyses revealed significant results for survey language on perceived self-efficacy F (1,120) = 4.28, p = .041, η2 = .034 (Table B3), perceived response-efficacy positive F (1,120) = 27.79, p = <.001, η2 = .188 (Table B3) and on reasons for exercise F (1,120) = 70.26, p = <.001, η2 = .369 (Table B 3). Perceived response-efficacy positive F (2,119) = 5.58, p = .005, η2 =
.086 (Table B4) and reasons for exercise F (2,119) = 17.70, p = <.001, η2 = .229 were also significantly influenced by nationality of participants as well as exercising behavior at Time1 F (2,119) = 4.36, p = .015, η2 = .068 (Table B4). Moreover, reasons for exercise were
significantly influenced by education F (1,120) = 11.28, p = .001, η2 = .086 (Table B5) and workload F (1,120) = 2.02, p = .097, η2 = .065 (Table B7). To increase statistical power of the results, those variables were treated as covariates in the main analyses. Striking was that neither correct recall nor the degree to which participants tried to act on their plans (Table B10) significantly influenced any outcome variable at Time1. As perceived self-efficacy, perceived response-efficacy (positive and negative) and reasons for exercise had negative directions, they were recoded before the main analyses.
Since the sample had an unusual high amount of exercising, a frequency analysis was conducted with individuals who were excluded from main analyses because they had a low to
neutral goal intention to exercise at Time1. This analysis revealed that exercising behavior was significantly different for those individuals with 76.5% (n = 68) who exercised at most one time a week.
After preparing the dataset, 10 RM AN(C)OVA’s were conducted as main analyses. Hypotheses were rejected/accepted based on p >.05 levels of significance.
Results
Table 2 presents frequencies and results of RM ANOVA’s at Time1 and Time2 of all outcome variables. It reveals that especially during Time1, participants had high scores on all outcome variables, which was particularly regarding positive consequences of exercising. Table2 indicates that exercising behavior had a small albeit significant increase over time but PMT-related variables (perceived self-efficacy, perceived response-efficacy negative and positive, reasons to exercise) significantly decreased over time.
Table 2
Frequencies and RM AN(C)OVA of Time 1-Time 2 on dependent variables (n = 122)
Time 1 Time 2 RM AN(C)OVA
Range Mean SD Range Mean SD F(1,120) p η2
Exercising Behavior 1-6 3.43 1.66 1-6 3.63 1.56 7.55 .007* .059 Goal Intention 3.33-5 4.03 0.52 2-5 4.01 0.64 0.15 .701 .001 Perceived severity neg.
Perceived severity pos. Perceived susceptibility neg. Perceived susceptibility pos. Perceived self-efficacy Perceived response-eff. neg. Perceived response-eff. pos.
1.33-5 2.67-5 1-5 2-5 1-5 1-5 1-5 3.70 4.48 3.38 4.09 3.18 3.30 3.67 0.93 0.49 0.98 0.63 0.84 1.00 1.01 1-5 2-5 1-5 2-5 1-5 1-5 1-5 3.81 4.42 3.46 4.07 3.08 3.00 2.86 0.95 0.55 0.95 0.70 0.94 1.00 1.22 1.76 1.98 0.91 0.12 0.58 4.79 47.97 .188 .161 .342 .731 .026* .031* <.001* .014 .016 .007 .001 .041 .038 .287 Reasons to exercise 1-5 3.67 1.34 1-5 2.65 1.39 8.62 .004* .068
Note. survey language was treated as a covariate on perceived self-efficacy and perceived response-efficacy negative,
nationality was treated as a covariate on perceived response-efficacy positive and reasons for exercise, workload and education were treated as covariates on reasons for exercise.
* p < .05.
Table 3 presents the results of RM AN(C)OVA’s comparing the effects of RII’s versus II’s versus the formulation of goal intentions on exercising behavior, PMT variables, reasons
Hypothesis 1A investigated whether RII’s were more likely to stimulate individuals to enhance their exercising (for at least 20 minutes per week) than II’s or the formulation of goal intentions. Hypothesis 1B tested whether there was a stronger growth in exercising behavior for participants in the II than in the goal intention condition. A RM ANCOVA (Table 3) revealed that there were significant albeit small effects over time F (2,118) = 4.88, p = .029, η2 =
.040, which, however did not significantly differ between conditions. Subsequently, Hypotheses 1A and B were rejected.
