{"Affiliation":[{"label":"Affiliation","value":"Business, Sauder School of","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","classmap":"vivo:EducationalProcess","property":"vivo:departmentOrSchool"},"iri":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","explain":"VIVO-ISF Ontology V1.6 Property; The department or school name within institution; Not intended to be an institution name."},{"label":"Affiliation","value":"Marketing and Behavioural Science, Division of","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","classmap":"vivo:EducationalProcess","property":"vivo:departmentOrSchool"},"iri":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","explain":"VIVO-ISF Ontology V1.6 Property; The department or school name within institution; Not intended to be an institution name."}],"AggregatedSourceRepository":[{"label":"Aggregated Source Repository","value":"DSpace","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider","classmap":"ore:Aggregation","property":"edm:dataProvider"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider","explain":"A Europeana Data Model Property; The name or identifier of the organization who contributes data indirectly to an aggregation service (e.g. Europeana)"}],"Campus":[{"label":"Campus","value":"UBCV","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeCampus","classmap":"oc:ThesisDescription","property":"oc:degreeCampus"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeCampus","explain":"UBC Open Collections Metadata Components; Local Field; Identifies the name of the campus from which the graduate completed their degree."}],"Creator":[{"label":"Creator","value":"Yi, Shangwen","attrs":{"lang":"","ns":"http:\/\/purl.org\/dc\/terms\/creator","classmap":"dpla:SourceResource","property":"dcterms:creator"},"iri":"http:\/\/purl.org\/dc\/terms\/creator","explain":"A Dublin Core Terms Property; An entity primarily responsible for making the resource.; Examples of a Contributor include a person, an organization, or a service."}],"DateAvailable":[{"label":"Date Available","value":"2025-03-27T16:09:58Z","attrs":{"lang":"","ns":"http:\/\/purl.org\/dc\/terms\/issued","classmap":"edm:WebResource","property":"dcterms:issued"},"iri":"http:\/\/purl.org\/dc\/terms\/issued","explain":"A Dublin Core Terms Property; Date of formal issuance (e.g., publication) of the resource."}],"DateIssued":[{"label":"Date Issued","value":"2025","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/issued","classmap":"oc:SourceResource","property":"dcterms:issued"},"iri":"http:\/\/purl.org\/dc\/terms\/issued","explain":"A Dublin Core Terms Property; Date of formal issuance (e.g., publication) of the resource."}],"Degree":[{"label":"Degree (Theses)","value":"Doctor of Philosophy - PhD","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#relatedDegree","classmap":"vivo:ThesisDegree","property":"vivo:relatedDegree"},"iri":"http:\/\/vivoweb.org\/ontology\/core#relatedDegree","explain":"VIVO-ISF Ontology V1.6 Property; The thesis degree; Extended Property specified by UBC, as per https:\/\/wiki.duraspace.org\/display\/VIVO\/Ontology+Editor%27s+Guide"}],"DegreeGrantor":[{"label":"Degree Grantor","value":"University of British Columbia","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeGrantor","classmap":"oc:ThesisDescription","property":"oc:degreeGrantor"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeGrantor","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the institution where thesis was granted."}],"Description":[{"label":"Description","value":"Many businesses routinely offer consumers price promotions, with or without restrictions, presented in various formats. Despite their prevalence in marketing, the psychological and behavioral mechanisms driving consumer responses to these promotions remain poorly understood. This thesis explores these mechanisms through two related essays, each addressing critical questions about price promotion effectiveness.\r\nEssay 1 examines the effectiveness of two comparable types of restricted price promotions: threshold promotions (conditional on spending more than a threshold amount; e.g., \u201cGet $5 off on orders of $10 or more\u201d) and capped promotions (limited to a maximum dollar value; e.g., \u201cGet 50% off, up to $5 per order\u201d). Results show that threshold promotions lead to higher purchase intentions, conversion rates, and overall sales than comparable capped promotions\u2014even though capped promotions are equivalent or better in economic savings\u2014when the trigger value (the spending amount at which the promotion activates or caps) is low. This effect occurs because consumers raise their promotion expectations for capped promotions and lower their spending expectations for threshold promotions, leading them to perceive the threshold promotion as fairer. However, this effect reverses when the trigger value is high, where consumers perceive capped promotions as fairer and prefer them to threshold promotions. \r\nEssay 2 investigates two types of percentage framings for price promotions. One is all-inclusive percentage framing (AIPF), and the other one is implicitly partitioned percentage framing (IPPF). We provide evidence supporting that IPPF may lead to higher purchase intention and deal evaluation than AIPF, but only when the promoted products are of high self-relevance to consumers. It is because IPPF leads to process-oriented mental simulation, and thus creates a \u201cprocess amplifier effect\u201d for deal evaluation.","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/description","classmap":"dpla:SourceResource","property":"dcterms:description"},"iri":"http:\/\/purl.org\/dc\/terms\/description","explain":"A Dublin Core Terms Property; An account of the resource.; Description may include but is not limited to: an abstract, a table of contents, a graphical representation, or a free-text account of the resource."}],"DigitalResourceOriginalRecord":[{"label":"Digital Resource Original Record","value":"https:\/\/circle.library.ubc.ca\/rest\/handle\/2429\/90534?expand=metadata","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO","classmap":"ore:Aggregation","property":"edm:aggregatedCHO"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO","explain":"A Europeana Data Model Property; The identifier of the source object, e.g. the Mona Lisa itself. This could be a full linked open date URI or an internal identifier"}],"FullText":[{"label":"Full Text","value":" TWO ESSAYS ON PROMOTION FRAMING  by  Shangwen Yi  B.S., Beijing Normal University, 2019  A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Business Administration - Marketing)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   March 2025  \u00a9 Shangwen Yi, 2025  ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: Two Essays on Promotion Framing  submitted by Shangwen Yi in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Business Administration - Marketing  Examining Committee: David Hardisty, Associate Professor, Sauder School of Business, UBC Supervisor  Dale Griffin, Professor, Sauder School of Business, UBC Co-Supervisor Darren Dahl, Professor, Sauder School of Business, UBC Supervisory Committee Member JoAndrea Hoegg, Professor, Sauder School of Business, UBC University Examiner Luke Clark, Professor, Professor, Department of Psychology, UBC University Examiner  Additional Supervisory Committee Members: Thomas Allard, Associate Professor, Lee Kong Chian School of Business, Singapore Management University Supervisory Committee Member iii  Abstract Many businesses routinely offer consumers price promotions, with or without restrictions, presented in various formats. Despite their prevalence in marketing, the psychological and behavioral mechanisms driving consumer responses to these promotions remain poorly understood. This thesis explores these mechanisms through two related essays, each addressing critical questions about price promotion effectiveness. Essay 1 examines the effectiveness of two comparable types of restricted price promotions: threshold promotions (conditional on spending more than a threshold amount; e.g., \u201cGet $5 off on orders of $10 or more\u201d) and capped promotions (limited to a maximum dollar value; e.g., \u201cGet 50% off, up to $5 per order\u201d). Results show that threshold promotions lead to higher purchase intentions, conversion rates, and overall sales than comparable capped promotions\u2014even though capped promotions are equivalent or better in economic savings\u2014when the trigger value (the spending amount at which the promotion activates or caps) is low. This effect occurs because consumers raise their promotion expectations for capped promotions and lower their spending expectations for threshold promotions, leading them to perceive the threshold promotion as fairer. However, this effect reverses when the trigger value is high, where consumers perceive capped promotions as fairer and prefer them to threshold promotions.  Essay 2 investigates two types of percentage framings for price promotions. One is all-inclusive percentage framing (AIPF), and the other one is implicitly partitioned percentage framing (IPPF). We provide evidence supporting that IPPF may lead to higher purchase intention and deal evaluation than AIPF, but only when the promoted products are of high self-relevance to consumers. It is because IPPF leads to process-oriented mental simulation, and thus creates a \u201cprocess amplifier effect\u201d for deal evaluation.  iv  Lay Summary The purpose of this dissertation is to examine how consumers react to price promotions used by firms. Essay 1 examines how people respond to promotions with restrictions. Although capped promotions with a high percentage but low cap (e.g., \u201c50% off with a maximum discount of $5\u201d) are economically superior or equivalent to threshold promotions (e.g., \u201c$5 off on orders of $10 or more\u201d) for consumers, consumers are more likely to make a purchase with threshold promotions.  Essay 2 investigates consumers\u2019 response to promotions highlighting each unit of product. I find that consumers think that the promotions are better when they are described as applying to each unit of the product (e.g., \u201c10% off each bottle\u201d) than when they are not (e.g., \u201c10% off\u201d or \u201c10% off in total\u201d).   v  Preface I am the primary author of the work presented in this Ph.D. dissertation. I was responsible for conducting the literature review, developing the hypotheses, designing the experiments, collecting the data, analyzing the data, and preparing the manuscript. Additional contributions for each essay are described below. I am the primary author of Chapter 1.  I am the primary author of Chapter 2. I designed the experiments, supervised data collection, conducted the analyses and prepared the manuscript. David Hardisty, Dale Griffin, and Thomas Allard assisted in designing the experiments and provided intellectual contributions. The work is intended for publication and has been submitted to a journal.  I am the primary author of Chapter 3. I designed the experiments, supervised data collection, conducted the analyses and prepared the manuscript. David Hardisty and Katherine White assisted in designing the experiments and provided intellectual contributions. I am the primary author of Chapter 4.  Generative artificial intelligence (ChatGPT) was used in the preparation of this dissertation solely to check and improve grammar and language clarity. It was not utilized to generate ideas, analyze data, interpret results, or draft substantive content. Ethical approval for all experimental studies was obtained from the UBC Office of Research Ethics Behavioural Review Board (Human Ethics) under the following certificates: Essay 1: H21-01024, Essay 2: H24-01073.   vi  Table of Contents Abstract ......................................................................................................................................... iii Lay Summary ............................................................................................................................... iv Preface .............................................................................................................................................v Table of Contents ......................................................................................................................... vi List of Tables ................................................................................................................................ xi List of Figures .............................................................................................................................. xii Acknowledgements .................................................................................................................... xiii Dedication ................................................................................................................................... xiv Chapter 1: Introduction ................................................................................................................1 Chapter 2: Promotion Architecture: The Perceived Fairness of Restricted Price Promotions..........................................................................................................................................................4 2.1 Introduction ..................................................................................................................... 4 2.2 Theoretical Development ................................................................................................ 7 2.2.1 Restricted Promotion Architectures: Capped versus Threshold ............................. 8 2.2.2 Fairness Perceptions and Expectation Discrepancies ........................................... 10 2.3 Empirical Studies .......................................................................................................... 19 2.3.1 Study 1: Food-Ordering Advertising Field Study ................................................. 19 2.3.1.1 Method .............................................................................................................. 20 2.3.1.2 Results ............................................................................................................... 21 2.3.1.3 Discussion ......................................................................................................... 21 2.3.2 Study 2: Price Promotions in Ride-Hailing .......................................................... 22 2.3.2.1 Method .............................................................................................................. 22 vii  2.3.2.2 Results and Discussion ..................................................................................... 23 2.3.3 Study 3: Unrestricted Promotion .......................................................................... 28 2.3.3.1 Method .............................................................................................................. 29 2.3.3.2 Results ............................................................................................................... 30 2.3.3.3 Discussion ......................................................................................................... 31 2.3.4 Study 4: Joint Evaluation ...................................................................................... 32 2.3.4.1 Method .............................................................................................................. 33 2.3.4.2 Results ............................................................................................................... 33 2.3.4.3 Discussion ......................................................................................................... 35 2.3.5 Study 5: Trigger Value as a Moderator ................................................................. 36 2.3.5.1 Method .............................................................................................................. 36 2.3.5.2 Results ............................................................................................................... 37 2.3.5.3 Discussion ......................................................................................................... 40 2.3.6 Study 6A & 6B: Overall Sales .............................................................................. 41 2.3.6.1 Study 6a: 40% Promotion Depth ...................................................................... 42 2.3.6.1.1 Method ........................................................................................................ 42 2.3.6.1.2 Results ......................................................................................................... 43 2.3.6.2 Study 6b: 20% Promotion Depth ...................................................................... 44 2.3.6.2.1 Method ........................................................................................................ 44 2.3.6.2.2 Results ......................................................................................................... 45 2.3.6.3 Discussion ......................................................................................................... 45 2.4 General Discussion ....................................................................................................... 47 2.4.1 Practical Implications ............................................................................................ 48 viii  2.4.2 Theoretical Contributions ..................................................................................... 50 2.4.3 Directions for Future Research ............................................................................. 52 Chapter 3: \u201c10% off Each\u201d: How Implicitly Partitioned Percentage Framing Affects Purchase Intentions .....................................................................................................................55 3.1 Introduction ................................................................................................................... 55 3.2 Theoretical Background ................................................................................................ 57 3.2.1 Percentage Framings and Consumer Evaluations ................................................. 57 3.2.2 Integration and Segregation .................................................................................. 59 3.2.3 Mental Simulation ................................................................................................. 62 3.3 Empirical Studies .......................................................................................................... 64 3.3.1 Study 1A: Main Effect of Percentage Framing ..................................................... 64 3.3.1.1 Method .............................................................................................................. 65 3.3.1.2 Results ............................................................................................................... 65 3.3.1.3 Discussion ......................................................................................................... 66 3.3.2 Study 1B: Main Effect of Percentage Framing with Quantity Discounts ............. 67 3.3.2.1 Method .............................................................................................................. 67 3.3.2.2 Results ............................................................................................................... 67 3.3.2.3 Discussion ......................................................................................................... 68 3.3.3 Study 2: Personal Relevance as a Moderator ........................................................ 68 3.3.3.1 Method .............................................................................................................. 68 3.3.3.2 Results ............................................................................................................... 69 3.3.3.3 Discussion ......................................................................................................... 72 3.3.4 Study 3: Mental Simulation as Mechanism .......................................................... 72 ix  3.3.4.1 Method .............................................................................................................. 73 3.3.4.2 Results ............................................................................................................... 73 3.3.4.3 Discussion ......................................................................................................... 74 3.3.5 Study 4: Perceived Number of Discounts ............................................................. 74 3.3.5.1 Method .............................................................................................................. 75 3.3.5.2 Results ............................................................................................................... 75 3.3.5.3 Discussion ......................................................................................................... 76 3.4 General Discussion ....................................................................................................... 76 3.4.1 Theoretical implications ........................................................................................ 77 3.4.2 Practical Implications ............................................................................................ 78 3.4.3 Limitations and Future Research Directions ......................................................... 79 Chapter 4: Conclusion .................................................................................................................82 4.1 Contributions ................................................................................................................ 83 4.2 Strengths and Limitations ............................................................................................. 84 4.3 Future Research Directions ........................................................................................... 85 4.4 Final Thoughts .............................................................................................................. 86 References .....................................................................................................................................87 Appendices ....................................................................................................................................98 Appendix A Chapter 2 Appendices .......................................................................................... 98 A.1 Examples of Promotional Messages ......................................................................... 98 A.2 Examples of the Two Promotion Types Used in Joint Evaluation ........................... 99 A.3 Prevalence of Three Types of Promotions for Food Delivery on Doordash .......... 100 x  A.4 Promotion Expectations, Spending Expectations, and Prevalence of Three Types of Promotions for Food Delivery ............................................................................................ 101 A.5 Prevalence of Low and High Trigger Values in the Marketplace .......................... 103 A.6 Tiktok Field Study Replication ............................................................................... 105 A.7 Restricted Promotions with Low Promotion Depth and Unrestricted Promotion .. 106 A.8 Charity Promotions in Grocery Shopping ............................................................... 109 A.9 The Role of Various Levels of Promotion Depth ................................................... 114 A.10 Trigger Value as a Moderator when Promotion Depth Kept Constant ................... 118 A.11 Capped Promotion versus No Promotion ............................................................... 120 Appendix B Chapter 3 Appendices ......................................................................................... 122 B.1 Examples of Implicit Partitioned Percentage Framing ........................................... 122    xi  List of Tables Table 2.1 Relationship Between Promotion Expectations and Perceptions ................................................ 15  xii  List of Figures Figure 2.1 Discounts for Comparable Capped and Threshold Promotions ................................................... 5 Figure 2.2 Conceptual Framework .............................................................................................................. 16 Figure 2.3 The Interaction Effect Among the Presence of Spending Information, Trigger Value, and Promotion Architecture in Study 2 .............................................................................................................. 27 Figure 2.4 Promotion Architecture and Purchase Intention in Study 3 ....................................................... 31 Figure 2.5 Moderation by Trigger Value in Study 5 ................................................................................... 38 Figure 3.1 The Effect of Framing and Personal Relevance on Purchase Intention ..................................... 71 Figure 3.2 The Effect of Framing and Personal Relevance on Deal Evaluation ......................................... 71 Figure 3.3 The Effect of Framing and Personal Relevance on Basket Size ................................................ 