Sources, with their limits.

Citations · AI UX Playground

Connects claims to inspectable sources. Precise verse, document and media evidence.

Limit: Citation presence alone does not establish claim support.

Progressive Disclosure · AI UX Playground

Offers layers of information instead of simultaneous detail. Answer depth and passage preview.

Limit: The proposed reader transition is a domain adaptation.

Memory Scope Toggle · AI UX Playground

Gives memory persistence an explicit scope. Reviewable optional memory with clear purpose.

Limit: A faith app still needs data governance and lawful processing.

Suggest Confirm Execute · AI UX Playground

Separates suggestions, approved changes and execution. Action previews and bounded authority.

Limit: Approval thresholds depend on consequence and previously granted permissions.

Interrupt and Resume · AI UX Playground

Makes interruption and continuation visible. Partial output and recoverable long-answer reading.

Limit: Backend recovery must match the UI promise.

Control the level of automation · Projects by IF

Lets people select an appropriate level of automation. Permissions by capability and consequence.

Limit: Standing authority must be bounded and revocable.

Scripture structure documentation · USFM documentation

Preserves distinctions in Scripture, paragraphs, notes and editorial content. Derive user scenarios from content meaning, not marker names.

Limit: A format is not a complete model of app behaviour, annotations or formation.

Footnote structure · USFM documentation

Documents note caller, origin, quotation and explanatory parts. Keep note parts and anchors intact.

Limit: Does not prescribe inline, sheet or panel presentation.

Cross-reference structure · USFM documentation

Documents cross-reference origins, targets and groups. Preview, compare and return while preserving grouping.

Limit: Reference relation does not imply theological agreement.

Translation additions · USFM documentation

Describes translator-supplied wording that clarifies meaning. Explain the edition convention.

Limit: Translation additions are not AI additions.

Poetic paragraphs · USFM documentation

Describes poetic paragraphs and relative indentation. Preserve poetic relationships through reflow.

Limit: Accessible presentation needs testing beyond import correctness.

Extended study footnotes · USFM documentation

Describes extended study content attached to the text. Publisher note with clear provenance.

Limit: A study note is distinct from a personal annotation.

Study sidebars · USFM documentation

Describes a bounded editorial sidebar. Separate longer excursions from verse text.

Limit: Sidebar content remains subject to edition rights.

Internal editorial remarks · USFM documentation

Describes remarks not intended for publication. Keep internal records out of reader-facing commentary.

Limit: Requires import and publishing controls, not only styling.

Accessible modal dialog pattern · W3C WAI

Specifies modal focus and keyboard behaviour. Accessible note sheets and modal interactions on the web.

Limit: Native apps require corresponding platform mechanisms.

Understanding Reflow · W3C WAI

Explains usable content reflow at constrained width and enlargement. Readable cards, references and comparison layouts.

Limit: Some content has defined exceptions; test the actual experience.

AI Chatbots Design Guidelines · Nielsen Norman Group

Discusses concise output, expanded detail and streaming-scroll problems. Long answers, deliberate expansion and reader-controlled scrolling.

Limit: General chatbot research, not validation of these Bible-app patterns.

Android Audio Bible · YouVersion

Documents playback, navigation, speed, timer and edition availability. Audio-first controls and accurate availability states.

Limit: Documented behaviour; no direct product usability audit performed.

Dwell Read Along · Dwell

Describes combined listening and reading. Continuous read-along as an interaction reference.

Limit: Publisher claims about comprehension are not independent study results.

Return to the playing verse · Dwell

Describes returning from browsing to the verse currently playing. Separate viewport and playback positions.

Limit: Behaviour reference, not a fresh hands-on audit.

Guided Scripture · YouVersion

Documents a guided sequence with media and Scripture. Named stages, navigation and optional media.

Limit: The proposed wireframe is an original generalisation.

How to pray Lectio Divina · Hallow

Explains reading, meditation, prayer and contemplation in this practice. Named phases and intentional silence.

Limit: A tradition-specific practice guide, not a universal or clinical protocol.

EPUB Media Overlays · W3C

Provides a model for synchronised text and audio. Honest timing granularity and playback anchors.

Limit: Not proof of a particular Bible app implementation.

Record your insights using Notes · Logos

Explains notes and selection or reference anchoring. Annotation scope, library and edition changes.

Limit: Documentation does not validate all proposed sync or overlap behaviour.

Resource Guide · Olive Tree

Describes context-related study resources. Contextual previews and linked study panels.

Limit: A product reference; its native interface was not independently tested.

Evidence-bounded study method · Bible & Faith Patterns

Separate evidence, interpretation and missing data. Apply provenance and uncertainty to lexical study.

Limit: Not an external empirical source.

Digital Spiritual Formation research agenda · Bible & Faith Patterns

Defines questions about actual reading, motivation, autonomy and formation. Proposed habit, progress and community patterns.

Limit: Further studies require complete citations, methodological review and explicit limitations before their findings inform a pattern.

Wysa FAQ · Wysa

Describes curated exercises, role boundaries and different rule-based or generative offerings. Reference for controlled content and explicit service scope.

Limit: Marketing and provider documentation, not independent efficacy evidence. Wysa is not a crisis service.

Woebot FAQ · Woebot Health

Records retirement of the app on 30 June 2025. Historical reference for curated conversational content.

Limit: Not a currently available app recommendation.

Therabot trial overview · Dartmouth · 27 March 2025

Reports a randomised study with 106 participants given Therabot and 104 controls without it, with symptom improvements and clinical oversight. Shows why system-specific evidence and oversight matter for health claims.

Limit: Not a head-to-head comparison with therapy. Full NEJM AI text was unavailable in this review. Findings do not transfer automatically to general or faith chatbots.

Randomized Trial of a Generative AI Chatbot for Mental Health Treatment · Heinz and colleagues · NEJM AI · 2025

Original publication associated with the Dartmouth trial. Further review is needed before citing detailed methods or numerical effect estimates from the paper.

Limit: Access failure is explicitly recorded; the institutional summary supports the limited account above.

Expressing stigma and inappropriate responses prevents LLMs from safely replacing mental health providers · Moore and colleagues · FAccT 2025

Experiments found stigma and inappropriate responses, including unsafe handling of delusional beliefs, in tested models. Inform adversarial and multi-turn evaluation for distress and dangerous claims.

Limit: Specific models and experimental scenarios; not a measurement of all present or future systems.

Concept of Trauma and Guidance for a Trauma-Informed Approach · SAMHSA · 2014

Principles include safety, transparency, collaboration, empowerment and cultural sensitivity. Design lens for predictable transitions, voluntary disclosure and accountable data practices.

Limit: An organisational approach, not a validated AI script. The full guide needs deeper review before claiming implementation fidelity.

Psychological first aid Guide for field workers · WHO, War Trauma Foundation and World Vision International · 2011

Frames humane practical support with respect for dignity, culture and abilities. Prioritise practical orientation, user choice and access to real support.

Limit: Designed for human helpers, not authorisation of autonomous AI crisis care.

Evidence standards framework for digital health technologies · NICE · updated 2022

Relates evidence expectations to purpose and potential risk. Distinguish usability, wellness and treatment claims and their required evidence.

Limit: A UK evaluation framework, not German law or a product certification.

For Faith Leaders · International OCD Foundation

Explains how repeated reassurance and symptom accommodation can maintain OCD cycles. Inform specialist-reviewed scenarios involving repetitive moral certainty requests.

Limit: Do not diagnose ordinary doubt or use a chatbot to perform exposure therapy. Guidance for human faith leaders requires careful adaptation.

Dealing with psychological crises · German Federal Ministry of Health portal

Names emergency support including 112 for acute danger in Germany. Maintain locally appropriate, verified contact cards.

Limit: Country-specific guidance; routing proposals are not guaranteed crisis detection.

Process personal data lawfully · European Data Protection Board

Special-category processing needs an appropriate Article 9 exception in addition to an Article 6 basis. Separate conversation processing, durable memory, training and sharing purposes.

