📚 Research Digest

Finding: This systematic literature review of 20 papers (2021–2026) finds dark patterns are pervasive across e-commerce, mobile, and social media platforms, manifesting as forced action, obstruction, urgency, and confirmshaming, and argues curbing them requires ethical design (bright patterns), regulation, and digital literacy.

Why it matters: Dark patterns sit squarely within behavior change design, persuasive technology, and ethical HCI — directly relevant to teaching choice architecture, manipulation vs. ethical nudging, and UX design ethics.

Method: Systematic literature review with predefined inclusion/exclusion criteria across peer-reviewed databases, covering 20 papers from 2021–2026.

Jargon: *Dark patterns* — UI/UX design tricks that deceive or coerce users into unintended actions; *confirmshaming* — guilt-inducing opt-out language (e.g., "No thanks, I don't want to save money"); *bright patterns* — ethical design counterparts that genuinely support user goals; *forced action* — requiring unrelated tasks to access a service.

Finding: People systematically misallocate risk-reduction investments under extinction risk (catastrophic, irrecoverable loss): they overinvest when risk is small and underinvest when risk is large, and fail to adjust across time, leading to worse outcomes than optimal strategy.

Why it matters: Directly relevant to behavioral science and decision-making under risk, especially cognitive biases around probability weighting and dynamic decision-making — useful for teaching behavioral economics and choice architecture.

Method: Dynamic programming used to derive optimal investment benchmarks, compared against 100-trial experimental task with real monetary stakes and extinction outcomes.

Jargon: *Extinction risk* = a catastrophic loss probability that ends all future decision-making (irreversible ruin); *dynamic programming* = mathematical optimization method for sequential decision problems.

Finding: Proposes "Algorithmic Trait Amplification" (ATA), a five-stage theoretical model where social media recommendation algorithms detect and reinforce narcissistic traits through iterative exposure to trait-congruent content, potentially intensifying those traits over time.

Why it matters: Directly relevant to behavioral science and persuasive technology interests — specifically algorithmic engagement mechanisms, addiction by design, dark patterns, and the ethical dimensions of recommendation systems shaping user psychology.

Jargon: *Subclinical narcissism* = narcissistic traits below clinical diagnosis threshold but still measurable; *trait-congruent content* = content that matches and validates a user's existing personality tendencies; *Digital Narcissus Effect* = the author's structural metaphor for how algorithms mirror users back to themselves, constraining corrective experiences.

Finding: People adaptively shift decision strategies based on environmental uncertainty, using simpler single-attribute heuristics under high uncertainty and integrating multiple attributes under low uncertainty, consistent with the bias-variance tradeoff.

Why it matters: Has marginal relevance to behavioral science and cognitive biases, but focuses on low-level decision strategy mechanics rather than applied behavioral design, nudges, or HCI contexts the professor prioritizes.

Method: Multi-attribute choice task with manipulated attribute reliability across conditions.

Jargon: *Bias-variance tradeoff*: the statistical principle that reducing model complexity (bias) increases sensitivity to noise (variance), so simpler models can outperform complex ones when data are noisy.

Finding: Central bias in decision-making (preferring middle/right-of-center options) is driven by physical reachability (least effort), not attentional bias or cognitive heuristics, as demonstrated using a magician's "placement force" trick with eye tracking.

Why it matters: Marginally relevant to choice architecture and nudges—reachability as a spatial nudge mechanism has implications for designing physical and digital interfaces.

Method: Clever use of a magician's card force combined with eye tracking (n=137) and Cognitive Reflection Test to disentangle attentional, cognitive, and motor explanations.

Jargon: *Placement force*: a magician's technique exploiting spatial bias to make a participant "freely choose" a pre-selected card. *Reachability hypothesis*: preference reflects minimization of physical motor effort to reach an option.