Hypothesis 2A examined whether PMT variables were rather enhanced by RII’s than by II’s or by formulating positive goal intentions and 2B investigated whether that those
variables were rather enhanced by II’s than by formulating positive goal intentions. Therefore, 8 RM AN(C)OVA’s were conducted to test whether there were significant differences
between Time1 and Time2 on PMT variables. As Table 3 shows, there were no significant main effects between conditions on either outcome variable. Accordingly, hypotheses 2a and 2b were rejected. There was a significant decline in perceived self-efficacy over time F (2,119) = 9.88, p = .029, η2 = .040 and a significant interaction effect of time and condition for susceptibility negative F (2,119) = 3.98, p = .021, η2 = .063. Perceived susceptibility positive did not show significant effects. Perceived response-efficacy negative F (2,119) = 4.85, p = .030, η2 = .039 as well as perceived response efficacy positive F (2,119) = 46.95, p = <.001, η2
=
.286 and reasons to exercise F (2,119) = 8.44, p = .004, η2 = .068 significantly decreased over time in all conditions. Subsequently, hypotheses 2A and 2B were rejected.
To test hypothesis 3 whether participants who received RII’s were more likely to enhance their goal intentions more than individuals who received II’s or formulated positive goal intentions, a RM ANCOVA revealed no significant effects over time. Neither were there significant effects between conditions. Hypothesis 3 was rejected.
Table 3
RM AN(C)OVA interaction effects of Time 1-Time 2 and condition on outcome variables
F (2,119) p η2
Exercising Behavior time 7.00 .009* .056 time*condition 0.48 .619 .008 Perceived Severity Negative time 1.64 .203 .014 time*condition 0.85 .429 .014 Perceived Severity Positive time 1.98 .162 .016 time*condition 0.87 .421 .014 Perceived Self-efficacy time 4.88 .029* .040 time*condition 0.11 .901 .002 Perceived Susceptibility Negative time 0.95 .331 .008 time*condition 3.98 .021* .063 Perceived Susceptibility Positive time 0.07 .797 .001 time*condition 1.01 .367 .017 Perceived Response-efficacy Negative time 4.85 .030* .039 time*condition 0.15 .864 .002 Perceived Response-efficacy Positive time
time*condition 46.95 0.04 <.001* .966 .286 .001 Reasons for exercise time
time*condition 8.44 0.10 .004* .905 .068 .002 Goal Intention time 0.11 .740 .001 time*condition 0.794 .394 .016
Note. survey language was treated as a covariate on perceived self-efficacy and perceived response-efficacy
negative, nationality was treated as a covariate on perceived response-efficacy positive and reasons for exercise, workload and education were treated as covariates on reasons for exercise.
* p < .05.
Discussion
The current study examined the relative strength of RII’s compared to II’s and goal
intentions to enhance exercising behavior as well as cognitions in form of threat and coping elements related to (not) exercising. The first aim was to determine if exercising would rather be enhanced in participants who use specific plans (RII’s, II’s) than a mere formulation of their goal intentions. The second aim was to determine if RII’s were more likely than II’s and formulating goal intentions to enhance perceived vulnerability, perceived severity, perceived self-efficacy, perceived response-efficacy and reasons for exercise. This was expected because RII’s seemingly combine motivational and volitional elements. Contradictorily with
the expectations, all hypotheses were rejected (see again Table 3). There were no significant developments between groups over the two weeks. Subsequently, RII’s did not have a
considerably stronger or weaker effect on outcome variables than II’s or the control condition. Neither did II’s have a significant stronger effect than the formulation of unspecific goal
intentions. Those results contradict a large body of research showing promising II effects (e.g., Dijkstra et al., 2006; Gollwitzer & Sheeran, 2006; Prestwich & Keller, 2014). However, they are in line with research showing no significant II effects (e.g., De Vet et al., 2009; Jackson et al., 2005). Further, the results of the current study contradict results of Prestwich et al., 2008 who found small but significant higher RII effects than in an II- or control cgroup on general fruit intake. Concluding, in the current study, RII’s were not more effective than II’s
in enhancing exercising, goal intentions or related cognitions (PMT) than II’s or formulation of a goal intention were. Neither were II’s more effective than the formulation of generalgoal intentions was.