72 xiii  Acknowledgements Completing this dissertation has been an incredible journey, and I am deeply grateful for the support, guidance, and encouragement I have received along the way. First and foremost, I would like to express my heartfelt gratitude to my supervisors, Dave Hardisty and Dale Griffin, for their invaluable intellectual guidance, mentorship, and unwavering support throughout this process. Your insights and encouragement have been instrumental in shaping this work. I am also deeply thankful to my supervisory committee members, Darren Dahl and Thomas Allard, for their thoughtful feedback and guidance, which have greatly enriched my research and personal growth. I extend my sincere thanks to Yann Cornil, Joey Hoegg, Lisa Cavanaugh, Kate White, Deepak Sirwani, Jen Park, Karl Aquino, Dan Skarlicki, Chunhua Wu, Yi Qian, Shin Oblander, Eddie Ning, Wenjia Ba, Baek Jung Kim, and Yanwen Wang for providing invaluable feedback during my research presentations. Your comments and suggestions have significantly enhanced the quality of this dissertation. To my collaborators, Johannes Boegershausen, Greg Nyilasy, Stephan Ludwig, and Dennis Herhausen: working with you has been an incredible privilege. I have learned so much through our collaborations, and I am grateful for the opportunities to grow as a scholar alongside you. I would also like to acknowledge my friends and peers who have brightened this journey with their company and support: Chuck, Rishad, Ekin, Wade, Fan, Qiyuan, Zining, Xixi, Sid, Julie, Guanzhong, Lucy, Will, Xilin, Shakti, Brooke, Leyao, and Neda. Thank you for making this journey cheerful and for ensuring that I never felt alone. Your friendship means the world to me. This dissertation is a testament to the collective efforts of an extraordinary group of people. Thank you all for your contributions and for walking this journey with me. xiv  Dedication This dissertation is dedicated to my parents, Lifang Pan and Yunbing Yi. 1  Chapter 1: Introduction Price promotions are an important tool for businesses aiming to stimulate consumer purchases and increase sales (Chandon, Wansink, and Laurent 2000). These promotions take various forms and may include different restrictions, each designed to appeal to specific consumer preferences and behaviors. Despite their ubiquity in marketing strategies, the psychological and behavioral mechanisms underlying consumer responses to different types of price promotions remain incompletely understood. This thesis aims to investigate these mechanisms through two distinct but related essays, each addressing key questions in the domain of price promotion effectiveness. Price promotions often include restrictions, such as minimum spending requirements or maximum savings limits, to guide consumer behavior and manage company margins. Two common types of restricted price promotions are threshold promotions (e.g., \u201cGet $5 off on orders of $10 or more\u201d) and capped promotions (e.g., \u201cGet 50% off, up to $5 per order\u201d). However, relatively little attention has been paid to the psychological drivers that influence consumer preferences for these promotions. Research on fairness perceptions (Xia, Monroe, and Cox 2004), reference points (Kahneman and Tversky 1979), and mental accounting (Thaler 1985) suggests that consumer expectations and evaluations of promotions play a crucial role in determining their effectiveness. Thus, a deeper exploration of how these factors vary across threshold and capped promotions is needed.  2  The framing of promotional discounts can significantly influence consumer perceptions. For instance, all-inclusive percentage framing (AIPF) and implicitly partitioned percentage framing (IPPF) represent two perceptually comparable but conceptually distinct approaches. AIPF conveys the discount as a single percentage off the total purchase (e.g., \u201cSave 10%\u201d), while IPPF breaks it into smaller, iterative units (e.g., \u201cSave 10% on every item\u201d). While prior studies have demonstrated the importance of framing in deal evaluations (DelVecchio, Krishnan, and Smith, 2007), the underlying mechanisms driving consumer preferences for IPPF versus AIPF remain underexplored.  Therefore, this thesis addresses two central research questions: 1) How do consumers evaluate and respond to threshold versus capped promotions, and what psychological mechanisms drive their preferences? 2) How does the framing of percentage promotions (AIPF vs. IPPF) influence purchase intentions and deal evaluations, and under what conditions are these effects most pronounced? These questions are explored through two essays. Essay 1 examines the relative effectiveness of threshold and capped promotions, focusing on consumer perceptions of fairness and the role of spending and promotion expectations. It seeks to determine how these effects vary across different trigger values. Essay 2 investigates the impact of AIPF and IPPF on purchase intentions and deal evaluations, highlighting the role of process-oriented (vs. outcome-oriented) mental simulations and self-relevance in moderating these effects. 3  This research aims to: 1) advance theoretical frameworks in price promotion research by integrating concepts of fairness perceptions, mental simulation, and self-relevance; 2) provide actionable insights for marketers to design more effective promotion strategies by understanding the psychological underpinnings of consumer responses. The remainder of this thesis is organized as follows. Essay 1 explores the comparative effectiveness of threshold and capped promotions, examining the roles of fairness perceptions and trigger values in Chapter 2. Essay 2 delves into the framing effects of AIPF and IPPF, emphasizing the process amplifier effect and the moderating role of self-relevance in Chapter 3. Chapter 4 provides a comprehensive understanding of how promotion design and framing influence consumer behavior, offering both theoretical contributions and practical implications for marketing practice.      4  Chapter 2: Promotion Architecture: The Perceived Fairness of Restricted Price Promotions 2.1 Introduction Imagine you are at home, debating whether to start cooking dinner or order food delivery. You have received a promotion from a well-known food-delivery platform: \u201cGet $5 off on orders of $10 or more.\u201d Would you take advantage of that promotion? Instead, imagine the same situation, but with a promotion offering \u201cGet 50% off, up to $5 per order.\u201d Would you make the same decision? In other words, are these two price promotions equally effective at prompting purchases?  Price promotions are temporary monetary incentives retailers use to motivate customers (Chandon, Wansink, and Laurent 2000), which can be classified as restricted or unrestricted. Examples of unrestricted price promotions include \u201c50% off\u201d or \u201c$5 off\u201d offers, which yield the same discount regardless of the purchase amount or unit quantity (Chen, Monroe, and Lou 1998). The two promotions in the previous paragraph are instead examples of restricted price promotions because they include a limit or precondition on the incentive. The present research compares two alternative forms of popular restricted price promotions and examines their impacts on consumers. For promotions such as \u201c$5 off if you spend $10 or more,\u201d we use the term threshold promotions, referring to promotions only applicable if the purchase amount exceeds a stated threshold value. Alternatively, for promotions such as \u201c50% off, up to $5 discount,\u201d we use the term capped promotions, referring to promotions with percentage-term reward caps at a maximum stated value. Consumers may see the two restricted promotions presented separately across service 5  providers (see Appendix A.1) or presented side by side on the same platform (e.g., promotions used by stores on apps or flyers; Appendix A.2). Notably, both capped and threshold promotions are as common as unrestricted promotions in the marketplace (see Appendices A.3 and A.4).  These two types of promotions\u2014capped and threshold\u2014can be compared because they carry the same maximum cash incentive (a $5 price discount) and trigger value ($10); trigger value refers to the critical spending level that triggers a change in the discount received. Stated differently, the trigger value is the minimum amount consumers must spend to receive the discount for threshold promotions and the maximum spending amount before a limit applies in capped promotions. We depict this relationship between the two promotion types in Figure 2.1. Notably, capped promotions are economically superior to threshold promotions for consumers until their spending reaches the promotional trigger value. Figure 2.1 Discounts for Comparable Capped and Threshold Promotions  $0$1$2$3$4$5$6$0 $5 $10 $15 $20 $25Discount AmountSpending AmountThreshold Promotion($5 off, min $10)Capped Promotion(50% off, up to $5)Trigger Value6  This paper examines how capped and threshold promotions affect consumers\u2019 purchase decisions. The current research finds that threshold promotions motivate consumers to purchase more than equivalent capped promotions when they perceive the trigger value as low (using the above example, $10 for food delivery) relative to their typical spending in that category (e.g., a $25 expenditure on average). We hypothesize that this pattern occurs because the high percentage in capped promotions (50%) leads consumers to expect more promotional benefits than they receive, while the low spending threshold in threshold promotions leads consumers to perceive the same discount as larger (\u201con orders of $10 or more\u201d). Although the capped offer seems attractive initially, it creates a high expectation for promotion percentage. Consumers ultimately perceive capped promotions (with low trigger values) as relatively poor deals and less fair than the equivalent threshold promotions. However, with high and hard-to-reach trigger values (e.g., a threshold of $50 for food delivery), consumers\u2019 expectations and reactions reverse, and capped promotions are seen as fairer and more effective than comparable threshold promotions. This research makes four major contributions. First, this research is the first to examine an emerging and increasingly common type of price promotion (capped promotions) and detail the psychological process and behavioral effects associated with this type of promotion. Second, this research contributes to the broader literature on restricted promotions (e.g., Sokolova and Li 2021). Past research has examined the effectiveness of restricted versus unrestricted promotions and found that restrictions convey information about good deals (e.g., Inman, Peter, and Raghubir 1997). The current work classifies promotion restrictions into two categories (precondition-type 7  and limit-type, i.e., threshold and capped). It is the first to examine the conditional effectiveness of different types of restricted price promotions and the role of expectations in shaping perceived fairness. Third, this research contributes to the percentage-dollar promotion literature (e.g., Chen, Monroe, and Lou 1998) by providing a parsimonious explanation that accounts for findings that the standard economic model and previously identified pricing mechanisms (e.g., format neglect; Sevilla, Isaac, and Bagchi 2018) have been unable to explain. For instance, this research identifies conditions under which discounts framed in small-dollar terms can be more effective than those framed in large-percentage terms. Fourth, this research provides a novel methodological example utilizing A\/B testing tools on the short-video platform TikTok to conduct a field study. Ultimately, this research provides managers with a framework for assessing the effectiveness of price promotions, guiding them in designing optimal promotion architectures regarding perceived fairness, cost-effectiveness, and impact on sales. In the following sections, we review the literature on price-promotion architecture and draw from research on expectation disconfirmation and fairness perception to motivate our hypotheses. We then present the results of seven pre-registered studies testing our hypotheses and discuss our findings\u2019 theoretical and managerial implications.  2.2 Theoretical Development Two of the most well-examined price promotion types are offers in dollar terms and percentage terms. When discounts in dollar terms and percentage terms are equivalent (50% off a $10 purchase vs. $5 off a $10 purchase), research shows that price discounts framed in dollar terms 8  are perceived as more substantial than same-amount alternatives framed in percentage terms when product prices are high, but this pattern reverses with low prices (Chen, Monroe, and Lou 1998). A meta-analysis further found that although the deal percentage and amount positively influence the perceived savings, the deal percentage has more impact (Krishna et al. 2002). Several variants of percentage and dollar terms discounts have recently become pervasive in marketing practice, including capped and threshold promotions. The current research compares these two types of restricted promotion architectures. 2.2.1 Restricted Promotion Architectures: Capped versus Threshold Promotion restriction is a tactic that constrains consumers\u2019 actions in some way when they purchase a market offering (Inman, Peter, and Raghubir 1997). Commonly applied promotion restrictions tend to either set a precondition or add a limit to the offer. The preconditions can be the minimum number of products purchased, the minimum purchase amount, or the fulfillment of a non-monetary condition (Bertini and Aydinli 2020; Cheng and Ross 2023; Dallas and Morwitz 2018; Lee and Ariely 2006; Sokolova and Li 2021). Precondition-type promotions can take the form of \u201cbonus packs,\u201d \u201cbuy-one-get-one-free\u201d (BOGO) deals, \u201cfree gifts,\u201d or \u201cbundle offers\u201d (Raghubir 2004). For example, some promotions offer consumers an additional free product, a bonus quantity, or a discounted price conditional on purchasing a focal product or reaching a minimum quantity (Ding and Zhang 2020; Palmeira and Srivastava 2013; Teng 2009). Alternatively, limit-type promotions can involve a maximum quantity one can buy under a discount or when an offer is available (Chandon, Wansink, and Laurent 2000; Cheema and Patrick 9  2008; Inman, Peter, and Raghubir 1997). Studies show that preconditions and limited sales restrictions affect consumers\u2019 likelihood of purchase (Gneezy 2005; Inman, Peter, and Raghubir 1997; Yoon and Vargas 2011). However, whether and how each type of restriction differentially influences consumer decisions remains unclear. The current research compares two common choice architectures (Thaler and Sunstein 2008) used in promotion contexts, which we call \u201cpromotion architectures.\u201d The first is a dollar-term promotion with a spending amount precondition, which we call a threshold promotion (e.g., \u201c$5 off on orders of $10 or more\u201d). The second is a capped promotion (e.g., \u201c50% off with a maximum discount of $5\u201d), a percentage-term promotion with a maximum value limit. Although the nature of the restrictions for threshold and capped promotions is quite different, if the thresholds are low, the economic value of the two promotions can be equivalent for an extensive range of spending amounts, specifically when the spending amount is at least equal to the trigger value (as seen in Figure 2.1). In the above examples, taking up either offer leads to the same outcome if the amount spent meets or exceeds the trigger value of $10. Notably, however, capped promotions apply to a broader range of purchases than threshold promotions: only capped promotions are applicable when consumers\u2019 spending is below the trigger value. Thus, the economic benefit for consumers from capped promotions is always equal to or greater than that of equivalent threshold promotions. However, consumers\u2019 preferences are affected not only by the financial benefits obtained from the promotion but also by its psychological aspects, such as the transaction utility generated 10  by perceiving a \u201cfair\u201d deal (Campbell 1999; Thaler 1985). Consumers are particularly critical of firms presenting overly optimistic claims and later making corrections in advertising (Darke, Ashworth, and Ritchie 2008; Darke and Ritchie 2007). Thus, it is possible that when consumers fully understand the benefit they will receive from a given promotion, they will judge the capped promotion as relatively unfair because of the limit to the discount\u2014a possibility that we explore in the next section. 2.2.2 Fairness Perceptions and Expectation Discrepancies One of the critical determinants of promotion evaluation is perceived fairness (Darke and Dahl 2003). Past literature suggests that fairness perceptions arise from comparing a target ratio (e.g., of an offer\u2019s outcome and input) with a reference ratio, which may be internal (e.g., one\u2019s outcome\u2013input ratio in the past) or external (e.g., others\u2019 outcome\u2013input ratio; Adams 1965; Thaler 1985; Xia, Monroe, and Cox 2004). We hence posit that the perceived promotion ratio (\u201cpromotion perception\u201d) is compared to an expected reference promotion ratio (\u201cpromotion expectation\u201d) to determine the perceived fairness of a promotion. For promotion perception, the outcome is considered the attainable discount, and the input is the expected spending amount (\u201cspending expectation\u201d). Indeed, research suggests that consumers have two a priori internal expectations from previous experiences when they evaluate a price promotion when no external references are available: 1) an expectation about the average or moderate promotion percentage (e.g., Grewal, Marmorstein, and Sharma 1996; Kalwani and Yim 1992) and 2) an expectation of their spending amount or shopping goal (e.g., Cheng and Ross 2023; Lee and Ariely 2006). We surveyed 11  consumers\u2019 internal promotion and spending expectations based on past experiences with food ordering (See Appendix A.4). Most respondents (85.1%) mentioned typically receiving discounts of less than or equal to 20% (M = 12.53%, SD = 13.71%). Their mean spending expectation was $33.99 (SD = 17.15), almost identical to the average spend per order reported on Statista ($33.94; 2022a). Therefore, assuming one\u2019s promotion expectation is 20% for food delivery, an unrestricted \u201c$5 off\u201d promotion for an order will be deemed as a fair and good deal if a consumer usually spends $25\u2014since the offer ratio ($5\/$25 = 20%) meets their expectation (20%; see Table 2.1 below). However, the same \u201c$5 off\u201d promotion will be perceived as relatively less fair if they usually spend $50 since the offer ratio ($5\/$50 = 10%) is lower than their expectation (20%). Indeed, research shows that consumers are more willing to put in effort to get a $5 discount on a product priced at $15 versus one priced at $125 (Tversky and Kahneman 1981). This reasoning is also consistent with Thaler\u2019s (1980) view that the \u201csearch for any purchase will continue until the expected amount saved as a proportion of the total price equals some critical value.\u201d  As noted, a reference amount or expectation is labile and shaped by multiple factors. It can be derived not only from an internal reference arising from previous experience in memory (e.g., past price; Kalyanaram and Winer 1995) but also from an external reference arising from information in the environment, such as the promotions that others receive (Darke and Dahl 2003). Studies show that extreme or exaggerated reference prices (e.g., exaggerated manufacturer\u2019s suggested retail price) can raise price expectations and thus improve the perceived offer value in 12  comparison (Urbany, Bearden, and Weilbaker 1988). However, mentioning a large promotion may increase expectations for the actual offer value and thereby undermine deal evaluation because of the disconfirmation stemming from the discrepancy between expectation and the actual savings outcome (Oliver, Balakrishnan, and Barry 1994; Spreng, MacKenzie, and Olshavsky 1996). Thus, a \u201c50% off\u201d price claim may increase consumers\u2019 promotion expectations. However, the promotion may appear unfair when consumers realize that it does not apply to their purchase or the obtained discount is lower than 50% (Mobley, Bearden, and Teel 1988). For the two types of restricted promotions, we propose that the claimed promotion percentages (\u201c50%\u201d) in capped promotions (e.g., \u201c50% off, max $5\u201d) act as a high external reference for promotion expectations and that the equivalent threshold levels (\u201c$10\u201d) in threshold promotions (e.g., \u201c$5 off on orders of $10 or more\u201d) act as a lower external reference for spending expectations. Since external references are available in the immediate environment where evaluation happens and thus are more salient and diagnostic to consumers than internal references, consumers may use accessible external references instead of internal ones when evaluating a price promotion (see Hamilton 2024). Accordingly, when assessing a restricted promotion, consumers may compare the perceived offer ratio created by a salient internal (external) spending expectation to a salient promotion expectation created by an external (internal) reference value for capped (threshold) promotions. In the following sections, we first build out our theory for the case where trigger values are below typical spending amounts and then examine when they are above.  13  When trigger values are low. For capped promotions (e.g., \u201c50% off, max $5\u201d), the large percentage discount value (50%) is an external reference point for consumers\u2019 promotion expectations. Once consumers consider that their usual spending amount (e.g., $25 in food ordering) is higher than the trigger value ($10), they realize they can only receive the capped offer amount ($5) as their discount. Thus, consumers would find the actual outcome\u2013input ratio ($5\/$25 = 20%) they receive is lower than the promotion expectation (50%). This negative expectation disconfirmation, a discrepancy generated from a worse-than-expected outcome (Oliver, Balakrishnan, and Barry 1994), leads to lower perceived fairness and purchase intention. In contrast, for comparable threshold promotions (e.g., \u201c$5 off on orders of $10 or more\u201d), the spending expectation is anchored on the low trigger value ($10; Tversky and Kahneman 1974). The threshold restriction of $10 also works as an external reference of spending amount (Cheng and Ross 2023), which consumers use to evaluate the size of the discount amount in percentage (Heath, Chatterjee, and France 1995). Since there is no external reference for the promotion percentage, consumers may rely on an internal reference in memory, such as previously encountered offers (Hamilton 2024). In this example, the promotion perception is derived from the ratio of the stated promotion amount ($5) against the anchored spending expectation ($10), and thus ($5\/$10 = 50%), it tends to be larger than the internal promotion expectation (e.g., 20% previously encountered in food ordering; see Appendix A.4) when the threshold is lower than consumers\u2019 usual spending amount. This positive expectation disconfirmation, a discrepancy 14  generated from a better-than-expected outcome (Oliver, Balakrishnan, and Barry 1994), tends to lead to higher perceived fairness and purchase intention.  We use mathematical notation to illustrate the relationship between such promotion expectations and the corresponding promotion perceptions (see Table 2.1). However, consumers may not explicitly recall their previously encountered promotions and do the calculations (see Kyung and Thomas 2016). Instead, the calculations we provide in Table 2.1 are \u201cas-if\u201d models approximating consumers\u2019 heuristic process. For example, consumers could also think about their spending amount first (e.g., $25), expecting to have $12.5 off (50% \u00d7 $25) and realize the actual discount ($5) is below expectation, which is a different thought process on the surface but mathematically equivalent to the model we proposed.    15  Table 2.1 Relationship Between Promotion Expectations and Perceptions  Unrestricted Promotions (e.g., \u201c$5 off\u201d) Promotion Perception:  !\"#$%&'()*\"# =\t +,%&%#-%)\t\/&%\")#0*')1-)2\t34*'$#5#-%) \t= \t0#5#'1\t+,%&%#-%)\t\/&%\")#67\"58\t0*')1-)2\t\/&%\")#\t(()#',)58) = ;<; = 20% Promotion Expectation: Previously Encountered Percentage = 20% (Internal) Low trigger values ($10) \u2013 lower than typical spending Capped Promotions (e.g., \u201c50% off, max $5\u201d)  Promotion Perception:  !\"#$%&'()*\"# =\t +,%&%#-%)\t\/&%\")#0*')1-)2\t34*'$#5#-%) \t= =5**'1\t\/&%\")#67\"58\t0*')1-)2\t\/&%\")#\t(()#',)58) = ;<; = 20% Promotion Expectation: Stated Promotion Percentage = 50% (External) Perception < Expectation (Lower fairness) Threshold Promotions (e.g., \u201c$5 off on orders of $10 or more\u201d) Promotion Perception: !\"#$%&'()*\"# =\t +,%&%#-%)\t\/&%\")#0*')1-)2\t\t34*'$#5#-%) \t= \t 0#5#'1\t+,%&%#-%)\t\/&%\")#>?,'7?%81\t\/&%\")#\t(34#',)58) =\t ;@A = 50%  Promotion Expectation: Previously Encountered Percentage = 20% (Internal) Perception > Expectation (Higher fairness)  Fairness comparison: Capped Promotions < Threshold Promotions High trigger values ($50) \u2013 higher than typical spending Capped Promotions (e.g., \u201c10% off, max $5\u201d)                Promotion Perception: !\"#$%&'()*\"# =\t +,%&%#-%)\t\/&%\")#0*')1-)2\t34*'$#5#-%) \t= \ud835\udc46\ud835\udc61\ud835\udc4e\ud835\udc61\ud835\udc52\ud835\udc51\t\ud835\udc43\ud835\udc52\ud835\udc5f\ud835\udc50\ud835\udc52\ud835\udc5b\ud835\udc61\ud835\udc4e\ud835\udc54\ud835\udc52 = 10%               Promotion Expectation: Stated Promotion Percentage = 10% (External) Perception = Expectation (Higher fairness) Threshold Promotions (e.g., \u201c$5 off on orders of $50 or more\u201d)                Promotion Perception: !\"#$%&'()*\"# =\t +,%&%#-%)\t\/&%\")#0*')1-)2\t34*'$#5#-%) \t= B',%>?,'7?%81\t\/&%\")#\t(34#',)58) = 0% Promotion Expectation: Previously Encountered Percentage = 20% (Internal) Perception < Expectation (Lower fairness) Fairness comparison: Capped Promotions > Threshold Promotions 16  Thus, even though the applicability of capped promotions is broader than threshold promotions (as capped promotions can be used on any purchase, and threshold promotions only apply above the threshold), the perceived fairness of capped promotions may be lower than that of threshold promotions. Accordingly, we put forth the following hypotheses conditional on low trigger values (summarized in Figure 2.2): Hypothesis 1 (Conditional main effect): When the trigger value is low (i.e., lower than consumers\u2019 usual spending), capped promotions lead to lower purchase intentions than comparable threshold promotions. Hypothesis 2 (Mediation): When the trigger value is low (i.e., lower than consumers\u2019 usual spending), capped promotions versus comparable threshold promotions create a more negative expectation disconfirmation, lowering perceptions of fairness and, in turn, lowering purchase intentions.  