Abstract
Previous findings suggest that patients choose physicians whose affective focus matches how they ideally want to feel (Sims et al., 2014). For instance, the more people wanted to feel excitement, the more likely they were to hypothetically choose a new physician who promoted excitement. What remains unknown is whether this match shapes how patients actually respond to physicians after being assigned to them (i.e., whether they adhere to physicians’ recommendations more and evaluate physicians more positively). To this end, community adults reported their global ideal affect and actual affect (how they ideally want to feel and actually feel during a typical week, respectively), and were randomly assigned to receive health recommendations from either a physician who expressed and promoted high arousal positive states (HAP) (e.g., excitement), or one who expressed and promoted low arousal positive states (LAP) (e.g., calm). For the next five days, participants reported their daily adherence to the recommendations and their daily ideal and actual affect. At the end of the week, participants evaluated their physician. As predicted, the more participants wanted to feel HAP, the more they adhered to the “HAP-focused” physician’s recommendations, and the more participants wanted to feel LAP, the more they adhered to the “LAP-focused” physician’s recommendations. Participants also evaluated their physician more positively when his affective focus matched their ideal affect. Neither global nor daily actual affect systematically predicted how patients responded to their physicians. These findings suggest that patients respond better to physicians whose affective focus matches their ideal affect.
Keywords: Ideal Affect, Adherence, Patient-Physician Communication, Health
In the United States, getting patients to follow their doctor’s advice and ensuring that they trust their doctor are major challenges to improving health care quality and health outcomes. On average, approximately 40% of American patients have trouble following their physicians’ recommendations (Martin, Williams, Haskard, & DiMatteo, 2005). Moreover, while the majority of patients report that their physicians provide adequate care, patients vary considerably in how positively they respond to their physicians (e.g., AHRQ, 2013; Gao, McCullough, Agarwal, & Jha, 2012; Hall, Blanch-Hartigan, & Roter, 2011). Multiple factors are associated with better health care experiences including physician characteristics (e.g., whether physicians have received training in communication skills; Haskard Zolnierek & DiMatteo, 2009) and patient characteristics (e.g., how conscientious patients are; Hill & Roberts, 2011). As we review below, a growing literature suggests that physicians’ and patients’ affective characteristics (i.e., their affective states, traits, and expressions) are especially critical for optimal health care. Much less is known, however, about how physicians’ affective characteristics interact with patients’ affective characteristics to influence patients’ health care experiences. In the current study, we examined whether the effect of physicians’ affective focus (i.e., whether they express and promote excited vs. calm states) on patients’ responses to their physicians varied as a function of whether patients ideally wanted to feel excitement vs. calm states. We focused on two responses that are critical to health: (1) patients’ adherence to physicians’ recommendations, and (2) patients’ evaluation of physicians. Based on affect valuation theory, we predicted that when a physician’s affective focus matches his/her patient’s ideal affect, the patient is more likely to adhere to the physician’s recommendations and to evaluate that physician positively (e.g., as more trustworthy).
The Effect of Physician Affect on Patient Response
Physicians play a key role in shaping their patients’ engagement in health-related behaviors; however, some physicians are more effective than others. In the United States, physicians who show more positive affect seem to be more effective with their patients than those who show less positive affect (for reviews, see Beck, Daughtridge, & Sloane, 2002; Blasi, Harkness, Ernst, Georgiou, & Kleijnen, 2001; DiMatteo, 1979; Levine & Ambady, 2013; Roter, Frankel, Hall, & Sluyter, 2006). For instance, patients report greater satisfaction, are more likely to keep follow-up appointments, and show better physical and cognitive outcomes after discharge when interacting with health care providers who express more positive affect (e.g., Ambady, Koo, Rosenthal, & Winograd, 2002; Cousin, Schmid Mast, & Roter, 2012; Griffith, Wilson, Langer, & Haist, 2003; Hall, Roter, & Rand, 1981; Schmid Mast, Hall, & Roter, 2008). Even when interacting with a virtual physician avatar, undergraduates made more emotional statements and were quicker to respond when the physician avatar communicated in a caring manner (e.g., made empathetic, reassuring statements) than when the physician avatar did not (Schmid Mast, et al., 2008). These findings suggest that patients’ perceptions of their clinical encounters and their actual engagement in health-promoting behavior depends in part on physicians’ affective focus.
The Effect of Patient Affect on Patient Response
Patients who experience more positive feelings in general may be more likely to engage in health-promoting behaviors, such as adhering to their physician’s recommendations (cf. Ong, 2010; cf. Pressman & Cohen, 2005). One meta-analysis revealed that the less depressed people were (or the more positive affect they experienced), the more they adhered to non-psychiatric treatments (DiMatteo, Lepper, & Croghan, 2000). Additionally, positive dispositional characteristics, like optimism, have been associated with health-related risk perception and health promoting behaviors (see Carver, Scheier, & Segerstrom, 2010, for review). For example, the more optimistic people were, the more likely they were to perform skin cancer self-examinations (Luo & Isaacowitz, 2007).
Patients’ positive feelings during clinical encounters have also been associated with how patients respond to their physicians. For instance, patients’ positive feelings toward their pediatricians predicted subsequent satisfaction, cooperation and perceived quality of care (Locke, 1996). Patients expressing more satisfaction and less anxiety during their appointments (based on recordings of their speech during clinical encounters) were more likely to return for their follow-up appointments (Hall, et al., 1981). Similarly, the more patients joked and laughed with their physicians during their appointments, the more they reported understanding their physicians, and the more confident patients were that they could manage their own health (Pawlikowska, Zhang, Griffiths, van Dalen, & van der Vleuten, 2012).
But does the effect of patients’ affective characteristics vary as a function of physicians’ affective characteristics? Although much less research has addressed this question, findings do suggest that patients adhere more to health recommendations that are presented in ways that match their motivational characteristics and goals (Aaker & Lee, 2001; Ledford, 2012; Lee & Aaker, 2004; Mann, Sherman, & Updegraff, 2004; Notthoff & Carstensen, 2014; Rothman, Bartels, Wlaschin, & Salovey, 2006 for review; Sherman, Mann, & Updegraff, 2006; Latimer et al., 2008; Zhao & Pechmann, 2007). For instance, patients who prefer that their doctors take charge were more satisfied with a dominant virtual physician avatar than were patients who did not have such a preference (Cousin, et al, 2012). In the present study, we test the hypothesis that patients are more likely to adhere to a physicians’ health recommendations when that physician’s affective focus is consistent with patients’ ideal affect. Although previous work has suggested that health recommendations presented in a positive manner may elicit more positive feelings, and that these positive feelings facilitate adherence (Agrawal, Menon, & Aaker, 2007; Hall, et al., 1981; Locke, 1996; Pawlikowska, et al., 2012; Rothman & Salovey, 1997; Shen & Dillard, 2007), no studies have considered whether these relationships depend on the degree to which people want to feel different types of positive states (e.g., excitement vs. calm).
Interactions Between Physician and Patient Affect: Contributions from Affect Valuation Theory
Affect valuation theory (Tsai, Knutson, & Fung, 2006; Tsai, 2007) distinguishes the states people ideally want to feel (“ideal affect”) from the states they actually feel (“actual affect”). Whereas actual affect is a feeling in response to an outcome, ideal affect is a feeling that people consciously or unconsciously strive to achieve. In this way, ideal affect and actual affect serve different functions: actual affect represents how someone is feeling (“I feel good”), whereas ideal affect provides a way of interpreting and evaluating that state (“Am I feeling how I want to feel?” or “Does this feel right?”). Indeed, across a variety of studies, reports of actual affect and ideal affect are only moderately correlated (Sims, Tsai, Jiang, et al., 2014; Tsai, 2007; Tsai, et al., 2006; Tsai, Louie, Chen, & Uchida, 2007; Tsai, Miao, & Seppala, 2007; Tsai, Miao, Seppala, Fung, & Yeung, 2007), and structural equation modeling has revealed that actual affect and ideal affect are distinct constructs (Tsai, et al., 2006; Koopmann-Holm & Tsai, 2014).
Across cultures, most people ideally want to feel positive more than negative, but people vary between and within cultures in the specific types of positive affect that they ideally want to feel (Barrett, 1996; Chow & Berenbaum, 2012; Kämpfe & Mitte, 2009; Koopmann-Holm, Sze, Ochs, & Tsai, 2013; Ruby, Falk, Heine, Villa, & Silberstein, 2012; Rusting & Larsen, 1995; Scheibe, English, Tsai, & Carstensen, 2012; Tsai, 2007; Tsai, et al., 2006; Tsai, Louie, et al., 2007; Tsai, Miao, & Seppala, 2007; Tsai, Miao, Seppala, et al., 2007; Tsai, Sims, Fung, & Jiang, 2014; Västfjäll, Gärling, & Kleiner, 2001). For instance, while some people want to feel excitement and other high arousal positive (HAP) states more, others want to feel calm and other low arousal positive (LAP) states more. Although not the main focus of the present paper, AVT predicts that cultural factors shape people’s ideal affect more than their actual affect, whereas temperamental factors shape people’s actual affect more than their ideal affect, and several studies have supported this prediction (e.g., Tsai, Knutson, & Fung, 2006).
Finally, AVT predicts that people’s ideal affect drives much of their behavior (Tsai, 2007). In the current study, we were interested in examining whether ideal affect shapes behavior in health care contexts. In particular, we examined whether people’s ideal affect predicts their adherence to physicians’ recommendations and their evaluations of physicians after receiving those recommendations. We focused on these two outcomes because previous research suggests that both are commonly assessed and reliable indicators of health care quality, and both exert independent effects on health and health care utilization (Jerant, Fenton, Bertakis, & Franks, 2014; Manary, Boulding, Staelin, & Glickman, 2013).
Adherence to physicians’ recommendations
As mentioned above, AVT predicts that because people strive to achieve their ideal affective states, they often make decisions based on how they want to feel, consciously or not. Indeed, people’s ideal affect predicts people’s vacation preferences, leisure activities, and music choices (Mogilner, Aaker, & Kamvar, 2012; Tsai, 2007, Tsai, Louie, et al., 2007; Tsai, Miao, Seppala, et al., 2007; for review, see Tsai, Chim, & Sims, 2014). People also prefer consumer products that are not directly mood producing but that are marketed as eliciting a specific affective state (e.g., “energizing” vs. “calming” lotions). For instance, in one study, participants were randomly assigned to value either excited or calm states (Tsai, Chim, et al., 2014). When presented with pairs of different consumer products (e.g., relaxing vs. energizing deodorant, chewing gum, lotions), regardless of how they actually felt, participants in the value excitement condition selected more exciting (vs. calm) consumer products than did participants in the value calm condition.