There are severalfactors that could have influenced the results. To begin with the control group: One contribution of the current study to existing II literature was the inclusion of a control group which also received an intervention. More specifically, participants were asked to formulate general goal intentions to exercise. Contrastingly, previous research (e.g., Milne et al., 2002) used control groups that did not receive topic-related information.
Consequently, participants were not reminded of exercising at all. The current study compared (R)II effects with a group that was also motivated to exercise. In this way, the failure to find significant differences between groups over time could be influenced by participants in the control group being asked to exercise more which reminded them of exercising. This assumption is supported by statistical results of the current study showing a significant increase of exercising behavior within subjects over time but not between
conditions. Speciically, all interventions equally influenced exercising behavior. This raises a question regarding the unique value of II’s. The main power of II’s lies in creating a mental
link between a healthy behavior (e.g., exercising) and a situation by specifying the goal intention to exercise (De Bruijn et al., 2017). Through this specification, II’s are supposed to provide individuals with self-regulative tools to perform their desired behavior. As the results of the current study did not show between-group effects, II effects could be influenced by
making exercising mentally accessible rather than by specifying the goal intention. This argumentation is in line with Jackson et al. (2005) who examined whether II’s were effective
in general fruit intake. They found a significant increase of fruit intake over time but not between conditions. Likewise, they particularly asked their control group to enhance their fruit intake. Concluding, changes in exercising behavior might as well result from merely formulating general goal intentions as creating a specified plan as it is done by II’s and RII’s.
Further, in Gollwitzer and Sheerans (2006) meta-analysis which is one of the most prominent papers in support of II’s, the included studies considerably differed in effect sizes.
For instance, Liebke and Ziegel (2002, unpublished) conducted an experimental study on II’s to increase exercising with an effect size of d. = .28, which is quite small (Cohen, 1992). Contrastingly, there was a huge effect of exercising (d = .68) in study results by Prestwich et al., (2003) who combined II’s with motivational information containing the advantages of exercising, based on PMT variables. Such divergent effect sizes on exercising-related behaviours indicate that II’s as enhancement of exercising seem to be only successful under certain circumstances. II effects may only appear in combination with motivation
strengthening elements as proposed by the model of action phases (Heckhausen & Gollwitzer, 1987). Based on this argumentation, it was expected that RIIs would be a stronger predictor for exercising than II’s as they were expected to enhance individuals’ goal intentions to exercise. Even if RII’s were based on these principles (combining motivational and volitional methods), the effect size of exercising in the current study was dramatically low with d = .06. This low effect size might strongly be influenced by sample characteristics which will be discussed in the limitation. The lack of finding significant differences between RII’s and other conditions might be due to adding the motivation to exercise may be too weak to enhance motivation to exercise and therefore goal intentions.
However, RII’s might still positively influence goal intentions in the long run. Rather than enhancing goal intentions, RII’s might stabilize them. Stabilization in this case means
maintain a positive goal intention over a long period of time. According to Prestwich et al., (2003) maintaining a positive goal intention to exercise is important as goal intentions often decrease over time. Therefore, through reminding the individual of its main motivation to exercise, RII’s could help stabilizing goal intentions. Future research is needed to clarify this
assumption.
Another characteristic that may have influenced the results is that participants had a high goal intention to exercise but did not seem to comply with the method. Firstly, there was a large dropout (n = 134). Further, only 19 of all participants who were included in the
analyses indicated to remember the structure of their plans. Subsequently, it was not likely that they could remember their (R)II’s. This raises the question if participants really tried to
implement their plans. Accordingly, (R)II’s studies should not solely require a positive goal intention to exercise, it should also be ensured that participants are motivated to apply the proposed method. Such motivation could be enhanced by collaboration (Prestwich & Kellar, 2014). In the context of II’s, collaboration means that specific plans are formed and conducted
in teams of two individuals. Prestwich, Powers, Koestmer and Topciu (2005) explored the effects of II’s formed in collaboration and found significant effects on behaviour, especially
because participants expected that performing the desired behaviour would be more fun together.