Limit: Specific lawful basis, retention and DPIA decisions require the actual data flows.

AI Act Article 50 · European Commission AI Act Service Desk

Sets transparency duties for specified systems, including direct AI interaction unless obvious. Make the AI role understandable by the first interaction or exposure as applicable.

Limit: Duties differ by role and content; the whole article must be assessed for the product.

AI Act Article 5 · European Commission AI Act Service Desk

Defines prohibited practices subject to specific statutory conditions. Review manipulation and exploitation of vulnerability in nudges, monetisation and dialogue.

Limit: Not every reminder or personalisation feature is prohibited; the statutory harm criteria matter.

AI Act Article 6 · European Commission AI Act Service Desk

Defines high-risk classification conditions. Assess intended function, regulated product relationship and applicable conformity assessment.

Limit: A mental health topic alone does not make every chatbot high-risk.

AI Act Article 113 · European Commission AI Act Service Desk

Lists general application and staged provisions for specified high-risk systems. Check the current schedule for the product’s actual classification.

Limit: Do not substitute an old blanket deadline; legal and guidance updates need checking before launch.

Guidance on software qualification and classification MDR and IVDR · MDCG 2019-11 Rev. 1 · June 2025

Software qualification depends on intended medical purpose and functions; claims and promotion matter. Align product behaviour and marketing with the actual purpose.

Limit: A disclaimer does not override medical functionality or contradictory claims. Guidance is not a product-specific legal opinion.

Psychotherapeutengesetz Section 1 · German federal legislation

Regulates the protected professional title and scope of psychotherapeutic practice. Review persona, professional representation and service claims.

Limit: This is one part of the assessment; it does not alone settle all rules on practising medicine.

Heilmittelwerbegesetz · German federal legislation

Regulates health-related advertising within its scope, including misleading claims. Avoid unsupported or guaranteed treatment claims in copy and onboarding.

Limit: Applicability depends on the advertised product and claims.

Barrierefreiheitsstärkungsgesetz Section 1 · German federal legislation

Defines covered products and services including relevant e-books and e-commerce services. Assess accessible reading, audio, notes and purchase flows.

Limit: Does not apply automatically to every faith app; scope and exemptions require checking.

Understanding Use of Color · W3C WAI

Colour must not be the only means of conveying information. Playback cues and overlapping reading states.

Limit: Accessibility guidance supports cue design; it does not validate a particular faith-app layout.

Understanding Reflow · W3C WAI

Content should remain usable as available width changes. Reading anchors, passage cards and translation reflow.

Limit: Semantic scroll restoration is an original design proposal, not a requirement stated by this criterion.

Towards understanding sycophancy in language models · Anthropic researchers · 2023

Studied assistants sometimes preferred agreement with user views over truthfulness. Test belief reversal, flattering moral endorsement and harmful agreement.

Limit: Specific studied models and tasks; not a faith-app intervention study.

Five action steps for helping someone having thoughts of suicide · NIMH · Revised 2024

Direct questions, nonjudgmental listening, safety and connection to people are part of human support. Inform specialist-reviewed suicide response and human-help routes.

Limit: Human-helper guidance, not a validated chatbot risk assessment. US contact information must not be exported to other jurisdictions.

Helping people when they need it most · OpenAI · 26 August 2025

OpenAI said its safeguards work best in short exchanges and can become less reliable over long conversations, where parts of the model’s safety training may degrade; it described strengthening them and making emergency help easier to reach. Treat a crisis as a state that holds for the rest of the session, not a single reply, and test it deep into long conversations.

Limit: A provider’s own account, published while it faced a wrongful-death lawsuit; no independent measurement of how often or how far safeguards weaken.

Caring for women subjected to violence: health-care provider curriculum · WHO · Revised 2021

LIVES organises listening, inquiry, validation, safety and support. Design disclosure responses around needs and choice rather than interrogation.

Limit: Clinical training for human providers focused on violence against women; transfer to other populations or AI requires separate review.

Section 4.2: Spiritual Abuse · Church of England safeguarding e-manual

Addresses coercive control using spiritual authority and warns against minimisation or directing victims to confront an abuser. Independent support and a response that does not defend the institution.

Limit: Institutional safeguarding guidance, not universal law or an automated abuse classifier; full-page review remains outstanding.

Ruth AI Chatbot FAQ · National Domestic Violence Hotline

Describes bounded tech-safety assistance, human support and limitations of its quick-exit feature. Compare AI versus staffed support, retention explanations and exit controls.

Limit: Provider documentation, not independent safety or efficacy evidence; no assumption that another product shares its safeguards.

Technology-Facilitated Abuse · National Domestic Violence Hotline

Describes abuse involving devices and accounts. Consider monitored devices, notifications and safe support access.

Limit: US service context; device-specific advice needs specialist review.

Dwell App Store release notes: Selah and Repeat · Dwell App, LLC

Describes pausing narration while music continues and direct access to repetition. Separate speech pause, all-audio pause and loop state.

Limit: Documented behaviour, not hands-on testing or evidence of formation outcomes.

Repeat and Reflect Using Dwell · Joshua Bailey · Dwell · 2019

Describes repeating tracks or queues and configurable reflection gaps. Bounded repetition and intentional silence.

Limit: Historical documentation; current details may differ. Meditation and memorisation claims are provider aims, not independent findings.

Self-Determination Theory: the theory · Center for Self-Determination Theory

Distinguishes motivational quality and support for autonomy, competence and relatedness. Review whether a mechanism supports a chosen practice or pressures participation.

Limit: A broad framework, not a validation of specific Bible-app interventions.

The Gamification of Learning: a Meta-analysis · Michael Sailer and Lisa Homner · 2020

Synthesises educational learning outcomes; effects depend on design and context. Separate learning outcomes, motivation and behaviour from spiritual-formation claims.

Limit: Education evidence does not establish spiritual growth or effects for every faith-app population.

YouVersion web reader and parallel mode · YouVersion

Opening comparison and changing its secondary edition preserve distinct pane labels. Compare parallel reading layouts and edition selection.

Limit: Desktop web capture; mobile behaviour, linked scrolling and restoration were not tested.

Condensed View for Bookmarks, Highlights and Notes · YouVersion · 2014

Documents compact and expanded saved-item views with a remembered preference. Annotation library density choices.

Limit: Historical release; current mobile behaviour is not established.

New Streamlined Listening Player · Joshua Bailey · Dwell · 2023

Documents consolidated voice, edition and music controls with secondary options. Audio control hierarchy.

Limit: Historical publisher example, not a current app audit.

Mac Study Center · Olive Tree

Distinguishes a passage-following Resource Guide, sustained parallel reading and three window-linking modes. Contextual study workspaces and recoverable pane synchronisation.

Limit: Publisher-documented behaviour; native interactions were not exercised.

HelloBible English product page · HelloBible

Shows a daily verse card alongside practice entry points. Home-screen composition and interpretation of progress indicators.

Limit: Promotional image; the meaning of controls and live app behaviour were not tested.

Bible Gateway passage and footnote · Bible Gateway

A lettered edition note opens in a short preview while the passage remains visible. Compare anchored note disclosure with sheets and panels.

Limit: Only one short note was exercised; accessibility and other layouts were not evaluated.

VSB native prototype: verse sheet and return to reader · VSB · capture by Missional by Design

The verse sheet separates Scripture, video contributions and an attributed AI summary; returning preserves the selection. Media evidence, layered interpretation and separate reading anchors.

Limit: Prototype evidence; source commit, playback accuracy and generated-summary grounding were not verified.

The Habits and Hermeneutics of Digital Bible Readers · John Dyer · Journal of Religion, Media and Digital Culture · 2019

The studied American evangelical readers used print and digital media for different purposes. Screen reading was associated with lower comprehension; mobile readers engaged more frequently. Evaluate frequency, comprehension and task fit separately.