These studies, however, have primarily focused on recreational or leisure choices. In the present study, we predicted that patients’ ideal affect would also influence how likely patients were to adhere to a physician’s recommended health behaviors when presented in line with their ideal affect. On a conscious level, patients might gauge how likely it is that the recommended behaviors will make them feel how they want to feel. On an unconscious level, patients might attend more to a physician’s claims about the benefits of a health-promoting behavior and feel more motivated to try the behavior if presented in a way that matches patients’ ideal affect.
Evaluations of physicians
Ideal affect also shapes how people perceive others. People often make inferences about others’ traits or abilities based on their affective expressions (e.g., Rule, Krendl, Ivcevic, & Ambady, 2013; Todorov, Pakrashi, & Oosterhof, 2009; for review, see Zebrowitz & Montepare, 2008). We propose that when evaluating a target person, people consciously or unconsciously evaluate how well that target person’s affective display matches their ideal. The more the affective display matches their ideal, the more positive of an impression people have of that target person. At an unconscious level, people may just like or “feel good” about the target person. At a conscious level, people may infer positive traits about the person. In support of this prediction, the more college students valued excitement states, the friendlier they rated targets who displayed excited (vs. calm) smiles. Conversely, the more students valued calm states, the friendlier they rated targets who displayed calm (vs. excited) smiles (Tsai, Chim, et al., 2014). At the cultural level, countries that value excitement states were more likely to have leaders who showed excited smiles in their official photos, and countries that valued calm states were more likely to have leaders who showed calm smiles in their official photos (Tsai et al., 2014).
Do similar processes occur in health care settings? In a recent study, we presented participants with a hypothetical scenario in which they were asked to imagine that their current physician was no longer available, and they needed to select a new physician. Participants evaluated and chose among physicians described as promoting HAP states, LAP states, or neither (Sims, Tsai, Koopmann-Holm, et al., 2014). The more participants wanted to feel HAP over the course of a typical week, the more likely they were to choose a HAP-focused physician. Similarly, the more participants wanted to feel LAP over the course of a typical week, the more likely they were to choose a LAP-focused physician. Participants also evaluated physicians whose affective expressions matched their ideals as more trustworthy, which in turn predicted their choice of physician. Importantly, participants’ actual trait-like experience of HAP and LAP did not predict physician choice or ratings of trustworthiness. Thus, patients’ ideal affect played a greater role than their actual affect when patients were deciding which physician to choose.
However, this study was limited in several ways. First, we used hypothetical scenarios. Thus, it remains unclear whether these findings generalize to patients’ actual responses to physicians in real life. Second, because participants viewed physicians once and did not have the opportunity to interact with them, it is unclear whether patients’ ideal affect continues to predict how patients respond to physicians after multiple contacts in which they are receiving some form of care. Third, participants were able to choose their preferred physician and therefore, it is unclear whether a similar process occurs when patients are assigned a physician. Fourth, the previous study was based on college students; thus, it is unclear whether these processes generalize to a sample of community adults.
The Present Study
To address gaps in our previous work and gaps in the larger literature, we recruited a community sample of adults, assessed their global ideal and actual affect (i.e., how they want to feel and actually feel in a typical week), and presented them with health recommendations from either a HAP-focused or LAP-focused physician on-line. For the following five days, we then assessed participants’ daily adherence to those recommendations as well as participants’ daily ideal and actual affect (i.e., how they wanted to feel and actually felt on a given day). Measuring ideal affect at both the global and daily levels allowed us to examine whether ideal affect would account for between as well as within person variation in adherence. At the end of the five days, participants also evaluated their physicians in terms of trustworthiness, competence, and knowledgeability.
Hypotheses
In terms of adherence, we predicted that: (1) the more patients valued HAP (ideal HAP), the more they would adhere to the HAP-focused physician’s recommendations, and the more patients valued LAP (ideal LAP), the more they would adhere to the LAP-focused physician’s recommendations, (2) ideal affect would predict adherence more than actual affect, and (3) ideal affect would predict both between-person and within-person variation in adherence.
In terms of patients’ evaluation of physicians, we predicted that (4) the more patients valued HAP, the more positively they would evaluate the HAP-focused physician, and the more patients valued LAP, the more positively they would evaluate the LAP-focused physician, and (5) ideal affect would predict evaluations of the physician more than actual affect.
Methods
Participants
A sample of community adults was recruited from the San Francisco Bay Area to “help evaluate a virtual health center pilot program.” Participants were recruited via newspaper advertisements, Craig’s list, and flyers distributed at public locations (e.g., libraries, grocery stores) and were required to live in the San Francisco Bay Area and have access to a computer with internet access. To minimize risk to participants, we screened them for any major health limitations that would prevent them from adhering to the recommended behaviors, as well as psychiatric distress, and cognitive impairment. There were 48 participants who expressed interest but were excluded for the following reasons: (1) failure to meet ethnic and age criteria (n = 33, see below), (2) major health limitations (n = 5; 2 had diabetes, 1 had kidney disease, 1 had a degenerative disk issue, and 1 had a leg injury), (3) lack of internet access (n = 1), or (4) residence outside of the San Francisco Bay Area (n = 8). One participant was excluded because of s/he attempted to take the study multiple times. No participants were excluded due to psychiatric distress or cognitive difficulties. The final sample consisted of 101 adults (age M = 56 years, SD = 12 years; Range = 27 - 87 years; 66% Female; 34% Male). Participants were compensated with a $30 gift card.
For the purpose of another study, participants were matched in terms of age and ethnicity (50 European Americans, 51 Asian Americans). While the effects of age and ethnicity were not the main focus of the current paper, we included both variables as potential covariates and moderators of our findings in all analyses.
Procedure
Participants were asked to “evaluate and provide feedback on a pilot test of a virtual health center.” Participants were randomly assigned to view either a physician emphasizing HAP states (“HAP-focused physician” condition, n = 50) or a physician emphasizing LAP states (“LAP-focused physician” condition, n = 51). Figure 1 illustrates the study procedures, which were approved by the Stanford University Institutional Review Board.
Figure 1.
Study procedure.
Each participant was scheduled for two phone sessions during which a research assistant guided them through portions of the virtual health center website to: (1) ensure that participants viewed the information provided on the website for similar amounts of time and in the same order, and (2) to provide support in case participants experienced any technical difficulties with the website.
During the first phone session, research assistants introduced participants to the “Virtual Bay Health Center” website, and then asked participants to complete the general health assessment on the website, which included measures of health status, typical engagement in health-promoting behaviors, and affect. To allow participants to focus on the assessment, research assistants hung up after providing instructions, but asked participants to call them if they had any questions. After completing the assessment, participants called the research assistant back; research assistants then helped participants navigate through the “Meet Your Virtual Physician” webpage (see Figure 2), where they viewed an introductory video presented by their assigned physician. Afterwards, participants evaluated the website and were scheduled for a second phone session.
Figure 2.
Screenshot of the “Meet the Virtual Physician” webpage on the Virtual Bay Health Center website. Left: Physician promoting high arousal positive (HAP) states. Right: Physician promoting low arousal positive (LAP) states. Photo is reprinted with permission.
Two days later, research assistants emailed participants a link to the “Feedback” webpage of the “Virtual Bay Health Center” and provided them with an ostensibly personalized login ID and password. Research assistants then called participants for the second phone session to help participants navigate through the “Feedback” page. On the “Feedback” page, participants read their health assessment results. All participants received a rating of “Good,” on a scale ranging from “Poor” to “Excellent,” in order to standardize what participants saw and so that participants were motivated to follow the physician’s recommendations. Underneath their health rating, all participants were shown the same six recommendations (described below) to follow over the next five days. Research assistants instructed participants to view the video on the “Feedback” page during which the physician described the benefits of his health recommendations.
For the next five days, participants were asked to report whether or not they engaged in each of the recommended behaviors using a daily log, which they accessed via a link on the feedback webpage. Because some of the recommendations involved behaviors in the evening (e.g., brisk walk after dinner) or before going to sleep (i.e., abstain from eating prior to bedtime), participants completed their daily logs in the morning about their behaviors during the previous day. At that time, participants also reported their ideal and actual affect for the previous day.
The day after participants completed their final daily log, they received a thank you email from their virtual physician with his photograph and a link to one last survey about their experience with the virtual health center. Thus, by the end of the week, participants had three contacts with the assigned physician (introduction, feedback/recommendations, thank you). Participants also had the opportunity to provide feedback or comments throughout the study, of which 93% did.1 Finally, all participants were provided with a written debriefing statement which disclosed the actual purpose of the study, and which explicitly stated that there was no virtual health center being piloted.
Virtual Physician Videos
Introductory Video (Day 1)
A 90-second introductory video of a “physician” expressing his views on patient care appeared on the “Meet Your Virtual Physician” webpage. Paid actors portrayed the physicians. Participants saw either the “HAP-focused” physician, who displayed a wide, toothy smile, and spoke in an enthusiastic manner, or the “LAP-focused” physician, who displayed a closed-mouth smile, and spoke in a soothing manner. The videos were pretested to ensure that the physicians were similar in terms of likeability, but that the HAP-focused physician conveyed more enthusiasm than calm compared to the LAP-focused physician (videos and pre-test data available upon request). As in previous work, descriptions of physicians’ medical training, views on patient care, and outside interests were also provided (Sims, Tsai, Koopmann-Holm, et al., 2014). HAP-focused and LAP-focused physicians were comparable in terms of medical training, but varied in terms of how HAP and LAP their views of patient care and outside interests were (e.g., whereas the HAP-focused physician wanted to increase “patients’ overall vitality,” the LAP-focused physician wanted patients to lead a “calm and tranquil lifestyle;” see Appendix A).
Because previous research suggests that patients prefer physicians of their same ethnicity (e.g., LaVeist & Nuru-Jeter, 2002), we recruited one European American actor and one Asian American actor to portray the HAP and LAP-focused physicians so that we could match participants and physicians by ethnicity.
