Effect of Psychological Distance on Intention in Self-Affirmation Theory

Abstract

In previous studies, self-affirmation has been shown to have no effect or have a negative impact on intention to change. We applied construal level theory to examine possible reasons for inconsistencies among findings in existing self-affirmation studies. In Study 1, when female college students in their 20s (N = 113) received health threat information with a psychologically proximal breast cancer message, it induced low-level construal and increased specific physical activity intention. In Study 2, when self-affirmed participants were exposed to health threat information with a psychologically distal Alzheimer’s disease message, there was no increase in specific physical activity intention. These results suggest that when a discrepancy exists between the psychological distance of health threat information and the intention reflecting the construal level, no change of intention occurs, even in self-affirmed individuals.

Keywords

Self-affirmation theory construal level theory psychological distance physical activity intention health threat message

Introduction

According to self-affirmation theory, people have an internal motivation to maintain their self-worth and integrity, and threats to personal integrity are met with attempts to restore damaged self-worth (Harris, 2011; Steele, 1988). For example, after losing a game a person might say, “I was unlucky this time” and move on, while others might go on a trip or watch a film to recuperate from their loss. People use different methods to change their moods, but self-affirmation theory posits that the goal of the different methods is the same, to restore self-worth and integrity. Therefore, the perception of self-worth is an important predictor of people’s behavior.

Research into self-affirmation began with the study of defensiveness and individual relevance in relation to health treats, finding that, for example, heavy smokers are more likely to strongly disagree with the message “smoking causes cancer” than are nonsmokers (Berkowitz & Cottingham, 1960). Previous research has also found that people with excessive caffeine intake are less likely to believe in the correlation between breast cancer occurrence and caffeine consumption (Kunda, 1987) and that generally, individuals who are at a greater risk of damaging their health are less likely to believe in relevant health threats (Liberman & Chaiken, 1992). In other words, there are differences in the extent to which important information is believed and trusted based on how relevant that information is to the individual, with people to whom the information is more relevant (e.g., heavy smokers, high caffeine consumers) less likely to believe and trust in the information compared with people to whom the information is not as relevant (e.g., occasional smokers, low caffeine consumers). This phenomenon occurs as a result of threats to the self-image triggering biased and defensive perspectives that negatively affect message acceptance (Ditto & Lopez, 1992; Dunning, Leuenberger, & Sherman, 1995; Kunda, 1990).

The first core feature of self-affirmation theory is the idea that when people receive threatening information, they react defensively to protect their positive self-perception. However, if the self-image of the threatened individual could be affirmed by other things (such as positive feedback, writing down natural strengths, etc.), the desire to react defensively in the presence of threatening information is diminished. This is the second core feature of self-affirmation theory (Steele, 1988). Essentially, people who are more self-affirmed (SA) are more likely to absorb situational threats and effectively use their mental resources to detach from threats (Steele, Spencer, & Lynch, 1993). Additionally, SA people can systematically examine and accept the threatening information, not just avoid it (Correll, Spencer, & Zanna, 2004).

According to Sherman, Nelson, and Steele (2000), when a SA individual with a coffee addiction sees validated research suggesting that “coffee causes breast cancer,” they are more likely to agree than are coffee addicts who are not SA. In research on heavy drinking in women, participants who gained self-affirmation became more open minded about the negative effects of alcohol on health, and their intention to drink was reduced (Harris & Napper, 2005). A recent meta-analysis found that self-affirmation leads to positive changes in intentions and behaviors, with a small or moderate effect size (Epton, Harris, Kane, van Koningsbruggen, & Sheeran, 2015; Sweeney & Moyer, 2015).

Although many studies (Armitage, Harris, Hepton, & Napper, 2008; Harris & Napper, 2005; Jessop, Sparks, Buckland, Harris, & Churchill, 2014; Pietersma & Dijkstra, 2011) have shown that self-affirmation has positive effects on intentions and behaviors, the relationship between enhanced intention and modified behavior is low. Several studies have failed to find significant differences between affirmed and nonaffirmed participants on measures of behavior (Harris & Napper, 2005) and intentions (Epton & Harris, 2008; Good & Abraham, 2011; Jessop, Simmonds, & Sparks, 2009). Surprisingly, one study found that nonaffirmed participants had higher intentions to change their behavior than affirmed participants (Reed & Aspinwall, 1998), and the results from a metaregression analysis (Sweeney & Moyer, 2015) also show the effect size of intention could not statistically predict the effect size of behavior in SA groups. These studies suggest that self-affirmation does not always lead to increased intention to change or positive behavior change.

The construal level theory (CLT) may explain the inconsistency in the results of self-affirmation research. CLT was suggested by Trope and Liberman (2003, 2010) and aims to predict and explain people’s behavior according to how they interpret objects and events in terms of perceived psychological distance. Entities perceived to be distant are thought of in an abstract sense; entities perceived to be closer are thought of in a concrete sense. CLT includes high and low construal. High-level construal represents an abstract “big-picture” way of thinking: this is a superordinate or central approach, thinking about the overall idea of the situation and extracting its main gist. Low-level construal involves more detail-oriented or subordinate thought processes, representing situations in concrete thought (Liberman & Trope, 2008). To prepare and plan for distant events, people engage in high-level construal; the process of cognitive abstraction to extract the general, global, and goal-relevant features likely to be apparent across all possible manifestations of the events (“WHY” process). As events become more proximal and specific details become increasingly available and reliable, people are able to engage in low-level construal, incorporating the idiosyncratic and secondary features that highlight the event’s uniqueness (“HOW” process).

The key point of CLT is that when people perceive objects as psychologically distal, their behavior is provoked by abstract or high-level construal, whereas if perceived psychological distance is proximal, behavior is provoked by concrete low-level construal. It is this congruency between perceived psychological distance and construal level that helps to explain health behavior. Inspired by CLT, we proposed that if the threatening health information is psychologically distal, it induces high-level construal or the “WHY” process; hence, the intention reflecting the abstract representation should be measured. However, if the threatening information is psychologically proximal, the intention reflecting the concrete representation should be measured because the threatening health message would induce low-level construal or the “HOW” process.