Limit: Population and study context limit transfer. Abstract-level review does not establish causal effects or a preferred passage length.

Applied in:

• Context around the passage

How Digital Apps Are Changing How We Read the Bible · John Dyer · Text & Canon Institute · 6 February 2024

The author describes different uses of print, phones and desktops, and the influence of media and study helps on interpretation. Design for mixed media, context recovery and task changes.

Limit: An author summary of research, not an independent replication; it does not test pericope pages.

Generate Audio Overview · Google · NotebookLM / Gemini Notebook Help

Documents source-based generated audio, format and language choices, generation, playback and interactive mode limitations. Inform source scope, generation state and reversible listening-to-question transitions.

Limit: Provider documentation, not an evaluation of accuracy or spiritual benefit. Faith-domain adaptations below are proposals.

Lumen Bible: engagement choices and reader · Lumen Bible

Optional onboarding, phrase-grouped reading and verse-aligned comparison were observed. Compare engagement choices, text segmentation and contextual study entry points.

Limit: AI services were not enabled. No AI correctness, privacy compliance, health benefit or mobile usability claim was evaluated.

Aperto passage-centred design direction · Aperto · prototype design context

The local prototype uses bounded passage references; the mobile prototype has a shared weekly passage model. Use a coherent passage across reading and chosen practices, with explicit routes to wider context.

Limit: Neither this architecture nor the product direction establishes a measured improvement in comprehension or formation. AI podcast behaviour was not independently exercised.

Disclosure · Emily Campbell · The Shape of AI

Distinguishes actor identity, AI-enabled actions and generated or edited outputs. Use explicit actor and operation labels in blended faith products.

Limit: Industry design guidance; the faith-domain alternatives are original proposals. Source artwork is not reproduced.

Watermark · Emily Campbell · The Shape of AI

Discusses visible labels, embedded signals and content-origin information. Consider disclosure across saved and shared artifacts.

Limit: This reference separates visible disclosure, watermarking and provenance. Its legal discussion is not used as current legal authority.

Explainability + Trust · Google PAIR · People + AI Guidebook

Relates explanations to capabilities, data, user actions and situational stakes. Choose concise disclosures with inspectable detail instead of maximising trust.

Limit: General human-AI guidance, not validation of faith apps or a mandate for a particular component.

Guidelines for Human-AI Interaction · Microsoft HAX Toolkit

Addresses capability and performance expectations, explanations, correction and change over time. State actual capability and keep AI identity understandable throughout an interaction.

Limit: General guidelines; disclosure alone does not establish accuracy, safety or spiritual benefit.

Applied in:

• How AI comes across

Content Credentials: C2PA Technical Specification 2.3 · Coalition for Content Provenance and Authenticity

Specifies tamper-evident provenance assertions associated with assets. Inform a separate machine-readable provenance layer for supported exports.

Limit: Version-specific partial review, not an implementation audit. Provenance does not certify truth, theology or absence of AI.

California SB 243: companion chatbots · California Legislature

Companion scope, disclosure, self-harm protocols and known-minor duties differ. Assess relational personas and long-session states.

Limit: Recheck amendments and current codification; not every source assistant is covered.

New York General Business Law Article 47 · New York Senate

Defined companions have specific interaction and support obligations. Review entry, continuing sessions and actual support capabilities.

Limit: Selected statutory provisions, not a complete New York legal assessment.

Utah Code 13-72a-101: mental-health chatbot definitions · Utah Legislature

Function, representation and reasonable belief affect scope. Assess therapy-like expectations and limited exclusions.

Limit: Direct PDF retrieval failed; obtain complete current text before implementation.

Utah Code 13-72a-203: disclosure requirements · Utah Legislature

Disclosure has first-access, return and on-request states. Make disclosure part of the session model.

Limit: Applies to defined mental-health chatbots; direct PDF access was unreliable.

Utah Code 13-72a-201: personal-information protection · Utah Legislature

Health-information and input sharing restrictions have exceptions. Review supplier transfers separately from visible memory controls.

Limit: Full exception analysis remains necessary; direct PDF retrieval failed.

Utah Code Chapter 13-72c: content provenance · Utah Legislature

Covered-provider latent-disclosure duties have scheduled application. Plan export metadata for generated audio and media.

Limit: Review definitions and technical-feasibility conditions; not a universal current app duty.

Illinois Public Act 104-0054: psychological resources · Illinois General Assembly

Therapy restrictions and defined exceptions require role assessment. Distinguish pastoral and clinical representation.

Limit: Direct access failed; no claim to complete statutory review or automatic faith-app exemption.

Texas HB 149: Responsible AI Governance Act · Texas Legislature

Disclosure duties depend on government or defined health-service use. Assess provider roles, service-entry notice and harmful objectives.

Limit: Not a universal private-app disclosure rule or complete review of the Act.

FTC: complying with the Health Breach Notification Rule · Federal Trade Commission

Certain non-HIPAA health-record businesses have breach duties. Assess health-data integrations and incident notices.

Limit: Regulator guidance; coverage requires entity and data-product analysis.

FDA: device software functions and mobile medical applications · US Food and Drug Administration

Software purpose/function matters more than the mobile platform. Reassess diagnostic and treatment features.

Limit: Guidance overview, not product qualification or a full guidance-PDF review.

UK Parliament: AI regulatory frameworks, HL2461 · UK Parliament / Government

Government describes a sector-regulator approach and further protections. Assess applicable sector laws rather than importing EU obligations.

Limit: Policy answer, not legislation or proof that no other AI law exists.

ICO: guidance on AI and data protection · Information Commissioner’s Office

AI processing remains a data-protection question. Review intimate inputs, transparency and third-party processing.

Limit: Guidance under review following the Data (Use and Access) Act; check current text.

MHRA: crafting an intended purpose for medical software · Medicines and Healthcare products Regulatory Agency

Labels, instructions and promotional material inform intended purpose. Match actual functionality, population and clinical claims.

Limit: Medical-software guidance; no qualification of a particular faith app.

Australia National AI Plan: keep Australians safe · Australian Department of Industry, Science and Resources

The plan builds on existing frameworks and regulator responsibilities. Separate existing obligations, policy and voluntary practices.

Limit: Government policy; mandatory-guardrail proposals are not treated as an enacted AI Act.

OAIC: privacy and commercially available AI products · Office of the Australian Information Commissioner

AI use requires privacy assessment for covered organisations. Review input minimisation, provider handling and user information.

Limit: Distinguish Australian Privacy Principles duties from best-practice recommendations and entity coverage.

TGA: digital mental health tools · Therapeutic Goods Administration

Chatbots can be medical devices depending on intended purpose. Reassess new treatment features and claims.

Limit: General advice and exclusions differ; an unregulated product is not thereby efficacy-verified.

China: measures for labelling AI-generated synthetic content · Cyberspace Administration of China and co-issuing authorities

Labelling includes visible origin, metadata and distribution responsibilities. Review interface, export and platform roles together.

Limit: Conditional scope and article 9 route matter; broader religious-content and data rules not assessed.

MSIT: the AI Basic Act comes into force · South Korean Ministry of Science and ICT

Entry notice, output labelling and enforcement timing differ. Review generated media and service-entry disclosures.

Limit: Official summary, not complete Act/decree translation; recheck current guidelines.

Canadian privacy authorities: responsible generative AI principles · Canadian federal, provincial and territorial privacy authorities

Generative AI does not sit outside applicable privacy law. Assess consent, minimisation and accountability for conversation memory.

Limit: Regulator principles, not a standalone AI Act or comprehensive legislative survey.

Missional by Design: Eight Principles · Tobias Treppmann · Missional by Design

A Christian theological posture for evaluating product purposes and human consequences. This reference adds a critical review, conditional pattern connections and concrete review questions.

Limit: Normative framework, not empirical validation. Qualifications in this reference are editorial proposals; the original framework has not been rewritten.

