Exploring the Effect of Time Horizon Perspective on Persuasion: Focusing on Both Biological and Embodied Aging
Round 1
Reviewer 1 Report
I read the paper and found it interesting and potentially relevant for both theory and practice. It is well-written, structured, and rooted in good theory. I don’t have very critical concerns about the theoretical development, but some points about the empirical part drew my attention and, I think, deserve reflection by the authors and substantial revision.
1) On page 4, the authors wrote that people can take one of two stances when reading a text: efferent or esthetic. Yet, they never defined what they meant by those two terms (i.e., efferent and esthetic) in that context.
2) As for Experiment 1, I am not sure whether 60-year old people can be considered older individuals. Scientific and social progress has been improving quality of life for all age groups and, as a result, it is unlikely that, today, a 60-year old individual feels and behaves as older people. Indeed, recent work on older consumers (see, for example, Amatulli et al. 2018, Journal of Consumer Research) only considers people aged 65 (not 60) and over. The authors should at least explain why they expanded this segment of consumers to include respondents between 60 and 65.
3) In all the experiments, it was not clear whether Spela was a real or fictitious brand name. Please specify.
4) Experiment 3 employed a too small sample. Given the 2 x 2 between-subjects design of that experiment, I’d suggest using around 40 respondents per cell, which means using a sample of around 160 respondents.
5) In addition, as Experiment 3 examined a moderated-mediation, it is important to report the index of moderated-mediation (the PROCESS macro by Hayes produces it automatically). That index represents that indirect effect of the interaction term on the dependent variable. Therefore, if that index is significant, then moderated mediation is established.
6) Finally, I see that all experiments basically use the same stimulus materials and experimental setting. This confers internal validity to the studies but reduces generalizability. The obtained results indeed emerge from the same experimental setting as well as the same brand name and advertisement. This implies that such results could be interpreted as brand- and context-specific. I’d suggest authors replace at least one of the studies with a new one that is conducted in a different experimental setting.
Author Response
Notes to Sustainability Reviewer A: #381115
General Overview of Revision
We are grateful to the Sustainability review team for the very insightful comments and suggestions. Following all comments and suggestions, we have revised the manuscripts overall. Broadly, we have made the following three major revisions to the paper.
First, we have conducted Experiment 3 again through Mturk (Amazon Mechanical Turk) by selecting new product category (e.g., coffee) and the number of participants in this experiment have increased from 75 (in original manuscript) to 172.
Second, in Experiment 2, we deleted data from older adults aged 60-64 because many of literature on aging has regard older adults as up to 65. Therefore, the results of this experiment have revised.
Third, in the general discussion section, we created both managerial implication and theoretical contribution areas including directions to future research and limitations.
We provide more detailed comments below. For clarity and convenience, we have paraphrased the Reviewer A’s comments in italics. The all revisions are in the manuscript, and the specific page numbers are indicated in revision notes.
Specific Comment
1. On page 4, the authors wrote that people can take one of two stances when reading a text: efferent or esthetic. Yet, they never defined what they meant by those two terms (i.e., efferent and esthetic) in that context.
Based on your comments, we described the definition of efferent and esthetic stances in the revised manuscript as follows:
Goal Pursuit [p. 4]
When people choose the efferent stance, they tend to analyze, abstract, and accumulate knowledge after the reading. On the other side, the aesthetic stance is involved primarily with attention to experiencing what is being evoked during the reading [35].
2. As for Experiment 1, I am not sure whether 60-year old people can be considered older individuals. Scientific and social progress has been improving quality of life for all age groups and, as a result, it is unlikely that, today, a 60-year old individual feels and behaves as older people. Indeed, recent work on older consumers (see, for example, Amatulli et al. 2018, Journal of Consumer Research) only considers people aged 65 (not 60) and over. The authors should at least explain why they expanded this segment of consumers to include respondents between 60 and 65.
1) First, in the original manuscript, we included older adults aged 60 to 64 as participants based on some of previous research on aging which has conducted experiments involving older adults aged 60 to 64 (see Table 1). However, as your comments, because many of literature on aging are considered older adults than 65 years, we included older adults over 65 only as participants in the revised manuscript.