Feedback Video (Day 3)
A 30-second video of each physician describing the benefits of following his recommendations appeared on the “Feedback” webpage. To maintain consistency with the physicians’ views on patient care, the HAP-focused physician told participants that following his recommendations would help them look and feel “energized,” whereas the LAP-focused physician told participants that following his recommendations would help them look and feel “rested.”
Physician Recommendations
The physician recommended that above and beyond their usual routine, patients should: (1) do at least 10 minutes of muscle strengthening activities per day (participants were given a link to a webpage with descriptions of these activities), (2) take a 30 minute brisk walk after dinner, (3) nap or rest every afternoon for at least 15-20 minutes, (4) drink no more than 1 caffeinated beverage per day, (5) drink at least 1 glass of water with every meal, and (6) abstain from eating at least 2 hours before bedtime. These recommendations were based on standard dietary, physical activity and sleep health recommendations made by various American health organizations (i.e., National Sleep Foundation, Mayo Clinic, U.S. Department of Health and Human Services). We also selected these behaviors because they could be construed as producing HAP or LAP (i.e., people could take a walk to feel “energized” or “rested”).
At the bottom of the “Feedback” webpage, participants were told explicitly that the “information and materials presented on the Virtual Student Health Center website, such as text, graphics, and other material” should not be considered a substitute for professional medical advice.
Instruments
Health
As part of participants’ initial health assessment, we administered the SF-36v2 health questionnaire (Ware, Kosinski, Dewey, & Gandek, 2000). We calculated physical and mental health composite scores standardized against population norms ranging from 0 to 100 with a mean of 50 and standard deviation of 10. Internal consistencies were high for the physical (α = .82) and mental (α = .86) health subscales. The subscales were not significantly correlated with each other (r = −.14), which is consistent with normative samples (Ware et al., 2000). Mean physical health (M = 51.71, SD = 8.75) and mental health (M = 49.40, SD = 10.80) ratings for the current sample were comparable with national averages. Participants also reported whether or not they regularly engaged in several health behaviors, including those recommended in the current study (e.g., walking, napping), with the exception of typical water consumption. On average, participants engaged in two of five of the recommended behaviors prior to the study (SD = 0.89). When we treated participants’ regular engagement in the recommended health behaviors as a covariate in our analyses, the results did not change, and therefore, we do not discuss this variable further.2
Ideal and Actual Affect
During the health assessment at the beginning of the week, we assessed global reports of ideal and actual affect (e.g., how often do you actually feel/ideally want to feel <state> in a typical week?). To maintain consistency with the daily measure (see below), we used a subset of the four HAP and four LAP terms from the Affect Valuation Index (AVI; Tsai et al., 2006). The HAP aggregate included “excited” and “enthusiastic,” and the LAP aggregate included “calm” and “relaxed.” Internal consistencies of aggregates were high (ordinal αs =.73 for ideal HAP, .92 for ideal LAP, .86 for actual HAP, and .85 for actual LAP).
We also assessed daily reports of ideal and actual affect (e.g., how much did you feel/ideally want to feel <state> yesterday?). We used only a subset of the HAP (excited, enthusiastic) and LAP (calm, relaxed) items from the AVI in order to reduce the number of items that participants had to rate each day. Across the five days, test-retest reliability for daily ideal and actual affect was high (ideal HAP α = .91, ideal LAP α = .92, actual HAP α = .85, actual LAP α = .88). Within each day, internal consistencies of ideal and actual affect aggregates were also high (ranges of ordinal αs =.79-.91 for ideal HAP, .88-.97 for ideal LAP, .77-.91 for actual HAP, and .88-.97 for actual LAP). To validate the global reports, we also created aggregated daily ideal and actual affect ratings by calculating the mean of daily ideal HAP, ideal LAP, actual HAP, and actual LAP across the five days and then assessed whether global and aggregated daily reports were correlated.
Adherence to Physician Recommendations
For five days, participants indicated the extent to which they engaged in each behavior (i.e., number of minutes doing muscle strengthening, walking, resting, eating prior to bedtime; number of caffeinated beverages and glasses of water taken with each meal that day). We coded reported adherence as either 0 = no adherence; 1 = some adherence (i.e., participants attempted to adhere to each behavior but did not meet the exact recommendations), or 2 = full adherence (i.e., participants adhered to each behavior at or above recommended levels). For each day, we created a daily adherence score calculated across all of the recommended behaviors; scores ranged from 0 (no adherence to any recommendation for a given day) to 12 (full adherence to all six recommendations for a given day).
At the end of the week, participants rated how each behavior made them feel in terms of valence (1 = Very Negative to 7 = Very Positive) and arousal (1 = Very Inactivated to 7 = Very Activated). Participants reported that all of the behaviors made them feel more positive than negative (M = 5.52, SE = .10) and more activated than inactivated (M = 5.13, SE = .10). These ratings did not significantly differ between the HAP and LAP conditions; across age, gender, ethnicity, or health; and were not associated with ideal or actual affect within conditions, suggesting that the behaviors were perceived similarly regardless of patients’ ideal affect.
Evaluation of Physician
At the end of the 5 days, in the final survey, participants were shown the same photo and description of their physician from the beginning of the week. They were then asked to rate how much they trusted their physician (1 = Extremely Trusted to 7 = Extremely Distrusted), how knowledgeable their physician was (1 = Extremely Unknowledgeable to 7 = Extremely Knowledgeable), and how competent their physician was (1 = Extremely Incompetent to 7 = Extremely Competent). We reverse scored trust ratings and aggregated across the three items to create an overall positive evaluation of the physician (α = 0.87).
Data Analyses and Results
Table 1 shows the bivariate correlations, means, and standard errors for all key variables in the study (means and standard errors by age group and ethnicity are available upon request). As illustrated in this table, global and daily ratings of ideal affect and of actual affect were moderately correlated, ranging from .47 to .60. Moreover, repeated measures analyses revealed that across conditions, there were no significant differences between global and aggregated daily reports of ideal affect (global vs. aggregated daily ideal HAP: F (1, 99) = 0.26, p = .61, partial η2 = .00; global vs. aggregated daily ideal LAP: F (1, 99) = 0.45, p = .51, partial η2 = .00) or actual affect (global vs. aggregated daily actual HAP: F (1, 99) = 1.65, p = .20, partial η2 = .02; global vs. aggregated daily actual LAP: F (1, 99) = 1.82, p = .18, partial η2 = .02). Findings did not change when controlling for the overlap between actual and ideal affect (i.e., covarying one when examining the other). These findings illustrate that global ratings accurately reflect how participants actually and ideally wanted to feel over the course of the week based on their daily reports.
Table 1. Bivariate between-subjects correlations among key study variables.
1. | 2. | 3. | 4. | 5. | 6. | 7. | 8. | 9. | 10. | 11. | 12. | 13. | 14. | 15. | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1. Global Ideal HAP | 1 | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
2. Global Ideal LAP | 0.36*** | 1 | - | - | - | - | - | - | - | - | - | - | - | - | - |
3. Global Actual HAP | 0.34*** | 0.15 | 1 | - | - | - | - | - | - | - | - | - | - | - | - |
4. Global Actual LAP | 0.1 | 0.25* | 0.45*** | 1 | - | - | - | - | - | - | - | - | - | - | - |
5. Average Daily Ideal HAP | 0.53*** | 0.23* | 0.32** | 0.05 | 1 | - | - | - | - | - | - | - | - | - | - |
6. Average Daily Ideal LAP | 0.13 | 0.48*** | 0.11 | 0.11 | 0.40*** | 1 | - | - | - | - | - | - | - | - | - |
7. Average Daily Actual HAP | 0.28** | 0.05 | 0.60*** | 0.22* | 0.60*** | 0.28** | 1 | - | - | - | - | - | - | - | - |
8. Average Daily Actual LAP | 0.14 | 0.17† | 0.34** | 0.47*** | 0.23* | 0.45*** | 0.50*** | 1 | - | - | - | - | - | - | - |
9. Average Daily Adherence | 0.05 | −0.41 | 0.17† | 0.19† | 0.08 | 0.01 | 0.25* | 0.28** | 1 | - | - | - | - | - | - |
10. Positive Physician Rating | 0.79† | 0.07 | 0.16 | 0.18† | 0.20* | 0.16 | 0.25* | 0.06 | 0.20† | 1 | - | - | - | - | - |
11. Gender (Female = 1) | 0.02 | −0.03 | −0.04 | −0.20† | −0.15 | −0.02 | −0.16 | −0.14 | −0.27** | −0.16 | 1 | - | - | - | - |
12. Age | −0.02 | 0.07 | −0.03 | 0.23* | −0.04 | 0.01 | −0.03 | 0.11 | 0.02 | 0.07 | −0.23* | 1 | - | - | - |
13. Ethnicity (EA = 0, AA = 1) | 0.05 | −0.01 | 0.14 | 0.21* | −0.04 | −0.05 | 0.10 | 0.08 | 0.17 | 0.22* | −0.06 | −0.16 | 1 | - | - |
14. Physical Health | 0.18† | 0.21* | 0.25* | 0.06 | 0.12 | 0.09 | 0.24* | 0.14 | 0.19† | 0.01 | 0.06 | −0.26** | 0.03 | 1 | - |
15. Mental Health | 0.03 | −0.01 | 0.54*** | 0.64*** | 0.01 | −0.01 | 0.36*** | 0.37*** | 0.19† | 0.08 | −0.23* | 0.35*** | 0.13 | −0.14 | 1 |
Mean (SD) | 3.56 (0.75) | 4.33 (0.72) | 2.92 (0.70) | 3.22 (0.71) | 3.63 (0.68) | 4.03 (0.61) | 2.92 (0.70) | 3.22 (0.71) | 6.63 (1.65) | 5.20 (0.96) | 0.65 (0.48) | 56.38 (11.94) | 0.51 (0.50) | 51.71 (8.62) | 49.40 (10.63) |
Note. HAP = high arousal positive affect, LAP = low arousal positive affect; Positive physician rating n = 99;
p < .001,
p < .01,
p < .05,
p < .10.
