---
title: Temporal Analysis of Dark Patterns (TADP)
url: https://www.emergentmind.com/topics/temporal-analysis-of-dark-patterns-tadp
type: topic
---

# Temporal Analysis of Dark Patterns (TADP)

Temporal Analysis of Dark Patterns (TADP) is a methodological framework for understanding, mapping, and measuring dark patterns as time-indexed events distributed throughout a user’s interaction journey. Rather than regarding dark patterns as isolated, static design elements, TADP emphasizes the cumulative, sequential, and often combinatorial effects of deceptive design across the entire flow of user engagement, from initial entry to task completion or abandonment. This analytic lens reveals how manipulative strategies amplify their impact over time, particularly in complex digital or hybrid physical–digital interfaces such as subscription flows or self-ordering kiosks [2309.09635][2603.03218].

## 1. Foundations and Conceptual Framework

TADP originated as a response to the limitations of static, screen-centric taxonomies of dark patterns. Gray et al. first defined TADP as a methodology to “map and measure” dark patterns as time-indexed events situated within the trajectory of a user’s journey [2309.09635]. The central construct is the user journey, formalized as a sequence of discrete interaction events,
$$J = \langle e_1,\, e_2,\, \ldots,\, e_N \rangle,$$
where each event $e_k$ occurs at time $t_k$. A dark pattern $p$ is operationalized as a label attached to one or more events: $(e_k, p)$. The TADP process consists of (a) identifying the set $\{p\}$ over time, (b) associating each $p$ with specific UI artifacts or state transitions, and (c) analyzing the temporal and combinatorial effects that arise as patterns recur or overlap within the journey.

The framework distinguishes three analytic levels:
- **Intra-page:** mapping patterns within a single screen or state,
- **Inter-page:** tracking interactions across consecutive screens within a session,
- **System:** situating the entire journey within the broader architectural or contextual environment [2603.03218].

## 2. Typology and Temporal Characterization

TADP leverages an ontology that distinguishes dark patterns at three hierarchical levels:
- **High level:** broad strategies (e.g., Obstruction, Interface Interference, Social Engineering, Sneaking)
- **Meso-level:** subclass patterns (e.g., Labyrinthine Navigation, Adding Steps, Bad Defaults, Pressured Selling)
- **Low level:** concrete UI manifestations (e.g., Misdirection, Confirmshaming, Visual Prominence, Hidden Information)

Temporal attributes assigned to each pattern instance include:
- **Onset time** $t_{\mathrm{on}}(p)$: the event index or timestamp when the user encounters $p$,
- **Duration** $\Delta t(p)$: how long $p$ persists (visibly or functionally),
- **State transition** $\sigma_{\mathrm{before}} \rightarrow \sigma_{\mathrm{after}}$: the alteration in task or mental state prompted by $p$.

Although the foundational papers do not introduce formal metrics, the temporal activity for a pattern $p$ may be represented as a binary time series,
\[
x_p(t) = \begin{cases}
1 & \text{if } p\text{ is active at } t \\
0 & \text{otherwise}
\end{cases}
\]
[2309.09635].

## 3. Accumulation, Interaction, and Amplification Effects

A central insight of TADP is that manipulative patterns do not merely stack additively; instead, they interact in ways that may amplify user confusion, cognitive load, or coerced compliance. Combinatorial analysis proceeds along two main axes:
- **Co-occurrence:** multiple pattern types may be implemented simultaneously within a single interface state (vertical stacking).
- **Sequential dependency:** exposure to certain patterns in earlier steps may prime or condition the user’s responses to subsequent patterns (horizontal flow).

These phenomena can be documented with constructs such as a co-occurrence matrix,
\[
C_{ij} = \sum_{k=1}^N \mathbf{1}\{p_i \text{ and } p_j \text{ both active at } e_k\},
\]
and, for sequential amplification, with the notion of an amplification factor,
\[
A_{ij} = f\bigl(\Delta t(p_i),\, \Delta t(p_j),\, \operatorname{gap}(p_i, p_j)\bigr),
\]
where $\operatorname{gap}(p_i, p_j)$ denotes the temporal distance between pattern onsets. In the case of the McDonald’s self-ordering kiosk, cumulative metrics such as
\[
E_H(k) = \sum_{i=1}^k |H(s_i)|
\]
track the growing exposure to high-level strategies as the user advances through the 12-screen ordering flow [2603.03218]. These cumulative and combinatorial properties reveal how momentum, time pressure, and asymmetrically distributed interaction costs (e.g., more clicks or navigation steps to refuse than accept) intensify manipulative pressure.