Table 1. Age Ranges of Participants in Experiments Conducted by Previous Researchers
Older adults | Younger adults | |
Current paper | ||
Original manuscript | 60–83 | 20–28 |
Revised manuscript | 65–83 | 20–28 |
Including aged 60-64 | ||
Einstein and Daniel (1995) | 60–78 | 18–27 |
Fung and Carstensen (2003) | 55–85 | 18–37 |
Hertzog and Touron (2011) | 61–81 | 18–24 |
Lockenhoff and Carstensen (2007) | 62–93 | 22–39 |
Rawson and Touron (2009) | 60–75 | 19–26 |
Excepting aged 60–64 | ||
Charles, Mather, and Carstensen (2003) | 65–80 | 18–29 |
Drolet, Williams, and Lau-Gesk (2007) | 65–88 | 18–25 |
Kave and Mashal (2012) | 70–86 | 18–29 |
Mather and Knight (2005) | 65–83 | 18–29 |
Reder et al. (1986) | 64–75 | 20–31 |
Yoon (1997) | 65–79 | 18–22 |
Williams and Drolet (2005) | 65–98 | 19–24 |
2) In Experiment 2, the results for the participants aged 65 to 83 (M = 71.81, SD = 4.82) in old condition had described by deleting data for participants aged 60 to 64 as follows:
Experiment 2 [p. 5]
Participants and design. Experiment 1 had 246 participants, 124 of whom were classified as older adults (age range: 65–83; M = 71.81, SD = 4.82) and 122 of whom were classified as younger adults (age range: 20–28; M = 23.66, SD = 6.20). This experiment was a 2 (age: old vs. young) × 2 (goal pursuit: goal priming vs. control) × 2 (message type: emotional vs. factual) between-subjects design. Attitude toward the product in the advertisement was the dependent variable.
Results
Manipulation checks. We excluded two participants who had heard of ‘Spela’ from the experiment. Participants in the old (vs. neutral) condition indicated that they imagined their future in terms of becoming older adults more vividly (Molder = 5.72 vs. Mcontrol = 2.30; F(1, 238) = 983.14, p < .001), participants who read the scenario for a specific goal pursuit (M = 5.21) perceived a higher level of goal relevancy than participants who did not read the scenario (M = 3.69; F(1, 238) = 423.73, p < .001), and participants confronted with the emotional (vs. factual) message rated the advertisement message as inducing less rational and more emotional feelings (Memotional = –1.63 vs. Mfactual = 1.18; F(1, 238) = 479.21, p < .001).
Experimental results. A 2 (age priming) × 2 (goal pursuit) × 2 (message type) ANOVA on the attitudes measure revealed main effects of age priming, goal pursuit, and message type. Specifically, participants primed with younger age (M = 4.69) held more favorable attitudes toward the product than those primed with older age (M = 4.57; F(1, 232) = 10.91, p < .005). Furthermore, participants in the goal priming (vs. control) condition held more positive attitudes (Mgoal = 4.95 vs. Mcontrol = 4.31; F(1, 232) = 266.52, p < .001) and those who read the factual message (M = 5.01) had more favorable attitudes toward the emotional message (M = 4.24; F(1, 232) = 379.55, p < .001). However, age priming × goal pursuit × message type was not significant (F < 1, see Figures 3A and 3B).
Figure 3. The effect of embodied aging on persuasion (Experiment 2)
A. Goal priming B. Control
3. I see that all experiments basically use the same stimulus materials and experimental setting. This confers internal validity to the studies but reduces generalizability. The obtained results indeed emerge from the same experimental setting as well as the same brand name and advertisement. This implies that such results could be interpreted as brand- and context-specific. I’d suggest authors replace at least one of the studies with a new one that is conducted in a different experimental setting.
We totally agree with what you pointed out. We changed the target product category of Experiment 3 from sun cream to coffee and created a new coffee advertisement. A coffee-related goal priming scenario was also created as following:
Experiment 3 [p. 9]
Figure 1. Goal priming scenario
Figure 2. Emotional message Figure 3. Factual message
4. In all the experiments, it was not clear whether Spela was a real or fictitious brand name. Please specify
In the revised manuscript, we described information of ‘Spela’ and ‘Con Soc’ that actually existed as following:
Experiment 1 [p. 5]
Procedure and materials. As a target product category, we selected a sun cream. We used an unfamiliar brand name 'Spela' which exists in reality to increase external validity. The data of participants who answered yes to the question of knowing or hearing ‘Spela’ were excluded from the analysis.
Experiment 3 [p. 9]
Procedure and materials. We selected a coffee as a target product category based on [10]. We used brand name 'Con Soc' that was not familiar to the participants and actually existed. The data of participants who answered yes to the question of knowing or hearing ‘Con Soc’ were excluded from the analysis.