Prior to testing our hypotheses, we examined whether there were any differences by condition (HAP-focused vs. LAP-focused physician) in terms of participants’ aggregated daily ratings of ideal and actual affect. We conducted a multivariate analysis of variance in which we entered condition as the between-subjects independent variable. We found no significant effect of condition on ratings of daily ideal and actual HAP and LAP, F (4, 96) = 1.87, p = .12, partial η2 = .07. Thus, during the week of the study, participants who had engaged with the HAP-focused physician were as likely to value and experience HAP and LAP states over the course of the five days as were those who had engaged with the LAP-focused physician.
Hypotheses 1-3: Does Ideal Affect Predict Self-Reported Adherence to Physician Recommendations?
Because we collected daily reports of adherence, we used hierarchical linear modeling software to test our hypotheses (HLM 7.01; Raudenbush, Bryk, Cheong, Congdon, & du Toit, 2011). We ran a series of analyses to build our model. First, at level 1, we entered daily adherence as the dependent variable. There was significant variation in daily adherence (α = .89), u0 = 3.47, SD = 1.86), χ2 (100) = 902.99, p < .001. To determine whether daily adherence differed between the HAP and LAP conditions, we entered condition (LAP-focused physician = −1, HAP-focused physician = 1) at level 2. If a significant interaction was revealed, we then coded the HAP-focused physician as 0 and the LAP-focused physician as 1 to examine effects within the HAP condition, and coded the HAP-focused physician as 1 and the LAP-focused physician as 0 to examine effects within the LAP condition. The effect of condition on daily adherence was not significant, B = 0.31, SE = .19, t(99) = 1.59, p = 0.114. On average, participants assigned to the HAP-focused physician adhered to his recommendations (M = 8.28, SD = 0.28) to a similar degree as those assigned to the LAP-focused physician (M = 7.66, SE = 0.27).
We then entered participants’ daily ideal HAP, ideal LAP, actual HAP, and actual LAP (grand-mean centered) as independent variables at level 1. Because the random effects of daily ideal and actual affect were not significant (see Appendix B), we constrained these effects to be fixed in the final model. At level 2, we added ethnicity (European American = 0; Asian American = 1), gender (Female = 0; Male = 1), age, and physical and mental health as covariates (grand-mean centered). When estimating overall adherence, there were marginal effects of physical health, B = 0.04, SE = .03, t(94) = 1.71, p = .091, and mental health, B = 0.03, SE = .02, t(94) = 1.72, p = .089: healthier participants reported greater adherence across conditions. There was also a significant effect of gender, B = 0.97, SE = .40, t(94) = 2.42, p = .017. Across conditions, men were more likely to adhere to physician recommendations (M = 7.23, SE = .26) than were women (M = 6.30, SE = .20). There were no significant effects of age, p = .736, and ethnicity, p = .473, on adherence or on the associations between daily affect (ideal or actual) and daily adherence, and thus, we excluded them from the final model to maintain parsimony (see Appendix B for final model).
Table 2 (top) displays beta coefficients, standard errors, and t statistics of effects within each condition and differences in effects between conditions. Consistent with Hypothesis 1, analyses revealed a significant effect of condition on the association between daily ideal HAP and daily adherence, p = .002, and the association between daily ideal LAP and daily adherence, p = .024. The more participants ideally wanted to feel HAP, the more participants reported adhering to the HAP-focused physician’s recommendations, p = .031. Also as predicted, the more participants ideally wanted to feel LAP, the more they reported adhering to the LAP-focused physician’s recommendations, p = .004. Interestingly, daily ideal HAP was significantly negatively associated with adherence to the LAP-focused physician’s recommendations, p = .028, suggesting that the more people value HAP, the less they adhere to a physician’s recommendations when that physician focuses on LAP. Daily ideal LAP, however, was not significantly associated with adherence to the HAP-focused physician’s recommendations, p = .548. Gender and health did not moderate effects of daily ideal affect on adherence behavior. In sum, consistent with Hypothesis 1, the likelihood of adherence on a given day for an individual increased when there was a match between physician’s affective focus and participants’ daily ideal affect (see Figure 3, top).
Table 2. Unstandardized beta coefficients, standard errors and t-statistics representing the magnitude of the effect of ideal and actual affect and the interaction between patient and physician affect on patients’ adherence to physicians’ recommendations and evaluations of their physician.
HAP Physician | LAP Physician | Patient X Physician Affect | ||||
---|---|---|---|---|---|---|
| ||||||
Adherence to Physician Recommendations |
B (SE) | t | B (SE) | t | B (SE) | t |
Ideal HAP | 0.53 (.24) | 2.16* | −0.54 (.24) | −2.21* | 0.53 (.17) | 3.11** |
Ideal LAP | −0.16 (.26) | −0.60 | 0.63 (.22) | 2.88** | −0.40 (.17) | −2.27* |
Actual HAP | −0.29 (.21) | −1.43 | 0.40 (.19) | 2.11* | −0.34 (.14) | −2.53** |
Actual LAP | 0.39 (.26) | 1.49 | 0.08 (.15) | 0.53 | 0.15 (.15) | 1.04 |
Evaluation of Physician | B (SE) | t | B (SE) | t | B (SE) | t |
---|---|---|---|---|---|---|
Ideal HAP | 0.77 (.23) | 3.39*** | −0.25 (.18) | −1.38 | 0.51 (.15) | 3.48*** |
Ideal LAP | −0.48 (.19) | −2.58** | 0.56 (.27) | 2.12* | −0.52 (.16) | −3.21** |
Actual HAP | −0.18 (.20) | −0.87 | 0.08 (.18) | 0.45 | −0.13 (.13) | −0.96 |
Actual LAP | 0.30 (.23) | 1.30 | 0.08 (.20) | 0.39 | 0.11 (.15) | 0.74 |
Note. df = 383 for analyses of adherence behavior; df = 87 for analyses of evaluations of physicians. HAP = high arousal positive affect; LAP = low arousal positive affect;
p ≤ .001,
p ≤ .01,
p ≤ .05.
Figure 3.
Top: Effect of participants’ daily ideal and actual affect on daily adherence to physician recommendations; bars represent mean within-person unstandardized beta coefficients. Bottom: Effect of participants’ global ideal and actual affect on positive evaluation of assigned physician at the end of the study; bars represent mean unstandardized beta coefficients. Note. HAP = high arousal positive affect, LAP = low arousal positive affect. Asterisks indicate differences between conditions, **p<.01; *p< .05.
To test Hypothesis 2 that ideal affect would predict adherence more systematically than actual affect, we examined the effects of daily actual affect. Analyses revealed a significant effect of condition on the association between daily actual HAP and daily adherence, p = .012. Daily actual HAP did not significantly predict adherence to the HAP-focused physician’s recommendations, p = .154, but did significantly predict adherence to the LAP-focused physician’s recommendations, p = .035. Moreover, this interaction varied as a function of physical health, B = 0.03, SE = .01, t(383) = 2.19, p = .029, and mental health, B = 0.033, SE = .014, t(383) = 2.47, p = .014. Post-hoc analyses revealed that actual HAP was more positively associated with adherence to the LAP-focused physician only for individuals whose health was above the sample median, B = 0.66, SE = .29, t(383) = 2.25, p = .025.
There was not a significant effect of daily actual LAP, p = .131, and no significant effect of condition on the association between daily actual LAP and daily adherence, p = .298. However, there was a significant effect of physical health on the association between daily actual LAP and daily adherence, B = −0.03, SE = .02, t(383) = −2.04, p = .042. A post-hoc analysis showed that daily actual LAP was more positively associated with adherence to both physicians, but only for individuals whose physical and mental health was rated lower than the sample median, B = 0.65, SE = .24, t(383) = 2.71, p = .007. In sum, although there were some effects of actual affect on adherence, these effects varied as a function of patients’ physical and mental health, and did not systematically vary as a function of the match between patients’ ideal affect and physicians’ affective focus. These findings provide some support for Hypothesis 2.
To test Hypothesis 3, we also examined a model at the between-person level in which we entered global ideal and actual affect (also grand-mean centered) as level 2 predictors of mean adherence across the week. We used the same covariates as in the within-person model. We found no significant main effects and no significant interactions with condition for global ideal HAP, global ideal LAP, global actual HAP, and global actual LAP on average adherence.3 Contrary to our prediction, these findings suggest that at the between person level, neither global ideal affect nor global actual affect predicted how much participants adhered on average to physicians’ recommendations.
Hypotheses 4-5: Does Ideal Affect Predict Positive Evaluations of Physicians?
Prior to testing Hypotheses 4-5, we examined whether participants varied in their evaluations of the HAP-focused and LAP-focused physicians. Two participants did not respond to the physician evaluation items and therefore were not included in these analyses. An independent samples t-test revealed that participants assigned to the HAP-focused physician evaluated their physician as positively (M = 5.09, SE = 0.15) as did participants assigned to the LAP-focused physician (M = 5.29, SE = 0.14), t(97) = −0.99, p = .327. Because we did not measure evaluations of the physician at the daily level,4 we only examined the links between global ideal and actual affect and evaluation of the physician.
To examine the effects of global ideal affect on evaluation of physicians within condition, we conducted moderation analyses using hierarchical regression (Aiken & West, 1991). We entered participants’ positive evaluation of the physician as the dependent variable and condition (using the same coding as in the previous analyses), global ideal HAP, global ideal LAP, global actual HAP, and global actual LAP as the independent variables at step 1. At step 2, we entered the interaction of condition and global ideal HAP, global ideal LAP, global actual HAP, and global actual LAP. Because participants’ adherence to physician’s recommendations during the week was marginally correlated with how positively they evaluated the physician, we also controlled for average adherence.
Additionally, we included ethnicity, gender, age, and health as covariates in the model. There was a significant effect of ethnicity such that Asian Americans rated both physicians more favorably (M = 5.41, SE = 0.13) than did European Americans (M = 4.97, SE = .15), B = 0.44, SE = .20, β = .22, t(98) = 2.23, p = .028. As expected, the more participants adhered to physician recommendations, the more positively they evaluated the physician, B = 0.14, SE = .06, β = .22, t(98) = 2.24, p = .030. The remaining covariates, however, did not have an effect on physician evaluation or alter observed results, and therefore, we excluded them from the final model for parsimony.