Limitations
The present study provides relevant insights but there are also limitations to
considerate when interpreting the results. The first and major limitation is regarding sample characteristics. The sample was extremely active as a huge part already exercised for more than three times a week and already met the WHO exercising requirements for individuals between 18-64 years (WHO, 2018). This could be a threat to external validity as exercising for more than three times a week is not a valid representation of Western society. Further,
there was little scope to enhance high exercising behavior. Subsequently, it was difficult to enhance outcomes in the current study. Individuals who would be responsible for the intention-behavior gap are inclined abstainers: individuals who are planning to exercise but cannot act accordingly (Sheeran, 2002). As the sample predominantly consisted of inclined actors (individuals who have a positive goal intention and act on this), there was no intention-behavior gap in the current study. Accordingly, a necessary condition of (R)II’s was violated. The present study could not control for this due to the fact that most participants who were not exercising did not have a positive goal intention to exercise and were therefore excluded from analyses.
The second limitation is the lack of a pilot study regarding the PMT scales that were used in the current study. The current study differed from most experimental studies
examining PMT variables as scales were formulated more generally. PMT variables needed to cover all possible benefits of exercising. For the current study such a formulation was of special importance as a major element of RII’s relies on self-persuasion. The scales should not be restricted to one benefit of exercising (e.g., losing weight). Most literature examined PMT in terms of specific threats. For example, Plotnikhoff and Higginbotham (2010) used PMT scales to examine the risk of coronary heart disease. Scales, like the ones used by Plotnikhoff and Higginbotham (2010), could be stronger in eliciting threat as they directly address harms of not-exercising. As the PMT did not cover positive exercising outcomes, this theory might not be the ideal framework for the current study. This leads to another measurement-related limitation. The broad formulation of exercising effects should have been tested in a pilot study, together with the modified, positively formulated items measuring PMT constructs. As the positive scales were newly created, a pilot study could have shown if participants
interpreted the items as desired by the researcher. Consequently, validity of the scales was not ensured, which could have influenced the results of the current study. Through a
due to time-limits a pilot study was not possible. Factor- and reliability analyses were conducted to compensate for this but did not recompense a pilot study.
A third limitation was that only a small percentage of participants could remember the structure of their formulated intention. The current study could not control for this, as there were only 19 participants who would have met those criteria. Therefore, it cannot be assumed with certainty that participants applied their created plans over the course of the experiment. A frequency analysis showed that 62% of the sample (n = 122) reported that their plan had the form of “I want to enhance my exercising”. This indicated that participants thought that their
plan would be their general positive intention to exercise, meaning that they might have interpreted the question in a wrong way. This problem is influenced by the lack of a pilot survey. Subsequently, there is no certainty whether participants forgot their plan or interpreted this question wrongly. However, it is important to check whether participants used the same plans they created at Time1 during the two weeks. Participants in the goal intention condition might have formed II’s or RII’s by themselves during the course of the experiment. This is a
limitation to be considered when interpreting the results.
Finally, the current study used self-reports and therefore underlies the common threats of self-reported data.
Future Research
The current study provided relevant implications for future research. Firstly, study results should be replicated by considering the limitations of the current study. Especially the sample should be changed for future research. It should be ensured that the sample a) consists of inclined abstainers, b) has a positive goal intention to exercise more and c) is motivated to comply with the proposed method (RII’s, II’s). To ensure a) a sample of inclined abstainers, future research should replicate this study with a greater sample to be able to exclude
second point b) a positive goal intention to exercise more might be achieved by the use of different measurement scales of goal intentions to exercise. For example Prestwich et al. (2003) assessed goal intention scores by four 7-point Likert scales which included items like “How likely or unlikely is it that you will exercise more during the next four weeks? (p. 713)”. It was explicitly asked for enhancing exercising behavior which was not the case in the
current study. To achieve the third necessary condition, which received remarkably little scientific attention so far c) motivation to comply with the proposed method, future research should investigate source-effects. For example, an employee in a gym or a professional health coach might be more reliable than a student researcher conducting a survey. Therefore, the study could be replicated by another source. Going along with this idea, another
argumentation is to test for participants’ compliance levels and compare (R)II effects of compliant and non-compliant individuals.