Participants who perceive health threat information to be psychologically proximal will have specific rather than abstract intention and the specific intention will have a positive effect on their health-related behaviors. In contrast, participants who perceive health threat information to be psychologically distal will have abstract rather than specific intentions and the abstract intention will have a positive effect on their health-related behavior. Depending on the psychological distance of the health threat information, the intention should also be measured as either a specific intention reflecting low-level construal or general intention reflecting high-level construal. However, if the measured intention reflects a different construal level from the psychological distance of the health threat information, then self-affirmation manipulation would not induce increased intention or behavior change and an inconsistency between intention and behavior will exist. Therefore, the reason intention change was not induced despite self-affirmation in previous studies may be a result of not considering the effect on intention of the threatening health message’s psychological distance.

Study 1 examined the effect of two health threat messages with different psychological distances (breast cancer vs. Alzheimer’s disease) on participants’ construal level (high level or “WHY” process vs. low level or “HOW” process). We also examined the relationship between health threat messages with different psychological distances and intention (specific physical activity intention (SPAI) vs. general physical activity intention (GPAI)). Study 2 examined the effect of self-affirmation on SPAI by type of threatening health message and self-affirmation.

Study 1

Overview

An important determinant of construal level as explained in CLT is the psychological distance of the event or target under consideration. Liberman and Trope (1998) identified how people’s representations change when they are given the information about temporal distance (tomorrow vs. a year later) of a “moving to a new apartment” event. In the study by Amit, Algom, and Trope (2009), the social distance manipulation involved presenting the message separately to subjects in domestic (in-group) or overseas (out-group) contexts. Probabilistic distance manipulation used the method of providing a probability for the occurrence of a given scenario (e.g., camping scenario, moving out). In this study, psychological distance was manipulated using the information contents of an event or message.

From the perspective of CLT research, the health threat message provided by self-affirmation includes information containing various psychological distances depending on how the participant perceives the situation or event, such that the psychological distance perceived by the participant may differ. Therefore, if women in their 20s perceive a health threat message to have psychological proximity, it will lead to a lower level construal (“HOW” process) and be associated with SPAI rather than GPAI; those who receive a health threat message with psychological distance will tend to use a high-level construal, where GPAI is higher than SPAI.

Hypotheses

Hypothesis 1-1: Female college students will perceive a breast cancer message to be more psychologically proximal than an Alzheimer’s disease message.

Hypothesis 1-2: Participants will have higher concrete representation scores for the advertisement emphasizing “HOW” and higher abstract representation scores for the advertisement emphasizing “WHY.”

Hypothesis 1-3: Participants given the breast cancer message will tend to use low-level construal, leading to higher scores on attitude toward the advertisement emphasizing specific representations, or the “HOW” process; participants given the Alzheimer’s disease message will tend to use high-level construal, leading to higher scores on attitude toward the advertisement emphasizing abstract representations, or the “WHY” process.

Hypothesis 1-4: Participants in the breast cancer group will display stronger SPAI than GPAI, and those in the Alzheimer’s disease condition will show stronger GPAI than SPAI.

Method

Ethical considerations

This study was approved by the university’s research ethics committee. All participants gave written informed consent to participate.

Psychological distance of health messages

This study uses two health messages. The first message, which was designed to seem psychologically distant, was “A lack of physical activity in your 20s can cause Alzheimer’s disease at 70 or 80 years of age.” The second message, which was designed to seem psychologically close, was “A lack of physical activity in your 20s can cause breast cancer in your 20s or 30s.” Both health messages took the form of published scientific research papers that were structured the same, including the same information regarding research institutes, number of participants, research year, effect size, and research method so that participants had the same belief in the study results. After participation, we explained to the students that the health messages had been modified for this study and we showed them the actual results (e.g., research institutes, number of participants, research year, effect size, and research method) from these studies regarding the links between physical activity and Alzheimer’s disease or breast cancer.

Construal level of advertising messages

Based on Trope and Liberman (2003, 2010), to manipulate the construal level of advertising messages about a fitness center, we created a message emphasizing “WHY” one should exercise at the fitness center (see Appendix 1) and a message emphasizing “HOW” one should exercise at the fitness center (see Appendix 2). In the “WHY” message, we used abstract words related to the outcome, emphasizing the effects and benefits of exercise, such as fun, health, youth, and happiness. In the advertising message emphasizing “HOW,” we used specific and process-centered words, such as body composition measurement, 3D spine stabilization system, and personal training program.

Participants

Participants in this study were randomly recruited from Ewha Womans University’s website. Since people who exercise regularly would perceive less threat from the health threat messages provided in this study, we recruited participants who did not exercise regularly. We defined regular exercise as a minimum of two to three times a week for 30 minutes, for at least the preceding six months. A total of 117 participants volunteered. The Transtheoretical Model (TTM) questionnaire (Marcus et al., 1992) was used to recheck for regular exercise participation among the 117 participants; 4 participants identified as engaging in regular exercise were excluded, resulting in 113 participants included in the analysis. Drink coupons worth about $5 were provided to all participants.

Measures

Preliminary questionnaire

Message belief was measured by a single item, “How much do you trust this message?” using a seven-point scale (1 = strongly disagree, 7 = strongly agree).

Physical activity level

To screen participants for exercise level, we used the TTM questionnaire developed by Marcus et al. (1992). One of the major contributions of TTM to the health field is the suggestion that behavior change unfolds slowly over time through a series of five stages: precontemplation (“I do not intend to begin exercising in the next six months”), contemplation (“I intend to begin exercising in the next six months”), preparation (“I intend to begin exercising regularly in the next 30 days”), action (“I have been exercising, but for less than six months”), and maintenance (“I have been exercising for more than six months”). The TTM questionnaire consists of identifying one of the five items that describe each stage. After having read five items, the respondents selected the stage. In Studies 1 and 2, participants in the precontemplation, contemplation, and preparation stages were classified as nonexercisers, whereas participants in the action and maintenance stages were classified as regular exercisers and were excluded from the analyses.