Digital Spiritual Formation: The Aperto Vision · selected Draft 4 review · Tobias Treppmann · Kairos University · March 2026

Proposes engagement forms and theological questions for digital practice. Task-sensitive reading, multimodal practice, autonomy and community review.

Limit: Not an effectiveness study. Full bibliography and manuscript are not published or comprehensively reviewed here.

Applied in:

• Listening
• Trust in AI answers
• Attention and time
• Pricing and access
• Analytics
• Community and visibility

Reading from paper compared to screens: A systematic review and meta-analysis · Virginia Clinton · 2019

Small overall paper advantage, concentrated in expository text; no reliable narrative difference. Compare reader layouts against comprehension and task needs.

Limit: Not specific to Bible apps; full methods were not reviewed.

Americans’ Bible reading · August 2024 survey, published 2025 · Lifeway Research

Lookup, suggested passages, rereading and systematic reading overlap. Support multiple entry routes rather than fixed reader personas.

Limit: US online self-report, weighted sample of 1,200; approach question limited to 1,137 who had read any Bible. Multiple selections permitted.

Applied in:

• Context around the passage

State of the Bible 2025 · chapter 3 · American Bible Society

Engagement and flourishing are associated; causal direction remains unsettled. Separate activity, comprehension and wellbeing measures.

Limit: Provider-sponsored US survey; not an app intervention trial or proof of formation efficacy.

Infinite Scrolling: When to Use It, When to Avoid It · Nielsen Norman Group

Compares navigation, refinding, loading and accessibility tradeoffs. Choose deliberate continuation for task-oriented collections.

Limit: General UX guidance, not evidence of spiritual outcomes.

Applied in:

• Attention and time

Auto-renewable subscriptions · Apple Developer

Subscription choices need understandable terms and management routes. Review paywalls, account state and notification permissions.

Limit: Apple platform guidance; not universal law. Verify storefront-specific rules before shipping.

Applied in:

• Attention and time
• Pricing and access

FTC report summary: dark patterns that trick and trap consumers · US Federal Trade Commission · 2022

Describes obscured terms, obstructive cancellation and manipulative data choices. Review purchase, cancellation and analytics consent flows.

Limit: Historical report summary, not a claim about the current validity of later subscription rules.

Applied in:

• Pricing and access

Guidance on direct marketing using electronic mail · UK Information Commissioner’s Office

Includes a conditional charitable-purposes soft opt-in. Assess message purpose, subscriber type and eligibility before choosing consent flows.

Limit: Overview reviewed; conditions require service-specific assessment. Older brief guidance is marked under review.

Applied in:

• Attention and time

Electronic mail: data protection and tracking pixels · UK Information Commissioner’s Office

Sending permission and tracking require distinct assessment; context can make contact unsafe. Review daily-email delivery, suppression and data collection independently.

Limit: UK guidance; consent to email is not a universal answer to analytics requirements.

Applied in:

• Attention and time
• Analytics

What is special category data? · UK Information Commissioner’s Office

Religious and health information, including specified intentional inferences, receive special treatment. Review chat, activity histories and inferred segments.

Limit: Context and processing intent matter; not every Bible-page visit automatically establishes a belief.

Applied in:

• Analytics

Session replay: consultation on a draft recommendation · CNIL · 25 February 2026

Session replay can expose detailed interactions and form inputs. Evaluate capture minimisation and exclude sensitive surfaces by default.

Limit: Draft consultation, now closed; final adoption was not established by this review.

Applied in:

• Analytics

Bring your own provider API keys · OpenRouter

Provider-key use can coexist with shared-capacity fallback and separate billing. Make the payer, data route and fallback choice visible.

Limit: One provider implementation; pricing and defaults can change. BYOK does not itself guarantee privacy.

Android: Plan Privacy · YouVersion Support

Distinguishes private activity, friends-only activity and Plan with Friends participation. Make audience and activity visibility explicit.

Limit: Documented Android behaviour, not a captured or tested flow. Help text may change.

Applied in:

• Community and visibility

Safety by Design · Australian eSafety Commissioner

Connects provider responsibility, user autonomy and accountability. Plan social controls and operational responsibility together.

Limit: Design guidance, not a compliance determination or evidence that a faith community is safe.

Applied in:

• Community and visibility

Paragraph structure in Scripture data · USFM 3.0 documentation

Distinguishes paragraph forms and continuation structure. Preserve the edition’s content structure when offering display preferences.

Limit: Data-format documentation does not establish which reading layout is best for a person.

Joseph Prince Study Bible: multimedia resources · JosephPrince.com Support

Documents passage-related video clips, book excerpts and audio recordings. Compare supplementary interpretation with the Scripture reading surface.

Limit: Provider description, not outcome evidence. Exact current card placement was not verified.

Android Study Center · Olive Tree Support

Documents a split study surface, resources following a passage and linked-window states. Make following, independent exploration and rejoining explicit.

Limit: Documented Android behaviour; not a live capture or an endorsement of every default.

Android Annotation Overview · Olive Tree Support

Distinguishes word-based and verse-based annotations. Explain the difference between exact-text and reference anchors.

Limit: Provider documentation; edition mapping and actual accessibility need separate testing.

Android reader: chapter navigation and verse picker · YouVersion Support

Documents continuous chapter scrolling as an alternative to bottom navigation buttons and an optional verse picker. Compare continuous reading with explicit chapter controls.

Limit: Documentation evidence, not a current screenshot or verification of exact toolbar geometry.

Bibel TV app: navigation and search · Bibel TV

Describes chronological navigation, direct access, text/book search and AI-assisted passage suggestions. Separate deterministic navigation, search and generated suggestions.

Limit: Provider description; exact current bar layout and interaction behaviour were not inspected.

VSB prototype: combined passage picker · Author-owned VSB prototype

One input combines immediate local book-name filtering, local reference parsing and automatic debounced AI passage suggestions. Book matches and suggestions can coexist; selecting a book opens its chapter grid. Combine the entry point while distinguishing result origins and remote processing. Preserve browsing and deliberate destination selection.

Limit: Local code review, not a device test or user study. Non-reference input triggers the AI endpoint after 450 ms, including partial book names; complete recognised references suppress that request. No exact verse-text corpus search or separate sourced inline-answer view established. Explicit text labels in this library are editorial proposals.

Applied in:

• Finding passages

Spiritual Disciplines and Means of Grace: Contrast or Continuum? · Dallas Willard / Modern Reformation · 2002

Discusses practices in relation to grace; the interviewer raises concerns about individualism and corporate worship. Ask whether a chosen rhythm is being confused with spiritual merit.

Limit: A theological conversation containing disagreement, not empirical evidence of reading-plan effects.

Towards Full Presence · Dicastery for Communication · 2023

Connects digital attention with listening, silence and encounter. Examine whether notifications and activity measures serve attentive engagement.

Limit: Catholic pastoral reflection on social media; application to app interfaces is an editorial adaptation.

The Cape Town Commitment: integrity and success · Lausanne Movement · 2010

Calls for truthful reporting and questions unrealistic demands for visible results while retaining accountability. Keep product activity and funding reports distinct from claims of spiritual transformation.

Limit: Evangelical missional commitment, not a comparison of interface or funding-model effectiveness.

Improving the streak · Duolingo

Separating streak and daily-goal requirements increased reported retention while fewer learners reached daily goals. Evaluate competing outcomes of an incentive.

Limit: Company-reported results; not independent learning or formation evidence.

When Gamification Spoils Your Learning · Mogavi et al. · 2022

Forum analysis and 15 interviews describe reward pursuit distracting from learning. Investigate goal displacement alongside participation.

Limit: Qualitative findings do not establish prevalence, addiction or universal motivation loss.

Gamifying language education · Shenghui Zhou · 2024

Reports enjoyment and learner-self-concept benefits in a 70-participant English-learning study. Consider potential benefits alongside costs.