Secondly, instead of relying on PMT variables, future research could investigate RII effects based on another theory. The TPB for example also focuses on building positive goal intentions but does not rely on threat elements. The advantage here is that the theory also covers participants’ motivations to exercise and not the fear of not-exercising.
Thirdly, (R)II’s rely on positive goal intentions to change behavior. However, high
goal intention scores leave little space for growth. Results of the current study indicated that RII’s might be too weak to enhance them. Future research should examine whether RII’s are
effective to stabilize those goal intentions rather than enhancing them.
Further, as the study results revealed several covariates, (R)II effects should be tested for mediating and moderating factors in future research. To strengthen the volitional element and the motivation to comply, RII’s could be combined with collaboration methods. Future research could examine the effects of RII’s versus RII’s formed and implemented in groups
and not by individuals alone. In this way, participants could be paired with others who have the same main motivation and try to exercise together.
Conclusion
The present study contributes to a growing body of research regarding the combination of volitional and motivational methods to achieve behavior change. Contradictory with
hypotheses, the only significant effects in the current study were differences between in time within participants and not between conditions. Perceived self-efficacy, perceived response-efficacy (negative and positive) and reasons for exercise significantly decreased over time of the experiment. Exercising behavior significantly increased. However, this did not depend on participants’ conditions. By considering limitations, the current study indicates that methods like II’s and RII’s alone not always seem to work for complex behaviors like exercising.
However, there is future research needed to clarify the role of RII’s and combine them with other methods to achieve optimal effects. Therefore, a focus should lie on strengthening the motivational part of RII’s to stabilize goal intentions as well as the volitional part to give individuals control over their own behavior. Further, the current study showed that motivation to exercise is not the only important motivational element. Participants should also be willing to comply with the proposed method. RII’s could be a promising method to enhance
exercising but there is still need to optimize them in order to a) motivate individuals to comply with the proposed method and b) help them turn their goal intentions into action.
Literature
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211.
Adriaanse, M. A., Vinkers, C. D., De Ridder, D. T., Hox, J. J., & De Wit, J. B. (2011). Do implementation intentions help to eat a healthy diet? A systematic review and meta-analysis of the empirical evidence. Appetite, 56, 183-193.
Arbour, K. P., & Martin Ginis, K. A. (2004). Helping middle-aged women translate physical activity intentions into action: Combining the theory of planned behavior and
implementation intentions. Journal of Applied Biobehavioral Research, 9, 172–187. Canning, E. A., & Harackiewicz, J. M. (2015). Teach it, don’t preach it: The differential
effects of directly-communicated and self-generated utility–value information. Motivation science, 1(1), 47.
Caudwell, K. M., Mullan, B. A., & Hagger, M. S. (2015). Combining motivational and volitional approaches to reducing excessive alcohol consumption in pre-drinkers: a theory-based intervention protocol. BMC public health, 16(1), 45.
Cohen, J. (1992). A power primer. Psychological bulletin, 112(1), 155.
DeBruijn, G.J., Nguyen, M. H., Rhodes, R. E., & van Osch, L. (2017). Effects of preparatory and action planning instructions on situation specific and general fruit and snack intake. Appetite,108,161-170.
Davis, C., Fox, J., Brewer, H., & Ratusny, D. (1995). Motivations to exercise as a function of personality characteristics, age, and gender. Personality and individual differences, 19(2), 165-174.
DeVet, E., Oenema, A., Sheeran, P., & Brug, J. (2009). Should implementation intentions interventions be implemented in obesity prevention: the impact of if-then plans on daily physical activity in Dutch adults. International Journal of Behavioral Nutrition and Physical Activity, 6(1), 11.
Gaston, A., & Prapavessis, H. (2014). Using a combined protection motivation theory and health action process approach intervention to promote exercise during pregnancy. Journal of Behavioral Medicine, 37(2), 173-184.
Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal achievement: A meta-analysis of effects and processes. Advances in Experimental Social Psychology, 38, 69 –119.