5. Experiment 3 employed a too small sample. Given the 2 x 2 between-subjects design of that experiment, I’d suggest using around 40 respondents per cell, which means using a sample of around 160 respondents.
We performed Experiment 3 again using Mturk (Amazon Mechanical Turk) with the new coffee stimuli presented above. The number of participants increased from 75 (in original manuscript) to 175, and the number of participants per condition was more than 40 as follows.
Table 2. The number of participants per condition in Experiment 3
Goal priming | |||
Goal priming | Control | ||
Message type | Emotional message | 43 | 43 |
Factual message | 43 | 43 | |
Experiment 3 [p. 8]
Participants and design. We recruited 172 participants from Amazon Mechanical Turk. The mean age of the workers was 27.20 years (age range: 25–35; SD = 3.57), and 97 were female.
Figure 4. Mturk screen shot during performing Experiment 3
6. As Experiment 3 examined a moderated-mediation, it is important to report the index of moderated-mediation (the PROCESS macro by Hayes produces it automatically). That index represents that indirect effect of the interaction term on the dependent variable. Therefore, if that index is significant, then moderated mediation is established.
The results of Experiment 3 with the coffee stimuli were described newly including moderated mediation index exactly as following:
Experiment 3 [pp. 9-10]
Experimental results. We expected that emotional attachment would mediate the moderating effect of goal pursuit, corresponding to Model 7 in Hayes’s [48] work. Thus, we employed a bootstrapping method, using the PROCESS macro with 10,000 resamples [48]. We coded each message type condition (1 = emotional message, –1 = factual message) as the independent variable, attitude as the dependent variable, the continuous goal relevancy scores as the moderator, and emotional attachment as the mediator of the moderator. Participants in the goal priming (vs. control) condition had more favorable attitudes (95% CI: .18, .53; β = .35, p < .001), and the effect of emotional attachment on attitudes also positively significant (95% CI: .04, .32; β = .18, p < .05). Both participants in control (vs. goal priming; 95% CI: –.63, –.30; β = –.47, p < .001) and emotional (vs. factual; 95% CI: –.50, –.18; β = –.34, p < .001) condition had more emotional attachment. The goal pursuit × message type interaction was significant (95% CI: .16, .48; β = .32, p < .001). More important, the mediation effect of emotional attachment was significant at emotional message (95% CI: –.31, –.01; β = –.14), but it was not significant at factual messages (95% CI: –.11, –.01; β = –.03). Taken together, the PROCESS results provide support for our assertion that emotional attachment mediates the moderating effect of goal pursuit on attitudes (95% CI: .01, .29; β = .11). In addition, there were no effects of gender (p > .10).
In this revision, we have tried to address all the concerns raised. We are very appreciative of the useful comments and suggestions, which have helped to significantly improve the quality of this paper. We look forward to your comments on the revised manuscript.
Reviewer 2 Report
In the assessment of the paper submitted for the review, I specifically focused on the discussed issues, applied research procedure, substantive content of the paper and its structure.
The considerations conducted in the paper are focused on such categories as: horizon perspective, message type, persuasion, goal pursuit, emotional attachment, biological and embodied aging.
The subject area discussed in the paper is important.
Presentation of the results of own empirical research is the value of the paper. However, deliberations conducted in the paper need to be expanded. Therefore, it is specifically recommended to:
- indicate the time the empirical research was conducted (in abstract),
- take into consideration other latest publications in the sphere of discussed subject matter,
- expand the managerial implications in the article (as a separate part of article),
- describe the limitations of conducted research and to indicate the trends for further empirical research (as a separate part of article),
- eliminate the appendix: “Goal priming scenario”.
Author Response
Notes to Sustainability Reviewer B: #381115
General Overview of Revision
We are grateful to the Sustainability review team for the very insightful comments and suggestions. Following all comments and suggestions, we have revised the manuscripts overall. Broadly, we have made the following three major revisions to the paper.
First, we have conducted Experiment 3 again through Mturk (Amazon Mechanical Turk) by selecting new product category (e.g., coffee) and the number of participants in this experiment have increased from 75 (in original manuscript) to 172.
Second, in Experiment 2, we deleted data from older adults aged 60-64 because many of literature on aging has regard older adults as up to 65. Therefore, the results of this experiment have revised.