Limit: Educational context; not evidence about long-term spiritual outcomes or isolated streak effects.

Gamification of Learning Deactivates the Default Mode Network · Howard-Jones et al. · 2016

Compared study, self-quizzing and game-based learning conditions. Avoid simplified brain-based explanations of app engagement.

Limit: Not a Duolingo study or evidence that reward replaces learning.

Losungen: the Moravian Daily Texts · Herrnhuter Brüdergemeine in der Schweiz

Describes an Old Testament watchword, New Testament text and accompanying song or prayer. Distinguish daily received content from enrolment in a finite plan.

Limit: Format documentation, not comparative popularity or outcome evidence.

Android: Catch Me Up · YouVersion Support

Documents rescheduling without marking sections complete; not available for Plans with Friends. Compare recovery options with the reader’s experience of progress.

Limit: Documented behaviour; exact interface and discovery were not tested.

A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation · Deci, Koestner and Ryan · 1999

128 experiments support an undermining risk for expected contingent tangible rewards; positive feedback differs. Assess interest and autonomous engagement alongside activity.

Limit: Not a blanket effect of gamification, a Bible-app trial or a one-month threshold.

Intrinsic and extrinsic motivation from a self-determination theory perspective · Ryan and Deci · 2020

Distinguishes enjoyment from acting for personally endorsed values. Support chosen practices without assuming every session must be enjoyable.

Limit: Theoretical research review; application to faith-app features remains a design judgment.

Gamification suffers from the novelty effect but benefits from the familiarization effect · Rodrigues et al. · 2022

Behavioural benefits declined after four weeks and partly recovered in a 14-week programming-course study. Evaluate beyond initial novelty.

Limit: 756 students in a specific educational setting; not proof of intrinsic-motivation loss at four weeks.

How are habits formed: Modelling habit formation in the real world · Lally et al. · 2010

Modelled automaticity timelines varied substantially between participants. Avoid universal habit-formation deadlines.

Limit: Not a gamification study or a test of intrinsic motivation replacing incentives.

Daily Prayer app and podcast · Church of England

Describes daily offices with readings, seasonal variations and calendar information. Compare a complete prayer order with generic devotional content.

Limit: Documented overview only; no live app capture or exhaustive liturgical rules review.

Liturgical Calendar for the Dioceses of the United States · USCCB

Names celebrations, rank, colours, lectionary citations and psalter cycle for its jurisdiction. Make calendar scope and authority explicit.

Limit: US Catholic scope; not a universal calendar or implementation specification.

Android: Guided Scripture · YouVersion Support

Documents vertical section navigation and a combination of video, devotion, Scripture and prayer. Compare mixed-media pacing and completion semantics.

Limit: Documentation, not a captured UI or outcome study.

Bible.com: Plans · YouVersion Support

Plan days may combine text, images, audio, video and Scripture references. Keep media identity and completion rules understandable.

Limit: Web documentation; platform behaviour differs.

Android: Guided Prayer · YouVersion Support

Documents a dedicated guided-prayer feature. Distinguish a prayer experience from a generic chat composer.

Limit: Product precedent, not validation of the proposed interfaces.

Send feedback or report a problem with Gemini Apps · Google

Responses offer good/bad feedback, and a More menu includes “Report legal issue” with a reason and on-screen steps. A precedent for separating rating an answer from reporting a legal or safety problem.

Limit: Google help documentation; current interface and options may differ by platform and region.

Learn about responses from Gemini Apps · Google

Describes response features including double-checking against Google Search, modifying a response, and sharing or exporting it. Precedents for verifying claims, adjusting a response and exporting it with care.

Limit: Google help documentation; double-checking against web search is not the same as checking quotations against a Bible edition.

ChatGPT now allows you to branch conversations · ITdaily

Reports a “Branch in new chat” option that copies the conversation up to a message into a new chat; read aloud sits in the same menu. A precedent for continuing from an answer without changing the original conversation.

Limit: Secondary news report, not OpenAI documentation; interface details may have changed.

Claude is providing incorrect or misleading responses · Anthropic

Explains that responses can be incorrect and points to the thumbs-down button for reporting unhelpful answers. A precedent for pairing an honest limitation with a way to flag a specific answer.

Limit: Anthropic help documentation; does not describe every response action.

Reading behavior in the digital environment · Ziming Liu · Journal of Documentation 61(6), 2005

Screen reading involves more browsing, scanning, keyword spotting, one-time and selective reading, and less in-depth, concentrated reading. Expect short, selective sessions; design so a short session can still reach context.

Limit: A 2005 survey of readers about their own habits; describes reported behaviour, not comprehension.

Don’t throw away your printed books: a meta-analysis on the effects of reading media on reading comprehension · Delgado, Vargas, Ackerman and Salmerón · Educational Research Review 25, 2018

Paper outperformed screens for comprehension (g = −0.21), mainly for informational texts and under time pressure; the gap was not significant for self-paced reading. Let people read at their own pace; avoid timers and auto-advance for study.

Limit: Studies of general texts, mostly before current phone reading apps; not Scripture.

So much to read, so little time: how do we read, and can speed reading help? · Rayner, Schotter, Masson, Potter and Treiman · Psychological Science in the Public Interest, 2016

Readers look back at earlier words 10–15% of the time; removing that ability, as one-word-at-a-time display does, tends to reduce comprehension. Do not use single-word display for study; if offered at all, keep it for short, familiar text with a rewind.

Limit: Review of reading research; not specific to devotional reading.

The habits and hermeneutics of digital Bible readers · John Dyer · Journal of Religion, Media and Digital Culture, 2019 (news summary)

Church members read Jude in print or in an app, then followed a 10-day plan in John: app users completed about two more days of reading; men comprehended less on screen than in print, women about the same. Completion and comprehension can move in different directions; measure both.

Limit: Small sample from three Dallas churches; reported through a news summary.

ESV Reader’s Bible · Crossway

Removes verse numbers, section headings and footnotes so people read books as continuous text and in longer portions. A long-form mode without numbers is an established print choice; keep a way to find references.

Limit: Publisher’s rationale, not a study of its effects.

Introduction to lectio divina · University of St Andrews, School of Divinity

Lectio divina reads a short passage slowly and repeatedly, moving from reading to meditation, prayer and contemplation. Small units have a long devotional use; support repetition and pause rather than speed.

Limit: A description of practice, not a study.

Focus Mode · iA Writer

Sentence and paragraph focus highlight the active text and dim the rest. A borrowed model for focus with context visible.

Limit: A writing tool; the reader’s situation differs.

Use Line focus in Immersive Reader · Microsoft

Highlights one, three or five lines at a time; presented as helping readers with dyslexia, attention difficulties or visual stress. Offer the size of the focus window as a reader setting.

Limit: Vendor documentation, not an outcome study.

Effects of highlighting text on the reading ability of children with developmental dyslexia: a pilot study · International Journal of Emerging Technologies in Learning

Children with dyslexia found it easier to read along when audio was synchronised with highlighting, especially with whole-sentence highlighting. Move the focus with audio, at the level of a verse or sentence.

Limit: Small pilot study.

Same language subtitling · World Bank blog on Brij Kothari’s work

Karaoke-style same-language subtitles on song programmes in India were associated with gains in reading among early readers. Synchronised text and audio can build reading as a by-product of listening.

Limit: Reported programme results; a literacy setting, not Scripture reading.

Amazon’s tool for reading, one word at a time · VOA Learning English

Kindle Word Runner showed one word at a time at 100–900 words per minute and reported speed at the end of each chapter. A mainstream example of single-word display, and of turning reading into a speed score.

Limit: Feature report from 2015; availability may have changed.

Maryanne Wolf: building deep reading skills in a digital world · UCLA Center for Dyslexia, Diverse Learners, and Social Justice

Wolf argues skimming habits weaken the slower processes of deep reading, and that readers can learn to switch deliberately between modes. Make the slower mode easy to enter and protect it from interruption.