Figure 2.2 Conceptual Framework   FairnessPromotionArchitecture(Capped vs Threshold)ExpectationDisconfirmation Purchase IntentionTrigger Value(Low vs. High)17  Thus, following H1 and H2, consumers may prefer a threshold promotion to a capped promotion when the trigger value is low.  When trigger values are high. Next consider the case where the trigger value exceeds consumers\u2019 usual spending amount. In this case, we predict that consumers generally perceive capped promotions as fairer than threshold promotions. For instance, for a capped promotion of \u201c10% off, max $5,\u201d the maximum trigger value ($50) is now higher than the typical spending amount ($25). Consumers spending an average amount can receive the full expected discount percentage (10%, leading to a rebate of $2.50) without triggering the cap. Accordingly, the promotion expectation ($2.50\/$25 = 10%) is equal to the indicated external reference for the promotion expectation of 10%. Thus, the capped promotion is perceived as fair.  When threshold promotions are set with a high trigger value (e.g., \u201c$5 off on orders of $50 or more\u201d), consumers\u2019 usual spending amount ($25) is now lower than the trigger value, thus the promotion amount is not applicable or relevant, and the actual outcome would be zero. Thus, when the trigger amount is higher than the consumers\u2019 usual spending amount, the outcome\u2013input ratio of the threshold promotion (0) is lower than the internal promotion expectation (20%). This negative expectation disconfirmation leads to low perceived fairness and purchase intention (see Table 2.1). Thus, (summarized in Figure 2.2) we hypothesize the following:  Hypothesis 3 (Conditional main effect): When the trigger value is high (e.g., higher than consumers\u2019 usual spending), capped promotions lead to higher purchase intention than comparable threshold promotions. 18  Hypothesis 4 (Mediation): When the trigger value is high, threshold promotions (versus capped promotions) create negative expectation disconfirmation, lowering perceptions of fairness and, in turn, purchase intentions. We examine the effect of promotion architectures when the trigger value is low and high (because both are prevalent in the marketplace; Appendix A.5). However, because threshold promotions apply to a narrower range of purchases than capped promotions with high trigger values, preferences between the promotion types can often be driven by differences in economic value. Thus, because it is theoretically more meaningful and interesting to examine the different psychological effects of two promotion architectures with similar economic values, our empirical studies mainly examine the former (H1 and H2; the promotions are closely comparable only when the trigger value is low; see Figure 2.1).  Upon consumers\u2019 intention to make purchases, managers endeavor to enhance their spending amount. Previous research has indicated that an attractive deal can increase purchase intention and spending (e.g., Hock, Bagchi, and Anderson 2020; Lam et al. 2001). Consequently, it might be assumed that increasing intentions associated with threshold promotions would also lead to higher spending amounts when the trigger value is low. However, consumers exposed to threshold promotions may intend to spend less because their spending amount is anchored to the low trigger value, which is the very mechanism that makes threshold promotions seem like better deals. Considering the possible opposing forces between the effect on purchase intention and on spending amount, we refrain from predicting the impact of promotion architecture on overall sales. 19  Therefore, in this paper, we focus on the effect of promotion architecture on purchase intention and explore its impact on purchase amount and overall sales to provide managerial guidance. We conducted seven pre-registered studies as well as several additional studies presented in appendices to test our hypotheses. All study materials, data, and pre-registrations can be found on OSF: https:\/\/osf.io\/tnhvb\/?view_only=0b5763ed25084a35a336ca91bd0218ae. Study 1 examines H1 in a real-world food-ordering setting using an online field study of digital discounts. Study 2 tests H1 and H3 with a price promotion in the ride-hailing context. Study 3 tests H1 and H2 using an unrestricted promotion to disentangle whether the observed difference between capped and threshold results from an aversion to capped promotions and a preference for threshold promotions. Study 4 tests H1 and H2 using a joint evaluation setting, which is also common in the real world. Studies 5 tests all hypotheses by examining the moderating effect of low versus high trigger value. Studies 6a and 6b test the robustness of H1 with lower promotion depths and further examine the impact of the two promotion types on overall sales. 2.3 Empirical Studies 2.3.1 Study 1: Food-Ordering Advertising Field Study This study examined the effects of promotion architecture on purchase interest when trigger values are low (H1). Specifically, we conducted a field study on TikTok, a popular short-form video platform, to compare responses to advertisements about equivalent threshold and capped types of promotions in a real-life food-ordering context. We examine whether a digital discount coupon for a capped or threshold promotion is more effective at increasing click-through 20  rate (CTR), a commonly used metric in digital advertising and marketing research (e.g., Hardisty and Weber 2020). We use this measure because clicks on a digital advertisement indicate consumers\u2019 interest in the advertised product, and we expect the CTR to be higher for threshold promotions than equivalent capped promotions.  2.3.1.1 Method We ran a coupon-based ad campaign for a food-delivery platform using the \u201csplit test\u201d tool in TikTok Ads Manager. This tool assigns the audience to one of two groups, each seeing only one advertisement, and measures the click-through rate. We used the \u201cSmart Video\u201d tool in Ads Manager to turn two images into two 30-second videos, one with a threshold-promotion message (\u201cEnjoy $3 off on an order of $5 or more\u201d) and the other with an equivalent capped-promotion message (\u201cEnjoy 60% off, $3 max discount per order\u201d) placed at the center of the screen throughout the video. Both ads presented the same background music recommended by TikTok and had the same call for action (\u201cOrder Now\u201d). We disallowed user comments to avoid any endogenous or unpredictable effects from the comments. We pre-registered to run the campaign among TikTok viewers above 18 years old in the United States for ten days, with an advertising cost of $400 for each ad. Finally, we directed the viewers who clicked on the advertisements to a website where we debriefed them and provided links to unrestricted \u201c$3 off\u201d Uber Eats coupon codes.  21  2.3.1.2 Results The threshold-promotion ad reached 61,895 viewers, and the capped-promotion ad reached 65,548 viewers. As pre-registered, we conducted a one-tailed two-proportion z-test on CTR. We found that viewers were more likely to click on the threshold-promotion advertisement (1.38%) than the capped-promotion one (1.25%), z = 2.04, p = .021. Exploratory analyses further revealed that viewers were more likely to play the threshold-promotion advertisement (13.59%) for at least two seconds compared to the capped-promotion one (11.77%), z = 9.38, p < .001.  2.3.1.3 Discussion Testing the effect of promotion architecture on click-through rates, a field study on TikTok examining real consumer choice supported H1, with threshold promotions being 10% more effective at generating clicks than capped promotions when trigger values were low. The finding that threshold promotions were 15% likelier to be kept on screen for two seconds or longer than capped promotions is also consistent with the notion that consumers are more interested in threshold promotions. Note that social media-based \u201cA\/B Tests\u201d are subject to \u201cdivergent delivery\u201d due to the optimization and targeting algorithms of the platform (Braun et al. 2024). Thus, although this study does not allow us to establish causality because of algorithmic confounds, it provides ecologically valid evidence supporting H1 (i.e., using A\/B tests is widespread in the advertising industry). Although the interference of the algorithm is unavoidable, we replicated our findings in Appendix A.6, suggesting a robust phenomenon in the real world 22  that can provide managerial guidance. Next, we follow up with fully randomized, lab-based experiments in the rest of the studies.  2.3.2 Study 2: Price Promotions in Ride-Hailing This study examined the effects of promotion architecture on coupon conversion and deal evaluation when the trigger value is lower than consumers\u2019 usual spending amount (H1) in a traditional experimental setting, this time using a ride-hailing context. We set the maximum discount amount for the two conditions at $5, with a trigger value of $10, a commonly used price incentive for ride-hailing service platforms such as Uber and Lyft. For exploratory purposes, we also examined 1) if the effect is moderated by whether participants\u2019 average spending amount is above or below the trigger value. As such, $10, a lower spending amount for most consumers, might still have been high for some participants, and thus, we expect that only participants whose average spending amount in real life was above the trigger value would likely prefer threshold promotions (H1). However, we expect those whose real-life spending amount was below the trigger value would be more likely to prefer capped promotions (H3); 2) whether the effect is robust to informing participants of the average spending on hailing taxis, which may shed light on whether consumers naturally consult their typical spending amount. 2.3.2.1 Method We pre-registered to recruit 400 Canadian residents from Prolific, and 403 participated in the study (Mage = 31.83, SDage = 20.22; 58.1% female). We used a 2 (promotion architecture: capped vs. threshold) x 2 (information about average spending: present vs. not) factorial between-23  subjects design. We asked participants to imagine going home from work and deciding whether to hail a ride or take the bus. Half of the participants were instructed about their typical purchase patterns (\u201cIf you hail a ride, you will spend $20 on average\u201d), and half were not. We randomly assigned participants to one of the two promotion-restriction conditions. Participants in the threshold-promotion condition read, \u201cEnjoy $5 off a ride, on a ride of $10 or more\u201d; those in the capped-promotion condition read, \u201cEnjoy 50% off a ride, $5 max discount per ride.\u201d  Next, we asked participants to indicate whether or not they would use the offer to hail a ride or take a bus (coupon conversion, binary), rate their strength of preference (1= Definitely hail a ride using Uber, 9 = Definitely take a bus; reverse coded in results) and the extent to which they perceived the promotions as bad or good deals (deal evaluation; 1 = A very bad deal, 9 = A very good deal). We also measured participants\u2019 typical spending on ride-hailing for a trip home, familiarity with Uber (1 = Not familiar at all, 9 = Very familiar), and whether they had an Uber account. 2.3.2.2 Results and Discussion  First, we confirmed that the trigger value we used ($10) was lower than the average spending amount (M = $21.15, SD = $16.94, range = [$1.00, $120.00]) on ride-hailing for a trip home, t(402) = 13.21, p < .001. Most participants (75%) typically spend over $10 on one ride. The majority of the participants (67%) had an Uber account and were familiar with Uber (M = 6.76, SD = 2.19), significantly higher than the scale midpoint (5), t(402) = 16.20, p < .001. 24  Conversion rate. Using logistic regressions, we examined the effect of promotion architecture and the provision of typical spending amounts on coupon conversion. Results revealed that respondents were significantly less likely to redeem the capped discount (56.22%) than the threshold discount (67.82%), Wald \u03c72(1) = 5.72, p = .017. There was neither a main effect of providing information on average spending, Wald \u03c72(1) = 0.23, p = .635, nor an interaction effect on coupon use, Wald \u03c72(1) = 0.37, p = .546.  Preference strength. An ANOVA with preference strength as the DV revealed that respondents had a stronger preference for hailing a ride (vs. taking a bus) when they saw a threshold discount (M = 5.63, SD = 2.55) than when they saw a capped discount (M = 5.12, SD = 2.45), F(1, 399) = 4.23, p = .040, \u03b7p2 = 0.01. There was, again, neither a main effect of mentioning the average spending, F(1, 399) = 0.16, p = .689, nor an interaction effect on preference strength, F(1, 399) = 0.47, p = .495. Deal evaluation. An ANOVA on deal evaluation revealed that respondents perceived the threshold discount (M = 6.72, SD = 1.66) as a significantly better deal than the capped discount (M = 5.46, SD = 1.93), F(1, 399) = 49.00, p < .001, \u03b7p2 = 0.11. Again, there was neither a main effect of providing the average spending, F(1, 399) = 0.15, p = .701, nor an interaction effect on deal evaluation, F(1, 399) = 0.43, p = .514. Consistent with H1, when the trigger value for a discount was low (i.e., lower than the typical purchase amount for most participants), consumers were more likely to choose to use the threshold discount coupon and have a higher deal evaluation than an equivalent capped discount. 25  This result is striking, given that the capped promotion is economically superior to the threshold promotion for spending amounts below the trigger value (about one-quarter of the sample). This effect also holds whether participants are assigned an average spending amount, suggesting that consumers in the control condition may have naturally thought of their typical spending amount (M = $21.15), which was very close to the $20 typical spending assigned in the other condition.  The role of trigger value. To provide an initial exploration of the role of the trigger value (H1 and H3), we separated participants into two groups: those whose self-reported spending amounts were higher than or equal to the trigger value (i.e., when the trigger value was low relative to a personal norm) and those whose self-reported spending amounts were lower than the trigger value (i.e., when the trigger value was high). An exploratory three-factor ANOVA showed a significant interaction effect on deal evaluation between spending information, trigger value, and promotion architecture, F(1, 395) = 7.07, p = .008, \u03b7p2 = 0.02 (see Figure 2.3). Consistent with H1 and H3, if the spending information was absent, there was a significant interaction effect between trigger value and promotion architecture, F(1, 197) = 9.91, p = .002, \u03b7p2 = 0.05. Consumers had a higher deal evaluation for the threshold promotion than the capped promotion when the trigger value was low, F(1, 197) = 26.76, p < .001, \u03b7p2 = 0.12. At the same time, the reverse was true when the trigger value was high, F(1, 197) = 3.30, p = .071, \u03b7p2 = 0.02. However, when spending information was given, there was no interaction effect between trigger value and promotion architecture, F(1, 198) = 0.32, p = .572. Instead\u2014and consistent with our instructions to assume a $20 typical spending\u2014consumers consistently had a higher deal evaluation for the threshold 26  promotion than the capped promotion regardless of their self-reported typical spending amount, F(1, 198) = 11.69, p < .001, \u03b7p2 = 0.06.    27  Figure 2.3 The Interaction Effect Among the Presence of Spending Information, Trigger Value, and Promotion Architecture in Study 2   Note: Error bars represent \u00b11 standard error. 28  In sum, the findings indicated that when the spending amounts were not provided, consumers naturally consult their real-life spending amounts and evaluate the deals based on whether they are above or below the trigger value. However, consumers no longer consulted their typical spending amount when spending amount instructions were provided. Since the given spending amount ($20) was above the trigger value ($10), which led to negative expectation disconfirmation for the capped promotion, consumers preferred the threshold promotion regardless of their actual spending amount. These findings were consistent with H1 and H3. They also supported our assumption that consumers consult their usual or typical spending amount to construct deal perceptions and spending expectations when evaluating deals. As this analysis was an exploratory test for H3, we formally examine it in study 5. 2.3.3 Study 3: Unrestricted Promotion   Thus far, we have compared the effect on purchase intention of two restricted price promotion types. In study 3, we include an unrestricted dollar-term promotion as a comparison condition to uncover whether the observed effect results from a negative expectation disconfirmation for capped promotions, a positive expectation disconfirmation for threshold promotions, or a combination when the trigger value is low. Based on our theorization (see Table 2.1), the spending expectation is anchored on a lower value for threshold promotions. In contrast, consumers have the same promotion expectations and perceived promotion amounts for unrestricted promotions and threshold promotions. Therefore, the perception-expectation ratio is higher for threshold promotions than unrestricted promotions, resulting in a positive expectation 29  disconfirmation for threshold promotions. In contrast, the promotion perceptions are the same for capped and unrestricted promotions, whereas the promotion expectations are higher for capped promotions. Therefore, the perception-expectation ratio is lower for capped promotions than unrestricted promotions, resulting in a negative expectation disconfirmation for capped promotions. Accordingly, we proposed that consumers would evaluate the threshold promotion (e.g., \u201c$3 off on an order of $5 or more\u201d) as better than the corresponding unrestricted dollar-term promotion (e.g., \u201c$3 off\u201d), which is evaluated as better than the capped promotion (e.g., \u201c60% off, $3 max on an order\u201d). 2.3.3.1 Method We pre-registered to recruit 600 participants residing in the US from Prolific, and 600 took part in the study (Mage = 35.13, SDage = 14.31; 69.7% female). Again, we asked participants to imagine themselves at home and deciding whether to order food from Uber Eats or cook dinner. Participants in the threshold-promotion condition saw a coupon: \u201cEnjoy $3 off on an order of $5 or more.\u201d Those in the capped-promotion condition read, \u201cEnjoy 60% off, $3 max discount on an order.\u201d Those in the unrestricted promotion condition read, \u201cEnjoy $3 off.\u201d Participants completed measures of their purchase intentions (\u201cAfter seeing this offer, how likely are you to order from Uber Eats?\u201d 1 = Very unlikely; 7 = Very likely) and promotion fairness perceptions (\u201cIs the offer shown above a fair deal?\u201d 1 = Not fair at all; 7 = Very fair; Campbell 1999) in a counterbalanced order. We also measured their expected spending amount per food order to confirm that our chosen trigger value ($5) was low for this category. 30  2.3.3.2 Results  The trigger value for the restricted discounts ($5) was lower than the average expected spending amount (M = $24.67, SD = $10.80, range = [$0.00, $80.00]) for one meal order, t(599) = 44.60, p < .001. Almost all participants (95%) expected to spend more than $5 on one meal order. Purchase intention. An ANOVA revealed a significant difference in purchase intentions among the three promotion-architecture conditions (see Figure 2.4), F(2, 597) = 21.90, p < .001, \u03b7p2 = 0.07. As expected, purchase intention was higher in the threshold promotion (M = 4.10, SD = 1.97) than in the capped promotion (M = 2.89, SD = 1.80), t(398) = 6.39, p < .001, Cohen\u2019s d = 0.64. Results also showed that purchase intentions with the unrestricted promotion (M = 3.61, SD = 1.72) were lower than with the threshold promotion, t(398) = 2.63, p = .009, Cohen\u2019s d = 0.26, but higher than with the capped promotion, t(398) = 4.09, p < .001, Cohen\u2019s d = 0.41.  Fairness perception. We also observed differences in fairness perceptions among the three promotion-architecture conditions, F(2, 597) = 54.86, p < .001, \u03b7p2 = 0.16. Results revealed that people perceived the threshold promotion (M = 5.02, SD = 1.67) as fairer than the capped promotion (M = 3.21, SD = 1.87), t(398) = 10.21, p < .001, Cohen\u2019s d = 1.02.  Indirect Effect. According to a standard mediation test, the indirect effect of threshold promotion (vs. capped promotion) on purchase intention through fairness perceptions is significant, b = 1.26, SE = 0.14, 95% CI = [1.00, 1.53]. Compared with the unrestricted promotion (M = 3.93, SD = 1.67), people perceived the threshold promotion as fairer, t(398) = 6.53, p < .001, 31  Cohen\u2019s d = 0.65, and the capped promotion as less fair, t(398) = 4.06, p < .001, Cohen\u2019s d = 0.41. Both the indirect effect of threshold promotion (vs. unrestricted promotion; b = 0.76, SE = 0.12, 95% CI = [0.53, 0.99]) and the indirect effect of capped promotion (vs. unrestricted promotion; b = -0.50, SE = 0.12, 95% CI = [-0.75, -0.26]) on purchase intention were mediated through fairness perceptions. Figure 2.4 Promotion Architecture and Purchase Intention in Study 3  Note: Error bars represent \u00b11 standard error. 2.3.3.3 Discussion This study reveals that consumers have the highest purchase intention for the threshold promotion and the lowest for the capped promotion (in support of H1), with the unrestricted promotion falling in the middle. Fairness perceptions mediate the differences in purchase intentions. These findings are also consistent with the theory that a negative expectation disconfirmation leads to lower purchase intentions for the capped promotion than the unrestricted promotion, and a positive expectation disconfirmation leads to higher purchase intentions for the 2.893.614.1012345Capped Unrestricted ThresholdPurchase Intention32  threshold promotion than the unrestricted promotion (supporting H2). See Appendix A.7 for a conceptual replication study examining both purchase intention and purchase amount of the three conditions with a lower promotion depth (20%): \u201cEnjoy 20% off ($3 max discount on an order)\u201d, \u201cEnjoy $3 off (On an order of $15 or more)\u201d, and \u201cEnjoy $3 off\u201d.  2.3.4 Study 4: Joint Evaluation The three studies presented so far show higher purchase intentions or interest for threshold promotions than capped promotions when participants evaluate only one of the two discounts (and the trigger value is low). However, these two types of restricted promotions can also be presented simultaneously in the real world (see Appendix A.2). We thus examine a phenomenon-driven moderator, whether the effect still holds when participants simultaneously evaluate both promotion types (i.e., a joint evaluation mode; Hsee 1996). In a joint evaluation mode, the promotional benefits can be directly compared, and consumers can quickly discover that the capped promotion is objectively better than (if they spend less than the threshold amount) or equivalent to (if they spend no less than the threshold amount) the threshold promotion, and thus the evaluability of economic values is higher (Hsee 1996). However, we hypothesize that consumers may still evaluate the two promotions against their respective promotion expectations, even in a more evaluable joint evaluation context. If so, fairness perceptions will be higher for threshold promotions, again driving higher purchase intentions for the threshold promotion over the capped promotion. Therefore, this study examines people\u2019s choice between two food-ordering apps using threshold versus capped promotion. 33  We also examine our proposed mechanism underlying preferences for threshold promotion by testing whether expectations and fairness perceptions serially mediate the effect (H2). Lastly, we assess whether consumers notice that capped promotions apply to a broader range of purchases and whether they consider this applicability when evaluating the promotions. 2.3.4.1 Method We pre-registered to recruit 200 US residents from Prolific, and 200 (Mage = 35.69, SDage = 13.71; 49.5% female) participated in the study. We asked participants to imagine they were considering two apps from food-delivering companies (KABU and APT). One company offered a threshold discount (\u201cEnjoy $5 off on an order of $10 or more\u201d). The other offered a capped discount (\u201cEnjoy 50% off, $5 max discount on an order\u201d). Whether the threshold discount or the capped discount was associated with a given company was counterbalanced.  Participants then indicated their choice (\u201cWhich app would you download and use the offer from?\u201d) and perceived fairness (\u201cThe offer from KABU\/APT is fair\u201d; Darke and Dahl 2003), negative expectation disconfirmation (\u201cThe offer from KABU\/APT is below my expectation\u201d), and applicability (\u201cThe offer from KABU\/APT applies to all orders\u201d) for each of the two companies using 5-point scales (1 = Strongly disagree; 5 = Strongly agree). 2.3.4.2 Results Logistic regression analysis showed that presentation position had no significant effect on the proportion of consumers choosing the app with the threshold promotion (capped promotion on the left: 58%; threshold promotion on the left: 65%), Wald \u03c72(1) = 1.03, p = .310. Similarly, 34  independent samples t-tests showed no significant position effect on any process measures, p > .10. Therefore, we collapsed the data across the two presentation-order conditions based on promotion architecture before conducting the main analyses. Choice. Consistent with H1, we find that with low promotional trigger values, consumers chose threshold promotions (61.5%) over capped promotions (38.5%), even when presented with both alternatives in a joint evaluation, \u03c72(1) = 10.58, p = .001.  