The following sections provide detailed insights into my PhD dissertation, where I design both field and online randomized control trials and experiments. My research employs causal methods and machine learning tools to analyze how farmers and consumers of shrimp products navigate and respond to perceived complexity in their decision-making processes.
This study examines how farmers perceive complexity in their farming practices and explores whether personalized training can reduce this complexity and promote technology adoption. Employing machine learning techniques, we quantify perceived complexity and conduct a randomized controlled trial, to assess the impact of personalized digital app training on complexity reduction and technology adoption. Our findings reveal a significant decrease in perceived complexity among trained farmers, with an average increase of 19% in app usage ability and a consequent 20% surge in actual app usage.Trained farmers experience approximately 80% higher usefulness, 64% higher usability, 57% higher desirability, and an impressive 73% higher credibility of the app interface compared to non-trained farmers. Our study reveals that targeted training to reduce perceived complexity can enhance user experience and boost adoption rates. However, our analysis of dynamic treatment effects suggests that farmers experience a 15.43% decline in app usage ability over time, indicating the importance of sustained training programs for long-term adoption success.
In the contemporary consumer landscape, individuals are often overwhelmed by an abundance of choices, which can complicate their decision-making processes. This research scrutinizes the effects of escalating complexity, quantified by the number of choice options, on the divergence between consumers’ stated preferences and their actual choices. We redefine ‘optimal choices’ by juxtaposing stated preferences against actual decisions under varying degrees of complexity. Our randomized controlled trial, involving 1,028 participants segregated into three groups representing low, medium, and high complexity, unveils a significant decrease of 0.6% in optimal choices correlating with increased complexity. Our findings suggest that the log-odds of making choices that deviate from stated preferences augment by 0.026 as complexity transitions from low to high. This research fills a gap in the existing literature by quantifying the degree of misalignment or sub-optimality in decision-making, a topic that has not been systematically explored in prior studies. We attribute the discrepancy between choices and preferences to heightened anxiety (8.3%), stress (10.1%), and confusion (12%), which impede cognitive function and result in sub-optimal decisions. Additionally, we note a 7% increase in reliance on heuristics such as brand recognition and lower prices, despite no significant change in the propensity to abstain from making a purchase. Our research indicates the necessity for policies that streamline choice environments to alleviate decision fatigue and foster alignment with genuine consumer preferences. This study not only contributes to the understanding of consumer behavior in complex choice environments but also provides actionable insights for policy-makers and businesses.