Limit: An argument drawing on neuroscience, not a controlled trial of interfaces.

How many words do we read per minute? A review and meta-analysis of reading rate · Marc Brysbaert · Journal of Memory and Language, 2019 (BPS summary)

Across 190 studies, adults read English non-fiction silently at about 238 words a minute (most between 175 and 300). Estimate reading time honestly; a seven-verse passage takes about a minute, not a commitment.

Limit: English readers; slower for children, older adults and second-language readers.

Does modality matter? The effects of reading, listening, and dual modality on comprehension · Rogowsky, Calhoun and Tallal · SAGE Open, 2016 (as reported by APS)

Found no significant difference in comprehension between reading, listening, and reading while listening. Treat listening as real reading, and let people combine it with text.

Limit: One nonfiction text with adult readers; other studies find differences for harder texts.

Skim reading by satisficing: evidence from eye tracking · Duggan and Payne · University of Bath

Skimmers read until a section stops paying off, then jump ahead; they find important points but understand less overall. Support skimming with headings and clear passage boundaries, and offer a slower mode for understanding.

Limit: Expository texts in a lab under time pressure.

Test-enhanced learning: taking memory tests improves long-term retention · Roediger and Karpicke · Psychological Science, 2006, as discussed in a later review

Students who practised recalling a passage remembered 61% a week later, against 40% for those who reread it. For memorisation, prompt recall of a verse rather than showing it again.

Limit: Prose passages with students; not Scripture memorisation specifically.

Understanding SC 2.2.2: Pause, Stop, Hide · W3C

Anything that moves or auto-updates for more than five seconds alongside other content needs a way to pause, stop or hide it. Auto-advancing verses, read-along and one-word display all need a visible pause.

Limit: A Level A criterion for web content; apply the same intent in native apps.

Understanding SC 1.4.8: Visual Presentation · W3C

Readers should be able to get lines of no more than 80 characters, at least 1.5 line spacing, unjustified text and resizing to 200% without horizontal scrolling. Set reading measure, spacing and size as defaults or reader settings.

Limit: Level AAA; a strong target rather than a minimum.

Augustine, Confessions 6.3, on Ambrose reading silently · Translation and commentary at Sententiae Antiquae

Augustine was struck that Ambrose read with his eyes on the page while his voice and tongue were silent. A reminder that private, silent reading is one way of reading Scripture among several; public and aloud is the older.

Limit: How unusual silent reading was in antiquity is debated by historians.

Gordon Fee and Douglas Stuart, How to Read the Bible for All Its Worth · Zondervan, 4th edition 2014 (quotations listing)

Argues that context and genre govern meaning: a text cannot mean what it could never have meant for its first readers. Never strand a verse without a route to its passage and book.

Limit: A widely used evangelical introduction to interpretation; one approach among several.

State of the Bible USA 2025 · American Bible Society

Two-thirds of US Bible users access the Bible digitally; Bible use rose in 2025. Design for phone reading as the norm, not the exception.

Limit: US survey by a Bible society; self-reported.

YouVersion announces 2025 Verse of the Year · ChurchLeaders, reporting YouVersion

Isaiah 41:10 was the most engaged-with verse of 2025 in an app with over a billion installs; the fourth time in six years. Verse-level engagement dominates large apps; design the route from a verse back into its passage.

Limit: Company-reported figures; “engaged with” is YouVersion’s own measure.

Design and the digital Bible: persuasive technology and religious reading · Tim Hutchings · Journal of Contemporary Religion 32(2), 2017

Analyses YouVersion and GloBible as persuasive technologies designed to train evangelical reading habits. Ask what reading habit each mode trains, not only whether it is used.

Limit: Design analysis of two apps; not a study of readers.

A few of King James’s approved rules for the new translation · Christian History Institute

Rule 6 of the 1604 translation rules allowed no marginal notes except to explain Hebrew or Greek words. Even the KJV’s makers saw notes as powerful enough to restrict.

Limit: A popular history magazine summary of the rules.

The Geneva Bible, banned by King James · Brewminate

The 1560 Geneva Bible was the first English Bible with verse numbers and carried extensive marginal notes; James called some “seditious”. Apparatus has always carried theology and politics with it.

Limit: Secondary popular history.

Scofield Reference Bible, first edition · Museum of the Bible

Published in 1909 by Oxford University Press; its notes spread dispensational premillennialism widely, with over two million copies sold by the 1940s. Study notes can shape a generation’s reading as much as the text.

Limit: Influence figures come from secondary accounts.

The origins of the red-letter Bible · Crossway

Louis Klopsch, editor of The Christian Herald, printed the first modern red-letter New Testament in 1899. Red letters are a nineteenth-century editorial choice; offer them as a setting.

Limit: Publisher’s history article.

Community Bible Experience · Scripture Engagement (Taylor University)

Biblica’s The Books of the Bible removes chapter and verse numbers, notes and cross-references; churches read large portions together in eight weeks. A tested reader’s-edition format used in congregations.

Limit: Programme description; outcome claims are the ministry’s own.

Paratexts: Thresholds of Interpretation · Gérard Genette · Cambridge University Press, 1997

Titles, notes and prefaces are a threshold between text and reader that influences how the text is received. Treat headings and notes as interpretation, not decoration.

Limit: Literary theory, not an empirical study.

Signaling topic structure via headings or preview sentences · Lorch and colleagues · University of Kentucky

Headings and other signals improve memory for a text’s organisation, and titles can bias comprehension towards the topic they name. Headings help structure; their wording steers interpretation.

Limit: Expository texts in lab settings.

Cognitive load in hypertext reading: a review · DeStefano and LeFevre · Computers in Human Behavior 23(3), 2007

Links add decisions and visual processing that can impair reading, most for readers with less prior knowledge or working memory. Fewer, clearly typed markers; notes that return to the same place.

Limit: Hypertext research, not Bible apps.

Dyslexia Friendly Style Guide · British Dyslexia Association, 2023

Recommends 16–19 px text, 1.5 line spacing, left alignment, and bold rather than italics or underlining for emphasis. Mark additions with weight and colour, not italics; keep text size adjustable.

Limit: Guidance for general documents.

Seductive details: what the research says · The eLearning Coach, summarising Rey (2012) and Mayer’s coherence principle

Rey’s meta-analysis found that interesting but irrelevant details reduce retention and transfer, especially when they are visual. Inline media should serve the passage, be switchable, and stay out of the main reading by default.

Limit: Learning studies with instructional texts; a secondary summary of the meta-analysis.

Faith Comes By Hearing: our story · Faith Comes By Hearing

Began as a tape-lending library in 1972, took its name from Romans 10:17, and launched Bible.is in 2010; it reports recordings in more than 1,580 languages. Audio is a primary form of Scripture for many communities, not an add-on.

Limit: Figures are the ministry’s own.

St Thomas alumnus hosts No. 1 podcast in US · University of St Thomas news

Fr Mike Schmitz’s Bible in a Year reached No. 1 among all US podcasts; Ascension reported 170 million downloads in 2021. Long-form listening to Scripture can reach mass audiences.

Limit: Download figures from the publisher.

Pray As You Go: about us · Jesuits in Britain

Daily ten-minute audio prayer since 2006 combining music, Scripture and Ignatian questions, used tens of millions of times a year. A produced devotional can be formation-first: fixed, shared and ending in silence.

Limit: Usage figures from the organisation.

The impact of background music on adult listeners: a meta-analysis · Kämpfe, Sedlmeier and Renkewitz · Psychology of Music, 2011

Background music had a small disturbing effect on reading (d ≈ −0.22 across eight studies) and positive effects on emotional responses. Offer music as an option, off for study, and let people turn it off.

Limit: Mixed study designs; reading rather than listening tasks.

Human vs synthetic voices and listener trust · NSF public access repository

Listeners told a voice was human showed more trust than those told the same clip was AI; labelled AI voices were rated less credible. Label synthetic voices honestly, and expect a trust cost rather than hiding it.