Fairness and negative expectation disconfirmation. Paired-sample t-tests show that the perceived fairness for threshold promotion (M = 4.32, SD = 0.71) is significantly higher than that for capped promotion (M = 3.74, SD = 1.11), t(199) = 7.39, p < .001, Cohen\u2019s d = 0.52; and the negative expectation disconfirmation for threshold promotion (M = 2.35, SD = 1.08) is significantly lower than that for capped promotion (M = 2.84, SD = 1.30), t(199) = -5.12, p < .001, Cohen\u2019s d = -0.36. A within-participant serial mediation analysis using MEMORE macro (Model 1; Montoya and Hayes 2017) shows results consistent with a process model where negative expectation disconfirmation and perceived fairness serially mediated the effect of promotion architecture (capped promotion = 0; threshold promotion = 1) on choice, b = 0.03, SE = 0.02, 95% CI = [0.00, 0.07], supporting H2. Applicability. A paired-sample t-test shows that the perceived applicability for threshold promotion (M = 2.42, SD = 1.49) is significantly lower than that for capped promotion (M = 3.88, SD = 1.37), t(199) = 10.04, p < .001, Cohen\u2019s d = 0.71, indicating that consumers did correctly notice that the applicability of the capped promotion was broader than that of the threshold 35  promotion. The mediation analysis demonstrates that the applicability also mediates the effect of promotion architecture on choice (capped promotion = 0; threshold promotion = 1), but its direction is opposite to that of fairness, b = -0.15, SE = 0.05, 95% CI = [-0.27, -0.06].  2.3.4.3 Discussion We find that consumers choose the threshold promotion over the capped promotion when the two promotions are presented side-by-side, making the economic equivalence or dominance of the capped promotion more salient, supporting a strong form of H1. That is, the mechanism of expectation disconfirmation we proposed is salient or robust enough to create the effect even in a direct-choice setting. The higher choice rate for the threshold promotion also indicates that our findings in previous studies are not artifacts of scale interpretation resulting from the between-subjects design but accurate representations of experienced value (McKenzie and Sher 2020; Sher and McKenzie 2014).  Consistent with H2, we find that the capped promotion leads to higher negative expectancy disconfirmation and is perceived as less fair than the threshold promotion. Notably, consumers preferred threshold promotions under low trigger values even though they observed that the capped promotion applies to a broader range of transactions. Moreover, we tested the robustness of the mechanism with a separate evaluation using a different promotion context: charitable appeals in grocery shopping (see Appendix A.8). We found that people were more likely to purchase with a threshold promotion (\u201cWe donate $5 per purchase, if you purchase $10 or more\u201d) than with an equivalent capped promotion (\u201cWe donate 50% of your purchase price, up to $5 per purchase\u201d) 36  and replicated the serial mediation model through negative expectation disconfirmation and fairness as well as the opposite effect of applicability. These findings support our theory that consumers consider both economic and psychological dimensions of promotions and that psychological fairness perception may prevail over the economic value. 2.3.5 Study 5: Trigger Value as a Moderator The previous studies determined that consumers prefer the objectively equivalent (and sometimes less applicable) threshold promotion to the comparable capped promotion when the trigger value is lower than consumers\u2019 usual spending amount. However, our proposed process implies that the opposite may be true when the trigger value is higher than consumers\u2019 usual spending amount because of an ineligibility for the threshold promotion. Also, when the trigger value is high, the consumer is less likely to exceed the cap for the capped promotions. Therefore, the perceived discounts are equal to the promotion expectations for most consumers. Thus, we expect a reversal for high (vs low) trigger values (H3 and H4), which we examine in Study 5. 2.3.5.1 Method  We pre-registered to recruit 400 US residents from Prolific, and a total of 398 (Mage = 33.58, SDage = 12.72; 57.5% female) participated in the study following a 2 (promotion architecture: threshold vs. capped) x 2 (trigger value: high vs. low) between-subjects design. We asked participants to imagine that they were deciding whether to order out or cook dinner at home. Participants in the low-trigger threshold-promotion condition read, \u201cEnjoy $5 off (On an order of $10 or more),\u201d while those in the low-trigger capped-promotion condition read, \u201cEnjoy 50% off 37  ($5 max discount on an order).\u201d In the high-trigger threshold-promotion condition, participants read, \u201cEnjoy $5 off (On an order of $50 or more).\u201d Those in the high-trigger capped-promotion condition read, \u201cEnjoy 10% off ($5 max discount on an order).\u201d We measured purchase intentions (\u201cAfter seeing this message, how likely are you to use this offer and order from UberEats?\u201d 1 = Extremely unlikely, 5 = Extremely likely), fairness perception (\u201cThe offer is fair.\u201d 1 = Strongly disagree, 5 = Strongly agree; same below), negative expectation disconfirmation (\u201cThe offer is below my expectation\u201d), and applicability (\u201cThe offer applies to all orders\u201d). We also measured participants\u2019 typical spending amount for one food-ordering delivery to confirm that the threshold settings used are comparatively high ($50) or low ($10). 2.3.5.2 Results  First, we confirmed that the typical spending in one meal order (M = $27.65, SD = $17.57) is higher than the low-level trigger value setting ($10), t(397) = 20.05, p < .001, and lower than the high-level trigger value setting ($50), t(397) = 25.38, p < .001. Results of an ANOVA on purchase intention revealed that consumers had higher purchase intentions for low-trigger promotions than for high-trigger promotions (see Figure 2.5), F(1, 394) = 43.95, p < .001, \u03b7p2 = 0.10. Notably, there was a significant interaction effect between promotion architecture and trigger values, F(1, 394) = 75.38, p < .001, \u03b7p2 = 0.16. Pairwise comparisons showed that consumers in the low-trigger condition expressed higher purchase intentions with threshold promotions (M = 3.73, SD = 1.11) than they did with capped promotions (M = 2.87, SD = 1.30), t(199) = 5.03, p < .001, Cohen\u2019s d = 0.71, supporting H1. However, consumers in the high-trigger condition 38  expressed lower purchase intentions with threshold promotions (M = 1.94, SD = 0.99) than they did with capped promotions (M = 3.11, SD = 1.24), t(195) = 7.34, p < .001, Cohen\u2019s d = 1.05, supporting H3. Figure 2.5 Moderation by Trigger Value in Study 5    Note: Error bars represent \u00b11 standard error. Results from an ANOVA revealed a significant interaction effect on negative expectation disconfirmation between promotion architecture and trigger values, F(1, 394) = 38.04, p < .001, \u03b7p2 = 0.09. Pairwise comparisons found that consumers in the low-trigger conditions had more negative expectation disconfirmation with capped promotions (M = 3.59, SD = 1.23) than with threshold promotions (M = 2.55, SD = 1.18), t(199) = 6.13, p < .001, Cohen\u2019s d = 0.87. However, consumers in high-trigger conditions had less negative expectation disconfirmation with capped promotions (M = 3.50, SD = 1.22) than with threshold promotions (M = 3.93, SD = 1.13), t(195) = 2.57, p = .011, Cohen\u2019s d = 0.37.  2.873.113.731.9412345Low Trigger Value High Trigger ValuePurchaseIntentionsCappedPromotionThresholdPromotion39  Results of an ANOVA on fairness perception showed a significant interaction effect between promotion architecture and trigger values, F(1, 394) = 78.00, p < .001, \u03b7p2 = 0.17. Pairwise comparisons revealed that consumers in low-trigger conditions had higher fairness perceptions for threshold promotions (M = 4.05, SD = 0.95) than for capped promotions (M = 3.12, SD = 1.15), t(199) = 6.27, p < .001, Cohen\u2019s d = 0.89. However, consumers in high-trigger conditions had lower fairness perceptions for threshold promotions (M = 2.55, SD = 1.18) than they did for capped promotions (M = 3.51, SD = 0.99), t(195) = 6.22, p < .001, Cohen\u2019s d = 0.89. Interestingly, the high-trigger capped promotion was perceived as fairer than the low-trigger capped promotion, t(198) = 2.59, p = .010, Cohen\u2019s d = 0.37. Results of an ANOVA on applicability revealed that participants recognized the promotion applicability as higher for capped promotions (M = 3.51, SD = 1.37) than threshold promotions (M = 2.19, SD = 1.44), F(1, 394) = 93.35, p < .001, \u03b7p2 = 0.19. We also found a significant interaction effect between promotion architecture and trigger values, F(1, 394) = 16.75, p < .001, \u03b7p2 = 0.04. Pairwise comparisons showed that consumers in both low-trigger and high-trigger conditions correctly noticed that capped promotions offered a broader range of applicability than threshold promotions, with the difference in perceived applicability larger when the threshold was high (vs. low), ps < .001. As pre-registered, we applied the Process Macro (Hayes 2017) to examine the whole theoretical framework (Figure 2.2) by conducting a moderated serial mediation analysis (Model 83) of the effect of promotion architecture on purchase intention with negative expectation 40  disconfirmation and fairness perception as serial mediators and the trigger value as the moderator. Consistent with H2, results showed that when the trigger value is low, the indirect effect of promotion architecture (coded as threshold = 0, capped = 1) through negative expectation disconfirmation and fairness perception is significant and negative, b = -0.23, SE = 0.05, 95% CI = [-0.35, -0.13]. Consistent with H4, results showed that when the trigger value is high, the indirect effect of promotion architecture (coded as threshold = 0, capped = 1) through negative expectation disconfirmation and fairness perception is significant and positive, b = 0.09, SE = 0.04, 95% CI = [0.02, 0.18]. Supporting the trigger value\u2019s moderating role, the moderated mediation index (difference between conditional indirect effects) is also significant, b = 0.32, SE = 0.08, 95% CI = [0.19, 0.49]. After applicability is controlled for, the results hold, and the moderated mediation index is still significant, b = 0.29, SE = 0.07, 95% CI = [0.16, 0.45]. 2.3.5.3 Discussion  Consistent with H1 and H3, results from Study 5 support our predictions that consumers\u2019 preferences for threshold promotions versus capped promotions depend on the trigger value. Consumers prefer threshold promotions to capped promotions when the trigger value is low (H1), and this preference reverses with high trigger values (H3). Consistent with our hypotheses H2 and H4, we find that negative expectation disconfirmation and fairness perceptions drive the effect of promotion architecture on purchase intentions.  Interestingly, the high-trigger capped promotion (\u201c10% off, max $5 discount\u201d) was also perceived as fairer than the low-trigger capped promotion (\u201c50% off, max $5 discount\u201d), although 41  its promotion depth was dominated by its low-trigger counterpart, and the capped values were the same ($5). This was also consistent with our proposed mechanism of expectation disconfirmation (H2). To further examine the mechanism and provide managerial insights, we explored how various levels of promotion depth affect evaluations of capped promotions in Appendix A.9. When the capped values were kept constant ($3), the promotion depth (10% to 60%) usually negatively affected fairness perception and deal evaluation. We also replicated that a high-trigger capped promotion (\u201cEnjoy 20% off, $3 max discount on an order\u201d) could be perceived as a fairer and better deal than a low-trigger one (\u201cEnjoy 60% off, $3 max discount on an order\u201d). All these findings support our proposed mechanism that an attractive offer could set a higher promotion expectation, leading to expectation disconfirmation and perceived unfairness. In study 5, we kept the maximum discount offered constant (i.e., $5). We varied the percentage level between high-trigger conditions (i.e., 10% off up to $5 max) and low-trigger conditions (i.e., 50% off up to $5 max). As a robustness check, we ran a study that kept the promotion depth or percentage level constant at 50% (see Appendix A.10) across high- (i.e., $25) and low-trigger conditions (i.e., $5). The results were again consistent with H1 and H3. 2.3.6 Study 6A & 6B: Overall Sales Studies 1-5 have supported the proposal that threshold promotions lead to higher interest, purchase intention, and conversion than capped promotions when the trigger value is low. Although deal evaluation and conversion effects are the focus of this research, it is also essential to examine total spending when evaluating the effectiveness of promotions (Hock, Bagchi, and 42  Anderson 2020; Lam et al. 2001). Considering that a low trigger value may evoke a goal of a smaller purchase amount for threshold promotions (Cheng and Ross 2023; Lee and Ariely 2006) and that it is unknown whether a low trigger value similarly affects purchase amounts for capped promotions, we are uncertain about which promotion architecture will maximize overall sales, a critical managerial outcome. Therefore, the first goal of the following two studies is to examine the effect of promotion architecture on overall sales when the trigger value is low. Secondly, we adopted a smaller percentage for capped promotions (i.e., 20% and 40%) to test the robustness of the findings across promotion depth levels. Third, we elicited consumers\u2019 spontaneous thoughts after seeing the promotional messages to explore potential mechanisms. Lastly, we examined the generalizability of the findings among a relevant and younger population, college students, considering that 43% of Generation Z use meal delivery services (Statista 2022b). 2.3.6.1 Study 6a: 40% Promotion Depth 2.3.6.1.1 Method A total of 278 students at a North American university participated in the study (Mage = 19.80, SDage = 1.28; 59.7% female). We asked participants to imagine that they were deciding whether to order out or cook dinner at home. Participants in the threshold-promotion condition saw a coupon: \u201cEnjoy $6 off (On an order of $15 or more).\u201d Those in the capped-promotion condition read, \u201cEnjoy 40% off ($6 max discount on an order).\u201d Participants were first asked to write down anything crossing their mind as they considered the offer. Then, they completed the same measure of deal evaluation used in Study 2 and indicated whether they would use the offer 43  and make an order (coupon conversion, binary). Participants who answered \u201cyes\u201d to the conversion question were further asked to indicate how much money they were likely to spend before the discount was subtracted.  2.3.6.1.2 Results An independent-sample t-test on deal evaluation revealed that respondents perceived the threshold discount (M = 6.94, SD = 1.35) as a significantly better deal than the capped discount (M = 5.03, SD = 1.93), t(276) = 9.66, p < .001, Cohen\u2019s d = 1.15. In addition, logistic regression analysis revealed that the conversion rate was significantly higher for the threshold promotion (85.61%) than for the capped promotion (50.36%), Wald \u03c72(1) = 35.90, p < .001.  Among those who would make an order, the purchase amount was smaller when the coupon was a threshold promotion (M = 18.59, SD = 3.87) than when the coupon was a capped promotion (M = 23.06, SD = 10.78), t(187) = 4.10, p < .001, Cohen\u2019s d = 0.62. However, after taking the proportion of those who would make an order into consideration (non-choosers coded as $0), a Mann-Whitney U test suggested that participants overall tended to spend more when it was the threshold promotion (M = 15.91, SD = 7.46) than when it was the capped promotion (M = 11.61, SD = 13.86), U = 7432.50, z = 3.42, p < .001. Consistent with this, the overall sales for the threshold promotion ($2,212.21) are larger than that for the capped promotion ($1,614.20). As a robustness check, using a bootstrapped method, we found that the 95% CI for the mean difference in individual spending (threshold - cap) is [$1.71, $6.80].   44  To give insight into the associated revenue analysis for managers, we also calculated the payment amount (purchase amount \u2013 discount amount) contingent on their purchase amount and the promotion type for each participant. After the appropriate discount was deducted, the payment amount was still larger for threshold promotions (M = 10.78, SD = 5.70) than that for capped promotions (M = 8.61, SD = 11.46), U = 7432.5, z = 3.42, p < .001. The bootstrapped 95% CI for the mean difference (threshold - cap) is [$0.03, $4.21]. We conducted an automated text analysis using the extended Moral Foundations Dictionary (Hopp et al. 2021) to determine the fairness perceptions implied by participants\u2019 thoughts. An independent t-test showed that the fairness sentiment conveyed in the text was significantly higher for threshold promotions (M = .02, SD = .04) than that for capped promotions (M = .01, SD = .02), t(276) = 2.53, p = .012, Cohen\u2019s d = 0.30. Mediation analysis showed that fairness sentiment mediated the effect of the promotion architecture on deal evaluation, b = 0.06, SE = 0.03, 95% CI = [0.02, 0.14].  2.3.6.2 Study 6b: 20% Promotion Depth 2.3.6.2.1 Method We recruited 139 university students to participate in the study (Mage = 19.89, SDage = 1.10; 56.0% female). The scenario and measures were the same as those used in study 6a except for the promotion messages. Participants in the threshold-promotion condition saw \u201cEnjoy $3 off (On an order of $15 or more),\u201d while those in the capped-promotion condition read, \u201cEnjoy 20% off ($3 max discount on an order).\u201d  45  2.3.6.2.2 Results Respondents perceived the threshold discount (M = 5.28, SD = 1.80) as a significantly better deal than the capped discount (M = 3.51, SD = 1.59), t(137) = 6.15, p < .001, Cohen\u2019s d = 1.04. Consistent with this, the conversion rate was significantly higher for the threshold promotion (61.76%) than for the capped promotion (21.13%), Wald \u03c72(1) = 22.00, p < .001. Interestingly, among those who chose to make an order, the average purchase amount in the threshold-promotion condition (M = 18.86, SD = 5.40) was found to be similar to that in the capped promotion (M = 18.57, SD = 4.30) in this study, t(55) = 0.19, p = .852. After taking those who would not make an order into consideration (coded as $0), a Mann-Whitney U test suggested that participants, on average, tended to spend more when it was the threshold promotion (M = 11.65, SD = 10.15) than when it was the capped promotion (M = 3.92, SD = 7.87), U = 1426.00, z = 4.69, p < .001. As a robustness check, using a bootstrapped method, we found that the 95% CI for the mean difference (threshold - cap) is [$4.75, $10.70].  After their discount amount was deducted, the payment amount for threshold promotions (M = 9.79, SD = 8.84) was nearly three times as large as that for capped promotions (M = 3.29, SD = 6.68), U = 1426.00, z = 4.69, p < .001. The bootstrapped 95% CI for the mean difference (threshold - cap) is [$3.93, $9.08]. 2.3.6.3 Discussion  Studies 6a and 6b show that when the trigger value is low, consumers have higher purchase intentions with threshold promotions than capped promotions even when the promotion depth level 46  is low (i.e., 20% or 40%), supporting the robustness of H1. The text analysis of participants\u2019 spontaneous thoughts suggests that fairness perception may drive the effect. Intriguingly, there is a possible trade-off between the effect of promotion architecture on purchase intention (or conversion rate) and its impact on purchase amount. Although consumers overall perceive the threshold promotion (versus capped promotion) as a better and fairer deal, and accordingly, the conversion rate is higher, buyers may spend less with a threshold promotion when the trigger value is low (Study 6a but not 6b). Descriptive statistics reveal that the inconsistency of the spending effect between Studies 6a and 6b is mainly driven by a major drop in the purchase amount for the capped promotions ($23.06 vs. $18.57) but not for the threshold promotions ($18.59 vs. $18.86) when the promotion depth changes from 40% to 20% and the maximum discount changes from $6 to $3. The results suggest that consumers\u2019 responses to purchase amounts with threshold promotions may have been similarly anchored on the threshold amount ($15) in both studies (Tversky and Kahneman 1974).  Overall, when both the conversion rate and the spending amount per consumer are considered, the sales are consistently higher for the threshold promotion than the capped promotion (when the trigger value is low). To further explore the practical implications of our research, in another study, we found that capped promotions backfired compared to having no promotion at all (See Appendix A.11). A capped promotion with a low trigger value led to both lower purchase intention and smaller purchase amount than a no-promotion message. After considering the 47  conversion rate, purchase amount, and discount, the no-promotion message boosted the final payment by 54.51% more than the capped promotion per capita.  2.4 General Discussion Across six controlled experiments, one field study as well as six studies in appendices, we examined consumers\u2019 attitudes toward two common restricted promotion types: capped and threshold. We find converging evidence that consumers have higher purchase intentions (and larger $ sales) when offered threshold promotions than equivalent capped promotions, as long as the trigger value is lower than their usual spending amount. We observe this effect even when the capped promotion applies to a broader range of purchases and is economically equivalent to or even better than the threshold promotion\u2014inconsistent with standard economic models. This phenomenon occurs because consumers heighten their promotion expectations for capped promotions and lower their spending expectations for threshold promotions using external references, which leads to overall negative expectation disconfirmation and lower perceived fairness for capped promotions. Consistently, we discover that the preference for threshold promotions over capped promotions can be decomposed into two sub-effects: we not only find an aversion to capped promotions but a preference in favor of threshold promotions over unrestricted promotions. Consistent with our framework, the preference between capped and threshold promotions reverses when the trigger value exceeds consumers\u2019 usual spending. Further, we present evidence supporting the generalizability of these findings. The capped versus threshold promotion architecture affects consumers\u2019 purchase intentions in everyday price-48  promotion contexts such as food ordering, ride-hailing, and grocery shopping. These findings are robust to different manipulations and measures: Across studies, these findings are robust to various percentage settings and promotion amounts and include continuous measures, such as purchase intentions and binary choices. The effect is also robust to providing an expected spending amount and using a joint evaluation mode that makes comparing the economic values of each promotion type more straightforward\u2014which should favor capped promotions. We show the effects in both lab studies and real-world advertising, further presenting evidence for the ecological validity of the findings. 2.4.1 Practical Implications Capped and threshold promotions are now as common in the marketplace as unrestricted promotions. Managers on many digital platforms must decide which type of restricted promotions to choose (e.g., DoorDash; see Appendices A.3 and A.4).  This research can help managers design cost-effective and successful price promotions and inform consumers about how they may be affected. According to Lam et al. (2001), price promotions attract more consumers to visit the store and purchase. On the other hand, managers also intend to increase consumers\u2019 spending amount. Our findings reveal that a trade-off between the two goals may exist when using threshold promotions (versus capped promotions).  Firstly, we suggest that marketers aiming to maximize purchase intention should adopt a threshold-promotion architecture with low or capped-promotion architecture with high trigger values. When marketing a new product, capturing early market share at the expense of profits may 49  be more critical. These strategies may also be beneficial for first-time users to develop product use habits or to attract dormant users to reorder. Additionally, increasing purchase intention is more relevant when consumers have a concrete shopping goal and a predetermined spending amount. For instance, when consumers decide whether to hail a ride with an offer or take a bus, their destination is fixed, and spending is inelastic. Research indicates that the promotion threshold does not affect consumer spending in a store (Lee and Ariely 2006). However, when the goal is to maximize the purchase amount, managers should be aware of the potential anchoring effect of threshold promotions, and they may prefer to adopt capped promotions with low trigger values and threshold promotions with high trigger values. Furthermore, when the goal is to maximize sales, both purchase intention and amount should be considered. Studies 6a and 6b, as well as Appendix A.7 and A.11, provide several cases where overall sales are examined, and we find that threshold promotions still lead to larger sales than capped promotions in these studies. However, we speculate that higher trigger values might sometimes result in more sales for capped promotions. Finally, we have provided managerial guidance on the more effective use of capped promotions and suggested less costly alternatives to this promotional strategy. When implementing a capped promotion, it is advisable to state a smaller percentage (e.g., 20%), which incurs lower costs, rather than a larger one to mitigate negative expectation disconfirmation and enhance purchase intention (Appendix A.9). Additionally, managers might consider using no-promotion 50  messages, which can result in both higher purchase intention and increased sales compared to capped promotions (Appendix A.11). 