PSYC 1500: In-class test

Comprehensive vocabulary flashcards covering privacy evolution, surveillance capitalism, educational impacts of generative AI, academic integrity, and AI applications in mental health care.

Surveillance Capitalism

An economic system, conceptualized by Shoshana Zuboff, in which human experience is extracted as behavioural data, analyzed into predictive models, and traded for profit.

Data Exhaust

The secondary trail of digital behavioural data, searches, logs, and activity left behind by users of digital platforms, recognized by tech firms in the early 2000s as a valuable resource for predictive modelling.

The Right to Privacy (Warren & Brandeis, 1890)

A foundational legal principle established by U.S. legal scholars Samuel Warren and Louis Brandeis arguing that privacy should be legally recognized as a distinct right, famously characterized as the 'right to be let alone.'

Universal Declaration of Human Rights (1948)

A landmark United Nations document adopted post-World War II that formally enshrined privacy as a fundamental, universal human right essential to human dignity and individual freedom.

Contextual Integrity

A privacy framework highlighted by Matz et al. (2020) asserting that privacy depends not merely on who gathers data or what is gathered, but on whether information flows and data uses are appropriate for a specific context (for example, $75\%$ of users find targeting acceptable for event suggestions, whereas only $37\%$ find it acceptable for political messaging).

Privacy Paradox

The observed contradiction, explored by Bandara et al. (2020), between individuals' expressed concerns regarding their digital privacy and their actual behaviour of routinely disclosing personal data when interacting with online services.

Cambridge Analytica Scandal

A major collective privacy violation where an academic personality quiz app created by Aleksandr Kogan harvested data from approximately $270{,}000$ participants and their Facebook friends without consent, allowing Cambridge Analytica to build psychographic Big Five profiles on around $87\text{ million}$ users to microtarget political ads in the 2016 U.S. presidential campaign and Brexit referendum.

Vastaamo Data Breach

A cyberattack and extortion case in Finland (2018–2020) where confidential psychotherapy notes and identifiers of approximately $36{,}000$ patients were stolen, resulting in ransom demands of $€450{,}000$ in Bitcoin from the provider and $€200\text{--}€500$ directly from individual patients.

Data Broker

A commercial entity that aggregates, packages, and sells consumer datasets—including sensitive personal health information like mental health diagnoses (e.g., depression, bipolar disorder, ADHD)—to third parties with minimal buyer verification.

BetterHelp FTC Settlement (2023)

A Federal Trade Commission enforcement ruling ordering the therapy platform to pay $\$7.8\text{ million}$ to users after an investigation revealed it disclosed sensitive customer health data to advertising and social media platforms despite explicit privacy guarantees.

AI Scribes

Automated clinical technologies that record and transcribe consultations, synthesize clinical interactions into structured session notes (such as SOAP or DAP notes), and integrate documentation directly into Electronic Health Records (EHRs).

Health Information Custodian (HIC)

Under Ontario's Personal Health Information Protection Act (PHIPA), a designated healthcare practitioner or organization with legal authority and responsibility over the collection, handling, disclosure, and retention of personal health information.

Agent (under PHIPA)

An entity authorized under PHIPA to handle personal health information on behalf of a health information custodian strictly for the custodian's defined clinical care purposes, precluding secondary activities such as commercial software training or external product development.

Unsolicited Suggestions (in AI Documentation)

An error identified in AI scribe evaluations where the model introduces clinical details, diagnoses, or treatment recommendations that were never mentioned or authorized during the provider-patient consultation.

Tokens

The basic discrete linguistic elements (words, sub-words, or characters) that large language models convert into mathematical vectors to calculate contextual and statistical relationships.

Vector

A multi-dimensional numerical coordinate that represents a token's semantic properties in an embedding space, grouping terms that appear in similar linguistic contexts close together.

Non Sequitur Fallacy (in Tech Hype)

An argumentative error that conflates what a technology makes possible (its affordances) with how individuals actually use it (behaviour) and the resulting impact on learning (outcomes).

Appeal to Novelty (in Tech Hype)

A logical fallacy that presumes an educational tool or technological development is superior or more beneficial solely because it is new or innovative, rather than relying on empirical evidence of its effectiveness.

Cognitive Offloading

The practice of delegating internal cognitive operations—such as memory retrieval, synthesis, planning, or analytical reasoning—to external digital tools, which reduces immediate mental effort at the expense of deep processing.

Desirable Difficulties

Structured learning conditions and challenges, identified by Bjork and Bjork (2011), that deliberately slow down immediate performance and fluency in order to foster durable long-term retention and flexible conceptual transfer.

Emotional Offloading

The reliance on generative AI tools to escape the discomfort, uncertainty, and cognitive friction naturally encountered when wrestling with complex or ambiguous academic tasks.

Fluency Illusion

A cognitive bias wherein reading clear, well-structured, AI-generated explanations leads a student to overestimate their personal mastery and competence regarding a subject.

Fundamental Values of Academic Integrity

The six core moral values underpinning scholarship: Honesty, Trust, Fairness, Respect, Responsibility, and Courage.

AI Disclosure Statement

A formal, transparent declaration informing an audience that generative AI tools were used to draft, edit, or generate content, clarifying the specific scope of tool usage and human oversight.

Critical Reader Prompting

A collaborative prompt strategy where generative AI is instructed to identify unsupported assumptions, evaluate logical arguments, and pose challenging questions rather than drafting or rewriting content for the user.

Covert AI Use in Psychotherapy

The undisclosed application of generative AI tools by a mental health practitioner during or between clinical encounters, which research shows significantly damages patient trust and therapeutic acceptability.

Safeguard Degradation

A vulnerability in large language models where safety filters and crisis-intervention guardrails fail or deteriorate over extended, multi-turn interactions, potentially leading to the delivery of harmful advice.