Psychological distance

Social distance used the same seven-point rating scale and was measured by the item, “This message (breast cancer or Alzheimer’s disease) is more relevant to women in their 50s than women in their 20s.” Temporal distance was measured using the item, “If you don’t engage in regular physical activity after reading these messages, you are likely to get breast cancer (or Alzheimer’s disease) after ___ years (please write time in years, with a maximum time of 80 years).” Probabilistic distance was measured by the question, “What do you think the probability of getting breast cancer (or Alzheimer’s disease) is in your 20s if you do not engage in regular physical activity? (Please write the probability as a percentage, with the maximum being 100%.)”

Construal level conditions

To assess whether the advertising messages shown in Appendix 1 emphasize why people should exercise at the fitness center or Appendix 2 emphasize how they should exercise at the fitness center, four items were administered using a seven-point scale (q1: The advertisement above explains in detail how the fitness center can help with exercise participation; q2: The message in the advertisement above is specific; q3: The advertisement above explains why exercise participation is necessary; and q4: The message in the advertisement above is abstract). Items 1 and 2 reflect the concrete representation (α = .79), while 3 and 4 reflect the abstract representation (α = .69); the mean of the two scores was used in the final analysis.

Attitude toward advertisement

To measure the dependent variable, attitude toward advertisement, five items (α = .87) using a seven-point scale were altered to be made suitable for the purpose of the present study, based on the measure of attitude toward advertisement suggested by MacKenzie and Lutz (1989) (“I trust this advertisement,” “I understand the content of this advertisement,” “I think the information in this advertisement is helpful,” “I can relate to the content of this advertisement,” and “This advertisement grabs my attention and interest”).

Physical activity intention

Physical activity intention (PAI) was measured with four items on a five-point scale (1 = strongly disagree, to 5 = strongly agree) by modifying and supplementing the Physical Activity Scale (PAS) of Armitage and Sprigg (2010). Physical activity is movement carried out by the skeletal muscles that requires energy. In other words, any movement one does is actually physical activity. Exercise, however, is planned, structured, repetitive, and intentional movement intended to improve or maintain physical fitness. Exercise is a subcategory of physical activity (Hausenblas & Rhodes, 2017). In this study, we determined that it was not appropriate to predict physical activity using an existing questionnaire of exercise intention; thus we constructed an exercise intention questionnaire based on the PAS to predict physical activity. The PAS consists of items measuring the amount of physical activity by a child using past tense (e.g., I ran around as much as possible in the last playtime). The PAI used future tense, as it predicts future exercise participation intention.

Before measuring PAI, SPAI and GPAI were differentiated based on whether any of six strategies for promoting physical activity were provided (walking to the bus stop, standing by public transportation, using the stairs from the first floor to the second floor, going downstairs or upstairs when using the common room, walking further to buy food, and shopping daily for food). For SPAI (α = .80), responses to the items (“I will be doing physical activities as hard as possible, based on the six strategies over the next week,” “I will be doing physical activities as steady as possible, based on the six strategies over the next week,” “I will be doing physical activities as consistently as possible, based on the six strategies over the next week,” and “I will be doing physical activities as much as possible, based on the six strategies over the next week”) were made after considering the strategies for promoting physical activity. In contrast, for the GPAI (α = .95), responses to the items (“I will be doing physical activities as hard as possible over the next week,” “I will be doing physical activities as steady as possible over the next week,” “I will be doing physical activities as consistently as possible over the next week,” and “I will be doing physical activities as much as possible over the next week”) were made without the strategies, with the items identical to the SPAI items except for the adverbial phrase “based on the six strategies.”

The method of measuring SPAI and GPAI by providing a physical activity strategy referred to implementation intentions, one of the strategies to reduce intention–behavior discrepancy (Webb & Sheeran, 2006). Implementation intentions reflect specific plans (when, where, and how) and the physical activity strategy provided in this study presented the required specific plans for performance intentions. Although SPAI and GPAI share the same measurement items, different intentions are measured depending on whether the strategy is provided so that provision of strategy serves as a way of manipulating the construal level of PAI.

Procedure

As participants’ own construal level could not be measured in this study, random assignment was used to compensate for this limitation. The 113 participants were randomly assigned to four groups (Group 1: breast cancer + strategies for promoting physical activity provided + HOW advertisements (n = 28), Group 2: breast cancer + strategies for promoting physical activity not provided + WHY advertisement (n = 29), Group 3: Alzheimer’s disease + strategies for promoting physical activity provided + HOW advertisement (n = 29), and Group 4: Alzheimer’s disease + strategies for promoting physical activity not provided + WHY advertisement (n = 27)). Participants were first provided with the health threat messages that differed in psychological distance and then measured on the perceived psychological distance of the provided health threat message as well as on PAI (specific/general). Finally, after reading the advertisements reflecting different construal levels, participants’ own construal level (concrete/abstract) and attitude toward the advertisement were measured.

Results

Manipulation check

Independent sample t tests were conducted between the Alzheimer’s disease message group and the breast cancer message group for message belief, social distance, temporal distance, and probabilistic distance. There were statistically significant differences in social distance (t(111) = 3.76, p < .001), temporal distance (t(111) = 24.81, p < .001), and probabilistic distance (t(111) = 26.77, p < .001) between the two groups, as shown in Table 1. In other words, women in their 20s perceived breast cancer messages to be more psychologically close than dementia messages. There were no statistically significant differences in message belief (t(111) = 0.06, p = .950) between the Alzheimer’s disease message group and the breast cancer message group. This suggests that participants had the same degree of belief for each message.