## 4. Detection Strategies and Methodological Workflow

TADP encompasses both expert-driven and automated (or semi-automated) detection methods. Expert evaluation relies on a standardized codebook and open coding using qualitative data analysis tools. Analysts identify pattern onset, map patterns to UI elements, reconstruct branching flows, and annotate state transitions [2309.09635][2603.03218].

Automated and human-in-the-loop detection pipelines utilize:
- Web instrumentation: capturing HAR logs and DOM state, parsing for elements such as buttons, banners, and modals,
- Mobile reverse engineering: extracting UI structure from APK/IPA packages using computer vision and NLP techniques,
- Rule-based heuristics: flagging visual-layout anomalies ($R_{\text{layout}}$), textual cues ($R_{\text{text}}$), and rapid modal reappearances ($R_{\text{temp}}$) indicative of dark patterns.

The recommended workflow for new TADP case studies proceeds as follows:
1. Select or build a dark pattern ontology,
2. Define user goals and states $\{\sigma_k\}$,
3. Record the complete journey (HAR/DOM for web; video frames for mobile),
4. Open-code each event $e_k$ with zero or more pattern labels,
5. Annotate onset times, durations, and UI transitions,
6. Build journey maps and pattern interaction matrices,
7. Report using standardized tables and annotated flowcharts,
8. Iterate annotation and ontology extension as new patterns are encountered [2309.09635].

## 5. Empirical Applications: Case Studies in Web and Hybrid Physical–Digital Contexts

TADP has been applied to diverse case studies across digital and hybrid contexts. Gray et al. analyzed Amazon Prime’s “Iliad Flow,” tracing dark pattern sequences during subscription cancellation, with explicit mapping of patterns to user events, UI states, and transitions [2309.09635]. Purohit et al. conducted a 12-step temporal audit of a McDonald’s self-ordering kiosk, showing the recurrence, layering, and temporal escalation of meso- and low-level dark patterns under conditions of social visibility and time pressure [2603.03218].

The following table summarizes the core findings from these two case studies:

| Case Study                           | Patterns Analyzed                     | Key Temporal Insights                       |
|--------------------------------------|---------------------------------------|---------------------------------------------|
| Amazon “Iliad Flow” [2309.09635]     | Obstruction, Interface Interference, Labyrinthine Navigation, Hidden Info | Recurrence and sequencing promote user frustration, amplify exit friction    |
| McDonald’s Kiosk [2603.03218]        | Adding Steps, Bad Defaults, Confirmshaming, Partitioned Pricing, Social Engineering | Linear flow accumulates manipulative pressure, increased susceptibility under time pressure         |

Contextually, these applications demonstrate that dark pattern detection and mitigation must account for temporal accumulation, particularly in linear or forced-choice systems where user autonomy is systematically eroded across multiple steps.

## 6. Regulatory and Design Implications

TADP evidences that regulatory frameworks focusing only on single-screen manipulations are insufficient, as manipulations frequently unfold as temporally distributed, multi-layered sequences. In hybrid systems, such as fast-food kiosks, this accumulation effect is further amplified by environmental factors—public queues, time pressure, and physical interface constraints. Best practices for mitigating dark patterns as revealed by TADP analysis include auditing complete interface flows, enforcing friction-balanced accept/refuse paths, making defaults transparent, and enabling bypass of secondary offers or upsells [2603.03218].

Regulators, especially under the EU’s Unfair Commercial Practices Directive (UCPD) and Digital Services Act, are urged to expand the scope of dark pattern prohibitions to flow-level and multi-event manipulations. A plausible implication is that legal definitions and compliance audits will increasingly require flow-tracing, temporal sequencing, and cumulative pattern mapping grounded in TADP principles.

## 7. Research Directions and Disciplinary Evolution

The temporal analytic lens of TADP emerges amidst broader socio-technical discourse about dark patterns. Longitudinal analyses of #darkpatterns discourse on Twitter from 2010–2021 illuminate the evolution toward transdisciplinary engagement—spanning design, law, policy, and activism—and underscore collective “socio-technical angst” focused on identifying and counteracting manipulative digital practices [2207.10563]. The research agenda outlined by both foundational and recent TADP work includes the development of interface auditing frameworks, specification of dark pattern definitions in law (e.g., under CPRA or DSA), and the formalization of “bright patterns” as countermodels of ethical design.

TADP intersects ongoing efforts in computational detection, regulatory compliance, and empirical HCI, suggesting a continued shift toward tools and audits capable of addressing both the temporal and structural complexities of dark pattern deployment across evolving digital and hybrid service architectures.

Source: https://www.emergentmind.com/topics/temporal-analysis-of-dark-patterns-tadp