Third, in the general discussion section, we created both managerial implication and theoretical contribution areas including directions to future research and limitations.
We provide more detailed comments below. For clarity and convenience, we have paraphrased the Reviewer B’s comments in italics. The all revisions are in the manuscript, and the specific page numbers are indicated in revision notes.
Specific Comment
1. Indicate the time the empirical research was conducted (in abstract)
According to your comments, we have described when each experiment was conducted in abstract as following. Especially, Experiment 3 with new stimuli has been re-conducted recently during the revision process.
Abstract [p. 1]
We investigated the moderating effect of goal pursuit on the relationship between biological aging and persuasion in October 2016 (Experiment 1). Results showed that both older and younger adults were more favorable toward factual (vs. emotional) messages when they pursue goal. Experiment 2 conducting in March 2017 examined whether embodied aging affects persuasion. The results indicated that younger adults with embodied aging did not show a similar persuasion pattern with actual older adults, while embodiment studies demonstrated that younger adults with embodied aging act like older adults. That means that only THP changes the angle leading to processing. In final experiment conducting in November 2018, younger adults primed with limited THP were prefer emotional to factual messages, which was similar to actual older adults’ responses. Moreover, we showed that emotional attachment mediated the effect of message types on persuasion of younger adults with limited THP.
2. Take into consideration other latest publications in the sphere of discussed subject matter
Recent papers related to our study are described in the literature review part newly as follows.
Socioemotional Selectivity Theory (SST) [p. 3]
Recent research showed that THP influences impulsive buying depending on type of relationship deficits [23]. They insisted that a lack of emotional attachments is associated with limited THP, whereas a lack of social connections is related to expansive THP. Therefore, emotionally (vs. socially) lonely participants with limited THP make more impulsive purchasing, whereas socially (vs. emotionally) lonely participants with expansive THP make more impulsive purchasing.
Embodied Aging [p. 4]
For example, Larson and Billeter [42] investigated that bodily states such as leaning back in a chair influenced consumer choice of compromise option by activating the concept of balance.
References
[23] Sinha, J.; Wang, J. How time horizon perceptions and relationship deficits affect impulsive consumption. J. Mark. Res. 2013, 50, 590–605, https://doi.org/10.1509/jmr.11.0246
[42] Larson, J.S.; Billeter, D.M. Consumer behavior in “equilibrium”: How experiencing physical balance increases compromise choice. J. Mark. Res. 2013, 50, 535–547, https://doi.org/10.1509/jmr.11.0455
3. Expand the managerial implications in the article (as a separate part of article)
In the revised manuscript, we have suggested practical implications more detail as a separate part of article as following:
Managerial implications [pp. 10-11]
Both older and younger adults have more favorable attitudes toward a factual (vs. emotional) message when they are primed with pursuing a specific goal. However, when they are not primed with pursuing a specific goal, older adults hold more favorable attitudes toward an emotional (vs. factual) message, whereas younger adults have more positive attitudes toward a factual (vs. emotional) message. These findings are consistent with prior research on SST. From these results, companies could use, for example, consumer age when establishing a marketing strategy, emphasizing an emotional message in advertisements geared to older adults or a factual message in advertisements intended for younger adults. However, although older adults were more persuaded by an emotional (vs. factual) message, there are other ways to establish a more effective marketing strategy. Specifically, to emphasize a factual message (e.g., a suggestion to use a specific utility company), marketers should activate specific goals for older adults by providing highly goal-relevant information to induce information processing of factual messages.
According to prior SST research, older adults hold more favorable attitudes toward emotional (vs. factual) messages, whereas the opposite is true for younger adults. However, we found that both older and younger adults were more persuaded by a factual (vs. emotional) message when they were primed to pursue a specific goal (Experiment 1). Furthermore, prior research has shown that people who view time as limited are more persuaded by an emotional (vs. factual) message, whereas people who view time as expansive are more persuaded by a factual (vs. emotional) message. However, our results indicate that people primed with a specific goal hold more favorable attitudes toward a factual (vs. emotional) message, regardless of the type of THP. With these results, we propose that when marketers working for a cosmetic company, for example, want to emphasize the specific utility of and provide a factual message about a cosmetic product (e.g., wrinkle prevention and improvement of skin elasticity), they also should include a highly goal-relevant message (i.e., antiaging) to activate consumers’ specific goals. Conversely, when emphasizing the emotional utility (e.g., self-satisfaction through beauty), they should focus on limited THP–related messages (e.g., “Life is short,” “Enjoy your short life”). This study provides instruction on the selection of message type to help marketers choose a message effectively and efficiently for use in an advertisement.