Table 1.

Responses to the outcome measures by message condition in Study 1.

Measure	AD (n = 56)		BC (n = 57)		t	p	95% CI
Measure	M	SD	M	SD	t	p	LL	UL
Message belief	5.46	0.76	5.47	0.82	0.06	.950	−0.287	0.306
Social distance	5.21	1.59	4.09	1.59	3.75	.000	0.53	1.72
Temporal distance	46.04	8.33	13.05	5.56	24.80	.000	30.35	35.35
Probabilistic distance	15.21	10.71	41.98	15.94	10.46	.000	21.70	31.84

AD: Alzheimer’s disease; BC: breast cancer; CI: confidence interval; LL: lower limit; UL: upper limit; SD: standard deviation.

To examine the construal levels (abstract or concrete) of the advertisements shown in Appendices 1 and 2, we compared the mean scores of abstract and concrete representations for each advertisement emphasizing “HOW” or “WHY.” The abstract representation mean score was 3.70 (standard deviation (SD) = 0.82) for the groups in which “HOW” was emphasized (n = 57) and 4.09 (SD = 0.61) for the groups in which “WHY” was emphasized (n = 56), showing a statistically significant difference (t(111) = 3.92, p < .001). The concrete representation mean score was 3.26 (SD = 1.27) for the groups in which “HOW” was emphasized and 2.34 (SD = 1.10) for the groups in which “WHY” was emphasized, also a statistically significant difference (t(111) = 2.89, p < .01). Since the advertisement emphasizing “HOW” reflects specific information and the advertisement emphasizing “WHY” reflects abstract information, it appears that the treatment using the advertisement that was provided was adequate for the purposes of the experiment (i.e., Hypothesis 1-2 was supported).

Effect of construal level on attitude toward advertisement by psychological distance of the health threat messages

Since the manipulation check indicated that female college students perceived health threat information about breast cancer to be more psychologically proximal than information about Alzheimer’s disease, participants who received the breast cancer message were expected to use low-level construal and thus to show higher scores on attitude toward advertisement and intention to register after viewing the advertisement emphasizing concrete representation or the “HOW” process. In contrast, for the Alzheimer’s disease, message participants were expected to use high-level construal and thus show higher scores in attitude toward advertisement and intention to register after viewing the advertisement emphasizing an abstract representation or “WHY” process.

To test this, we analyzed participants’ attitude toward the advertisement as a psychological distance of the health threat message (breast cancer vs. Alzheimer’s disease) and the construal level of advertising messages (“HOW” vs. “WHY”) using a 2 × 2 between-subjects analysis of variance (ANOVA). Participants (M = 2.77, SD = 1.14) who were provided the breast cancer-threatening health message scored lower on attitude toward the advertisement than those (M = 3.31, SD = 0.98) who were provided the Alzheimer’s disease-threatening health message of (F(1, 109) = 7.71, p < .01, $η_{p}^{2}$ = .06). The advertisement emphasizing “HOW” was associated with higher scores on attitude toward the advertisement (M = 3.29, SD = 1.03) than those emphasizing “WHY” (M = 2.79, SD = 1.10) (F(1, 109) = 6.48, p < .05, $η_{p}^{2}$ = .05).

On whether the interaction between the type of health threat message and “HOW” or “WHY” emphasis in the advertisements affects attitude toward advertisement, a statistically significant interaction effect was found (F(1, 109) = 8.65, p < .01, $η_{p}^{2}$ = .07). As shown by the interaction effect graph in Figure 1, participants who received the (psychologically proximal) breast cancer message scored higher on attitude toward the “HOW” process advertisement, a concrete representation, compared to when given the “WHY” process advertisement, an abstract representation. In contrast, participants who received the (psychologically distal) Alzheimer’s disease message scored similarly on attitude toward the “HOW” or “WHY” advertisements. Thus, the findings partially support Hypothesis 1-3.

Figure 1.

Attitude toward advertisement (+SE) by type of health threat information and construal level condition. SE: standard error.

Effect of strategies promoting physical activity on PAI by psychological distance of the health threat message

After providing the psychologically proximal breast cancer message and strategies for promoting physical activity, SPAI was expected to be stronger than GPAI measured by the absence of strategies for promoting physical activity, whereas in the psychologically distal Alzheimer’s disease message, GPAI was expected to be stronger than SPAI.

To analyze the difference in PAI depending on providing the physical activity promotion strategy (yes vs. no) and psychological distance of the health threat information (breast cancer vs. Alzheimer’s disease), a 2 × 2 between-subjects ANOVA was performed. There was no statistically significant difference in physical activity between breast cancer (M = 3.89, SD = 0.73) and Alzheimer’s disease messages (M = 4.07, SD = 0.69) (F(1, 109) = 1.82, p = .180, $η_{p}^{2}$ = .01). However, PAI was higher when the physical activity promotion strategy was provided (M = 4.12, SD = 0.67) than when it was not (M = 3.84, SD = 0.73) (F(1, 109) = 4.31, p < .04, $η_{p}^{2}$ = .03).

The interaction between provision of strategies for promoting physical activity and the psychological distance of the health threat messages had a statistically significant effect on PAI (F(1, 109) = 10.51, p < .01, $η_{p}^{2}$ = .08). As shown by the interaction effect graph in Figure 2, while participants in the breast cancer message group had higher SPAI and lower GPAI, those in the Alzheimer’s disease message group had similar levels of GPAI and SPAI, partially supporting Hypothesis 1-4.

Figure 2.

Physical activity intention (+SE) by type of health threat information and provision of a physical activity promotion strategy (yes vs. no). SE: standard error.