Table 2 (bottom) contains beta coefficients, standard errors, and t statistics of effects within each condition and differences of effects between conditions. In support of Hypothesis 4, the interactions between condition and global ideal HAP, β = .38, p = .001, and between condition and global ideal LAP, β = −.37, p = .002, were significant: the more participants wanted to feel HAP on average, the more they positively they evaluated the HAP-focused physician at the end of the study, β = .58, p = .001. Similarly, the more participants wanted to feel LAP on average, the more positively they evaluated the LAP-focused physician at the end of the study, β = .40, p = .037. Global ideal HAP was not significantly associated with evaluations of the LAP-focused physician, β = −.19, p = .172; however, global ideal LAP was significantly negatively associated with evaluations of the HAP-focused physician, β = −.34, p = .012: the more people wanted to feel LAP on average, the less positively they evaluated the HAP-focused physician (See Figure 3 bottom).
In support of Hypothesis 5, the interactions between condition and global actual HAP, β = −.11, p = .341, and between condition and global actual LAP, β = .08, p = .463, were not significant. There were also no significant main effects of global actual HAP, β = −.04, p = .715, or global actual LAP, β = .14, p = .214. Thus, participants’ global ideal affect systematically predicted their evaluation of physicians whereas their global actual affect did not.5 Therefore, participants’ ideal affect not only predicts their evaluation and selection of hypothetical physicians after a single contact (as shown in our previous work), but also predicts patients’ evaluation of assigned physicians after multiple contacts in real life.
Discussion
Although previous findings suggest that for American samples, physicians’ and patients’ affective characteristics independently predict patient outcomes, few studies have examined how the two interact. The present study is the first to examine how patients’ ideal affect interacts with physicians’ affective focus to influence their actual (i.e., non-hypothetical) responses to physicians. As predicted, we observed that patients were more likely to adhere to physicians’ recommendations and to evaluate their physicians positively if their physician focused on the type of positive affect that patients valued.6
In some cases, we also observed that patients were less likely to adhere and to evaluate their physicians positively when the physician did not focus on their ideal affective state. However this pattern was not consistent across outcomes or affective states, and therefore, further research is needed to determine whether “mismatches” result in more negative patient outcomes.
Daily Ideal Affect Predicts Daily Adherence to Physicians’ Recommendations
Consistent with our hypotheses, at the daily level, the more participants wanted to feel HAP, the more they adhered to the HAP-focused physician’s recommendations. Similarly, the more people wanted to feel LAP, the more they adhered to the LAP-focused physician’s recommendations. Daily actual affect did not systematically predict daily adherence to the HAP-focused versus LAP-focused physician. Instead, among the healthier participants, the more they actually felt HAP, the more they adhered to recommendations made by the LAP-focused physician. In contrast, among less healthy participants, the more they actually felt LAP, the more they adhered to recommendations made by either physician. Future work should examine why the effects of actual HAP and LAP vary depending on the health status of the participant. One possibility is that HAP states are too physically stimulating for unhealthy participants, and therefore, do not facilitate engagement in health-promoting behavior for these individuals. In contrast, LAP states may not be physically stimulating enough for healthy American participants, and therefore, may not facilitate engagement in health-promoting behavior for these individuals. Together, these findings suggest that while both daily ideal and actual affect shape adherence behavior, they do so in distinct ways. Whereas the effect of daily ideal affect on health-related behavior and cognitions may depend more on features of the physician, the effect of daily actual affect on health-related behavior and cognitions may depend more on characteristics of the individual (i.e., patients’ health status).
In addition, we found that men were more likely to adhere to physicians’ recommendations overall. In part, this may be because we used male actors to portray the physicians, and studies suggest that female patients prefer female doctors (e.g., Garcia, Paterniti, Romano, & Kravitz, 2003). While we did not observe gender differences in ideal affect or the effects of ideal affect on adherence, a promising direction for future research would be to investigate whether patients perceive the affective qualities of physicians differently depending on the physician’s gender (cf. Hall, et al, 2000).
Whereas participants’ daily ideal affect predicted how likely they were to adhere to physicians’ recommendations on a given day relative to the rest of their week, contrary to our expectation, participants’ global ideal affect did not predict their average adherence. Instead, other factors such as gender and physical health predicted average adherence regardless of the assigned physician. These findings suggest that the links between affect and actual health behavior may be better understood when considering daily variation and processes within an individual (e.g., Hooker, Choun, Mejia, Pham, & Metoyer, 2013).
Future work should explore how ideal affect leads to greater adherence. One possibility is that ideal affect enhances how people perceive the health behaviors recommended by the physician. For instance, by recommending engagement in health behaviors in order to feel their ideal state, physicians may be helping their patients see the efficacy or benefit of those behaviors in their daily lives (cf., Azjen, 2002; Bandura, 20004; Becker, 1974; Carpenter, 2011; McEachan, Conner, Taylor, & Lawton, 2011).
Global Ideal Affect Predicts Positive Evaluations of Physicians
As predicted, at the global level, the more participants valued HAP, the more positively they evaluated the HAP-focused physician at the end of the five days. Also as predicted, the more participants valued LAP, the more positively they evaluated the LAP-focused physician at the end of the five days. These findings are consistent with Sims et al. (2014), which demonstrated that when asked to imagine that their primary care physician was no longer available, college students more positively evaluated and were more likely to choose physicians whose affective focus matched their ideal affect. In that study, participants did not actually interact with the HAP or LAP-focused physicians, but made their evaluations and choices based on information about how they thought the physician would be (i.e., after a single encounter). The current study expands upon these findings by demonstrating that community adults more positively evaluated physicians whose affective focus matched their ideal affect even after the physician provided feedback and made health recommendations (i.e., after multiple encounters). In other words, the match between a patient’s ideal affect and a physician’s affective focus matters not only when patients are selecting a physician (as demonstrated in previous work), but also when they are actually interacting with and responding to an assigned physician (as demonstrated in the present work). In future work, we plan to explore possible mechanisms underlying the link between ideal affect and physician evaluation. For example, one possibility is that patients selectively attend to physicians who promote their ideal affect, which in turns makes them appear more knowledgeable, competent, and trustworthy. Indeed, people are more likely to attend to cues that are congruent with their motivational goals and attitudes (e.g., Forster & Higgins, 2005; Hamamura, Meijer, Heine, Kamaya, & Hori, 2009; Sakaki, Nga, & Mather, 2013; Smith, Fabrigar, Powell, & Estrada, 2007). Another possibility is that patients find it more rewarding (e.g., enjoyable, satisfying) to interact with physicians whose affective focus matches their ideal affect and, as a result,, they evaluate those physicians as more trustworthy, knowledgeable and competent. Indeed, patients report feeling more satisfied with physicians whose preferences aligns with their own (see Kiesler & Auerbach for review, 2006).
Limitations and Future Directions
The current study design had several limitations that should be addressed in future research. First, our findings were based on participants’ self-reports of adherence, raising the possibility that they were due to experimenter demand. In other words, individuals may have reported greater adherence because they wanted to please the physician who matched their affective ideals. If this were the case, participants would have also reported increased experience or valuation of the affective states promoted by their physician; our data, however, show that participants experienced and valued HAP and LAP states similarly across conditions. Moreover, reports of global ideal and actual affect assessed prior to viewing the physician did not differ from daily reports assessed after three encounters with the physician (“meeting” the virtual physician, receiving feedback and recommendations from the physician, being thanked by the physician), suggesting that participants did not change their reports of affect to be more consistent with their assigned physicians. Regardless, future work is needed to assess whether the observed effects generalize to more objective behavioral assessments of adherence.
Second, due to the correlational design of our study, we cannot definitively conclude that how patients ideally want to feel shapes how they respond to physicians. It may be that physicians’ affective foci directly influence patient adherence and judgments, which in turn influences their ideal affect. However, the prospective design of our study suggests otherwise. That is, patient global ideal affect sampled prior to viewing their physician predicted patient evaluations of the physician at the end of the five-day period. Additionally, as described above, global reports of ideal affect assessed prior to viewing the physician were similar to aggregated daily reports of ideal affect provided after viewing the physician, suggesting that physicians’ affective foci did not influence patient ideal affect over the course of a week. However, it would be interesting to examine in future work whether over longer periods of time, repeated exposure to a trusted physician whose affective focus differs from his or her patients’ initial ideal affect does indeed change how his or her patients subsequently ideally want to feel.
Third, physicians differed in their “affective focus,” which included differences in their facial expressions and vocal tone, their views on patient care, their interests, and the affective outcomes of their health recommendations. We altered multiple characteristics of each physician both to strengthen the effect of our manipulation, as well as to increase ecological validity. In other words, we assumed that physicians in the real world would be consistent in their affective focus across these characteristics, and that inconsistencies (e.g., a calm physician promoting feeling energetic) would seem odd to participants. However, more work is needed to examine whether one characteristic is more necessary than the other to shape adherence and evaluation. Indeed, in a previous study, we observed that participants’ ideal affect predicted their evaluations of physicians even when they read only about physicians’ views on patient care, suggesting that images of the physicians were not necessary to influence patients’ evaluations of physicians (Sims, Tsai, Koopmann-Holm, et al., 2014). On the other hand, research has demonstrated that people make significant inferences about people based only on their facial expressions (Rule et al., 2013; Todorov et al., 2009; see Zebrowitz et al., 2008 for review). Future work should examine the incremental impact of each characteristic on patients’ responses to their physicians as well as the mechanisms underlying these effects.
Finally, future research should examine whether our findings generalize to other health recommendations and to other samples. In order to accommodate an age diverse sample, we did not examine more vigorous health recommendations. However, it would be important to include a variety of other health recommendations in future studies. Relatedly, to ensure we did not risk participants’ health, we selectively recruited a relatively healthy sample comprised of European Americans and Asian Americans. It would be important to examine whether a match between patients’ ideal affect and physicians’ affective focus is also relevant among chronically ill populations and among other ethnic groups. Furthermore, because we matched participants and physicians by ethnicity, it would also be important to determine whether similar processes occur in situations when physicians and patients are of different ethnicities.
Implications for Patient-Physician Communication
Despite these limitations, our findings have important implications for the role of affect in the context of health care. Our work demonstrates that patients’ and physicians’ affective characteristics interact to influence two clinically significant patient outcomes: adherence to health recommendations and evaluations of physicians. Physicians should recognize that their patients have affective ideals that might influence how they respond to physicians. Moreover, physicians could assess their patients’ ideal affect and then consider whether there are particular health care providers or treatments that are consistent with patients’ ideals. If none exist, physicians might explicitly discuss with patients how their ideal affect might hinder their adherence to particular recommendations and their interactions with particular health care providers. Along these lines, in our current research, we are developing and assessing the effectiveness of interventions aimed at educating clinicians about the importance of ideal affect in health care.