2.4.2 Theoretical Contributions This research makes several theoretical contributions. First, by drawing on expectation discrepancy theory (Oliver and Swan 1989), we propose a framework of promotion architecture to explain how two types of restrictions (a threshold precondition vs. a cap limit) affect consumers\u2019 purchase intentions. Specifically, we highlight the roles of two types of expectations, spending expectations and promotion expectations, both of which are affected by external references in the offer messages. We propose that the relationship between promotion perceptions (which are affected by spending expectations) and promotion expectations determines a promotion\u2019s perceived fairness. When the trigger value is low, threshold promotions decrease consumers\u2019 spending expectations, which increases the perception of promotion fairness. In contrast, capped promotions increase promotion expectations, which decreases fairness perceptions because of the high external references provided.  Consistent with the dual-sided explanation, our findings demonstrate that a threshold promotion (e.g., \u201cEnjoy $3 off when spending $5 or more\u201d) may result in a higher purchase intention compared to an unrestricted promotion (e.g., \u201cEnjoy $3 off\u201d; see Study 3 and Appendix A.7). Conversely, a large-percentage capped promotion (e.g., \u201cEnjoy 60% off, up to $3\u201d) not only leads to a lower purchase intention than the comparable threshold promotion, but also performs worse than a small-percentage capped promotion (e.g., \u201cEnjoy 20% off, up to $3\u201d; see Appendix 51  A.9) or even a no-promotion message (e.g., \u201cEnjoy your meal with us, order now\u201d; see Appendix A.11). Notably, in all these scenarios, the promotions that result in lower purchase intentions are economically equivalent or superior to their counterparts.  Our findings also contribute to the literature on restricted promotions, answering a call to examine the effect of framing on promotion restrictions (Inman, Peter, and Raghubir 1997). We classify promotion restrictions into two categories: preconditions and limits. We build on past restricted promotions research that examined the effects of restrictions (e.g., threshold precondition; time limit) versus non-restrictions (Gneezy 2005; Inman, Peter, and Raghubir 1997; Yoon and Vargas 2010) to conduct\u2014to our knowledge\u2014the first comparison between two types of promotion restrictions. As such, a general preference or aversion to restrictions in the previous literature cannot explain our effects parsimoniously.  Furthermore, our findings show when restricted promotions will be attractive (or not) to consumers, which helps to reconcile inconsistent guidance from previous literature regarding threshold promotions. For instance, studies find that adding a restriction (including a trigger) typically positively affects deal evaluation because consumers infer good deals from restrictions (Inman, Peter, and Raghubir 1997). Meanwhile, Gneezy (2005) reports that participants had a more negative attitude toward restricted coupons (minimum $20 purchase) than unrestricted ones ($10 coupon from the university\u2019s bookstore) because consumers infer that the store wants them to spend on unnecessary items. The current research finds that restriction architecture matters, and we propose a theoretical framework to predict when different price promotions will be most 52  effective. For instance, the trigger values of $20 used in Gneezy (2005) may have been higher than participants\u2019 usual spending amounts in bookstores at that time, leading them to perceive the threshold promotion as unfair. Therefore, our framework enables us to reconcile seemingly disparate findings in the literature.  The present research also broadly builds on the literature on price-promotion framings. A long-lasting debate in price promotion research is whether to frame discounts in percentage or dollar terms (Chen, Monroe, and Lou 1998; DelVecchio, Krishnan, and Smith 2007; Hardesty and Bearden 2003). Previous work indicates that people may rely predominantly on the nominal value and neglect other information when making evaluations, which is referred to as \u201cbase value neglect\u201d (Chen et al. 2012), \u201cformat neglect\u201d (Sevilla, Isaac, and Bagchi 2018), or \u201cpricing focalism\u201d (Allard, Hardisty, and Griffin 2019). Thus, a large percentage number (e.g., 50% off) may make capped promotions more attractive to consumers than a small absolute dollar number ($5 off) in threshold promotions. However, our findings support the opposite in restricted-promotions architecture, revealing a solid aversion to capped promotions in dollar terms (vs. threshold percentage promotions) because of negative expectation disconfirmation and lower fairness perception, which seems to dominate the previously discovered heuristic preference for large nominal values.   2.4.3 Directions for Future Research Although our research has examined the effect of promotion architecture on purchase intention and amount when consumers see the offers at the moment, the negative expectation 53  disconfirmation and lower fairness perception of capped promotions (vs. threshold promotions) may last longer and create a chain of influence over time, as is implied in a study where capped promotions may create more negative inferences about the brand (Appendix A.8). A better-perceived deal may form user habits and lead to more purchases in the long run. In contrast, a worse deal may lead to broader side effects on the effectiveness of later advertising (Darke, Ashworth, and Main 2010). We leave this question open and encourage future research to examine the long-term and carryover effects of promotion architecture. Our framework suggests that consumers often consult their usual spending amount when evaluating promotions with no external reference point for spending amount. It would be interesting to examine how consumers would evaluate threshold promotions versus capped promotions for a new product category where the usual spending amount is uncertain. Moreover, this research has examined threshold promotions in explicit nominal values such as \u201c$5 off if you spend more than $10.\u201d However, broad threshold promotions also include free delivery or a free item whose monetary value is not explicit. Similarly, the threshold can also be non-nominal, such as \u201c$5 off if you buy two items\u201d. Future research can test the effectiveness of non-nominal threshold promotions to extend the scope of the threshold promotions.  Methodologically, our study introduces the possibility of running field experiments on a new and popular digital marketing platform, TikTok. Many previous online field experiments were conducted on Facebook (Meta; e.g., Hardisty and Weber 2020), yet, to our knowledge, this is the first field experiment in marketing research conducted on TikTok. Compared to Facebook 54  experiments, where many participants may not notice the ads on the screen, TikTok ads occupy the entire screen of the devices, and users must either express their interest in the ads by clicking the link or express their lack of interest by scrolling down explicitly. Thus, the click-through rate measures on TikTok are more accurate indicators of people\u2019s interest. While managers widely utilize the newly developed restricted promotions, this research highlights the need for careful attention to promotion architecture design. Specifically, capped percentage promotions are not as well-received as intended when a significant discrepancy exists between the actual discount percentage and the stated percentage. Instead, a comparable threshold promotion can be more cost-effective for managers and more appealing to consumers, mainly when the goal is to boost purchase intentions.   55  Chapter 3: \u201c10% off Each\u201d: How Implicitly Partitioned Percentage Framing Affects Purchase Intentions 3.1 Introduction Imagine seeing a price promotion message outside a liquor store that states, \u201cReceive 5% off.\u201d Would you be motivated to visit the store? Now consider a slightly different promotion: \u201cReceive 5% off each bottle.\u201d Would your decision change? If so, would you be more or less likely to visit the store? Percentage-off discounts are among the most common forms of price promotions. Researchers have extensively examined the effectiveness of percentage-off framing (e.g., Chen, Monroe, and Lou 1998; DelVecchio, Krishnan, and Smith 2007). Promotions like \u201cReceive 5% off\u201d or \u201cReceive 5% off in total\u201d utilize all-inclusive percentage framing (AIPF), as these discounts are applied proportionally to the total order value. In contrast, we refer to promotions such as \u201cReceive 5% off each bottle\u201d as implicitly partitioned percentage framing (IPPF) (a real example can be found in Appendix B.1). IPPF suggests that the proportional discount applies to each unit of the product individually rather than to the total amount. This framing is distinct from explicitly partitioned percentage framing (EPPF), which emphasizes the discount on each product explicitly (e.g., Greenleaf et al. 2016; Morwitz, Greenleaf, and Johnson 1998). For instance, a comparable EPPF might state: \u201cReceive 5% off on the first bottle, 5% off on the second bottle, 5% off on the third bottle.\u201d While AIPF, IPPF, and 56  EPPF provide identical discounts to consumers, IPPF is perceptually closer to AIPF than to EPPF due to its subtle framing. In this research, we investigate how the framing of percentage-off promotions influences consumers\u2019 purchase intentions and deal evaluations. Our findings reveal that consumers exhibit higher purchase intentions and more favorable deal evaluations for promotions framed as IPPF compared to AIPF, despite both offering the same discount and having minimal perceptual differences. This study extends the research on price partitioning by exploring the effects of discount partitioning, introducing a novel and implicit partitioning practice for managerial applications (e.g., Greenleaf et al. 2016; Morwitz, Greenleaf, and Johnson 1998). Furthermore, it contributes to the theoretical literature on the comparison of segregated versus integrated gains by examining the efficacy of a subtler form of segregation framing (e.g., Gourville 1998; Kahneman and Tversky 1979; Thaler 1985). We present evidence supporting process-oriented mental simulation as the underlying mechanism for the observed effects, which advances the literature on mental simulation by demonstrating an induced mental simulation process driven by framing (\u00dclk\u00fcmen and Thomas 2013). We further identify self-relevance as a key moderator of the effect. In the following sections, we review existing literature on percentage framing and the integration versus segregation of gains. We then present a series of studies examining the effectiveness of AIPF and IPPF. Finally, we discuss the theoretical contributions and managerial implications of our findings. 57  3.2 Theoretical Background 3.2.1 Percentage Framings and Consumer Evaluations  Price framings\u2014the manner in which price offerings are presented\u2014significantly influence consumer evaluations (Krishna et al. 2002). A common framing in price promotions involves the use of percentage terms, such as \u201c20% off.\u201d Prior research has compared percentage framings with other types of promotions, such as \u201cbuy-one-get-one-free\u201d deals (Gordon-Hecker et al. 2020) and dollar-off deals (e.g., Chen, Monroe, and Lou 1998; Hardesty and Bearden 2003). A meta-analysis by Krishna et al. (2002) highlights that the percentage discount often has a greater impact than the absolute deal amount, though both positively affect perceptions of deal savings. However, limited attention has been given to examining variations within percentage-based framings. In this research, we differentiate three types of percentage framings. Percentage framings generally refer to discounts proportional to the total amount, such as \u201cReceive 5% off\u201d or \u201cReceive 5% off in total,\u201d which we refer to as All-Inclusive Percentage Framings (AIPF). As DelVecchio, Krishnan, and Smith (2007) noted, consumers typically calculate the discount by multiplying the percentage by the total base price. For example, if consumers see an AIPF and purchase n bottles of beer (where B\u2099 represents the original price of the nth bottle), the savings calculation would be: (B1+B2+\u2026+Bn) \u00d7 5% = (B \u00d7 n) \u00d7 5%. Another type of framing is Explicitly Partitioned Percentage Framings (EPPF), where the price or discount components are itemized for each product. For instance, partitioned pricing might 58  specify \u201c18.5% for shipping and handling\u201d and \u201c12% for tax\u201d (Morwitz, Greenleaf, and Johnson 1998). In some regions, taxes may be further divided into provincial sales tax and federal goods and services tax, both of which are proportional to the total order amount. A comparable EPPF to AIPF in price promotions, such as \u201cReceive 5% off in total,\u201d could break it down into \u201cReceive 5% off on the first bottle, 5% off on the second bottle, 5% off on the third bottle,\u201d and so on. In this case, if consumers purchase n bottles of beer, the savings calculation would be: B1 \u00d7 5% + B2 \u00d7 5% + \u2026 + Bn \u00d7 5%.  EPPF is often employed when the discount percentages vary across different categories (e.g., 5% off beer, 10% off wine). When the percentages are consistent for each item or category, managers may opt for Implicitly Partitioned Percentage Framings (IPPF). IPPF indicates that a proportional discount applies to each unit, such as \u201cReceive 5% off each bottle.\u201d It is perceptually similar to AIPF. Unlike EPPF, it does not require listing each percentage for every item. For instance, for n bottles of beer, \u201cReceive 5% off each bottle\u201d is mathematically equivalent to: \u2211 (\ud835\udc35! \u00d7 5%!\"#$ ) = (B \u00d7 5%) \u00d7 n.  The unit of partitioning in IPPF is not restricted to individual items\u2014it can apply to other categories as well. For example, a credit card promotion might use IPPF like \u201cEarn 2.5% back every day,\u201d as opposed to \u201cEarn 2.5% back\u201d (AIPF). IPPF can even extend beyond pricing to contexts such as sustainability. For example, an energy-efficient light bulb might be marketed with IPPF, stating \u201cSave 10% energy each night\u201d rather than \u201cSave 10% energy\u201d (AIPF).  59  Since both IPPF and AIPF are interchangeable when discounts are consistent across units, it is worth comparing their effectiveness. These framings, while mathematically equivalent, could influence consumer perceptions differently due to the distinct ways they are presented. 3.2.2 Integration and Segregation One common question that marketers face is whether to use integration or segregation in their pricing or promotions. In the context of pricing, marketers must decide whether to emphasize partitioned prices or present a single all-inclusive price (Morwitz, Greenleaf, and Johnson 1998). Similarly, when it comes to promotional discounts, marketers can choose to display either the total savings or the savings broken down by individual components, as illustrated in the example above. Several seminal theories have offered insights into the problems. Firstly, as extensionality principles in mathematics suggest, (B \u00d7 n) \u00d7 5% = B1 \u00d7 5% + B2 \u00d7 5% + \u2026 + Bn \u00d7 5% = (B \u00d7 5%) \u00d7 n. Standard economic models indicate that preferences should be invariant across framings since the values in the three framings are equivalent.  Prospect theory (Kahneman and Tversky 1979) and the theory of mental accounting (Thaler 1985) suggest that the value function is concave for gains and convex for losses. This means consumers prefer a series of small gains over a single large gain and an integrated loss over multiple small losses. If prices are perceived as losses and discounts as gains, these theories indicate that consumers should prefer segregated framings for discounts and integrated framings for prices. Research supports this view, showing that consumers have more positive evaluations when component prices are bundled into a single price, and when discounts are debundled into 60  components (Johnson, Herrmann, and Bauer 1999). Relatedly, support theory in probability judgment (Tversky and Koehler 1994) suggests that unpacking an event's description (e.g., a plane crash) into components (e.g., accidental versus nonaccidental) increases its overall probability judgment, as it highlights possibilities that might otherwise be overlooked, making these possibilities more salient. However, all this research compared explicitly partitioned framings (e.g., a gain of $2 plus a gain of $3) with all-inclusive dollar framings (e.g., a gain of $5). Given that the segregation in IPPF is implicit and less salient, and the percentage size (e.g., 5%) is equivalent in both framings, IPPF and AIPF are perceptually comparable and free from numerosity bias (Pelham, Sumarta, and Myaskovsky 1994). The difference between (B \u00d7 n) \u00d7 5% and (B \u00d7 5%) \u00d7 n is merely the order of calculation. Therefore, comparing IPPF and AIPF provides a more conservative test for a mere segregation effect.  However, other studies suggest the opposite pattern may hold. Research on partitioned pricing indicates that consumers may overlook separate surcharges, leading them to perceive lower costs, especially when the surcharges are presented in percentage framings (Morwitz, Greenleaf, and Johnson 1998; Xia and Monroe 2004). Product bundles are evaluated more favorably when their components are presented with partitioned prices rather than integrated prices (Chakravarti et al. 2002). According to Gourville (1998), reframing an aggregate one-time expense as a series of small ongoing expenses (i.e., the \"pennies-a-day\" strategy) reduces the overall perceived cost because the large expense is assimilated into several more affordable payments.  61  In the domain of gains, recent research reveals that quantity discounts framed with IPPF can lead to lower purchase likelihood and smaller perceptions of savings compared to consolidated promotional framing (Yang and Chakravarti 2024). For example, consumers are less likely to purchase when presented with the offer \"Buy two, get 26% off each bottle\" compared to \"Buy two, get 52% off the second bottle.\" This body of research suggests that consumers may perceive savings as smaller when proportional discounts are framed as applying to each small unit, as opposed to being integrated. The lower purchase intention for IPPF is explained by two factors: a diminished savings perception due to the smaller numerical value (e.g., 26% versus 52%) and a lower quality perception of the product due to the uniform discount applied to every item. Therefore, when the promotion depth is substantively small, there is no significant difference in purchase intention or savings perception between \"Buy two, get 2.6% off each bottle\" and \"Buy two, get 5.2% off the second bottle.\" However, the numeric values are identical, and the discount is consistently applied to the same products in both IPPF and AIPF. Since the two factors are consistent between IPPF and AIPF, they would not explain any observed differences between the two framings, and we do not expect that promotion depth to attenuate our effect.  Notably, Yang and Chakravarti (2024) have suggested that consumers might not perceive IPPF as representing multiple gains, highlighting the need for further research to understand how consumers process IPPF. Overall, the existing literature on segregation versus integration and partitioning versus consolidation leaves the question unresolved as to whether IPPF would outperform AIPF. In the next section, we introduce a novel perspective to address this gap. 62  3.2.3 Mental Simulation In this research, we investigate the effectiveness of IPPF through the lens of mental simulation. According to mental simulation theory, there are two distinct types of simulation: process-oriented simulation and outcome-oriented simulation (Pham and Taylor 1999). Process-oriented simulation involves elaborating on the step-by-step process leading to a desired outcome, whereas outcome-oriented simulation focuses on envisioning the end state one aims to achieve (Pham and Taylor 1999).  Past research has suggested that the two types of mental simulation influence consumers\u2019 judgments and decision-making processes differently. For example, studies have shown that process-oriented thinking increases decision difficulty when consumers face trade-offs, as it leads to greater focus on both the means and the end benefits. In contrast, outcome-oriented simulation focuses solely on the end benefits (Escalas and Luce 2004; Thompson, Hamilton, and Petrova 2009). Similarly, focusing on the process or specific goals heightens the perceived difficulty of achieving those goals (\u00dclk\u00fcmen and Cheema 2011; \u00dclk\u00fcmen and Thomas 2013). In general, the literature indicates that process-oriented thinking demands and fosters more attention and elaboration compared to outcome-oriented thinking. Building on this, we propose that the heightened elaboration resulting from process-oriented thinking may amplify the evaluation of a given attribute\u2019s valence. This mechanism is akin to how unpacking events increases the perceived support for each sub-event in probability judgments (Tversky and Koehler 1994). We term this phenomenon the \u201cprocess amplifier effect.\u201d 63  Since IPPF emphasizes that discounts are applied to each unit of the purchase (e.g., \u201ceach bottle\u201d), consumers are likely to perceive each unit as separate by default. They initially evaluate the discount magnitude for a single unit (e.g., \u201cHow much is 5% off for the first bottle?\u201d) and then repeat this process for subsequent units (e.g., \u201cWhat about the second, third, etc.?\u201d). Thus, IPPF encourages consumers to simulate the process of how the percentage discount applies through the whole process and across all units. In contrast, AIPF indicates that discounts are applied to the total price, leading consumers to view the entire purchase as a single entity. In this case, they are more likely to calculate the total price first and then apply the percentage discount to the final outcome. As a result, while the percentage information in IPPF is elaborated multiple times during the evaluation process, in AIPF it is elaborated only once at the outcome stage. Accordingly, we propose that consumers are more likely to engage in process-oriented simulation with IPPF and outcome-oriented simulation with AIPF. Given the process amplifier effect associated with process-oriented simulation, we hypothesize the following: Hypothesis 1: IPPF leads to higher purchase intention and deal evaluation than AIPF. Hypothesis 2: IPPF leads to more process-oriented simulation than AIPF. A factor closely related to process-oriented mental simulation is self-relevance. High self-relevance products are consistent with consumers\u2019 needs and values and are more likely to motivate consumers to process the information about the products (Zaichkowsky 1985). Previous research has shown that personal relevance amplifies the effects of process-oriented simulation. 64  For instance, \u00dclk\u00fcmen and Thomas (2013) demonstrated that individuals concerned about their BMI perceived a 365-day diet plan as more difficult\u2014and were less likely to adopt it\u2014compared to an equivalent 12-month diet plan. This framing effect did not occur among individuals for whom dieting was not personally relevant. Similarly, if IPPF fosters process-oriented simulation, we expect the percentage framing effect to be contingent on consumers\u2019 relevance to the products. Specifically, we hypothesize the following: Hypothesis 3: Personal relevance moderates the percentage framing effect. High-personal-relevance consumers perceive the deal as better with IPPF than with AIPF, while low-personal-relevance consumers perceive the deals as similar. We conducted five studies to test our hypotheses. All study materials and data are available on OSF: https:\/\/osf.io\/yg3fh\/?view_only=1a2e340d92ba40c0aeb60f9000a04648. Studies 1A and 1B examine the main effect of framing on deal evaluation and purchase intention (H1), while Study 2 investigates the moderating role of personal relevance in the process amplifier effect (H3). The first three studies also explore how framing influences basket size. Finally, Studies 3 and 4 provide evidence for the mechanism of process-oriented mental simulation (H2). 3.3 Empirical Studies 3.3.1 Study 1A: Main Effect of Percentage Framing This study examines the effects of percentage framing on purchase intention and deal evaluation (H1). Building on prior research suggesting that promotion depth may moderate the 65  effects of IPPF (Yang and Chakravarti 2024), we explore whether promotion depth (e.g., 1% versus 10%) influences these outcomes. As noted earlier, distinct from Yang and Chakravarti (2024), the numeric values in IPPF and AIPF are always identical, and the distance between these values remains constant regardless of the promotion depth. Therefore, we do not anticipate that promotion depth will attenuate our effect. Furthermore, we include basket size as an exploratory measure to evaluate its potential impact on consumer responses. 3.3.1.1 Method We recruited four hundred participants from Prolific (pre-registration: https:\/\/aspredicted.org\/p88c-6r4k.pdf). They were asked to imagine that they passed by a grocery store and saw one of the following messages from the store. The study is 2 (framing: AIPF versus IPPF) x 2 (promotion depth: 10% versus 1%) between-subject design. In AIPF condition, participants saw \u201cBuy soda and receive 10%\/1% off in total!\u201d In IPPF condition, participants saw \u201cBuy soda and receive 10%\/1% off each bottle!\u201d We measured deal evaluation (\u201cIs the promotion a bad or good deal?