Discussion

According to CLT, construal level can change depending on how people perceive the distance of an object (Trope & Liberman, 2010). The theory divides psychological distance into four subdimensions of temporal distance, spatial distance, social distance, and probabilistic distance (Liberman & Trope, 2014; Trope & Liberman, 2010). In this study, we examined the physiological distance of breast cancer and Alzheimer’s disease messages based on three measures of probabilistic distance, social distance, and temporal distance. Preliminary results indicated that female college students in their 20s perceived the breast cancer message to have less psychological distance than the Alzheimer’s disease message in terms of temporal, social, and probabilistic distance (supporting Hypothesis 1-1). In addition, Appendix 1 had high abstract representation scores and Appendix 2 had high concrete representation scores (supporting Hypothesis 1-2).

Our results showed that participants who received the psychologically proximal health threat information (on breast cancer) had a more positive attitude toward the “HOW” advertisements emphasizing specific methods, while participants who received the psychologically distal health threat information (on Alzheimer’s disease) had similar attitudes toward the “WHY” advertisement, emphasizing the purpose or value of exercise (partially supporting Hypothesis 1-3). In addition, participants in the breast cancer group had stronger SPAI than GPAI, while those in the Alzheimer’s disease group had stronger GPAI than SPAI, but the difference was slight (partially supporting Hypothesis 1-4). In conclusion, while the breast cancer message induced low-level construal, which led to improved attitude toward the advertisement emphasizing the “HOW” process and in SPAI, the Alzheimer’s disease message appears to have not induced enough high-level construal to sufficiently improve attitude toward the advertisement emphasizing the “WHY” process, and GPAI. In general, the link between attitude and behavior can be weak, so it is not desirable to overemphasize the idea that strong attitudes toward the advertisement lead to actual participatory behavior.

These results partially support the CLT and previous research suggesting that, when participants perceive greater psychological distance, high-level construal is induced so that messages emphasizing abstract values are more effective, while when the perceived psychological distance is less, low-level construal is induced so that messages emphasizing specific characteristics are more effective (Ledgerwood, Wakslak, & Wang, 2010; Trope & Liberman, 2003, 2010).

In this study, for participants who received the psychologically proximal breast cancer message, low-level construal was induced (a concrete, “HOW” process), which influenced SPAI. However, for those who received the psychologically distal Alzheimer’s disease message, high-level construal (an abstract, “WHY” process) was not induced to a sufficient degree, failing to increase GPAI.

One reason that the results differed from the predictions for the Alzheimer’s disease message may be an item priming effect, which refers to the phenomenon where items presented first influence the processing of items presented subsequently (Podsakoff, Mackenzie, Lee, & Podsakoff, 2003). When measuring GPAI, the phrase “next week” implies the near future, which differs in temporal distance and abstract construal from GPAI, which reflects temporally distant, abstract representations. Therefore, the adverbial phrase, “next week,” was perhaps associated by participants with the near future, resulting in imprecise measurement of GPAI, which should reflect abstract representation.

A second reason appears to be the result of the health threat information reinforcing participants’ self-change motives in high-level construal. Self-change motives, such as self-assessment and self-improvement, are tuned to diagnosing one’s strengths and weaknesses and using this information to better one’s self (Sedikides & Hepper, 2009). Belding, Naufel, and Fujita (2015) suggest that negative feedback in high-level construal increases self-change motives to diagnose and improve one’s weakness. Therefore, the Alzheimer’s disease message, which led to high-level construal, would have reinforced self-change motives, resulting in acceptance of information that is threatening to oneself, increasing not only GPAI but also SPAI.

Study 2

Overview

Study 1 showed a relationship between the induction of low-level construal by the psychologically proximal breast cancer message and SPAI. Study 2 examined the effect of self-affirmation on SPAI depending on health threat messages that differ in psychological distance. This may explain the case in which self-affirmation does not lead to increased intention. For example, among participants who received the psychologically proximal message, intention increase would likely be observed when SPAI is measured in SA groups. In contrast, participants who received the psychologically distal Alzheimer’s disease message would not show an increase in SPAI, despite being SA.

Hypotheses

Hypothesis 2-1: For the psychologically proximal breast cancer message, self-affirmation in female college students will have a positive effect on SPAI.

Hypothesis 2-2: For the psychologically distal Alzheimer’s disease message, self-affirmation in female college students will not have a positive effect on SPAI.

Method

Self-affirmation and control tasks

The most common method of self-affirmation manipulation is a writing task. In Study 2, the SA group was asked to write about their experiences of pride, self-esteem, or natural strengths (Klein, Blier, & Janze, 2001). Participants in the SA group were asked to write about their talents, competencies, and achievement experiences, such as receiving a high score, or getting an award or a promotion (Schimel, Amdt, Banko, & Cook, 2004). For the non-SA (NSA) group, participants were asked to write a description of their journey to a store, describing the places that they visited (Napper, Harris, & Epton, 2009).

Participants

Female college students in their 20s who did not engage in regular exercise were openly recruited on the Ewha Womans University website. There were 85 participants for Study 2. Regular exercisers were not appropriate for this study, as they would not perceive the health messages as being threatening. Thus, participants suspected of participating in regular exercise based on the TTM questionnaire (nine participants) were excluded. Participants were randomly assigned to four experimental groups, as shown in Table 2. Participants received drink coupons worth $5.

Table 2.

Number of participants and average age in each experimental group.

Experimental group	n	Age
Experimental group	n	M	SD
NSA, breast cancer	21	20.71	2.14
SA, breast cancer	17	21.06	1.71
NSA, Alzheimer’s disease	19	20.95	2.39
SA, Alzheimer’s disease	18	21.50	2.04
Total	75	21.04	2.07

SA: self-affirmed; NSA: non-SA; SD: standard deviation.

Measures

Affirmation manipulation

We used the five-item scale developed by Napper et al. (2009), which contains the following items: “The writing task reminded me of my positive qualities,” “The writing task helped me to understand who I am,” “The writing task was an opportunity to see what my value was,” “The writing task helped me identify my natural strengths,” and “The writing task reminded me of my overall value.” All items were scored on a five-point scale (1 = strongly disagree to 5 = strongly agree).

