Abstract
Today consumers are increasingly interacting with algorithms and artificial intelligence (AI) technologies instead of or in addition to humans. Holding everything else constant, would merely framing a decision (e.g., an application outcome made about a consumer, a personalized recommendation) as made by an algorithm or a human still change the way consumers react? The goal of this dissertation is to examine whether and why consumers react to algorithms and humans differently. Offering a counterpoint on the pervasive algorithms-are-bad rhetoric in contemporary marketing literature, this dissertation adopts a nuanced perspective on consumers’ reactions towards algorithms and humans and introduces three contextual factors that impact consumers’ reactions. Specifically, it reveals how consumers’ reactions towards algorithms and humans depends on what the outcome of the decision is (Chapter 2), who the consumer is (Chapter 3) and on what type of complexity the decision possesses (Chapter 4).
Chapter 2 examines how consumers react to favorable versus unfavorable decision outcomes made about themselves (e.g., acceptances vs. rejections) that are framed to be made by algorithms versus humans. Ten studies reveal that, in contrast to managers’ predictions, consumers react less positively when a favorable decision is made by an algorithmic (vs. a human) decision maker. This difference, however, is mitigated for an unfavorable decision.
Chapter 3 tests how consumers’ subjective knowledge in a focal domain affects their reactions towards algorithmic versus human-based recommendations. Seven studies reveal that consumers with high subjective knowledge value recommendations from algorithms (vs. human experts) more, whereas this greater valuation of algorithmic recommendations is mitigated for consumers with low subjective knowledge.
Chapter 4 studies the role of two types of decision complexity (i.e., emotional vs. technical) on individuals’ perceptions towards algorithms versus humans making legal decisions. Two experiments and an internal meta-analysis demonstrate that individuals trust algorithmic (vs. human) judges less and have lower intentions to go to court when algorithms adjudicate. Trust for algorithmic judges is especially penalized when cases involve emotional complexities (vs. simple or technically complex cases). This chapter also reveals two relative advantages of algorithms that policy-makers could consider emphasizing when communicating with citizens: algorithms’ perceived speed and cost.
Chapter 2 examines how consumers react to favorable versus unfavorable decision outcomes made about themselves (e.g., acceptances vs. rejections) that are framed to be made by algorithms versus humans. Ten studies reveal that, in contrast to managers’ predictions, consumers react less positively when a favorable decision is made by an algorithmic (vs. a human) decision maker. This difference, however, is mitigated for an unfavorable decision.
Chapter 3 tests how consumers’ subjective knowledge in a focal domain affects their reactions towards algorithmic versus human-based recommendations. Seven studies reveal that consumers with high subjective knowledge value recommendations from algorithms (vs. human experts) more, whereas this greater valuation of algorithmic recommendations is mitigated for consumers with low subjective knowledge.
Chapter 4 studies the role of two types of decision complexity (i.e., emotional vs. technical) on individuals’ perceptions towards algorithms versus humans making legal decisions. Two experiments and an internal meta-analysis demonstrate that individuals trust algorithmic (vs. human) judges less and have lower intentions to go to court when algorithms adjudicate. Trust for algorithmic judges is especially penalized when cases involve emotional complexities (vs. simple or technically complex cases). This chapter also reveals two relative advantages of algorithms that policy-makers could consider emphasizing when communicating with citizens: algorithms’ perceived speed and cost.
Original language | English |
---|---|
Awarding Institution |
|
Supervisors/Advisors |
|
Award date | 12 May 2022 |
Place of Publication | Rotterdam |
Print ISBNs | 978-90-5892-630-2 |
Publication status | Published - 12 May 2022 |