\u201d; 1 = A very bad deal; 7 = A very good deal), purchase intention (\u201cAfter seeing this message, how likely are you to make a purchase?\u201d; 1 = Extremely unlikely; 7 = Extremely likely), and purchase basket size (\u201cAfter seeing this message, how many bottles of soda would you buy?\u201d).  3.3.1.2 Results  Results from a two-factor ANOVA on deal evaluation revealed that consumers perceived the deal as significantly better for IPPF (M = 3.10, SD = 1.62) than AIPF (M = 2.77, SD = 1.55), 66  F(1, 396) = 6.70, p = .010, \u03b7p2 = 0.02, which is consistent with H1. Consumers also perceived the deal as significantly better for 10% promotion depth (M = 3.86, SD = 1.36) than for 1% depth (M = 2.02, SD = 1.24), F(1, 396) = 200.64, p < .001, \u03b7p2 = 0.34. However, there is no interaction effect between promotion framing and promotion depth, p = .949. Results from a two-factor ANOVA on purchase intention revealed that consumers had directionally higher purchase intention for IPPF (M = 2.88, SD =1.82) than AIPF (M = 2.68, SD =1.75), F(1, 396) = 1.38, p = .242, significantly higher purchase intention for 10% promotion depth (M = 3.53, SD = 1.78) than 1% depth (M = 2.04, SD = 1.45), F(1, 396) = 83.71, p < .001, \u03b7p2 = 0.18, with no significant interaction effect, p = .797.  Results from a two-factor ANOVA on basket revealed that consumers had directionally larger basket sizes for 10% promotion depth (M = 1.53, SD = 2.44) than 1% depth (M = 0.94, SD = 3.86), F(1, 396) = 3.31, p = .070, \u03b7p2 = 0.01. Neither the effect of framing (p = .743) nor the interaction effect between framing and promotion depth (p = .720) was significant. 3.3.1.3 Discussion  The findings of this study suggest that consumers perceive the deal more favorably with IPPF than with AIPF and exhibit a directionally higher purchase intention when the offer is framed as IPPF, partially supporting H1. The lack of a significant effect of framing on purchase intention could be attributed to the fact that factors beyond deal evaluation, such as personal relevance influence purchase intention. As a result, the impact of framing on purchase intention may be smaller, and the current study design may have insufficient power to detect this effect. 67  3.3.2 Study 1B: Main Effect of Percentage Framing with Quantity Discounts This study examines the effects of percentage framing on purchase intention and deal evaluation (H1) within a quantity-purchase context. We hypothesize that IPPF will be perceived as a better deal and lead to higher purchase intention than AIPF. We again measure whether framing affects basket size as an exploratory measure. 3.3.2.1 Method We recruited two hundred and one participants from Prolific who drink alcohol. They were asked to imagine that they passed by a liquor store and saw one of the following messages from the store. In AIPF condition, participants saw \u201cBuy 6 bottles of beer or more and receive 10% off in total!\u201d In IPPF condition, participants saw \u201cBuy 6 bottles of beer or more and receive 10% off each bottle!\u201d We measured purchase intention (\u201cAfter seeing this message, how likely are you to visit the liquor store?\u201d; 1 = Extremely unlikely; 5 = Extremely likely), purchase basket size (\u201cAfter seeing this message, how many bottles of beer would you buy?\u201d), and deal evaluation (\u201cIs the promotion a bad or good deal?\u201d; 1 = A very bad deal; 5 = A very good deal).  3.3.2.2 Results  Results showed that consumers had significantly higher purchase intention for IPPF (M = 3.54, SD =1.03) than AIPF (M = 3.16, SD =1.09), t(199) = 2.55, p = .012, Cohen\u2019s d = 0.36. Consumers also perceived the deal as significantly better for IPPF (M = 3.49, SD = 0.84) than AIPF (M = 3.13, SD = 0.95), t(199) = 2.87, p = .005, Cohen\u2019s d = 0.41, which is consistent with 68  H1. Interestingly, we also found that consumers had marginally larger basket size for IPPF (M = 7.03, SD = 4.17) than AIPF (M = 6.04, SD = 3.46), t(199) = 1.83, p = .068, Cohen\u2019s d = 0.26. 3.3.2.3 Discussion  Consistent with H1, the findings of this study indicate that consumers had a higher purchase intention and perceived the deal as better when the framing was IPPF (vs. AIPF). Furthermore, despite the presence of a quantity purchase anchor (6 bottles), consumers were more likely to purchase 7 bottles under the IPPF framing compared to 6 bottles under the AIPF framing. This suggests that consumers viewing IPPF perceive each unit separately, making them less susceptible to the anchor effect of quantity requirement. 3.3.3 Study 2: Personal Relevance as a Moderator Although Study 1A and 1B provided initial support for H1, the underlying mechanism of the framing effect remained unclear. Building on \u00dclk\u00fcmen and Thomas (2013), we investigated whether personal relevance moderated this effect. Specifically, we hypothesized that consumers with high personal relevance for the products would exhibit higher purchase intention and perceive the deal as better when presented with IPPF rather than AIPF. However, this effect was not observed for consumers with low personal relevance to the products. Additionally, we explored whether the presence or absence of a quantity threshold influenced the framing effect. 3.3.3.1 Method We recruited 600 US participants from Prolific who had indicated that they drank alcohol. They were asked to imagine that they passed by a liquor store and saw one of the following 69  messages from the store. When the quantity threshold was present, participants in AIPF condition saw \u201cBuy 6 bottles or more and receive 10% off in total!\u201d In IPPF condition, participants saw \u201cBuy 6 bottles or more and receive 10% off each bottle!\u201d When the quantity threshold was absent, participants in AIPF condition saw \u201cBuy our beverages and receive 10% off in total!\u201d In IPPF condition, participants saw \u201cBuy our beverages and receive 10% off each bottle!\u201d We used the same scales as in Study 1 to measure purchase intention, purchase basket size, and deal evaluation. We measured personal relevance by asking, \u201cHow many units of alcohol do you drink on average per week? (1 unit of alcohol = 1 small glass of wine; half pint of beer; pub measure of spirits; 1 = 0; 2 = 1-4; 3 = 5-9; 4 = 10-13; 5 = 14+)\u201d 3.3.3.2 Results  Purchase Intention. Results of ANOVA on purchase intention showed that consumers had marginally higher purchase intention for promotions with IPPF (M = 3.45, SD = 1.05) than promotions with AIPF (M = 3.31, SD = 1.09), F(1, 596) = 2.84, p = .093, \u03b7p2 = 0.01. Consumers also had significantly higher purchase intention when the quantity threshold was absent (M = 3.64, SD = 0.93; vs. when the quantity threshold was present, M = 3.11, SD = 1.15), F(1, 596) = 37.98, p < .001, \u03b7p2 = 0.06. There was no interaction effect between framing and quantity threshold, F(1, 596) = 0.04, p = .836.  Deal evaluation. Similarly, results of ANOVA on deal evaluation showed that consumers perceived promotions with IPPF as better deals (M = 3.41, SD = 0.88) than promotions with AIPF (M = 3.17, SD = 0.94), F(1, 596) = 10.90, p = .001, \u03b7p2 = 0.02. Consumers also perceived the 70  promotions as better deals when the quantity threshold was absent (M = 3.48, SD = 0.86; vs. when the quantity threshold was present, M = 3.10; SD = 0.93), F(1, 596) = 26.83, p < .001, \u03b7p2 = 0.04.  There was no interaction effect between framing and quantity threshold, F(1, 596) = 0.28, p = .597.  Basket Size. Results of basket size only showed that consumers would buy more when the quantity threshold was present (M = 4.81, SD = 3.68; vs. absent, M = 3.87, SD = 4.31), F(1, 596) = 8.15, p = .004, \u03b7p2 = 0.01. Neither a main effect of framing (p = .514) nor an interaction effect (p = .159) was found. Personal Relevance. Results of regression analyses showed that personal relevance moderated the effects of framing on purchase intention (see Figure 3.1), b = 0.21, SE = 0.09, t = 2.29, p = .023, deal evaluation (see Figure 3.2), b = 0.17, SE = 0.08, t = 2.17, p = .031, and basket size (see Figure 3.3), b = 0.71, SE = 0.34, t = 2.06, p = .040. Specifically, simple slopes analyses showed that increasing personal relevance led to higher purchase intention (b = 0.31, SE = 0.07, t = 4.69, p < .001), higher deal evaluation (b = 0.16, SE = 0.06, t = 2.79, p = .005), and larger basket size (b = 1.11, SE = 0.25, t = 4.50, p < .001), when the framing was IPPF. However, when the framing was AIPF, personal relevance did not have significant effect on purchase intention (p = .11), deal evaluation (p = .83), or basket size (p = .08). In other words, consumers with high personal relevance (+1 SD from the mean) have higher purchase intention (b = 0.35, SE = 0.12, t = 2.89, p = .004), higher deal evaluation (b = 0.40, SE = 0.10, t = 3.87, p < .001) and similar basket size (p = .281) for IPPF than AIPF. When the personal relevance was low (-1 SD from the mean), 71  framing did not have a significant effect on purchase intention (p = .728), or deal evaluation (p = .422), and a marginal (reversed) effect on basket size (p = .067). Figure 3.1 The Effect of Framing and Personal Relevance on Purchase Intention    Figure 3.2 The Effect of Framing and Personal Relevance on Deal Evaluation   72  Figure 3.3 The Effect of Framing and Personal Relevance on Basket Size  3.3.3.3 Discussion  Consistent with H3, we found that consumers who consumed alcohol more frequently were more susceptible to the process amplifier effect, whereas those who drank less showed similar purchase intention and deal evaluations across both framings. The findings also indicated that the framing effect persisted regardless of the presence of a quantity threshold. This suggests that simply including the phrase \"each bottle\" is sufficient to prompt consumers to mentally apply the discount to each unit, effectively simulating the process.  3.3.4 Study 3: Mental Simulation as Mechanism Consistent with the mental simulation account, which suggests that IPPF encourages a more process-oriented (vs. outcome-oriented) approach (H2), we predict that consumers exposed to IPPF (vs. AIPF) will engage in repeated elaboration of the discount throughout the process, 73  rather than evaluating it only once as an outcome. Additionally, we propose that the perceived number of discounts mediates the effect of framing on deal evaluation. We also explore whether the presence or absence of a product quantity threshold moderates this effect. 3.3.4.1 Method A total of 385 college students from a North American university participated in the study (pre-registration: https:\/\/aspredicted.org\/GVM_BZK). Participants were asked to imagine purchasing candy for an upcoming get-together. They came across a candy store offering a promotional deal. The study employed a 2x2 between-subjects design, manipulating framing (AIPF vs. IPPF) and quantity threshold (presence vs. absence) in the offer message: \u201cBuy four bars of candy (or candy bars) and get 10% off (each bar)!\u201d We measured deal evaluation (\u201cIs the offer a bad or good deal?\u201d; 1 = A very bad deal; 7 = A very good deal) and perceived number of discounts (\u201cHow many discounts do you feel like you are receiving? 1 = I feel like I am receiving a single discount; 7 = I feel like I am receiving multiple discounts\u201d).  3.3.4.2 Results  Deal Evaluation. Results of ANOVA on deal evaluation showed that consumers perceived the deal with IPPF (M = 4.35, SD = 1.41) as significantly better than the deal with AIPF (M = 3.84, SD = 1.35), F(1, 381) = 13.21, p < .001, \u03b7p2 = 0.03. Consumers also perceived the deals as better when the quantity threshold was absent (M = 4.25, SD = 1.29; vs. when the quantity threshold was present, M = 3.94; SD = 1.48), F(1, 381) = 5.19, p = .023, \u03b7p2 = 0.01. The interaction between framing and threshold is not significant, F(1, 381) = 2.58, p = .109. 74  Perceived Number of Discounts. Results of ANOVA on perceived number of discounts showed that consumers perceived the deal with IPPF (M = 3.41, SD = 2.14) to have significantly more discounts than the deal with AIPF (M = 2.22, SD = 1.59), F(1, 381) = 37.84, p < .001, \u03b7p2 = 0.09, while neither the main effect of quantity threshold (p = .576) nor the interaction between framing and threshold (p = .731) is significant. Perceived number of discounts mediates the effect of framing on deal evaluation, b = 0.28, SE = 0.06, 95% CI = [0.17, 0.42]. 3.3.4.3 Discussion  The findings of this study further confirmed that consumers exposed to IPPF (vs. AIPF) perceived the offer as better and to have more discounts, consistent with the notion that IPPF prompted consumers to engage in process-oriented mental simulation (H1 and H2). Once again, we observed that the effect of IPPF was independent of the presence or absence of a threshold. 3.3.5 Study 4: Perceived Number of Discounts To further support the idea that consumers exposed to IPPF (vs. AIPF) engage in multiple elaborations of the discount throughout the process (as opposed to a single elaboration for the outcome), we employed a design similar to the previous study. In this study, participants were explicitly asked to indicate the number of discounts they perceived receiving. We retained the two conditions with thresholds to control for any potential effects of imagined basket size on deal evaluation. 75  3.3.5.1 Method A total of 202 UK residents from Prolific participated in the study (pre-registration: https:\/\/aspredicted.org\/KBR_DZL). Participants were asked to imagine purchasing candy for an upcoming get-together. They came across a candy store offering a promotional deal, which is either \u201cBuy four bars of candy and get 10% off!\u201d (AIPF) or \u201cBuy four bars of candy and get 10% off each bar!\u201d (IPPF). We measured deal evaluation (\u201cIs the offer a bad or good deal?\u201d; 1 = A very bad deal; 7 = A very good deal), perceived magnitude of the offer (\u201cHow large is this offer?\u201d; 1 = Very small; 7 = Very big), and perceived number of discounts (\u201cHow many discounts does it feel like you would receive from the offer? Enter a number only\u201d).  3.3.5.2 Results  Deal Evaluation. Results of an independent-sample t-test on deal evaluation showed that consumers perceived the deal with IPPF (M = 4.39, SD = 1.34) as significantly better than the deal with AIPF (M = 3.86, SD = 1.30), t(200) = 2.82, p = .005, Cohen\u2019s d = 0.40.  Perceived Magnitude. Results of an independent-sample t-test on the perceived magnitude of the offer showed that consumers perceived the deal with IPPF (M = 3.54, SD = 1.20) as significantly larger than the deal with AIPF (M = 2.94, SD = 1.13), t(200) = 3.69, p < .001, Cohen\u2019s d = 0.52.  Perceived Number of Discounts. Most participants (84.16%) perceived one or four discounts. However, a significantly higher proportion of participants indicated \u201c4\u201d (compared to \u201c1\u201d) in the IPPF condition than in the AIPF condition, despite the presence of \u201cfour bars\u201d in both 76  scenarios, Mann-Whitney U = 3040.00, z = 5.53, p < .001. Specifically, 53.5% of participants in the IPPF condition reported \u201c4,\u201d while 34.7% reported \u201c1.\u201d In contrast, only 4.0% of participants in the AIPF condition reported \u201c4,\u201d with 76.2% indicating \u201c1.\u201d 3.3.5.3 Discussion  The findings of this study supported the idea that IPPF led consumers to elaborate on the discount multiple times and perceive the offer as larger and better, consistent with H1 and H2. Specifically, although the same purchase threshold of \"four bars of candy\" was indicated in both IPPF and AIPF, consumers were more likely to perceive four discounts with IPPF compared to just one discount with AIPF. 3.4 General Discussion This research examined consumer deal evaluations for two perceptually similar framings: the commonly used AIPF and the novel IPPF. Our findings across four studies demonstrate that IPPF can lead to higher purchase intention and better deal evaluation compared to AIPF (H1), especially when the promoted products hold high self-relevance to consumers (H3). The moderation effect of personal relevance on framing in study 2 aligns with the mental simulation account, suggesting that IPPF encourages consumers to engage in process-oriented mental simulations of the discount (H2). To further substantiate this mechanism, we conducted studies 3 and 4 to test the hypothesis that framing effects are driven by a \"process amplifier effect.\" The findings support that IPPF promotes multiple elaborations of the discount for each product unit, resulting in enhanced deal evaluation. 77  3.4.1 Theoretical implications Our findings contribute to the literature on aggregation versus segregation by introducing a nuanced paradigm to test prospect theory while shedding light on the role of the process amplifier effect as an alternative mechanism. Our results demonstrate that the segregation inherent in IPPF leads to higher deal evaluation, primarily among consumers with higher personal relevance to the products and are thus more motivated to process promotional information. Conversely, this effect diminishes for consumers who find the products irrelevant, highlighting the moderating role of personal relevance. This mechanism also helps to explain discrepancies observed in previous studies, such as the pennies-a-day effect (Gourville 1998). Specifically, we hypothesize that process-oriented mental simulation, which fosters deeper elaboration, may inversely affect consumer perceptions when prices are framed in trivial amounts. Repetitive simulation of these small amounts in the IPPF condition could make them appear even less significant, ultimately reducing the perceived magnitude of the deal. This insight underscores the importance of understanding how mental simulation shapes consumer decision-making and offers a framework for reconciling prospect theory with real-world consumer responses to various framing effects. This research also advances the literature on mental simulation by demonstrating an induced mental simulation process driven by framing (\u00dclk\u00fcmen and Thomas 2013). Unlike previous work, which predominantly examines mental simulation related to product consumption or usage (Cornil and Chandon 2016; Dahl, Chattopadhyay, and Gorn 1999; Schlager, de Bellis, and Hoegg 2020; Zhao, Hoeffler, and Dahl 2009) or product progression (Cian, Longoni, and 78  Krishna 2020), our findings reveal a distinct form of simulation. Here, process-oriented simulation involves consumers imagining how the discount applies to each unit of the product. While consumption-oriented simulations often influence basket size (Cornil and Chandon 2016), the process-oriented simulation in our studies primarily impacts deal evaluation. 3.4.2 Practical Implications This research offers valuable managerial implications, demonstrating that transitioning from AIPF to IPPF is both practical and advantageous across various promotional contexts, provided the framing effectively employs small, meaningful units. Beyond improving the effectiveness of price promotions, this approach can also enhance promotions in other domains. For example, temporal units could be utilized in credit card advertisements by reframing \u201cEarn 2.5% back\u201d as \u201cEarn 2.5% back every day,\u201d or in marketing energy-efficient light bulbs by presenting \u201cSave 10% energy each night\u201d instead of the generic \u201cSave 10% energy.\u201d Such reframing not only makes the offer more relatable to consumers but also enhances its perceived value by encouraging repeated mental simulation of the benefits, thereby increasing consumer engagement and the overall appeal of the promotion.  It is noteworthy that the scope of our findings extends beyond the quantity promotion context explored in previous research. For instance, the discount consolidation effect examined IPPF within the framework of quantity thresholds, such as \"Buy two, get 30% off each bottle,\" compared to consolidated promotions like \"Buy two, get 60% off the second bottle\" (Yang and Chakravarti 2024). This prior research was therefore limited to quantity discounts. In contrast, our 79  study demonstrates that IPPF leads to higher deal evaluations than AIPF regardless of whether a quantity requirement is present. This broader applicability allows our conclusions to be generalized across diverse promotional contexts, making them relevant for a wider range of marketing strategies beyond just quantity-based promotions.  Furthermore, IPPF proves to be particularly impactful for consumers with high personal relevance to the product, who are also the most likely to make a purchase. By appealing more strongly to this segment, IPPF has the potential to boost brand loyalty, fostering stronger connections between consumers and the brand over time.  3.4.3 Limitations and Future Research Directions This research has several limitations that open avenues for future investigation. First, while our findings provide valuable insights into consumer responses to different percentage framing strategies, the data is based on controlled experimental designs rather than real purchase behavior. Future research should employ field experiments or analyze transactional data to validate whether the observed effects translate to actual consumer behavior in real-world settings. Second, while we present evidence supporting process-oriented mental simulation as the underlying mechanism for the observed effects, the measures used primarily capture consumers\u2019 self-reported perceptions of discounts. While this demonstrates the partitioning of discounts, it remains unclear whether consumers are actively processing the discounts multiple times. Future studies should directly test the proposed process amplifier effect by manipulating mental 80  simulation processes. For example, examining whether cognitive load moderates the framing effect could provide more definitive evidence (Cian, Longoni, and Krishna 2020). If the framing effect diminishes under high cognitive load\u2014where consumers\u2019 ability to engage in process-oriented mental simulation is impaired\u2014it would strongly support mental simulation as a key mechanism. Another promising avenue involves facilitating process-oriented mental simulations. By prompting participants to imagine how percentage discounts apply to each unit of a product, researchers could test whether AIPF can also enhance deal evaluations under such conditions. This would deepen our understanding of the framing effect and suggest strategies for mitigating or amplifying its impact depending on specific promotional objectives. Additionally, future research should examine how the format of promotions\u2014such as percentage discounts versus dollar discounts\u2014interacts with framing. For instance, consumers may perceive \"Buy 10 bottles of beer and receive 5% off each bottle\" as a better deal than \"Buy 10 bottles of beer and receive 5% off in total.\" However, the reverse may occur when promotions involve trivial monetary amounts, such as \"Buy 10 bottles of beer and receive 10 cents off each bottle\" versus \"Buy 10 bottles of beer and receive $1 off in total.\" Investigating this interaction would clarify how mental simulation influences consumer responses to differently framed promotions across various contexts. Moreover, the application of IPPF extends beyond price promotions and could be leveraged in other domains such as charitable donations and sustainable behavior campaigns. For example, 81  instead of Patagonia\u2019s \"1% for the Planet\" campaign emphasizing a single percentage donation from total sales, framing it as \"donate 1% of the sales of each piece of clothing\" might resonate more strongly with consumers by making the donation seem more immediate and tangible. Similarly, promoting sustainable behavior through IPPF could highlight incremental benefits, such as \u201csave 10% energy every day\u201d with energy-efficient appliances, rather than \u201csave 10% energy.\u201d These applications demonstrate the broader potential of IPPF to influence consumer attitudes and behaviors across a wide range of contexts. Lastly, extending this research to a loss context, such as price increases or inflation, could provide valuable insights. For example, consumers might react differently to \"Prices will increase by 10% for each bottle\" versus \"Prices will increase by 10% in total.\" Understanding how framing effects operate in negative contexts could guide marketers and policymakers in designing communications that align with consumer perceptions and preferences. By addressing these limitations and exploring these directions, future research can build a more comprehensive understanding of framing effects and their implications for consumer decision-making and marketing theory and practice. 82  Chapter 4: Conclusion This dissertation contributes to the understanding of consumer responses to price-promotion architectures and framings by examining two distinct but interrelated research streams. Across two essays, we explored how promotion restrictions (Essay 1) and percentage framing strategies (Essay 2) influence purchase intentions, deal evaluations, and broader managerial and theoretical implications. Together, these essays highlight the nuanced interplay between consumer psychology and promotional strategies, offering actionable insights for marketing practitioners and enriching the academic discourse on promotion effectiveness. In Essay 1, I investigated consumers\u2019 attitudes toward two common restricted promotion types: capped and threshold promotions. Across six controlled experiments and one field study, I found robust evidence that consumers exhibit higher purchase intentions and generate larger sales when offered threshold promotions compared to economically equivalent capped promotions\u2014provided the trigger value is below their usual spending amount. This preference is driven by consumers\u2019 differential expectations: capped promotions heighten promotion expectations, leading to negative expectation disconfirmation, while threshold promotions lower spending expectations, enhancing perceived fairness. The results demonstrate the robustness of these effects across diverse contexts, measures, and manipulations, including real-world scenarios such as food ordering, ride-hailing, and grocery shopping. I also highlight the nuanced interplay of trigger values, revealing that the preference reverses when the trigger value exceeds consumers\u2019 usual spending. 