Acceptance of message conclusions

The message acceptance (Sherman et al., 2000) was modified to properly fit the two items used in this study and consisted of a nine-point scale (Cronbach’s α = .87): “To what extent do you agree or disagree that there is an association between low physical activity and breast cancer (or Alzheimer’s disease)?” and “How important do YOU think it is that women participate in physical activity to avoid breast cancer (or Alzheimer’s disease)?”

The physical activity level, psychological distance, message belief, and the items used to measure SPAI were identical to those used in Study 1.

Procedure

The self-affirmation task was performed. Participants in the SA group were given a writing task to confirm their own self-worth, and the participants in the NSA group were asked to complete the journey task. Both tasks took 15 minutes. Next, all participants completed the five-item self-affirmation questionnaire. In addition, participants were provided with health threat messages (breast cancer or Alzheimer’s disease, the same as in Study 1) that showed a difference in psychological distance and were surveyed regarding perceived psychological distance, message belief, acceptance of the provided health threat message, and SPAI.

Results

Manipulation check

To determine whether the self-affirmation task was successful, we conducted an independent sample t test between the writing task group and the journey task group on the five items confirming their worth. Results revealed that the writing task group (SA) had statistically higher values of self-affirmation than the journey task group (NSA) in all five items, as shown in Table 3. Therefore, the writing task was useful for increasing self-affirmation.

Table 3.

Responses to the outcome measures by self-affirmation condition.

Measure	WT (n = 38)		JT (n = 38)		t	p	95% CI
Measure	M	SD	M	SD	t	p	LL	UL
The questionnaire made me
Think about positive aspects of self	4.45	0.60	2.97	1.26	6.49	<.001	1.02	1.92
Focus my attention on who I am	3.87	0.84	2.66	1.07	5.46	<.001	.76	1.65
Aware of things I valueabout myself	3.95	0.73	2.79	1.06	5.50	<.001	.73	1.57
Think about things personallyimportant to me	4.29	0.73	2.63	0.97	8.40	<.001	1.26	2.05
Think about my values	4.16	0.71	2.68	1.16	6.64	<.001	1.03	1.91

WT: writing task; JT: journey task; CI: confidence interval; LL: lower limit; UL: upper limit; SD: standard deviation.

Independent sample t tests were conducted between the Alzheimer’s disease message group and the breast cancer message group for message belief, social distance, temporal distance, and probabilistic distance. There were statistically significant between-group differences in social distance (t(73) = 3.12, p < .01), temporal distance (t(73) = 17.95, p < .001), and probabilistic distance (t(73) = 7.99, p < .001), as shown in Table 4. There were no statistically significant differences in message belief (t(73) = 0.06, p = .950) between the Alzheimer’s disease message group and the breast cancer message group.

Table 4.

Responses to the outcome measures by message condition in Study 2.

Measure	AD (n = 37)		BC (n = 38)		t	p	95% CI
Measure	M	SD	M	SD	t	p	LL	UL
Message belief	5.65	0.82	5.82	0.86	0.86	.395	−0.22	0.56
Social distance	5.27	1.43	4.21	1.51	3.12	.003	0.38	1.74
Temporal distance	45.68	10.94	11.08	4.59	17.95	.000	30.75	38.44
Probabilistic distance	16.68	8.14	42.24	17.73	7.99	.000	19.18	31.94

AD: Alzheimer’s disease; BC: breast cancer; CI: confidence interval; LL: lower limit; SD: standard deviation; UL: upper limit.

Self-affirmation and acceptance of the threatening information

An independent sample t test revealed that affirmed participants (M = 7.77, SD = 0.72) were more accepting of the article’s conclusions than were nonaffirmed participants (M = 6.96, SD = 1.05), t(73) = 3.84, p < .001. Thus, for the women participating in low amounts of physical activity, self-affirmation reduced the defensive processing of the threatening message.

Influence of self-affirmation on SPAI by type of threatening health message

To test the hypothesized difference in type of self-affirmation influence on SPAI according to type of threatening health message, a 2 × 2 between-subjects ANOVA was performed. There was a statistically significant difference in SPAI between breast cancer (M = 4.21, SD = 0.61) and Alzheimer’s disease messages (M = 3.81, SD = 0.91) (F(1, 71) = 5.59, p = <.05, $η_{p}^{2}$ = .07). There was no SPAI difference between affirmed participants (M = 3.99, SD = 0.99) and nonaffirmed participants (M = 4.04, SD = 0.58), (F(1, 71) = 0.05, p = .824, $η_{p}^{2}$ = .00). The interaction effect was qualified by the predicted type of message × SA/NSA interaction variables, F(1, 71) = 5.26, p < .05, $η_{p}^{2}$ = .07.

Interpreting the interaction effect

As shown in Figure 3, in the breast cancer group, SPAI was higher for SA than NSA (supporting Hypothesis 2-1). However, in the Alzheimer’s disease group, SPAI was higher for NSA than SA (supporting Hypothesis 2-2).

Figure 3.

Specific physical activity intention (+SE) by type of health threat information and self-affirmation condition. SE: standard error.

Discussion

In previous self-affirmation research, self-affirmation has not shown significant effects on intention and behavioral change (Good & Abraham, 2011; Jessop et al., 2014; Sweeney & Freitas, 2016). Based on the results of our study, such as with regard to Hypothesis 2-2, previous findings that have not supported the self-affirmation theory can be explained by the mismatch between psychological distance of the threatening health information and construal level.