Our findings also have implications for theories of health decision-making and health behavior change. A large body of research has shown that actual affect influences people’s health-related decisions and behaviors. While we found that actual affect played a role in patient behavior in some cases, ideal affect predicted our outcomes more consistently, robustly and systematically. This work raises the possibility that knowing how someone ideally wants to feel may be as important as knowing how someone actually feels, if not more so in certain contexts. Therefore, current models of affect and health-related behavior and cognition should include how people ideally want to feel. Second, this work has implications for research on health communication and message framing. Using message framing to promote health behaviors was initially born out of prospect theory (Kahneman & Tversky, 1979; Rothman et al., 1997), and focuses on the utility of emphasizing gains (i.e., benefits of engaging in a behavior) versus losses (i.e., costs of not engaging in a behavior) for promoting healthy behavior. Our findings indicate that emphasizing certain gains will be more or less effective depending on what the patient values. Finally, several studies have observed significant cultural and age-related variation in ideal affect (Scheibe, et al., 2011; Sims, Tsai, Jiang, et al. 2014; Tsai, Koopmann-Holm, et al., 2014; Tsai, 2007; Tsai, Knutson, et al., 2006; Tsai, Miao, et al., 2007; Tsai, Louie, et al., 2007; Ruby, et al., 2012). Thus, considering patients’ ideal affect may be useful in communicating health information to diverse populations and ultimately help reduce ethnic disparities in health care utilization across the life span.
Acknowledgments
Research reported in this publication was supported by the Stanford Graduate Research Opportunity fund, Stanford Diversity Dissertation Research Opportunity fund, Stanford Enhancing Diversity in Graduate Education Research Award awarded to the first author, NIA P03AG024957 awarded to the first and second authors, and NIA R03AG023302 and NIMH R01MH068879 awarded to the second author. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or other affiliated organizations. The authors thank the Stanford Culture and Emotion Lab, Stanford Culture Collaboratory, Stanford Life Span Development Lab, Ewart Thomas, Hazel Markus, Laura Carstensen, and Mary Goldstein for their helpful suggestions, and Jin Song, Annabel Sun, Henry Young, Daniel Virtheim, J.D. Haddon, Ana Markda Silva, Deja Stockdale, Cyrus Sarrafpour, and Jazmin Youngblood for their research assistance.
Appendix A
HAP Physician Video Script
Introduction
“Hi, I’m Dr. Jason Smith (Lee). I am excited to be assigned as your personal virtual physician. As your physician, my goal is to enhance your well-being by increasing your activity levels and overall vitality so you can lead a more dynamic, energetic lifestyle! I believe that the key to a long, healthy life is to have an enthusiastic attitude and approach each day with passion and zest! Finding exhilaration will lead to good health. Engagement in vigorous, strengthening activities while keeping in high spirits can help you get fit and feeling great! Today, we will begin by giving you a general health assessment. Once I review the results, I will provide you with a health program to follow for the week. Then, we can meet again to review how your week went and determine whether we are meeting your health goals. “
Feedback
Thank you for completing our health assessment survey. Here are your results and my recommendations. Please bookmark this page, so you can return any time. Based on my review, I have prescribed a personalized health plan for the upcoming week. This plan is intended to be in addition to your regular daily activities over the next 5 days. It consists of aerobic exercise, muscle resistance training, and healthy practices to ensure an upbeat and stimulating lifestyle. Descriptions of the muscle strengthening exercises can be found at the link below. By following this plan as closely as possible, I am confident that you will look and feel energized! I hope you will approach this week with much enthusiasm. I look forward to checking in with you soon.
HAP Physician Written Description
Specialty: Internal Medicine
Facility: Virtual Bay Health Center
Department of Internal Medicine
Education: Doctor of Medicine, University of California San Francisco, 1973
Award for Outstanding Healthcare Provider, Virtual Bay Health Center
Views on Patient Care: “My goal as a physician is to enhance patient well-being by increasing their activity levels and overall vitality so my patients can lead dynamic lifestyles.”
Outside interests: “I volunteer in a youth home where we help adolescents to feel passionate about their education.”
LAP Physician Video Script
Introduction
“Hi, I’m Dr. Steven Jones (Wong). I am pleased to be assigned as your personal virtual physician. As your physician, my goal is to ensure your well-being by promoting peace of mind when it comes to your health so you can live a more calm and tranquil lifestyle. I believe that the key to a long, healthy life is to have a carefree attitude and approach each day cool and collected. Finding serenity will lead to good health. Engagement in soothing, strengthening activities can help you get fit and feel rested. Today, we will begin by giving you a general health assessment. Once I review the results, I will provide you with a health program to follow for the week. Then, we can meet again to review how your week went and determine whether we are meeting your health goals.”
Feedback
“Thank you for completing our health assessment survey. Here are your results and my recommendations. Please bookmark this page, so you can return any time. Based on my review, I have prescribed a personalized health plan for the upcoming week. This plan is intended to be in addition to your regular daily activities over the next 5 days. It consists of aerobic exercise, muscle resistance training, and healthy practices to ensure a calm and peaceful lifestyle. Descriptions of the muscle strengthening exercises can be found at the link below. By following this plan as closely as possible, I am confident that you will look and feel rested. I hope you can relax as you approach this week. I look forward to checking in with you soon.”
LAP Physician Written Description
Specialty: Internal Medicine
Facility: Virtual Bay Health Center
Department of Internal Medicine
Education: Doctor of Medicine, University of California San Francisco, 1973
Award for Outstanding Healthcare Provider, Virtual Bay Health Center
Views on Patient Care: “My goal as a physician is to ensure that my patients have peace of mind when it comes to their health by promoting a calm and relaxed lifestyle.”
Outside interests: “I volunteer in a youth home where we help adolescents feel at ease about their education.”
Appendix B
Multilevel model equation in which participants’ daily self-reported adherence was regressed onto daily reports of ideal and actual affect at level 1 and the effect of condition (physician type), gender, and health included at level 2 to examine their effects on the within-person association between daily adherence and ideal and actual affect. Random effects of daily and actual affect were not significant and constrained to be fixed in the final model: Daily Ideal HAP: r1 = 0.73, SD = 0.85, χ2 (34) = 34.62, p = .44; Daily Ideal LAP: r3 = 0.53, SD = 0.73, χ2 (34) = 45.35, p = .09; Daily Actual HAP: r2 = 0.59, SD = 0.77, χ2 (34) = 24.53, p > .50; Daily Actual LAP: r4 = 0.36, SD = 0.60, χ2 (34) = 42.85, p = .14
Level-1 Model
Daily Adherenceij = β0j + β1j*(Daily Ideal HAPij) + β2j*(Daily Actual HAPij) + β3j*(Daily Ideal LAPij) + β4j*(Daily Actual LAPij) + rij
Level-2 Model
β0j = γ00 + γ01*(Conditionj) + γ02*(Genderj) + γ03*(Physical Healthj) + γ04*(Mental Healthj) + u0j
β1j = γ10 + γ11*(Conditionj) + γ12*(Genderj) + γ13*(Physical Healthj) + γ14*(Mental Healthj)
β2j = γ20 + γ21*(Conditionj) + γ22*(Genderj) + γ23*(Physical Healthj) + γ24*(Mental Healthj)
β3j = γ30 + γ31*(Conditionj) + γ32*(Genderj) + γ33*(Physical Healthj) + γ34*(Mental Healthj)
β4i = γ40 + γ41*(Conditionj) + γ42*(Genderj) + γ43*(Physical Healthj) + γ44*(Mental Healthj)
Mixed Model
Daily Adherenceij = γ00 + γ01*Conditionj + γ02*Genderj + γ03*Physical Healthj
+ γ04* Mental Healthj + γ10*Daily Ideal HAPij + γ11*Conditionj*Daily Ideal HAPij + γ12*Genderj*Daily Ideal HAPij + γ13*Physical Healthj*Daily Ideal HAPij + γ14*Mental Healthj*Daily Ideal HAPij
+ γ20*Daily Actual HAPij + γ21*Conditionj*Daily Actual HAPij + γ22*Genderj*Daily Actual HAPij +
γ23*Physical Healthj*Daily Actual HAPij + γ24*Mental Healthj*Daily Actual HAPij +
γ30*Daily Ideal LAPij + γ31*Conditionj*Daily Ideal LAPij + γ32*Genderj*Daily Ideal LAPij +
γ33*Physical Healthj*Daily Ideal LAPij + γ34*Mental Healthj*Daily Ideal LAPij +
γ40*Actual Ideal LAPij + γ41*Conditionj*Actual Ideal LAPij + γ42*Genderj*Actual Ideal LAPij +
γ43*Physical Healthj*Actual Ideal LAPij + γ44*Mental Healthj*Actual Ideal LAPij + u0j+ rij
Footnotes
Importantly, none of the participants questioned the purpose of providing feedback on the virtual health center. When we explicitly asked at the end of the study whether they thought the study was about anything else, two participants expressed doubt about the validity of the virtual health center; however, excluding these participants from data analysis did not alter the results.
Only one participant engaged in all five behaviors at baseline. Because the virtual physician recommended engagement in health behaviors “above and beyond their regular routine,” we did not exclude this participant from our analyses.
There were also no significant main effects and no significant interactions with condition for aggregated daily ratings of ideal HAP, ideal LAP, actual HAP, and actual LAP on average adherence across the week.
We did not measure evaluation of physicians at the daily level for two reasons. First, we wanted to keep the daily questionnaire brief to minimize participant fatigue. Second, we were concerned that participants would find rating how they felt about the physician on a daily basis unnatural.
Findings were similar when we included aggregated daily ideal and actual affect as predictors of evaluation of physician. Model fit statistics were comparable for aggregated daily ideal affect, F(11, 87) = 2.71, p < .01, R2 = .26, and global ideal affect, F(11, 87) = 2.90, p < .01, R2 = .27, demonstrating that both aggregated daily and global reports of ideal affect predict participants’ evaluation of their physicians to a similar degree.