83  In Essay 2, we examined how consumers evaluate two perceptually similar framing strategies: All-Inclusive Percentage Framing (AIPF) and Implicitly Partitioned Percentage Framing (IPPF). Across four studies, we found that IPPF leads to higher purchase intentions and better deal evaluations than AIPF, particularly when the promoted products are highly self-relevant. This effect is mediated by consumers\u2019 enhanced perceptions of the number of discounts in IPPF. The results underscore the role of mental simulation in shaping consumer deal evaluations, particularly when the promoted products are of high self-relevance. The identification of the \u201cprocess amplifier effect\u201d provides a novel perspective on how framing can enhance perceived value, building on existing literature on framing effects (DelVecchio, Krishnan, and Smith 2007). 4.1 Contributions This research addresses the questions of how consumers evaluate threshold versus capped promotions and how the framing of percentage promotions influences purchase intentions. Its contributions are both theoretical and practical. Theoretically, this research advances the understanding of price promotion effectiveness by identifying conditions under which different types of promotions succeed or fail. It also deepens the field\u2019s knowledge of framing effects by uncovering the psychological mechanisms, such as the expectations disconfirmation and the process-oriented mental simulation, that drive consumer preferences. By introducing a framework based on expectation discrepancy theory, Essay 1 provides a dual-sided explanation for how threshold and capped promotions influence consumer behavior. The decomposition of the effects into an aversion to capped promotions and preference 84  for threshold promotions over unrestricted promotions offers a more nuanced understanding of restricted promotion effectiveness. Essay 2 enriches the literature on promotional percentage framing by demonstrating the advantages of IPPF over AIPF in specific contexts. The findings complement prior research by identifying mental simulation as an important mechanism in framing effects. Practically, the insights derived from this thesis provide valuable guidance for marketers. Marketers should carefully consider the architecture of restricted promotions. Threshold promotions are recommended for increasing purchase intentions and sales, particularly when the trigger values align with consumers\u2019 usual spending patterns. For capped promotions, managers should mitigate negative expectation disconfirmation by setting lower percentage caps and emphasizing attainable benefits. Moreover, IPPF emerges as a promising alternative promotion strategy to traditional AIPF. Managers should leverage IPPF for products with high self-relevance to consumers to maximize deal evaluations and purchase intentions. 4.2 Strengths and Limitations This dissertation has several strengths. First, the studies employ robust experimental designs and methodologies, ensuring the reliability of the findings. Second, the research addresses practical, real-world issues in promotion design, making the results highly applicable to marketing practice. Third, the integration of diverse theoretical perspectives\u2014expectation disconfirmation, fairness perceptions, and mental simulation\u2014adds depth and breadth to the analysis of price promotions strategies. 85  However, the research is not without its limitations, as outlined in the general discussion sections of each essay. Most notably, the studies are conducted in controlled settings, which may not fully capture the complexities of real-world purchase behavior. Future research should validate these findings through field experiments and longitudinal studies. Additionally, while the proposed mechanisms are supported by experimental evidence, direct measures of mental simulation remain limited, necessitating further investigation. 4.3 Future Research Directions This dissertation opens several avenues for future research. First, further studies are needed to examine real-world purchase behavior, particularly through field experiments that test the effects of promotions across diverse retail contexts. Additionally, future research could delve into the carryover effects of promotion architecture and framing strategies on brand perceptions, user habits, and long-term purchase behavior. Such investigations could yield valuable insights into both long-term purchase intentions and overall purchase volumes. Another promising direction is the exploration of alternative framing techniques beyond percentage discounts. For instance, how do consumers respond to dollar-based partitioned frames compared to percentage-based ones? Additionally, examining the role of cultural and individual differences in promotion evaluation could provide a more nuanced understanding of consumer behavior across diverse markets. Lastly, future research should explore how the principles identified in this thesis apply to non-monetary promotions, such as loyalty programs, product bundles, and subscription models. 86  By broadening the scope of inquiry, researchers can further uncover the underlying principles of effective promotion design and framing. 4.4 Final Thoughts This dissertation underscores the critical role of promotion architecture and framing in shaping consumer behavior. By identifying the psychological mechanisms underlying consumer responses to restricted promotions and framing strategies, this research provides a robust foundation for designing effective marketing practices. The insights not only enhance our theoretical understanding of promotion effectiveness but also offer actionable guidance for marketers navigating the complexities of consumer decision-making in dynamic marketplaces.   87  References Adams, J. Stacy (1965), \u201cInequity in Social Exchange,\u201d in Advances in Experimental Social Psychology, Elsevier, 267\u201399. Allard, Thomas, David J. Hardisty, and Dale Griffin (2019), \u201cWhen \u2018More\u2019 Seems Like Less: Differential Price Framing Increases the Choice Share of Higher-Priced Options,\u201d Journal of Marketing Research, 56 (5), 826\u201341. Bertini, Marco and Aylin Aydinli (2020), \u201cConsumer Reactance to Promotional Favors,\u201d Journal of Retailing, 96 (4), 578\u201389.  Braun, Michael, Bart De Langhe, Stefano Puntoni and Eric M. Schwartz (2024), \u201cLeveraging Digital Advertising Platforms for Consumer Research,\u201d Journal of Consumer Research, 51 (1), 119\u201328. Campbell, Margaret C. (1999), \u201cPerceptions of Price Unfairness: Antecedents and Consequences,\u201d Journal of Marketing Research, 36 (2), 187. Chakravarti, Dipankar, Rajan Krish, Pallab Paul, and Joydeep Srivastava (2002), \u201cPartitioned Presentation of Multicomponent Bundle Prices: Evaluation, Choice and Underlying Processing Effects,\u201d Journal of Consumer Psychology, 12 (3), 215-29. Chandon, Pierre, Brian Wansink, and Gilles Laurent (2000), \u201cA Benefit Congruency Framework of Sales Promotion Effectiveness,\u201d Journal of Marketing, 64 (4), 65\u201381. 88  Cheema, Amar and Vanessa M. Patrick (2008), \u201cAnytime versus Only: Mind-sets Moderate the Effect of Expansive versus Restrictive Frames on Promotion Evaluation,\u201d Journal of Marketing Research, 45 (4), 462\u201372. Chen, Haipeng (Allan), Howard Marmorstein, Michael Tsiros, and Akshay R. Rao (2012), \u201cWhen More is Less: The Impact of Base Value Neglect on Consumer Preferences for Bonus Packs over Price Discounts,\u201d Journal of Marketing, 76 (4), 64\u201377. Chen, Shih-Fen S., Kent B. Monroe, and Yung-Chien Lou (1998), \u201cThe Effects of Framing Price Promotion Messages on Consumers\u2019 Perceptions and Purchase Intentions,\u201d Journal of Retailing, 74 (3), 353\u201372. Cheng, Andong, and Gretchen R. Ross (2023), \u201cTiered discounts as multiple reference points for spending,\u201d Journal of Consumer Psychology, 33 (2), 424-31. Cian, Luca, Chiara Longoni, and Aradhna Krishna (2020), \u201cAdvertising a Desired Change: When Process Simulation Fosters (vs. Hinders) Credibility and Persuasion,\u201d Journal of Marketing Research, 57 (3), 489\u2013508. Cornil, Yann and Pierre Chandon (2016), \u201cPleasure as a Substitute for Size: How Multisensory Imagery Can Make People Happier with Smaller Food Portions,\u201d Journal of Marketing Research, 53 (5), 847\u201364. Dahl, Darren W., Amitava Chattopadhyay, and Gerald J. Gorn (1999), \u201cThe Use of Visual Mental Imagery in New Product Design,\u201d Journal of Marketing Research, 36 (1), 18\u201328. 89  Dallas, Steven K. and Vicki G. Morwitz (2018), \u201c\u2018There Ain\u2019t No Such Thing as a Free Lunch\u2019: Consumers\u2019 Reactions to Pseudo-Free Offers,\u201d Journal of Marketing Research, 55 (6), 900\u201315. Darke, Peter R., Laurence Ashworth, and Kelley J. Main (2010), \u201cGreat Expectations and Broken Promises: Misleading Claims, Product Failure, Expectancy Disconfirmation and Consumer Distrust,\u201d Journal of the Academy of Marketing Science, 38, 347\u201362. Darke, Peter R., Laurence Ashworth, and Robin J.B. Ritchie (2008), \u201cDamage from Corrective Advertising: Causes and Cures,\u201d Journal of Marketing, 72 (6), 81\u201397. Darke, Peter R. and Darren W. Dahl (2003), \u201cFairness and Discounts: The Subjective Value of a Bargain,\u201d Journal of Consumer Psychology, 13 (3), 328\u201338. Darke, Peter R. and Robin J.B Ritchie (2007), \u201cThe Defensive Consumer: Advertising Deception, Defensive Processing, and Distrust,\u201d Journal of Marketing Research, 44 (1), 114\u201327. DelVecchio, Devon, H. Shanker Krishnan, and Daniel C. Smith (2007), \u201cCents or Percent? The Effects of Promotion Framing on Price Expectations and Choice,\u201d Journal of Marketing, 71 (3), 158\u201370. Ding, Yu and Yan Zhang (2020), \u201cHiding Gifts Behind the Veil of Vouchers: On the Effect of Gift Vouchers versus Direct Gifts in Conditional Promotions,\u201d Journal of Marketing Research, 57 (4), 739\u201354.  90  Escalas, Jennifer Edson and Mary Frances Luce (2004), \u201cUnderstanding the Effects of Process-Focused versus Outcome-Focused Thought in Response to Advertising,\u201d Journal of Consumer Research, 31 (2), 274\u201385. Gneezy, Ayelet (2005), \u201cAttitudes and Promotions,\u201d in NA - Advances in Consumer Research Volume 32, ed. Geeta Menon and Akshay R. Rao, Duluth, MN: Association for Consumer Research, 485\u201386.  Gordon-Hecker, Tom, Andrea Pittarello, Shaul Shalvi, and Marieke Roskes (2020), \u201cBuy-one-get-one-free deals attract more attention than percentage deals,\u201d Journal of Business Research, 111, 128\u201334. Gourville, John T. (1998), \u201cPennies\u2010a\u2010Day: The Effect of Temporal Reframing on Transaction Evaluation,\u201d Journal of Consumer Research, 24 (4), 395\u2013403. Greenleaf, Eric A., Eric J. Johnson, Vicki G. Morwitz, and Edith Shalev (2016), \u201cThe price does not include additional taxes, fees, and surcharges: A review of research on partitioned pricing,\u201d Journal of Consumer Psychology, 26 (1), 105\u201324. Grewal, Dhruv, Howard Marmorstein, and Arun Sharma (1996). \u201cCommunicating Price Information Through Semantic Cues: The Moderating Effects of Situation and Discount Size,\u201d Journal of Consumer Research, 23 (2), 148\u201355. Hamilton, Ryan (2024), \u201cConsumer Price Evaluation Strategies: Internal References, External References, and Price Images in Consumer Price Perception,\u201d Consumer Psychology Review, 7 (1), 58-74. 91  Hardesty, David M. and William O. Bearden (2003), \u201cConsumer Evaluations of Different Promotion Types and Price Presentations: The Moderating Role of Promotional Benefit Level,\u201d Journal of Retailing, 79 (1), 17\u201325. Hardisty, David J. and Elke U. Weber (2020), \u201cImpatience and Savoring vs. Dread: Asymmetries in Anticipation Explain Consumer Time Preferences for Positive vs. Negative Events,\u201d Journal of Consumer Psychology, 30 (4), 598\u2013613.  Heath, Timothy B., Chatterjee, S., and France, Karen R. (1995), \u201cMental Accounting and Changes in Price: The Frame Dependence of Reference Dependence,\u201d Journal of Consumer Research, 22 (1), 90-97  Hock, Stefan J., Rajesh Bagchi, and Thomas M. Anderson (2020), \u201cPromotional Games Increase Consumer Conversion Rates and Spending,\u201d Journal of Consumer Research, 47 (1), 79-99. Hopp, Frederic R., Jacob T. Fisher, Devin Cornell, Richard Huskey, and Ren\u00e9 Weber (2021), \u201cThe extended Moral Foundations Dictionary (eMFD): Development and Applications of a Crowd-Sourced Approach to Extracting Moral Intuitions from Text.\u201d Behavior Research Methods, 53, 232\u201346. Hsee, Christopher K. (1996), \u201cThe Evaluability Hypothesis: An Explanation for Preference Reversals between Joint and Separate Evaluations of Alternatives,\u201d Organizational Behavior and Human Decision Processes, 67 (3), 247\u201357. 92  Inman, J. Jeffrey, Anil C. Peter, and Priya Raghubir (1997), \u201cFraming the Deal: The Role of Restrictions in Accentuating Deal Value,\u201d Journal of Consumer Research, 24 (1), 68\u201379. Johnson, Michael D, Andreas Herrmann, and Hans H Bauer (1999), \u201cThe effects of price bundling on consumer evaluations of product offerings,\u201d International Journal of Research in Marketing, 16 (2), 129\u201342. Kahneman, Daniel and Amos Tversky (1979), \u201cProspect Theory: An Analysis of Decision under Risk,\u201d Econometrica, 47 (2), 263. Kalwani, Manohar U. and Chi Kin Yim (1992), \u201cConsumer Price and Promotion Expectations: An Experimental Study,\u201d Journal of Marketing Research, 29 (1), 90\u2013100. Kalyanaram, Gurumurthy and Russell S. Winer (1995), \u201cEmpirical Generalizations from Reference Price Research,\u201d Marketing Science, 14 (3_supplement), G161\u201369. Krishna, Aradhna, Richard Briesch, Donald R. Lehmann, and Hong Yuan (2002), \u201cA Meta-Analysis of the Impact of Price Presentation on Perceived Savings,\u201d Journal of Retailing, 78 (2), 101\u201318.  Kyung, Ellie J., & Manoj Thomas (2016), \u201cWhen Remembering Disrupts Knowing: Blocking Implicit Price Memory,\u201d Journal of Marketing Research, 53 (6), 937-953.  Lam, Shun Yin, Mark Vandenbosch, John Hulland, and Michael Pearce (2001), \u201cEvaluating Promotions in Shopping Environments: Decomposing Sales Response into Attraction, Conversion, and Spending Effects,\u201d Marketing Science, 20 (2), 194-215. 93  Lee, Leonard and Dan Ariely (2006), \u201cShopping Goals, Goal Concreteness, and Conditional Promotions,\u201d Journal of Consumer Research, 33 (1), 60\u201370. McKenzie, Craig R.M. and Shlomi Sher (2020), \u201cGamble Evaluation and Evoked Reference Sets: Why Adding a Small Loss to a Gamble Increases Its Attractiveness,\u201d Cognition, 194, 104043. Mobley, Mary F., William O. Bearden, and Jesse E. Teel (1988), \u201cAn Investigation of Individual Responses to Tensile Price Claims,\u201d Journal of Consumer Research, 15 (2), 273-79.  Morwitz, Vicki G., Eric A. Greenleaf, and Eric J. Johnson (1998), \u201cDivide and Prosper: Consumers\u2019 Reactions to Partitioned Prices,\u201d Journal of Marketing Research, 35 (4), 453\u201363. Oliver, Richard L., PV (Sundar) Balakrishnan, and Bruce Barry (1994), \u201cOutcome Satisfaction in Negotiation: A Test of Expectancy Disconfirmation,\u201d Organizational Behavior and Human Decision Processes, 60 (2), 252\u201375. Oliver, Richard L. and John E. Swan (1989), \u201cEquity and Disconfirmation Perceptions as Influences on Merchant and Product Satisfaction,\u201d Journal of Consumer Research, 16 (3), 372\u201383. Palmeira, Mauricio M. and Joydeep Srivastava (2013), \u201cFree Offer \u2260 Cheap Product: A Selective Accessibility Account on the Valuation of Free Offers,\u201d Journal of Consumer Research, 40 (4), 644\u201356. 94  Pelham, B.W., T.T. Sumarta, and L. Myaskovsky (1994), \u201cThe Easy Path From Many To Much: the Numerosity Heuristic,\u201d Cognitive Psychology, 26 (2), 103\u201333. Pham, Lien B. and Shelley E. Taylor (1999), \u201cFrom Thought to Action: Effects of Process-Versus Outcome-Based Mental Simulations on Performance,\u201d Personality and Social Psychology Bulletin, 25 (2), 250\u201360. Raghubir, P. (2004), \u201cCoupons in Context: Discounting Prices or Decreasing Profits?\u201d Journal of Retailing, 80 (1), 1\u201312. Schlager, Tobias, Emanuel de Bellis, and JoAndrea Hoegg (2020), \u201cHow and when weather boosts consumer product valuation,\u201d Journal of the Academy of Marketing Science, 48 (4), 695\u2013711. Sevilla, Julio, Mathew S. Isaac, and Rajesh Bagchi (2018), \u201cFormat Neglect: How the Use of Numerical versus Percentage Rank Claims Influences Consumer Judgments,\u201d Journal of Marketing, 82 (6), 150\u201364. Sher, Shlomi and Craig R. M. McKenzie (2014), \u201cOptions as Information: Rational Reversals of Evaluation and Preference,\u201d Journal of Experimental Psychology: General, 143 (3), 1127\u201343. Sokolova, Tatiana and Yi Li (2021), \u201cMix or Match? Consumer Spending Decisions in Conditional Promotions,\u201d Journal of Consumer Psychology, 31 (1), 91\u2013102. Spreng, Richard A., Scott B. MacKenzie, and Richard W. Olshavsky (1996), \u201cA Reexamination of the Determinants of Consumer Satisfaction,\u201d Journal of Marketing, 60 (3), 15\u201332.  95  Statista (2022a), \u201cAverage Spend Per Order on Selected Online Food Delivery Services in the United States in 2022\u201d (accessed May 13, 2024), https:\/\/www.statista.com\/statistics\/1358873\/average-food-delivery-spend-per-order-us\/. Statista (2022b), \u201cShare of Meal Delivery Service Users in Australia in 2022, by Generation\u201d (accessed December 22, 2023), https:\/\/www.statista.com\/statistics\/1249873\/ australia-breakdown-meal-delivery-service-users-by-generation\/. Teng, Lefa (2009), \u201cA Comparison of Two Types of Price Discounts in Shifting Consumers\u2019 Attitudes and Purchase Intentions,\u201d Journal of Business Research, 62 (1), 14\u201321. Thaler, Richard H. (1980), \u201cToward a Positive Theory of Consumer Choice,\u201d Journal of Economic Behavior & Organization, 1 (1), 39-60. Thaler, Richard H. (1985), \u201cMental Accounting and Consumer Choice,\u201d Marketing Science, 4 (3), 199\u2013214. Thaler, Richard H. and Cass R. Sunstein (2008), Nudge: Improving Decisions About Health, Wealth, and Happiness, New Haven, CT: Yale University Press. Thompson, Debora Viana, Rebecca W. Hamilton, and Petia K. Petrova (2009), \u201cWhen Mental Simulation Hinders Behavior: The Effects of Process-Oriented Thinking on Decision Difficulty and Performance,\u201d Journal of Consumer Research, 36 (4), 562\u201374. Tversky, Amos and Daniel Kahneman (1974), \u201cJudgment under Uncertainty: Heuristics and Biases,\u201d Science, 185 (4157), 1124\u201331. 96  Tversky, Amos and Daniel Kahneman (1981), \u201cThe Framing of Decisions and the Psychology of Choice,\u201d Science, 211 (4481), 453\u201358. Tversky, Amos and Derek J. Koehler (1994), \u201cSupport theory: A nonextensional representation of subjective probability.,\u201d Psychological Review, 101 (4), 547\u201367. \u00dclk\u00fcmen, G\u00fclden and Amar Cheema (2011), \u201cFraming Goals to Influence Personal Savings: The Role of Specificity and Construal Level,\u201d Journal of Marketing Research, 48 (6), 958\u201369. \u00dclk\u00fcmen, G\u00fclden and Manoj Thomas (2013), \u201cPersonal Relevance and Mental Simulation Amplify the Duration Framing Effect,\u201d Journal of Marketing Research, 50 (2), 194\u2013206. Urbany, Joel E., William O. Bearden, and Dan C. Weilbaker (1988), \u201cThe Effect of Plausible and Exaggerated Reference Prices on Consumer Perceptions and Price Search,\u201d Journal of Consumer Research, 15 (1), 95.  Xia, Lan, Kent B. Monroe, and Jennifer L. Cox (2004), \u201cThe Price is Unfair! A Conceptual Framework of Price Fairness Perceptions,\u201d Journal of Marketing, 68 (4), 1\u201315. Yang, Haiyang, and Dipankar Chakravarti (2024), \u201cThe discount consolidation effect: How brands can present quantity discounts more effectively,\u201d Journal of Consumer Psychology, 34 (4), 620-31. Yoon, Sukki and Patrick T. Vargas (2011), \u201c\u2018No More\u2019 Leads to \u2018Want More,\u2019 but \u2018No Less\u2019 Leads to \u2018Want Less\u2019: Consumers\u2019 Counterfactual Thinking when Faced with Quantity Restriction Discounts,\u201d Journal of Consumer Behavior, 10 (2), 93\u2013101. 97  Yoon, Sukki and Patrick T. Vargas (2010), \u201cFeeling Happier When Paying More: Dysfunctional Counterfactual Thinking in Consumer Affect,\u201d Psychology & Marketing, 27 (12), 1075\u20131100. Zaichkowsky, Judith Lynne (1985), \u201cMeasuring the Involvement Construct,\u201d Journal of Consumer Research, 12 (3), 341\u201352. Zhao, Min, Steve Hoeffler, and Darren W. Dahl (2009), \u201cThe Role of Imagination-Focused Visualization on New Product Evaluation,\u201d Journal of Marketing Research, 46 (1), 46\u201355.     98  Appendices Appendix A  Chapter 2 Appendices A.1 Examples of Promotional Messages  Capped Promotions   DoorDash Coupon KABU Coupon Threshold Promotions   UberEats Coupon UberEats Coupon    99  A.2 Examples of the Two Promotion Types Used in Joint Evaluation    DoorDash Coupon in the US Uber Eats Coupon in the UK 100  A.3 Prevalence of Three Types of Promotions for Food Delivery on Doordash  To examine the prevalence of the above promotions in the marketplace, blind coders collected information from 100 stores that had provided offers on the most popular US food-delivery platform, DoorDash, in the two largest national markets, New York and Los Angeles (50 stores around a university in each city).  Results showed that 55% of the stores used capped promotions, 24% used threshold promotions, and the rest, 21%, were free-delivery promotions. Thus, restricted promotions (capped and threshold types) are the dominant types used on DoorDash. Our results also suggest that both restricted promotion types appear commonly across many industries (e.g., ride-hailing, grocery shopping, and food delivery; see Appendix A.1).  Since DoorDash does not provide the specific regular spending amount, we use the price indicator, the number of dollar signs, as a proxy for spending amount, and we don\u2019t not find a significant correlation between the spending amount and the choice of promotion architecture (0 = capped, 1 = threshold), r = .057, p = .619, N = 79, which may indicate that managers are not aware of the relationship between the two.   101  A.4 Promotion Expectations, Spending Expectations, and Prevalence of Three Types of Promotions for Food Delivery As a pretest, we conducted a survey to examine consumers\u2019 internal promotion expectations and spending expectations based on past experiences with food ordering, considering that most of our studies\u2019 scenarios are in the food delivery context. Moreover, we also examined the prevalence of threshold, capped, and unrestricted promotions.  A total of 101 (Mage = 35.34, SDage = 11.84; 68.3% female) US residents from Prolific completed the survey. To measure promotion expectations, we asked participants to indicate the discount in percentage terms that they usually receive on online food ordering platforms. To measure spending expectations, we asked participants to indicate the amount in US dollars they usually spend for one food-ordering delivery. Next, we asked whether they have seen an unrestricted, threshold, or capped promotion before using three binary choice (yes\/no) questions in a random order. Results show that the mean of promotion expectations in percentage is 12.53% (SD = 13.71), and its median is 10%. The majority (85.1%) of the participants indicate a percentage below or equal to 20%. If those who indicate 0 are excluded, the mean is 17.83% (SD = 13.15), and its median is 15%. The mean of spending expectations is $33.99 (SD = 17.15), and its median is $30.00. Moreover, we found that 86% of the participants have seen threshold promotions, 76% have seen 102  capped promotions, and 73% have seen unrestricted promotions, which indicates that the two restricted promotions are as common as (if not more common) unrestricted promotions nowadays.   103  A.5 Prevalence of Low and High Trigger Values in the Marketplace We set out to test the prevalence of price promotion with low and high trigger values (when the trigger value is lower than people\u2019s usual spending amount versus when it is higher) because we hypothesize that the effect of the promotion architecture on purchase intention depends on the relation between the trigger value and the consumer\u2019s typical spending amount.  We ask naive coders to collect all the current and historic coupons accessible for Instacart (grocery delivery) and Uber Eats (primarily restaurant delivery) from couponfollow.com. Results showed that Instacart set its promotional thresholds between $10 and $80, below its average order value ($113). Similarly, Uber Eats typically set their promotional thresholds at $20, lower than the average order value ($34; also see Appendix A.4). We also partner with Fantuan, an Asian cuisine food and grocery delivery platform present in the North American, Australian, and UK markets. To our knowledge, Fantuan is the only platform on which each restaurant\u2019s exact average spending amounts are shown. Two naive coders were asked to collect information from 50 restaurants in a major North American city to examine the prevalence of high versus low threshold promotions. Results showed that 59% of stores set their promotional thresholds higher than the corresponding average order spending amount, while a meaningful 40% of stores set their promotional thresholds below this value.  These observations about the prevalence of thresholds both higher and lower than typical spending suggest that managers may not only use high thresholds to stretch consumers\u2019 spending amounts (i.e., spend more to meet the promotion requirement), but also use low thresholds to 104  increase purchase intention and conversion. Notably, in the case of Fantuan abovementioned, the correlation between threshold and average spending amount is positive and significant but relatively small, r = 0.28, p = .006.  