This conclusion can be explained by Jessop et al. (2014), who classified intentions into implementation intentions and no implementation intentions. Implementation intention was measured after concretely considering how, where, or when people participate in physical activity, so it was similar to the “how” representations of the low-level construal suggested by CLT. The results of Jessop et al. (2014) showed that SA and implementation intention groups have lower rates of participation in physical activity than an NSA and implementation intention group. In our view, considering the male participants in this study, the threatening health information (risk of breast cancer) was psychologically distal information, which would not induce low-level construal. It seems that, although male students were SA, low-level construal was not induced, which failed to increase implementation intention reflecting concrete representations.

In a study by Reed and Aspinwall (1988), female college students were provided with the threatening health message (e.g., fibrocystic breast disease) as psychologically close information. However, the intention to reduce caffeine consumption was measured with two items that seemingly intended to describe the general situation about caffeine consumption rather than the participants’ specific situation. If the researchers had measured the intention using a specific situation such as implementation intention, self-affirmation may have had a positive effect on behavioral change.

The efficacy information presented in the study by Good and Abraham (2011) was a way to increase the intention of using sunscreen and reduce ultraviolet (UV) exposure by adding the sentence (“you can carry a small bottle, sachet, or pouch of sunscreen so you can’t forget” and “sunscreen can block up to 96% of UV radiation”) before the intention was measured in their study. The intention measured after providing efficacy information would have reflected specific representation. As a result, the intention to use sunscreen and to reduce UV exposure was increased when the skin cancer message was provided along with efficacy information and self-affirmation. However, in the skin aging message, the intention did not improve even though efficacy information and self-affirmation were provided. Our results may help to explain the contradictory result of that study. The intention measured after providing efficacy information would have reflected specific representation; therefore, if young women perceived the skin aging message to have distal psychological distance, the intention would not increase even though it was SA.

According to Sweeney and Freitas (2016), general intention to increase consumption of fruit and vegetables showed a positive influence through the interaction between health risk and self-affirmation, when information on the health benefits of fruit and vegetable consumption was provided to 203 healthy college students. However, there was no effect on specific intention to increase consumption of fruit and vegetables. Based on our findings, this result also suggests that the participants would perceive health-threatening information as psychologically distal.

After controlling for the covariate variables (e.g., response-efficacy), the previous study analyzed data to resolve the inconsistency of the effect of self-affirmation on intention and behavior (Epton & Harris, 2008). In this study, the consistency of psychological distance and construal level of threatening health messages through CLT appears to overcome the low consistency of self-affirmation research. Therefore, to expect a positive effect on intention and behavior change in self-affirmation research in the future, intention reflecting specific or abstract representation should be measured according to the psychological distance of the threatening health information.

In future studies, empirical research based on an integrated approach, such as combining self-affirmation theory and CLT, are needed in the various domains of health behavior (drinking alcohol, smoking, vegetable intake, etc.) to address the low consistency of findings in existing self-affirmation studies.

Appendix 1: Advertisement Emphasizing High-Level Construal or “WHY” Process

Appendix 2: Advertisement Emphasizing Low-Level Construal or “HOW” Process

Footnotes

Article Notes

References

Amit

Algom

Trope

(2009). Distance-dependent processing of pictures and words. Journal of Experimental Psychology: General, 138, 400–415. doi:10.1037/a0015835

Armitage

C. J.

Harris

Hepton

Napper

(2008). Self-affirmation increases acceptance of health-risk information among UK adult smokers with low socioeconomic status. Psychology of Addictive Behaviors, 22, 88–95. doi:10.1037/0893-164X.22.1.88

Armitage

C. J.

Sprigg

C. A.

(2010). The roles of behavioral and implementation intentions in changing physical activity in young children with low socioeconomic status. Journal of Sport and Exercise Psychology, 32, 359–376. doi:10.1123/jsep.32.3.359

Belding

J. N.

Naufel

K. Z.

Fujita

(2015). Using high-level construal and perceptions of changeability to promote self-change over self-protection motives in response to negative feedback. Personality and Social Psychology Bulletin, 41, 822–838. doi:10.1177/0146167215580776

Berkowitz

Cottingham

D. R.

(1960). The interest value and relevance of fear arousing communications. Journal of Abnormal and Social Psychology, 60, 37–43. doi:10.1037/h0045247

Correll, J., Spencer, S. J., & Zanna, M. P. (2004). An affirmed self and an open mind: Self-affirmation and sensitivity to argument strength. Journal of Experimental Social Psychology, 40, 350–356. doi: 10.1016/j.jesp.2003.07.001

Ditto

P. H.

Lopez

D. F.

(1992). Motivated skepticism: Use of differential decision criteria for preferred and nonpreferred conclusions. Journal of Personality and Social Psychology, 63, 568–584. doi:10.1037/0022-3514.63.4.568

Dunning

Leuenberger

Sherman

D. A.

(1995). A new look at motivated inference: Are self-serving theories of success a product of motivational forces?. Journal of Personality and Social Psychology, 69, 58–68. doi:10.1037/0022-3514.69.1.58

Epton

Harris

P. R.

(2008). Self-affirmation promotes health behavior change. Health Psychology, 27, 746–752. doi:10.1037/0278-6133.27.6.746

10.

Epton

Harris

P. R.

Kane

Van Koningsbruggen

G. M.

Sheeran

(2015). The impact of self-affirmation on health-behavior change: A meta-analysis. Health Psychology, 34, 187. doi:10.1037/hea0000116

11.

Good

Abraham

(2011). Can the effectiveness of health promotion campaigns be improved using self-efficacy and self-affirmation interventions? An analysis of sun protection messages. Psychology and Health, 26, 799–818. doi:10.1080/08870446.2010.495157

12.

Harris

P. R.

(2011). Self-affirmation and the self-regulation of health behavior change. Self and Identity, 10, 304–314. doi:10.1080/15298868.2010.517963

13.

Harris

P. R.

Napper

(2005). Self-affirmation and the biased processing of threatening health-risk information. Personality and Social Psychology Bulletin, 31, 1250–1263. doi:10.1177/0146167205274694

14.