We also found that when people who viewed time as limited were primed to pursue a specific goal, the mediated moderating role of emotional attachment was not significant on the effect of message type on persuasion. However, when they were not primed to pursue a specific goal, they were more persuaded by an emotional (vs. factual) message because it allowed them to form an emotional attachment. We inspected in depth the effect of age on persuasion by directly measuring emotional attachment and by analyzing the mediating role of emotional attachment. Therefore, when marketers use emotional messages to communicate with older adults, they should increase emotional intimacy. For example, they could instill high emotional attachment in older adults by using anthropomorphic wording.
Age discrimination is a form of social discrimination and prejudice, especially for the elderly. Social norms that connect aging with mental and physical weakness alienate the elderly by restricting social participation of the elderly, no matter how healthy and healthy the spirit and body of the elderly individual is. Despite the individuality of individual elderly people and their differentiation from other elderly people, society recognizes and treats all elderly people equally based on the elderly’s high age. Such age discrimination is carried out both on a personal level and on a social level, and it is also important to consider the effects of repeated stereotypes, myths, direct expression through hate speech, indirect expression through refusal of contact and exchange. In recent years, information and knowledge has emerged as a major social mechanism to replace existing capital and commodities. Unlike conventional general merchandise (such as clothes, shoes, food, etc.), information and knowledge are circulating rapidly through creation, utilization, and extinction, and both quantitative and qualitative changes are beyond predictions. Therefore, social inequality (hierarchy, class, sex, education, region, etc.) is also widespread, and age is also a major factor that differentiates access and use of information. The results of this study provide implications for understanding the mechanisms and patterns of how older people process their environmental information, which is expected to be useful that society maintains sustainable welfare beyond ‘ageism’ in terms of information and knowledge.
4. Describe the limitations of conducted research and to indicate the trends for further empirical research (as a separate part of article)
Based on your comments, we have created section of Theoretical contribution including direction to future research and limitations as following:
Theoretical contribution [p. 11]
This study investigated the effect of embodied aging on persuasion (Experiment 2) and found that both participants primed with terminology associated with older adults and those not primed in this way held more favorable attitudes toward a factual (vs. emotional) message, regardless of goal pursuit priming. This result differs from the results of Experiment 1, in which older adults in the control condition held more favorable attitudes toward the emotional than factual message. We inferred that the reason for this was that younger adults’ THP was unaltered when they were primed with terminology associated with older adults. Younger adults cannot vividly imagine the maturity and expansive THP that results from the long life experienced by older adults. Therefore, the difference between the results of Experiments 1 (biological age) and 2 (embodied aging) indicate that THP does not influence the effect of embodied aging on persuasion, though biological age does have an effect on persuasion.
Further research could investigate the mediating role of mortality salience on the effect of message type on persuasion when people have limited THP. Prior research on mortality salience has indicated that older (vs. younger) adults have higher mortality salience and that they tend to focus on emotional rather than factual or material attributes [49]. Therefore, we assume that people with a limited THP will process emotional messages from a standpoint of high mortality salience. If so, the mediating role of mortality salience would be significant. However, when primed to pursue a specific goal, people’s fear of death could decrease and the value of life could become clearer, such that the approach, rather than inhibition, system is activated. In this case, the mediating role of mortality salience would not be significant. In this article, we examined the effects of THP and message type on persuasion in a various ways.
Potential limitations of this study include the fact that THP and message typed were manipulated simultaneously in one advertisement based on Williams and Drolet [10] in Experiment 3. In future research, THP could be manipulated in a different way instead of being manipulated through advertising such as sentence unscramble task [50]. For example, participants in limited THP condition are allowed to complete the sentence using words such as short, limited, and present, whereas those in expansive THP condition are presented with words such as long, extended, and future. As an alternative method, participants could be exposed these words repeatedly at regular intervals these words [45]. We hope that further research will bring more insights to this issue.
5. Eliminate the appendix: “Goal priming scenario”
Thank you for pointing this out in detail. We have deleted the goal priming scenario in revised manuscript.
In this revision, we have tried to address all the concerns raised. We are very appreciative of the useful comments and suggestions, which have helped to significantly improve the quality of this paper. We look forward to your comments on the revised manuscript.