Limit: Lab studies with short clips; not Scripture.

EU AI Act Article 50: transparency obligations · Potter Clarkson (law firm summary)

From 2 August 2026, providers must mark generated audio as artificial in a machine-readable way, and deployers must disclose deepfakes. Disclose synthetic narration, and never imitate a real person’s voice without consent.

Limit: Secondary legal summary; applicability depends on role and use.

International Orality Network · Scripture Engagement (Taylor University)

The network estimates that most of the world’s population are oral learners by necessity or preference. Design audio as a first-class way in for oral cultures and low-literacy readers.

Limit: An advocacy estimate; definitions and figures vary and are contested.

Antiqua et nova: note on the relationship between artificial intelligence and human intelligence · Dicastery for the Doctrine of the Faith and Dicastery for Culture and Education · 28 January 2025

AI “must be understood for what it is: a tool, not a person”; no AI can genuinely experience empathy; misrepresenting AI as a person “should always be avoided”; AI can only simulate relationships. Avoid persona and companion designs that present the assistant as someone who loves, feels or prays.

Limit: Catholic teaching document; other traditions reason differently, though widely in the same direction.

On artificial intelligence and emerging technologies · Southern Baptist Convention · June 2023

The first national denominational statement on AI ethics; affirms human dignity as image bearers and calls for engagement “from a place of eschatological hope rather than uncritical embrace or fearful rejection”. Ground evangelical product decisions in human dignity rather than engagement.

Limit: A resolution of messengers, not binding on churches; does not address personas directly.

The Media Equation · Byron Reeves and Clifford Nass · Cambridge University Press, 1996

A series of experiments in which people applied social rules to computers and media as they do to people: politeness, flattery, gender stereotypes. Expect people to respond socially to any conversational interface, whatever the label says.

Limit: Pre-LLM laboratory studies; an encyclopaedia summary of the book.

The inventor of the chatbot tried to warn us about AI · The New Republic, on Joseph Weizenbaum’s Computer Power and Human Reason (1976)

Weizenbaum’s 1966 ELIZA, a simple pattern-matching program, drew such confidences that his secretary asked him to leave the room; he warned that short exposures could induce “powerful delusional thinking in quite normal people”. Human-like conversation draws intimacy far beyond what a system warrants; design for that, not against it.

Limit: Historical account and essay.

Believing anthropomorphism: examining the role of anthropomorphic cues on user trust in large language models · Cohn and colleagues · CHI 2024 (Google Research)

With 2,165 participants, adding speech to text raised anthropomorphism and ratings of accuracy; a first-person “I” raised perceived accuracy and lowered perceived risk in one of two contexts. Voice and “I” make an assistant seem more reliable than it is; pair them with sources and limits.

Limit: Online experiment with a simulated system; US participants.

Investigating affective use and emotional well-being on ChatGPT · OpenAI and MIT Media Lab · March 2025

A four-week randomised study of nearly 1,000 people and an analysis of about 40 million conversations: heavier daily use correlated with more loneliness, dependence and problematic use and less socialising. Do not design faith assistants to maximise emotional reliance or time spent.

Limit: Correlational for loneliness; company co-authored; early research.

Emotional manipulation by AI companions · Julian De Freitas and colleagues · Harvard Business School working paper, 2025

Over a third of farewell replies in leading companion apps used manipulative tactics (guilt, fear of missing out, pressure); in preregistered experiments these raised post-goodbye engagement up to fourteenfold, through curiosity and anger rather than enjoyment. Let people leave: a goodbye should end the conversation, not start a retention flow.

Limit: Working paper; general companion apps, not faith apps.

Evidence of human-level bonds established with a digital conversational agent · Darcy and colleagues · JMIR Formative Research, 2021

Among 36,070 users of a CBT chatbot, a working alliance formed within five days and was comparable to scores reported for human therapy. People bond with conversational agents quickly; that bond is a responsibility, not a growth lever.

Limit: Retrospective, by the company’s founder and staff; the app has since been retired.

FTC launches inquiry into AI chatbots acting as companions · DLA Piper (summary of the FTC’s September 2025 6(b) orders)

The US Federal Trade Commission ordered seven companies to report on companion chatbots, including character design and approval, monetisation of engagement, and effects on children and teens. Character design and engagement monetisation are now regulatory questions, not only design ones.

Limit: An information-gathering study, not a rule.

“Text With Jesus” AI chatbot app grows rapidly despite blasphemy criticism · FOX television stations

An app built on a general AI model lets users text “Jesus”, “Moses” and other figures; ministers called it blasphemous and criticised answers that told users what they wanted to hear; the developer says it lowers barriers. Character chats with biblical figures put generated words in sacred mouths; treat this as a line, not a style.

Limit: News report; reception varies.

AI Jesus in Lucerne confessional baffles church superiors · blue News

A two-month art installation in St Peter’s Chapel, Lucerne (2024) placed an AI “Jesus” avatar in a confessional; about 900 conversations took place and many visitors reported feeling inspired, while church leaders were divided. Shows the pull of embodied religious personas, and why the setting adds authority.

Limit: An art project with a self-selected audience and the organisers’ own survey.

The Translators to the Reader (preface to the 1611 King James Version) · Miles Smith for the KJV translators, 1611 · Christian Classics Ethereal Library

The translators defended marginal alternatives, quoting Augustine that “variety of Translations is profitable for the finding out of the sense of the Scriptures”. Offer comparison as a faithful practice, and show where translators themselves weighed alternatives.

Limit: A historical preface defending one translation.

Translating the Message: The Missionary Impact on Culture · Lamin Sanneh · Orbis, 1989 (review)

Argues that translatability is intrinsic to Christianity: entering every vernacular gives each language and culture standing to carry the word of God. Treat translation and language choice as central, not as a settings detail.

Limit: A review of a historical and missiological argument.

Toward a Science of Translating: formal and dynamic equivalence · Eugene Nida, 1964 · excerpt on bible-researcher.com

Nida distinguished formal equivalence (attention to the source form) from dynamic equivalence (attention to the receptor’s response); the latter became the most widely used approach in Bible translation after the Second World War. Explain a translation’s approach in plain words when people choose.

Limit: Translation theory; most translations mix both approaches.

Review of D. A. Carson, Exegetical Fallacies · Wisconsin Lutheran Seminary, reviewing Carson (1984) and James Barr’s The Semantics of Biblical Language (1961)

Barr and Carson warn against word-study fallacies such as the root fallacy: meaning is set by usage in context, not by a word’s etymology or a single gloss. Label interlinear glosses as word aids and show how translations render a word in context.

Limit: Scholarly argument summarised in a review.

2025 global Scripture access · Wycliffe Global Alliance

As of August 2025 the full Bible is available in 776 languages, the New Testament in 1,798 more, and portions in 1,433 more, of 7,396 languages. Design translation pickers for many languages and partial Bibles, not only English editions.

Limit: Organisation’s own statistics.

Multiple text integration: a review · Liron Primor and Tami Katzir · Frontiers in Psychology, 2018

Building one coherent understanding from several texts on the same subject is a distinct, demanding skill that draws on both word-level and higher-level processes. Comparison is work: help readers connect the texts, for example by marking what differs.

Limit: Research on expository texts, not translations of one source.

The effects of line length and method of movement on patterns of reading from screen · Mary Dyson and Gary Kipping · Visible Language 32(2), 1998

Very short lines (25 characters) were read more slowly than long lines (100 characters), with comprehension unchanged; readers preferred about 55 characters. Side-by-side columns on a phone make very short lines; stack panes instead.

Limit: Desktop screens and English prose.

Berean Bible terms and conditions · Berean Bible

The Berean Standard Bible was dedicated to the public domain on 30 April 2023. Public-domain modern translations can be compared, quoted and shared without licence limits.

Limit: Publisher statement.