Hausenblas

H. A.

Rhodes

R. E.

(2017). Exercise psychology: Physical activity and sedentary behavior. Burlington, MA: Jones & Bartlett Learning.

15.

Jessop

Simmonds

Sparks

(2009). Motivational and behavioral consequences of self-affirmation interventions: A study of sun- screen use among women. Psychology & Health, 24, 529–544. doi:10.1080/08870440801930320

16.

Jessop

Sparks

Buckland

Harris

P. R.

Churchill

(2014). Combining self-affirmation and implementation intentions: Evidence of detrimental effects on behavioral outcomes. Annals of Behavioral Medicine, 47, 137–147. doi:10.1007/s12160-013-9536-0

17.

Klein

W. M.

Blier

H. K.

Janze

A. M.

(2001). Maintaining positive self-evaluations: Reducing attention to diagnostic but unfavorable social comparison information when general self-regard is salient. Motivation and Emotion, 25, 23–40. doi:10.1023/A:1010607821908

18.

Kunda

(1987). Motivated inference: Self-serving generation and evaluation of causal theories. Journal of Personality and Social Psychology, 53, 636–647. doi:10.1037/0022-3514.53.4.636

19.

Kunda

(1990). The case for motivated reasoning. Psychological Bulletin, 108, 480–498. doi:10.1037/0033-2909.108.3.480

20.

Ledgerwood

Wakslak

C. J.

Wang

M. A.

(2010). Differential information use for near and distant decisions. Journal of Experimental Social Psychology, 46, 638–642. doi:10.1016/j.jesp.2010.03.001

21.

Liberman

Chaiken

(1992). Defensive processing of personally relevant health messages. Personality and Social Psychology Bulletin, 18, 669–679. doi:10.1177/0146167292186002

22.

Liberman, N., & Trope, Y. (1998). The role of feasibility and desirability considerations in near and distant future decisions: A test of temporal construal theory. Journal of Personality and Social Psychology, 75, 5–18. doi: 10.1037/0022-3514.75.1.5

23.

Liberman

Trope

(2008). The psychology of transcending the here and now. Science, 322, 1201–1205. doi:10.1126/science.1161958

24.

Liberman

Trope

(2014). Traversing psychological distance. Trends in Cognitive Sciences, 18, 364–369. doi:10.1016/j.tics.2014.03.001

25.

MacKenzie

S. B.

Lutz

R. J.

(1989). An empirical examination of the structural antecedents of attitude toward the ad in an advertising pretesting context. The Journal of Marketing, 53, 48–65. doi:10.2307/1251413

26.

Marcus

B. H.

Banspanch

S. W.

Lefebvre

R. C.

Rossi

J. S.

Carleton

R. A.

Abrams

D. B.

(1992). Using the stage of change model to increase the adoption of physical activity among community participants. American Journal of Health Promotion, 6, 424–429. doi:10.4278/0890-1171-6.6.424

27.

Napper

Harris

P. R.

Epton

(2009). Developing and testing a self-affirmation manipulation. Self and Identity, 8, 45–62. doi:10.1080/15298860802079786

28.

Pietersma

Dijkstra

(2011). Do behavioral health intentions engender health behavior change? A study on the moderating role of self-affirmation on actual fruit intake versus vegetable intake. British Journal of Health Psychology, 16, 815–827. doi:10.1111/j.2044-8287.2011.02018.x

29.

Podsakoff

P. M.

MacKenzie

S. B.

Lee

Y. J.

Podsakoff

N. P.

(2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879–903. doi:10.1037/0021-9010.88.5.879

30.

Reed

M. B.

Aspinwall

L. G.

(1998). Self-affirmation reduces biased processing of health-risk information. Motivation and Emotion, 22, 99–132. doi:10.1023/A:1021463221281

31.

Schimel

Amdt

Banko

Cook

(2004). Not all self-affirmations were created equal: The cognitive and social benefits of affirming the intrinsic (vs. extrinsic) self. Social Cognition, 22, 75–99. doi:10.1521/soco.22.1.75.30984

32.

Sedikides

Hepper

E. G. D.

(2009). Self-improvement. Social & Personality Psychology Compass, 3, 899–917. doi:10.1111/ j.1751-9004.2009.00231.x

33.

Sherman

D. A.

Nelson

L. D.

Steele

C. M.

(2000). Do messages about health risks threaten the self? Increasing the acceptance of threatening health messages via self-affirmation. Personality and Social Psychology Bulletin, 26, 1046–1058. doi:10.1177/01461672002611003

34.

Steele

C. M.

(1988). The psychology of self-affirmation: Sustaining the integrity of the self. Advances in Experimental Social Psychology, 21, 261–302. doi:10.1016/S0065-2601(08)60229-4

35.

Steele

C. M.

Spencer

S. J.

Lynch

(1993). Self-image resilience and dissonance: The role of affirmational resources. Journal of Personality and Social Psychology, 64, 885–896. doi:10.1037/0022-3514.64.6.885

36.

Sweeney

A. M.

Freitas

A. L.

(2016). Self-affirmation impacts behavioral intentions but not preferences for delayed outcomes. Journal of Experimental Social Psychology, 67, 34–42. doi:10.1016/j.jesp.2015.09.005

37.

Sweeney

A. M.

Moyer

(2015). Self-affirmation and responses to health messages: A meta-analysis on intentions and behavior. Health Psychology, 34, 149–159. doi:10.1037/hea0000110

38.

Trope

Liberman

(2003). Temporal construal. Psychological Review, 110, 403–421. doi:10.1037/0033-295X.110.3.403

39.

Trope

Liberman

(2010). Construal-level theory of psychological distance. Psychological Review, 117, 440–463. doi:10.1037/a0018963

40.

Webb

T. L.

Sheeran

(2006). Does changing behavioral intentions engender behavior change? A meta-analysis of the experimental evidence. Psychological Bulletin, 132, 249. doi:10.1037/0033-2909.132.2.249