We performed a multilevel mediation analysis using Mplus and found that patients’ evaluation of physician did not mediate the link between ideal affect and adherence. This result and the finding that patients’ evaluation and adherence ratings were only marginally correlated support previous findings that these are two distinct clinical outcomes.
References
- Aaker JL, Lee AY. “I” Seek Pleasures and “We” Avoid Pains: The Role of Self-Regulatory Goals in Information Processing and Persuasion. Journal of Consumer Research. 2001;28(1):33–49. doi: 10.1086/321946. [Google Scholar]
- Agrawal N, Menon G, Aaker JL. Getting emotional about health. Journal of Marketing Research. 2007;44(1):100–113. [Google Scholar]
- Aiken LS, West SG. Multiple regression: Testing and interpreting interactions. Sage Publications, Inc; Thousand Oaks, CA, US: 1991. [Google Scholar]
- Ajzen I. Perceived Behavioral Control, Self-Efficacy, Locus of Control, and the Theory of Planned Behavior1. Journal of Applied Social Psychology. 2002;32(4):665–683. doi: 10.1111/j.1559-1816.2002.tb00236.x. [Google Scholar]
- Ambady N, Koo J, Rosenthal R, Winograd CH. Physical therapists’ nonverbal communication predicts geriatric patients’ health outcomes. Psychology and Aging. 2002;17(3):443–452. doi: 10.1037/0882-7974.17.3.443. doi: 10.1037/0882-7974.17.3.443. [DOI] [PubMed] [Google Scholar]
- Bandura A. Health Promotion by Social Cognitive Means. Health Education & Behavior. 2004;31(2):143–164. doi: 10.1177/1090198104263660. doi: 10.1177/1090198104263660. [DOI] [PubMed] [Google Scholar]
- Barrett LF. Hedonic Tone, Perceived Arousal, and Item Desirability: Three Components of Self-reported Mood. Cognition & Emotion. 1996;10(1):47–68. doi: 10.1080/026999396380385. [Google Scholar]
- Beck RS, Daughtridge R, Sloane PD. Physician-patient communication in the primary care office: a systematic review. The Journal of the American Board of Family Practice. 2002;15(1):25–38. [PubMed] [Google Scholar]
- Becker MH, editor. The health belief model and personal health behavior. Health Education Monographs. 1974;2:324–473. [Google Scholar]
- Blasi ZD, Harkness E, Ernst E, Georgiou A, Kleijnen J. Influence of context effects on health outcomes: a systematic review. The Lancet. 2001;357(9258):757–762. doi: 10.1016/s0140-6736(00)04169-6. [DOI] [PubMed] [Google Scholar]
- Carpenter CJ. A Meta-Analysis of the Effectiveness of Health Belief Model Variables in Predicting Behavior. Health Communication. 2010;25(8):661–669. doi: 10.1080/10410236.2010.521906. doi: 10.1080/10410236.2010.521906. [DOI] [PubMed] [Google Scholar]
- Carver CS, Scheier MF, Segerstrom SC. Optimism. Clinical Psychology Review. 2010;30(7):879–889. doi: 10.1016/j.cpr.2010.01.006. doi: http://dx.doi.org/10.1016/j.cpr.2010.01.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chow PI, Berenbaum H. Perceived utility of emotion: The structure and construct validity of the Perceived Affect Utility Scale in a cross-ethnic sample. Cultural Diversity and Ethnic Minority Psychology. 2012;18(1):55–63. doi: 10.1037/a0026711. doi: 10.1037/a0026711. [DOI] [PubMed] [Google Scholar]
- Cousin G, Schmid Mast M, Roter DL, Hall JA. Concordance between physician communication style and patient attitudes predicts patient satisfaction. Patient Education and Counseling. 2012;87(2):193–197. doi: 10.1016/j.pec.2011.08.004. doi: 10.1016/j.pec.2011.08.004. [DOI] [PubMed] [Google Scholar]
- DiMatteo M, Lepper HS, Croghan TW. Depression is a risk factor for noncompliance with medical treatment: Meta-analysis of the effects of anxiety and depression on patient adherence. Archives of Internal Medicine. 2000;160(14):2101–2107. doi: 10.1001/archinte.160.14.2101. doi: 10.1001/archinte.160.14.2101. [DOI] [PubMed] [Google Scholar]
- DiMatteo MR. A social psychological analysis of physician patient rapport: Toward a science of the art of medicine. Journal of Social Issues. 1979;35(1):12–33. [Google Scholar]
- Förster J, Higgins ET. How Global Versus Local Perception Fits Regulatory Focus. Psychol Sci. 2005;16(8):631–636. doi: 10.1111/j.1467-9280.2005.01586.x. doi: 10.1111/j.1467-9280.2005.01586.x. [DOI] [PubMed] [Google Scholar]
- Gao GG, McCullough JS, Agarwal R, Jha AK. A changing landscape of physician quality reporting: analysis of patients’ online ratings of their physicians over a 5-year period. Journal of medical Internet research. 2012;14(1):e38. doi: 10.2196/jmir.2003. doi: 10.2196/jmir.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Garcia JA, Paterniti DA, Romano PS, Kravitz RL. Patient preferences for physician characteristics in university-based primary care clinics. Ethn Dis. 2003;13(2):259–267. [PubMed] [Google Scholar]
- Griffith CH, Wilson JF, Langer S, Haist SA. House staff nonverbal communication skills and standardized patient satisfaction. Journal of General Internal Medicine. 2003;18(3):170–174. doi: 10.1046/j.1525-1497.2003.10506.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hall JA, Blanch-Hartigan D, Roter DL. Patients’ satisfaction with male versus female physicians: a meta-analysis. Medical care. 2011;49(7):611–617. doi: 10.1097/MLR.0b013e318213c03f. [DOI] [PubMed] [Google Scholar]
- Hall JA, Roter DL, Rand CS. Communication of affect between patient and physician. Journal of Health and Social Behavior. 1981;22(1):18–30. [PubMed] [Google Scholar]
- Hamamura T, Meijer Z, Heine SJ, Kamaya K, Hori I. Approach—Avoidance Motivation and Information Processing: A Cross-Cultural Analysis. Personality and Social Psychology Bulletin. 2009;35(4):454–462. doi: 10.1177/0146167208329512. doi: 10.1177/0146167208329512. [DOI] [PubMed] [Google Scholar]
- Haskard Zolnierek KB, DiMatteo MR. Physician Communication and Patient Adherence to Treatment: A Meta-Analysis. Medical Care. 2009;47(8):826–834. doi: 10.1097/MLR.0b013e31819a5acc. 810.1097/MLR.1090b1013e31819a31815acc. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hill PL, Roberts BW. The role of adherence in the relationship between conscientiousness and perceived health. Health Psychology. 2011;30(6):797–804. doi: 10.1037/a0023860. doi: 10.1037/a0023860. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hooker K, Choun S, Mejía S, Pham T, Metoyer R. A microlongitudinal study of the linkages among personality traits, self-regulation, and stress in older adults. Research in Human Development. 2013;10(1):26–46. [Google Scholar]
- Jerant A, Fenton JJ, Bertakis KD, Franks P. Satisfaction With Health Care Providers and Preventive Care Adherence: A National Study. Medical Care. 2014;52(1):78–85. doi: 10.1097/MLR.0000000000000021. 10.1097/MLR.0000000000000021. [DOI] [PubMed] [Google Scholar]
- Kahneman D, Tversky A. Prospect theory: An analysis of decision under risk. Econometrica: Journal of the Econometric Society. 1979:263–291. [Google Scholar]
- Kämpfe N, Mitte K. What you wish is what you get? The meaning of individual variability in desired affect and affective discrepancy. Journal of Research in Personality. 2009;43(3):409–418. doi: 10.1016/j.jrp.2009.01.007. [Google Scholar]
- Kiesler DJ, Auerbach SM. Optimal matches of patient preferences for information, decision-making and interpersonal behavior: Evidence, models and interventions. Patient Education and Counseling. 2006;61(3):319–341. doi: 10.1016/j.pec.2005.08.002. doi: http://dx.doi.org/10.1016/j.pec.2005.08.002. [DOI] [PubMed] [Google Scholar]
- Koopmann-Holm B, Sze J, Ochs C, Tsai JL. Buddhist-inspired meditation increases the value of calm. Emotion. 2013;13(3):497–505. doi: 10.1037/a0031070. doi: 10.1037/a0031070. [DOI] [PubMed] [Google Scholar]
- Holm B, Tsai JL. Cultural variation in avoided affect: Implications for responses to suffering. Journal of Personality and Social Psychology. in press. [Google Scholar]
- Latimer AE, Williams-Piehota P, Katulak NA, Cox A, Mowad L, Higgins ET, Salovey P. Promoting fruit and vegetable intake through messages tailored to individual differences in regulatory focus. Annals of Behavioral Medicine. 2008;35(3):363–369. doi: 10.1007/s12160-008-9039-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- LaVeist TA, Nuru-Jeter A. Is Doctor-Patient Race Concordance Associated with Greater Satisfaction with Care? Journal of Health and Social Behavior. 2002;43(3):296–306. doi: 10.2307/3090205. [PubMed] [Google Scholar]
- Ledford CJW. Exploring the interaction of patient activation and message design variables: Message frame and presentation mode influence on the walking behavior of patients with type 2 diabetes. Journal of Health Psychology. 2012;17(7):989–1000. doi: 10.1177/1359105311429204. doi: 10.1177/1359105311429204. [DOI] [PubMed] [Google Scholar]
- Lee AY, Aaker JL. Bringing the frame into focus: the influence of regulatory fit on processing fluency and persuasion. Journal of personality and social psychology. 2004;86(2):205–218. doi: 10.1037/0022-3514.86.2.205. [DOI] [PubMed] [Google Scholar]
- Levine CS, Ambady N. The role of non-verbal behaviour in racial disparities in health care: implications and solutions. Medical education. 2013;47(9):867–876. doi: 10.1111/medu.12216. [DOI] [PubMed] [Google Scholar]
- Locke K. A Funny Thing Happened! The Management of Consumer Emotions in Service Encounters. Organization Science. 1996;7(1):40–59. doi: 10.1287/orsc.7.1.40. [Google Scholar]