Hagger, M. S., & Luszczynska, A. (2014). Implementation intention and action planning interventions in health contexts: State of the research and proposals for the way forward. Applied Psychology: Health and Well‐Being, 6, 1-47.
Heckhausen, H., & Gollwitzer, P. M. (1987). Thought contents and cognitive functioning in motivational versus volitional states of mind. Motivation and emotion, 11(2), 101-120. Jackson, C., Lawton, R., Knapp, P., Raynor, D. K., Conner, M., Lowe, C., & Closs, S. J.
(2005). Beyond intention: Do specific plans increase health behaviours in patients in primary care? A study of fruit and vegetable consumption. Social Science & Medicine, 60(10), 2383-2391.
Koka, A. (2016). An intervention crossing two types of action planning with mental simulation for the promotion of physical activity among adolescents. International journal of adolescent medicine and health, 29(3).
Lang, A. (2000). The limited capacity model of mediated message processing. Journal of Communication, 50(1), 46-70.
Lippke, S., & Ziegelmann, J. P. (2002). Self-regulation and exercise: A Study on Stages of Change and Successful Ageing. Unpublished manuscript, Free University of Berlin, Berlin, Germany.
McGowan, L., Cooke, L.J., Gardner, B., Beeken, R.J., Croker, H.,& Wardle, J.(2013).Healthy feeding habits: efficacy results from a cluster-randomized, controlled exploratory trial of a novel, habit-based intervention with parents. Am J Clin Nutr, 98(3), 769-777. Milne, S., Sheeran, P., & Orbell, S. (2000). Prediction and intervention in health‐related
behavior: A meta‐analytic review of protection motivation theory. Journal of Applied Social Psychology, 30(1), 106-143.
Milne, S., Orbell, S., & Sheeran, P. (2002). Combining motivational and volitional interventions to promote exercise participation: Protection motivation theory and implementation intentions. British Journal of Health Psychology, 7, 163-184. Papies, E. K., & Hamstra, P. (2010).Goal priming and eating behavior: enhancing
self-regulation by environmental cues. Health Psychology, 29, 384-388.
Plotnikoff, R. C., & Higginbotham, N. (2002). Protection Motivation Theory and exercise behaviour change for the prevention of heart disease in a high-risk, Australian representative community sample of adults. Psychology, health & medicine, 7(1), 87-98.
Topciu, R.A. Powers, T.A., Koestner, R. (2005). Implementation intentions, perfectionism, and goal progress: perhaps the road to hell is paved with good intentions. Personality and Social Psychology Bulletin, 31, pp. 902-912
Prestwich, A., Perugini, M., & Hurling, R. (2009). Can the effects of implementation intentions on exercise be enhanced using text messages? Psychology & Health, 24, 677-687.
Prestwich, A., Ayres, K., & Lawton, R. (2008). Crossing two types of implementation intentions with a protection motivation intervention for the reduction of saturated fat intake: A randomized trial. Social Science & Medicine, 67, 1550-1558.
Prestwich, A., & Kellar, I. (2014). How can the impact of implementation intentions as a behaviour change intervention be improved?. Revue Européenne de Psychologie Appliquée/European Review of Applied Psychology, 64, 35-41.
Prestwich, A., Lawton, R., & Conner, M. (2003). The use of implementation intentions and the decision balance sheet in promoting exercise behaviour. Psychology and Health, 18(6), 707-721.
Rhodes, R. E., & Bruijn, G. J. (2013). How big is the physical activity intention–behaviour gap? A meta‐analysis using the action control framework. British journal of health psychology, 18, 296-309.
Rogers, R. W. (1983). Cognitive and psychological processes in fear appeals and attitude change: A revised theory of protection motivation. Social psychophysiology: A sourcebook, 153-176.
Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American psychologist, 55(1), 68. Sheeran,P. (2002). Intention-behavior relations: A conceptual and empirical review. European
Review of Social Psychology, 12(1), 1-20.
Tiro, J. A., Lee, S. C., Marks, E. G., Persaud, D., Skinner, C. S., Street, R. L., Wiebe D.J., Farrell, D., Pechoro Bischop, W., Fuller, S. & Baldwin, A. S. (2016). Developing a tablet-based self-persuasion intervention promoting adolescent HPV vaccination: protocol for a three-stage mixed-methods study. JMIR research protocols, 5(1). Witte, K. Cameron, K., A., McKeon, J.K. & Berkowitz, J.M. (1996). Predicting risk
behaviors: Development and validation of a diagnostic scale. Journal of health communication, 1(4), 317-342.