Together, these findings suggest that it is quite common to see threshold promotions with low thresholds.    105  A.6 Tiktok Field Study Replication  Considering that the online platform A\/B tests may be affected by algorithm, we ran another coupon-based ad campaign on TikTok to replicate Study 1. This study\u2019s design and data analysis approach is the same as Study 1, with a budget of $200 for each ad. The threshold-promotion ad reached 35,539 viewers, and the capped-promotion ad reached 38,525 viewers. Same as Study 1, we conducted a one-tailed two-proportion z-test on CTR and found that viewers were more likely to click on the threshold-promotion advertisement (1.45%) than the capped-promotion one (1.29%), z = 1.79, p = .037. We also replicated our exploratory analysis in Study 1 that viewers were more likely to play the threshold-promotion advertisement (22.36%) for at least two seconds than the capped-promotion one (21.82%), z = 1.80, p = .036. This result suggests that our finding is a solid and replicable phenomenon in the real world and can provide managerial guidance.    106  A.7 Restricted Promotions with Low Promotion Depth and Unrestricted Promotion Consistent with our expectation disconfirmation account, we proposed that consumers would evaluate the threshold promotion (e.g., \u201c$3 off on an order of $15 or more\u201d) as better than the corresponding unrestricted dollar-term promotion (e.g., \u201c$3 off\u201d) which is evaluated as better than the capped promotion (e.g., \u201c20% off, $3 max discount on an order\u201d). A total of 387 college students participated in the study. Participants in the threshold-promotion condition saw a coupon: \u201cEnjoy $3 off (On an order of $15 or more).\u201d Those in the capped-promotion condition read, \u201cEnjoy 20% off ($3 max discount on an order).\u201d Those in the unrestricted promotion condition read, \u201cEnjoy $3 off.\u201d The scenario, the procedure, and the measures are the same as in Studies 6a and 6b. An ANOVA revealed a significant difference in deal evaluations among the three promotion-architecture conditions (see Figure A.1), F(2, 384) = 13.52, p < .001, \u03b7p2 = 0.07. As expected, deal evaluation was higher in the threshold promotion (M = 4.29, SD = 1.57) than in the capped promotion (M = 3.19, SD = 1.68), p < .001. Results also showed that deal evaluations with the unrestricted promotion (M = 3.81, SD = 1.87) were lower than with the threshold promotion, p = .023, but higher than with the capped promotion, p = .004. The results replicated our findings in Study 3 and supported our expectation disconfirmation account.    107  Figure A.1 Promotion Architecture and Deal Evaluation   Note: Error bars represent \u00b11 standard error.  Consistently, logistic regression analysis revealed that the conversion rate for the threshold promotion (41.54%) is higher than that for the capped promotion (25.00%), Wald \u03c72(1) = 7.81, p = .005, and marginally higher than that for the unrestricted promotion (31.01%), Wald \u03c72(1) = 3.09, p = .079. The conversion rate for the unrestricted promotion is directionally higher than that for the capped promotion, Wald \u03c72(1) = 1.15, p = .284. Among participants who decided to order, we did not find a significant difference in purchase amount among the three promotion-architecture 3.193.814.2912345Capped Unrestricted ThresholdDealEvaluation108  conditions, F(2, 123) = 1.80, p = .169. However, we found that the purchase amount was marginally higher for the capped promotion (M = 23.53, SD = 11.52) than the threshold promotion (M = 20.13, SD = 7.50), p = .062, which may suggest the existence of an anchoring effect for the threshold promotion. However, after taking those who would not make an order into consideration (coded as 0), a Mann-Whitney U test suggested that participants overall tended to spend more when it was the threshold promotion (M = 8.36, SD = 11.06) than when it was the capped promotion (M = 5.88, SD = 11.71), U = 7146, z = 2.34, p = .019, with unrestricted promotion condition in the middle (M = 6.53, SD = 10.14). The overall sales for the threshold promotion ($1087) are larger than that for the unrestricted promotion ($842) and that for the capped promotion ($753). We also conducted a cost-benefit analysis by calculating the payment amount (spending amount \u2013 discount amount) contingent on their spending amount and the promotion type for each participant. After the appropriate discount was deducted, the payment amount was still larger for threshold promotions (M = 7.16, SD = 9.76) than that for capped promotions (M = 5.15, SD = 10.59), U = 7140, z = 2.35, p = .019, with unrestricted promotion condition in the middle (M = 5.60, SD = 8.81).    109  A.8 Charity Promotions in Grocery Shopping This study extends previous results by testing the robustness of our promotion-architecture effect with low trigger values (H1) using a different promotion context: charitable appeals. We predict that when the trigger value is low, people are more likely to purchase with a threshold promotion (\u201cWe donate $5 per purchase, if you purchase $10 or more\u201d) than with an equivalent capped promotion (\u201cWe donate 50% of your purchase price, up to $5 per purchase\u201d). We also examine our proposed mechanism underlying preferences for threshold promotion by testing whether expectations and fairness perceptions serially mediate the effect (H2). Additionally, since unfair promotions may have a broader effect on a firm\u2019s reputation (Darke, Ashworth, Ritchie 2008), we examine whether the restricted promotion type affects a firm\u2019s moral image and hypothesize that a firm that uses a \u201cfair\u201d threshold promotion will be perceived as more moral. Lastly, we determine whether consumers correctly notice that capped promotions apply to a broader range of purchases and whether they consider this applicability when evaluating the promotions. We recruited 400 US residents from Prolific, and 403 participated in the study (Mage = 39.50, SDage = 14.90; 49.6% female). We used a two-condition (promotion architecture: threshold vs. capped promotion) between-subjects design. We asked participants to imagine they were about to buy groceries for next week and that while thinking of going to a supermarket, they received a notification from the Amazon Whole Foods Market app.  110  Participants in the threshold-promotion condition read, \u201cWe donate $5 per purchase (if you purchase $10 or more) on the app to support the Organic Farming Research Foundation,\u201d whereas participants in the capped-promotion condition read, \u201cWe donate 50% of your purchase price (up to $5 per purchase) on the app to support the Organic Farming Research Foundation.\u201d We measured purchase intentions (\u201cAfter seeing this message, how likely are you to download the Amazon Whole Foods Market app and make a purchase?\u201d 1 = Extremely unlikely, 5 = Extremely likely) and moral image (\u201cHow moral is Amazon Whole Foods Market?\u201d 1 = Not moral at all, 5 = Extremely moral). We also included process measures to test H2: fairness (\u201cThe donation amount is fair\u201d; 1 = Strongly disagree, 5 = Strongly agree), negative expectation discrepancy (\u201cThe donation amount is below my expectation\u201d; 1 = Strongly disagree, 5 = Strongly agree), and applicability (\u201cThe donation applies to all orders\u201d; 1 = Strongly disagree, 5 = Strongly agree). The order of the three process items was randomized. Purchase intention. As expected in a low-trigger context, consumers\u2019 purchase intentions were higher with the threshold promotion (M = 3.01, SD = 1.20) than with the capped promotion (M = 2.75, SD = 1.22), t(401) = 2.07, p = .039, Cohen\u2019s d = 0.21. Furthermore, consumers rated the company as fairer when it used the threshold promotion (M = 3.94, SD = 0.99) compared to the capped promotion (M = 3.23, SD = 1.20), t(401) = 6.51, p < .001, Cohen\u2019s d = 0.65. They also had less negative expectation disconfirmation with the threshold promotion (M = 2.25, SD = 1.15) than with the capped promotion (M = 3.10, SD = 1.25), t(401) = 7.14, p < .001, Cohen\u2019s d = 0.71. A mediation analysis (see Figure A.2) using PROCESS Macro (Model 6; Hayes 2017) revealed 111  that our results were consistent with a model in which negative expectation disconfirmation and fairness perception serially mediate the effect of promotion architecture on purchase intention; b = -0.23, SE = 0.05, 95% CI = [-0.34, -0.14].   Figure A.2 Mediation Effect Through Expectation Disconfirmation and Fairness Perceptions  \t  Although the threshold promotion led to higher purchase intentions than the capped promotion, participants correctly noticed that the capped promotion applied to a broader range of transactions (M = 3.72; SD = 1.23) than the threshold promotion (M = 2.61; SD = 1.42), t(401) = 8.44, p < .001, Cohen\u2019s d = 0.84. A mediation analysis (see Figure A.3) using PROCESS Macro (Model 4; Hayes 2017) reveals that our data are consistent with a model in which individual differences in perceived applicability mediate the effect of promotion architecture on purchase intentions; b = 0.13, SE = 0.06, 95% CI = [0.03, 0.25].   FairnessCapped (1) vsThreshold (0)B = .47***NegativeExpectationDisconfirmationB = -.57***B = .86***Purchase IntentionB = -.22* B = .17**B = -.06*  p < .05 ** p < .01 *** p < .001 112  Figure A.3 Mediation Effect of Applicability \t Moral image. Consumers perceived the company as more moral when it used a threshold promotion (M = 3.12, SD = 1.04) compared to a capped promotion (M = 2.91, SD = 1.09), t(401) = 1.97, p = .050, Cohen\u2019s d = 0.20. Negative expectation disconfirmation and fairness perceptions serially mediated the effect of promotion architecture on perceptions of moral image; b = -0.23, SE = 0.05, 95% CI = [-0.34, -0.14]. We also analyzed moral image as a mediator of purchase intention and found support for a model in which moral image mediated the effect of promotion architecture on purchase intention; b = 0.11, SE = 0.06, 95% CI = [0.00, 0.23]. This study reveals that given a low promotion trigger value, consumers have higher purchase intention in a charity promotion context with a threshold promotion than with a comparable capped promotion, supporting H1 again. Consistent with H2, we find that the capped promotion leads to higher levels of negative expectancy disconfirmation and is perceived as less fair than the threshold promotion. These effects on fairness perceptions arising from the promotion type affected the firm\u2019s moral image more broadly, suggesting that firms using restricted price promotions should choose their promotion architecture carefully. Notably, consumers preferred ApplicabilityPurchase IntentionCapped(1) vsThreshold (0)B = .12**B = 1.11***B = -.38***  p < .05 ** p < .01 *** p < .001 113  threshold promotions under low promotion triggers even though they correctly observed that the capped promotion applies to a broader range of transactions.    114  A.9 The Role of Various Levels of Promotion Depth To further examine the mechanism of expectation disconfirmation and provide managerial insights, we explored how various levels of promotion depth affect evaluations of capped promotions.  A total of 599 US residents from Prolific (Mage = 41.76, SDage = 32.74; 56.9% female) participated in the study with a between-subjects design. The scenario was the same as those used in study 5. Participants were randomly assigned into one of the six conditions, reading a promotion message among \u201cEnjoy 10% off ($3 max discount on an order)\u201d, \u201cEnjoy 20% off ($3 max discount on an order)\u201d, \u201cEnjoy 30% off ($3 max discount on an order)\u201d, \u201cEnjoy 40% off ($3 max discount on an order)\u201d, \u201cEnjoy 50% off ($3 max discount on an order)\u201d, and \u201cEnjoy 60% off ($3 max discount on an order)\u201d. We measured fairness perception (\u201cIs the offer shown above a fair deal?\u201d 1 = Not fair at all, 7 = Very fair) and deal evaluation (\u201cIs the offer shown above a bad or good deal?\u201d 1 = A very bad deal, 7 = A very good deal). We found that the promotion depth (from 10% to 60%) in general has a significantly negative effect on fairness perception, \u03b2 = -0.10, t = 2.44, p = .015 (see Figure A.4). Results of an ANOVA on fairness perception were not significant, F(1, 593) = 1.44, p = .208, but pairwise comparisons revealed that \u201cEnjoy 20% off ($3 max discount on an order)\u201d was perceived as significantly fairer than \u201cEnjoy 50% off ($3 max discount on an order)\u201d, p = .033, and \u201cEnjoy 60% off ($3 max discount on an order)\u201d, p = .041.  115  Similarly, the promotion depth also has a marginally negative effect on deal evaluation, \u03b2 = -0.08, t = 1.92, p = .055 (see Figure A.5). Results of an ANOVA on deal evaluation were not significant, F(1, 593) = 1.21, p = .301, but pairwise comparisons revealed that \u201cEnjoy 20% off ($3 max discount on an order)\u201d was perceived as a marginally better deal than \u201cEnjoy 50% off ($3 max discount on an order)\u201d, p = .099, and a significantly better deal than \u201cEnjoy 60% off ($3 max discount on an order)\u201d, p = .026. The findings support our proposed mechanism that an offer that seemed more attractive could set a higher promotion expectation, leading to expectation disconfirmation and perceived unfairness.   116  Figure A.4 Fairness Perception Among Capped Promotions  Note: Error bars represent \u00b11 standard error.   3.884.023.813.693.48 3.51234510% 20% 30% 40% 50% 60%FairnessPerception117  Figure A.5 Deal Evaluation Among Capped Promotions  Note: Error bars represent \u00b11 standard error.   3.63.843.65 3.653.423.281234510% 20% 30% 40% 50% 60%DealEvaluation118  A.10 Trigger Value as a Moderator when Promotion Depth Kept Constant A total of 403 US residents from Prolific (Mage = 38.89, SDage = 13.55; 50.4% female) participated in the study with a 2 (promotion architecture: threshold vs. capped) by 2 (trigger value: high vs. low) between-subjects design. Participants in the two low-trigger conditions read the same messages as the ones shown in study 5, while those in the high-trigger threshold-promotion condition read, \u201cEnjoy $25 off (On an order of $50 or more).\u201d Those in the high-trigger capped-promotion condition read, \u201cEnjoy 50% off ($25 max discount on an order).\u201d The scenario and the purchase intention measure were the same as those used in Study 5. Results of an ANOVA on purchase intention revealed that consumers had higher purchase intentions for high-trigger promotions than for low-trigger promotions (see Figure A.6), F(1, 399) = 8.46, p = .004, \u03b7p2 = 0.02. Again, there was a significant interaction effect between promotion architecture and trigger values, F(1, 399) = 17.50, p < .001, \u03b7p2 = 0.04. Pairwise comparisons showed that consumers in the low-trigger condition expressed higher purchase intentions with threshold promotions (M = 3.59, SD = 1.22) than they did with capped promotions (M = 3.15, SD = 1.36), t(201) = 5.03, p = .018, Cohen\u2019s d = 0.33, supporting H1. Consumers in the high-trigger condition expressed lower purchase intentions with threshold promotions (M = 3.43, SD = 1.36) than they did with capped promotions (M = 4.03, SD = 0.98), t(198) = 3.59, p < .001, Cohen\u2019s d = 0.51, supporting H3.    119  Figure A.6 Moderation by Trigger Value   Note: Error bars represent \u00b11 standard error.   3.154.033.593.4312345Low Trigger Value High Trigger ValuePurchaseIntentionCappedPromotionThresholdPromotion120  A.11 Capped Promotion versus No Promotion To further explore the practical implications of our research, we investigated whether a capped promotion would result in lower conversion and sales compared to a no-promotion condition. A total of 600 US residents from Prolific (Mage = 37.12, SDage = 13.16; 59.2% female) participated in the study with a between-subjects design. Participants in the capped-promotion condition read, \"Craving something special? Explore a wide variety of cuisines and discover new favorites. Enjoy 50% off, $3 max discount on an order.\" Participants in the no-promotion condition read, \"Craving something special? Explore a wide variety of cuisines and discover new favorites. Enjoy your meal with us, order now.\" Then they indicated their attitude towards this message (1 = Very bad, 7 = Very good) and whether they would make an order (conversion, binary). Participants who answered \u201cyes\u201d to the conversion question were further asked to indicate how much money they were likely to spend.   We found that the participants perceived the capped promotion (M = 4.02, SD = 1.72) as significantly worse than the no-promotion message (M = 4.73, SD = 1.25), t(598) = 5.74, p < .001, Cohen\u2019s d = 0.47. Consistently, logistic regression analysis revealed that the conversion rate was significantly lower for the capped promotion (42.86%) than for the no-promotion condition (52.51%), Wald \u03c72(1) = 5.61, p = .018.  Among those who would make an order, the purchase amount was also marginally smaller for the capped promotion (M = 27.02, SD = 17.39) than for the no-promotion condition (M = 30.34, 121  SD = 14.04), t(284) = 1.79, p = .075, Cohen\u2019s d = 0.21. After taking the proportion of those who would make an order into consideration (non-choosers coded as $0), a Mann-Whitney U test suggested that participants overall tended to spend less when it was the capped promotion (M = 11.58, SD = 17.56) than when it was the no-promotion message (M = 15.93, SD = 18.26), U = 38531.00, z = 3.30, p < .001. Consistent with this, the overall sales for the capped promotion ($3,485.50) are smaller than that for the no-promotion condition ($4,763.00).  To give insight into the associated revenue analysis for managers, we also calculated the payment amount (purchase amount \u2013 discount amount) contingent on their purchase amount and the promotion type for each participant. After the appropriate discount was deducted, the payment amount was even smaller for the capped promotion (M = 10.31, SD = 16.44) than that for the no-promotion condition (M = 15.93, SD = 18.26), U = 37168.00, z = 3.99, p < .001. The overall payment amount for the capped promotion ($3,103.50) are smaller than that for the no-promotion condition ($4,763.00).  The results indicate that a promotion message, such as a capped promotion, can actually backfire compared to having no promotion at all. Changing from a capped promotion with a low trigger value to a no-promotion message may boost both purchase intention and purchase amount. Notably, in this study, the no-promotion message leads to 54.51% more payment amounts than the capped promotion per capita.    122  Appendix B  Chapter 3 Appendices B.1 Examples of Implicit Partitioned Percentage Framing  ","attrs":{"lang":"en","ns":"http:\/\/www.w3.org\/2009\/08\/skos-reference\/skos.html#note","classmap":"oc:AnnotationContainer"},"iri":"http:\/\/www.w3.org\/2009\/08\/skos-reference\/skos.html#note","explain":"Simple Knowledge Organisation System; Notes are used to provide information relating to SKOS concepts. There is no restriction on the nature of this information, e.g., it could be plain text, hypertext, or an image; it could be a definition, information about the scope of a concept, editorial information, or any other type of information."}],"Genre":[{"label":"Genre","value":"Thesis\/Dissertation","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/hasType","classmap":"dpla:SourceResource","property":"edm:hasType"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/hasType","explain":"A Europeana Data Model Property; This property relates a resource with the concepts it belongs to in a suitable type system such as MIME or any thesaurus that captures categories of objects in a given field. It does NOT capture aboutness"}],"GraduationDate":[{"label":"Graduation Date","value":"2025-05","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#dateIssued","classmap":"vivo:DateTimeValue","property":"vivo:dateIssued"},"iri":"http:\/\/vivoweb.org\/ontology\/core#dateIssued","explain":"VIVO-ISF Ontology V1.6 Property; Date Optional Time Value, DateTime+Timezone Preferred "}],"IsShownAt":[{"label":"DOI","value":"10.14288\/1.0448248","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/isShownAt","classmap":"edm:WebResource","property":"edm:isShownAt"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/isShownAt","explain":"A Europeana Data Model Property; An unambiguous URL reference to the digital object on the provider\u2019s website in its full information context."}],"Language":[{"label":"Language","value":"eng","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/language","classmap":"dpla:SourceResource","property":"dcterms:language"},"iri":"http:\/\/purl.org\/dc\/terms\/language","explain":"A Dublin Core Terms Property; A language of the resource.; Recommended best practice is to use a controlled vocabulary such as RFC 4646 [RFC4646]."}],"Program":[{"label":"Program (Theses)","value":"Business Administration - Marketing","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeDiscipline","classmap":"oc:ThesisDescription","property":"oc:degreeDiscipline"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeDiscipline","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the program for which the degree was granted."}],"Provider":[{"label":"Provider","value":"Vancouver : University of British Columbia Library","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/provider","classmap":"ore:Aggregation","property":"edm:provider"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/provider","explain":"A Europeana Data Model Property; The name or identifier of the organization who delivers data directly to an aggregation service (e.g. Europeana)"}],"Publisher":[{"label":"Publisher","value":"University of British Columbia","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/publisher","classmap":"dpla:SourceResource","property":"dcterms:publisher"},"iri":"http:\/\/purl.org\/dc\/terms\/publisher","explain":"A Dublin Core Terms Property; An entity responsible for making the resource available.; Examples of a Publisher include a person, an organization, or a service."}],"Rights":[{"label":"Rights","value":"Attribution-NonCommercial-NoDerivatives 4.0 International","attrs":{"lang":"*","ns":"http:\/\/purl.org\/dc\/terms\/rights","classmap":"edm:WebResource","property":"dcterms:rights"},"iri":"http:\/\/purl.org\/dc\/terms\/rights","explain":"A Dublin Core Terms Property; Information about rights held in and over the resource.; Typically, rights information includes a statement about various property rights associated with the resource, including intellectual property rights."}],"RightsURI":[{"label":"Rights URI","value":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/","attrs":{"lang":"*","ns":"https:\/\/open.library.ubc.ca\/terms#rightsURI","classmap":"oc:PublicationDescription","property":"oc:rightsURI"},"iri":"https:\/\/open.library.ubc.ca\/terms#rightsURI","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the Creative Commons license url."}],"ScholarlyLevel":[{"label":"Scholarly Level","value":"Graduate","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#scholarLevel","classmap":"oc:PublicationDescription","property":"oc:scholarLevel"},"iri":"https:\/\/open.library.ubc.ca\/terms#scholarLevel","explain":"UBC Open Collections Metadata Components; Local Field; Identifies the scholarly level of the author(s)\/creator(s)."}],"Supervisor":[{"label":"Supervisor","value":"Hardisty, David J.","attrs":{"lang":"","ns":"http:\/\/purl.org\/dc\/terms\/contributor","classmap":"vivo:AdvisingRelationship","property":"dcterms:contributor"},"iri":"http:\/\/purl.org\/dc\/terms\/contributor","explain":"A Dublin Core Terms Property; An entity responsible for making contributions to the resource.; Examples of a Contributor include a person, an organization, or a service."},{"label":"Supervisor","value":"Griffin, Dale W.","attrs":{"lang":"","ns":"http:\/\/purl.org\/dc\/terms\/contributor","classmap":"vivo:AdvisingRelationship","property":"dcterms:contributor"},"iri":"http:\/\/purl.org\/dc\/terms\/contributor","explain":"A Dublin Core Terms Property; An entity responsible for making contributions to the resource.; Examples of a Contributor include a person, an organization, or a service."}],"Title":[{"label":"Title ","value":"Two essays on promotion framing","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/title","classmap":"dpla:SourceResource","property":"dcterms:title"},"iri":"http:\/\/purl.org\/dc\/terms\/title","explain":"A Dublin Core Terms Property; The name given to the resource."}],"Type":[{"label":"Type","value":"Text","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/type","classmap":"dpla:SourceResource","property":"dcterms:type"},"iri":"http:\/\/purl.org\/dc\/terms\/type","explain":"A Dublin Core Terms Property; The nature or genre of the resource.; Recommended best practice is to use a controlled vocabulary such as the DCMI Type Vocabulary [DCMITYPE]. To describe the file format, physical medium, or dimensions of the resource, use the Format element."}],"URI":[{"label":"URI","value":"http:\/\/hdl.handle.net\/2429\/90534","attrs":{"lang":"","ns":"https:\/\/open.library.ubc.ca\/terms#identifierURI","classmap":"oc:PublicationDescription","property":"oc:identifierURI"},"iri":"https:\/\/open.library.ubc.ca\/terms#identifierURI","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the handle for item record."}],"SortDate":[{"label":"Sort Date","value":"2025-12-31 AD","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/date","classmap":"oc:InternalResource","property":"dcterms:date"},"iri":"http:\/\/purl.org\/dc\/terms\/date","explain":"A Dublin Core Elements Property; A point or period of time associated with an event in the lifecycle of the resource.; Date may be used to express temporal information at any level of granularity. Recommended best practice is to use an encoding scheme, such as the W3CDTF profile of ISO 8601 [W3CDTF].; A point or period of time associated with an event in the lifecycle of the resource.; Date may be used to express temporal information at any level of granularity. Recommended best practice is to use an encoding scheme, such as the W3CDTF profile of ISO 8601 [W3CDTF]."}]}