- Luo J, Isaacowitz DM. How optimists face skin cancer information: Risk assessment, attention, memory, and behavior. Psychology and health. 2007;22(8):963–984. [Google Scholar]
- Manary MP, Boulding W, Staelin R, Glickman SW. The Patient Experience and Health Outcomes. New England Journal of Medicine. 2013;368(3):201–203. doi: 10.1056/NEJMp1211775. doi: doi:10.1056/NEJMp1211775. [DOI] [PubMed] [Google Scholar]
- Mann T, Sherman D, Updegraff J. Dispositional Motivations and Message Framing: A Test of the Congruency Hypothesis in College Students. Health Psychology. 2004;23(3):330–334. doi: 10.1037/0278-6133.23.3.330. doi: 10.1037/0278-6133.23.3.330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Martin LR, Williams SL, Haskard KB, Dimatteo MR. The challenge of patient adherence. Therapeutics and clinical risk management. 2005;1(3):189–199. [PMC free article] [PubMed] [Google Scholar]
- McEachan RRC, Conner M, Taylor NJ, Lawton RJ. Prospective prediction of health-related behaviours with the theory of planned behaviour: A meta-analysis. Health Psychology Review. 2011;5(2):97–144. [Google Scholar]
- Mogilner C, Aaker J, Kamvar SD. How Happiness Affects Choice. Journal of Consumer Research. 2012;39(2):429–443. doi: 10.1086/663774. [Google Scholar]
- National Healthcare Quality & Disparities Reports. Agency for Healthcare Research and Quality; Rockville, MD: Jun, 2014. http://www.ahrq.gov/research/findings/nhqrdr/index.html. [Google Scholar]
- Notthoff N, Carstensen LL. Positive Messaging Promotes Walking in Older Adults. Psychology and aging. doi: 10.1037/a0036748. in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ong AD. Pathways linking positive emotion and health in later life. Current Directions in Psychological Science. 2010;19(6):358–362. [Google Scholar]
- Pawlikowska T, Zhang W, Griffiths F, van Dalen J, van der Vleuten C. Verbal and non-verbal behavior of doctors and patients in primary care consultations – How this relates to patient enablement. Patient Education and Counseling. 2012;86(1):70–76. doi: 10.1016/j.pec.2011.04.019. doi: http://dx.doi.org/10.1016/j.pec.2011.04.019. [DOI] [PubMed] [Google Scholar]
- Pressman SD, Cohen S. Does Positive Affect Influence Health? Psychological Bulletin. 2005;131(6):925–971. doi: 10.1037/0033-2909.131.6.925. doi: 10.1037/0033-2909.131.6.925. [DOI] [PubMed] [Google Scholar]
- Raudenbush S, Bryk A, Cheong Y, Congdon R, du Toit M. HLM 7. Scientific Software International Inc; Lincolnwood, IL: 2011. [Google Scholar]
- Roter DL, Frankel RM, Hall JA, Sluyter D. The expression of emotion through nonverbal behavior in medical visits. Journal of General Internal Medicine. 2006;21(S1):S28–S34. doi: 10.1111/j.1525-1497.2006.00306.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rothman AJ, Bartels RD, Wlaschin J, Salovey P. The Strategic Use of Gain- and Loss-Framed Messages to Promote Healthy Behavior: How Theory Can Inform Practice. Journal of Communication. 2006;56:S202–S220. doi: 10.1111/j.1460-2466.2006.00290.x. [Google Scholar]
- Rothman AJ, Salovey P. Shaping perceptions to motivate healthy behavior: The role of message framing. Psychological Bulletin. 1997;121(1):3–19. doi: 10.1037/0033-2909.121.1.3. doi: 10.1037/0033-2909.121.1.3. [DOI] [PubMed] [Google Scholar]
- Ruby MB, Falk CF, Heine SJ, Villa C, Silberstein O. Not all collectivisms are equal: Opposing preferences for ideal affect between East Asians and Mexicans. Emotion. 2012;12(6):1206–1209. doi: 10.1037/a0029118. doi: 10.1037/a0029118. [DOI] [PubMed] [Google Scholar]
- Rule NO, Krendl AC, Ivcevic Z, Ambady N. Accuracy and consensus in judgments of trustworthiness from faces: Behavioral and neural correlates. Journal of personality and social psychology. 2013;104(3):409–426. doi: 10.1037/a0031050. doi: 10.1037/a0031050. [DOI] [PubMed] [Google Scholar]
- Rusting CL, Larsen RJ. Moods as sources of stimulation: Relationships between personality and desired mood states. Personality and Individual Differences. 1995;18(3):321–329. doi: 10.1016/0191-8869(94)00157-n. [Google Scholar]
- Sakaki M, Nga L, Mather M. Amygdala Functional Connectivity with Medial Prefrontal Cortex at Rest Predicts the Positivity Effect in Older Adults’ Memory. Journal of Cognitive Neuroscience. 2013;25(8):1206–1224. doi: 10.1162/jocn_a_00392. doi: 10.1162/jocn_a_00392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Scheibe S, English T, Tsai JL, Carstensen LL. Striving to feel good: Ideal affect, actual affect, and their correspondence across adulthood. Psychology and Aging. 2013;28(1):160–171. doi: 10.1037/a0030561. doi: 10.1037/a0030561. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schmid Mast M, Hall J, Roter D. Caring and Dominance Affect Participants’ Perceptions and Behaviors During a Virtual Medical Visit. Journal of General Internal Medicine. 2008;23(5):523–527. doi: 10.1007/s11606-008-0512-5. doi: 10.1007/s11606-008-0512-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shen L, Dillard JP. The influence of behavioral inhibition/approach systems and message framing on the processing of persuasive health messages. Communication Research. 2007;34(4):433–467. [Google Scholar]
- Sherman D, Mann T, Updegraff J. Approach/avoidance motivation, message framing, and health behavior: understanding the congruency effect. Motivation and Emotion. 2006;30(2):164–168. doi: 10.1007/s11031-006-9001-5. doi: 10.1007/s11031-006-9001-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sims T, Tsai JL, Jiang D, Wang I, Fung HH, Zhang X. Wanting to Maximize the Positive and Minimize the Negative: Implications for Affective Experience in American and Chinese Contexts. Manuscript under review. 2014 doi: 10.1037/a0039276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sims T, Tsai JL, Koopmann-Holm B, Thomas EAC, Goldstein MK. Choosing a physician depends on how you want to feel: The role of ideal affect in health-related decision making. Emotion. 2014;14(1):187–192. doi: 10.1037/a0034372. doi: 10.1037/a0034372. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Smith SM, Fabrigar LR, Powell DM, Estrada M-J. The Role of Information-Processing Capacity and Goals in Attitude-Congruent Selective Exposure Effects. Personality and Social Psychology Bulletin. 2007;33(7):948–960. doi: 10.1177/0146167207301012. doi: 10.1177/0146167207301012. [DOI] [PubMed] [Google Scholar]
- Todorov A, Pakrashi M, Oosterhof NN. Evaluating Faces on Trustworthiness After Minimal Time Exposure. Social Cognition. 2009;27(6):813–833. doi: 10.1521/soco.2009.27.6.813. [Google Scholar]
- Tsai JL. Ideal Affect: Cultural Causes and Behavioral Consequences. Perspectives on Psychological Science. 2007;2(3):242–259. doi: 10.1111/j.1745-6916.2007.00043.x. doi: 10.1111/j.1745-6916.2007.00043.x. [DOI] [PubMed] [Google Scholar]
- Tsai J, Ang J, Blevins E, Fung H, Jiang D, Elliott J, Uchida Y, Lee YC, Zhang X, Fernandez H, Kaiser A, Govindama Y, Haddock L. Leader’s emotional expressions reflect nation’s ideal affect. Manuscript under review. 2014 [Google Scholar]
- Tsai JL, Chim L, Sims T. Understanding Affect and Consumer Behavior Across Cultures: The Role of Ideal Affect. In: Lee AY, Ng S, editors. Handbook of Culture and Consumer Behavior. in press. [Google Scholar]
- Tsai JL, Knutson B, Fung HH. Cultural variation in affect valuation. Journal of Personality and Social Psychology. 2006;90(2):288–307. doi: 10.1037/0022-3514.90.2.288. doi: 10.1037/0022-3514.90.2.288. [DOI] [PubMed] [Google Scholar]
- Tsai JL, Louie JY, Chen EE, Uchida Y. Learning What Feelings to Desire: Socialization of Ideal Affect Through Children’s Storybooks. Personality and Social Psychology Bulletin. 2007;33(1):17–30. doi: 10.1177/0146167206292749. doi: 10.1177/0146167206292749. [DOI] [PubMed] [Google Scholar]
- Tsai JL, Miao FF, Seppala E. Good Feelings in Christianity and Buddhism: Religious Differences in Ideal Affect. Personality and Social Psychology Bulletin. 2007;33(3):409–421. doi: 10.1177/0146167206296107. doi: 10.1177/0146167206296107. [DOI] [PubMed] [Google Scholar]
- Tsai JL, Miao FF, Seppala E, Fung HH, Yeung DY. Influence and adjustment goals: Sources of cultural differences in ideal affect. Journal of personality and social psychology. 2007;92(6):1102–1117. doi: 10.1037/0022-3514.92.6.1102. doi: 10.1037/0022-3514.92.6.1102. [DOI] [PubMed] [Google Scholar]
- Västfjäll D, Gärling T, Kleiner M. Does it make you happy feeling this way? A core affect account of preference for current mood. Journal of Happiness Studies. 2001;2(4):337–354. doi: 10.1023/a:1013934417046. [Google Scholar]
- Ware JE, Kosinski M, Dewey JE, Gandek B. SF-36 health survey: manual and interpretation guide. Quality Metric Inc; 2000. [Google Scholar]
- Zebrowitz LA, Montepare JM. Social Psychological Face Perception: Why Appearance Matters. Social and Personality Psychology Compass. 2008;2(3):1497–1517. doi: 10.1111/j.1751-9004.2008.00109.x. doi: 10.1111/j.1751-9004.2008.00109.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhao G, Pechmann C. The Impact of Regulatory Focus on Adolescents’ Response to Antismoking Advertising Campaigns. Journal of Marketing Research. 2007;44(4):671–687. doi: 10.1509/jmkr.44.4.671. [Google Scholar]