World Health Organization (2018). Information sheet: global recommendations on physical activity for health 18 - 64 years old. Retrieved from:
http://www.who.int/dietphysicalactivity/factsheet_adults/en/
World Health Organization (2017). Obesity and Overweight. Retrieved from: http://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight Institute for Health Metrics Evaluation (2016). Measuring what matters. Netherlands.
Retrieved from: http://www.healthdata.org/netherlands
Renew Bariatrics (2017). Report: Obesity rates by country -2017. Retrieved from: https://renewbariatrics.com/obesity-rank-by-countries/
Appendix
A1. Survey Time 1 English and German A1A. Time 1 English
Introduction
What you can expect from participating Welcome to a study examining the effectiveness of a method which is supposed to help you enhance the amount of exercising. My name is Clio Carlotta Rosebrock and I am
conducting this survey as part of my Master’s degree at the Graduate School of Communication at the University of Amsterdam, supervised by Prof. Dr. Bas van den Putte.
In the following study, you are asked to fill in two surveys, two weeks apart. Within the first survey, you need to answer several general questions and I will introduce you to the method that is supposed to help you start exercising. Filling in this survey will cost you around 10 minutes. Within the second survey (two weeks later), you are again asked to answer several general questions. This survey will approximately take 3 minutes to complete.
Furthermore, I would like to request you to answer the following questions honestly; there is no “right” or “wrong” answer. All data is processed anonymously, meaning that the information will not be linked to your name or other identity-related elements. However, please notice that I will ask for your mail address at the end of both surveys. This is necessary, as I need to send you the follow-up questionnaire and connect both surveys for my analyses. Note that your personal information will never be passed to third parties without explicit
permission. At the end of the study, all data that can identify you, such as your email address, will be deleted. If you completely fill in both questionnaires, you have the opportunity to win an Amazon voucher of
20Euros.
As the results of the study are only used for scientific purposes, you have the right to receive them. If you would like to receive the results, please contact me under following email address:
Clio.rosebrock@student.uva.nl
By clicking the button below, you acknowledge that your participation in the study is voluntary, you are 18 years of age, and that you are aware that you may choose to terminate your participation in the study at any time and for any reason. Within 7 days after answering the second survey, you may withdraw your permission to use the data.
Should you have any complaints or comments about this research, you can contact the designated member of the Ethics Committee representing the Department of Communication Science, at the following address: ASCoR Secretariat, Ethics Committee, University of Amsterdam, Postbus 15793, 1001NG Amsterdam; 020‐525 3680; ascor‐secr‐fmg@uva.nl.
Any complaints or comments will be treated in strictest confidence.
Thanks a lot for your time and success with exercising!
o
Agree with terms and conditionsGender
What is your gender?
o
Maleo
Femaleo
OtherAge
Please indicate your age
18 28 38 48 58 68 78 88 98 100 in years ()
Nationality
What is your Nationality?
o
Germano
Dutcho
Other, namely ________________________________________________Education
What is your highest educational level?
o
Primary schoolo
High schoolOccupation
What is your main occupation?
o
Studento
Workingo
Home duties/ not employedo
OtherWorkload
What is the estimated amount of hours weekly spend for your main occupation or studies?
o
0-20 hourso
21-36 hourso
37-42 hourso
43-50 hoursDV1. Exercising behavior
Please indicate your current amount of exercising per week.
o
I currently don't exerciseo
20 minutes one time a weeko
20 minutes two times a weeko
20 minutes three times a weeko
20 minutes four times a weeko
20 minutes more than four times a weekGoal intention 1-4
Please indicate to what extent you agree with following statements
Never 1 2 Sometimes 3 4 Very often 5 How often do
you tell yourself to get adequate
exercise
o
o
o
o
o
Please indicate to what extent you agree with following statements
Very unlikely 1 2 Neutral 3 4 Very likely 5 How likely or
unlikely is it that you will get
adequate exercise during
the next six months