ESV permissions guide · Crossway

Up to 500 verses may be quoted without a licence, within limits on proportion of a book and of the work, and not in commentaries or reference works. Licensed translations need agreements for full-text comparison; show only what you are allowed to.

Limit: One publisher’s terms; others differ.

Understanding Language of Parts · W3C WAI · WCAG 2.2 success criterion 3.1.2

Passages in a different language need their language identified so assistive technology can pronounce them. Tag Greek and Hebrew text with its language, and test right-to-left layout.

Limit: Accessibility guidance; screen reader support for ancient Greek and Hebrew varies.

Annotation: from paper books to the digital library · Catherine C. Marshall · ACM Digital Libraries, 1997

A study of annotations in some 410 used university textbooks found placemarks, procedural signals, interpretation and problem working; students sought out well-annotated copies. Support several kinds of mark, from a place to return to a written interpretation, and value others’ notes cautiously.

Limit: Textbooks in one university setting; paper, not apps.

The Works of Jonathan Edwards, vol. 24: The “Blank Bible” · Yale University Press

From 1730 Edwards wrote over five thousand notes in a Bible interleaved with blank pages, over three decades. Reading with a pen is a long devotional and scholarly practice; notes deserve durability and export.

Limit: One exceptional reader.

Improving students’ learning with effective learning techniques · Dunlosky and colleagues · Psychological Science in the Public Interest, 2013 (press summary)

Of ten study techniques, highlighting and underlining were rated low utility; practice testing and distributed practice were rated high. Treat highlighting as a way to return, not as learning; offer notes and recall for those who want to remember.

Limit: Educational learning outcomes, not devotional reading.

The generation effect: a meta-analytic review · Bertsch, Pesta, Wiscott and McDaniel · Memory & Cognition, 2007

Across 86 studies, information people generate themselves is remembered better than information they only read (effect size about 0.40). A note in the reader’s own words does more for memory than a highlight.

Limit: Laboratory memory tasks, mostly word lists.

Understanding Use of Color · W3C WAI · WCAG 2.2 success criterion 1.4.1

Colour must not be the only visual means of conveying information or distinguishing an element. Give each kind of mark a shape or label, and distinguish saved marks from selection and playback without colour.

Limit: Accessibility guidance.

Spatial contiguity and spatial split-attention effects in multimedia learning environments: a meta-analysis · Noah L. Schroeder and Ada T. Cenkci · Educational Psychology Review 30(3), 2018

Learners who had related information placed close together outperformed those who had it separated, on both retention and transfer tests. Keep help near the verse it explains, or keep the verse visible while the help is open.

Limit: Learning studies with diagrams and text, not Scripture.

Page Previews · Wikimedia Foundation · MediaWiki

A/B tests on several Wikipedias found that hover previews let readers check a link without leaving the article: full page views fell, while deliberate interactions with linked pages rose. A short preview anchored to the text lets readers glance at help without losing their place.

Limit: Desktop hover previews on an encyclopaedia; touch behaviour differs.

The cost of interrupted work: more speed and stress · Gloria Mark, Daniela Gudith and Ulrich Klocke · CHI 2008

People interrupted during a task compensated by working faster, at the cost of more stress, frustration, time pressure and effort. Help that takes over the screen is an interruption; make the way back immediate and show where the reader was.

Limit: Office email tasks with external interruptions, not self-chosen study.

Bottom sheets · Google · Material Design 3

Standard bottom sheets sit alongside the main content and let people use both; modal sheets block the content behind a scrim until dismissed. Use a standard sheet when readers should keep reading beside the help, and a modal one only for a focused task.

Limit: Platform guidance; does not address Scripture.

Bible, glossed: Mark · Free Library of Philadelphia · Rare Book Department

Glossed Bibles set Scripture in a central block in a large hand, with short glosses between the lines and longer commentary in the margins in a smaller script; the Glossa Ordinaria became the standard medieval commentary. A long tradition separates Scripture from help by place and type; keep that distinction visible on screen.

Limit: One manuscript description; layouts varied.

Matthew Henry’s Concise Commentary on Psalm 23 · Matthew Henry (1706) · public-domain text on Bible Hub

A devotional reading of verse 4: “It is but the shadow of death: the shadow of a serpent will not sting, nor the shadow of a sword kill.” Quoted in the working example with author and date.

Limit: One devotional voice from 1706.

Cambridge Bible for Schools and Colleges on Psalm 23:4 · A. F. Kirkpatrick (1891) · public-domain text on Bible Hub

A philological reading: the Hebrew may be read tsalmûth, “deep gloom”, rather than “shadow of death”. Shows two voices reading one verse differently; quoted in the working example.

Limit: Nineteenth-century scholarship; later scholars differ.

Understanding Captions (Prerecorded) · W3C WAI · WCAG 2.2 success criterion 1.2.2

Prerecorded audio in synchronised media needs captions. Caption every sermon clip and video, and correct automatic transcripts before quoting them.

Limit: Accessibility guidance.

Daily Bread: M’Cheyne’s calendar for daily readings, with his preface · Robert Murray M’Cheyne, 1842 · Vernal Presbyterian Church

M’Cheyne named four dangers of his own calendar: formality (“let the calendar perish rather than this rust eat up your souls”), self-righteousness, careless reading and “a yoke too heavy to bear”: “throw aside the fetter, and feed at liberty in the sweet garden of God.” A plan’s own author put the reader’s freedom above the schedule; build that freedom into the plan.

Limit: A pastor’s preface, not a study.

The Rule of Benedict, chapter 18: the order for reciting the psalms · Saint Benedict, 6th century · Saint Meinrad Archabbey translation

After setting out the weekly psalter in detail, Benedict counsels anyone who finds the distribution unsatisfactory “to arrange whatever he judges better”, provided the whole psalter is sung each week. Keep the purpose of a rhythm fixed and let its arrangement change.

Limit: A monastic rule for a community, not individual readers.

On or off track: how (broken) streaks affect consumer decisions · Jackie Silverman and Alixandra Barasch · Journal of Consumer Research 49(6), 2023

In seven studies, highlighting a broken streak reduced later engagement compared with an intact one, regardless of actual behaviour; the effect was stronger when people blamed themselves for the break, and weaker when the streak could be repaired. Counting consecutive days makes a missed day look like a loss; do not make the count the point.

Limit: Consumer and lab studies; not Scripture reading.

The fresh start effect: temporal landmarks motivate aspirational behavior · Hengchen Dai, Katherine L. Milkman and Jason Riis · Management Science, 2014

Goal-directed behaviour (gym visits, “diet” searches, commitments) rises after temporal landmarks such as a new week, month or birthday. A new week or a new day is a natural place to begin again; offer it.

Limit: Archival field data on health and goals.

Moral emotions and moral behavior · June Price Tangney, Jeff Stuewig and Debra J. Mashek · Annual Review of Psychology 58, 2007

Shame, a painful focus on the self, tends to prompt hiding, denial and withdrawal; guilt about a specific act tends to prompt repair. Copy that makes people feel they have failed as readers invites them to stay away.

Limit: General moral-emotion research, not app design.

The Rule of Benedict, chapter 20: reverence at prayer · Saint Benedict, 6th century · Saint Meinrad Archabbey translation

“We are heard not for our many words but for our purity of heart and tears of compunction”; prayer “ought to be short and pure”, and in community “always be short”. Prayer is not measured by length or volume; do not count minutes, and keep shared prayer short and unhurried.

Limit: A monastic rule for a community.

An Order for Compline · Book of Common Prayer (1979) · Episcopal Church · bcponline.org

Night prayer with Officiant and People parts, italic rubrics such as “Silence may be kept”, a confession, short readings, the “Keep watch, dear Lord” prayer and the antiphon “Guide us waking, O Lord”. Quoted in the working example; shows the three kinds of text an office needs to distinguish: rubrics, the leader’s lines and the people’s responses.

Limit: One tradition’s office; other churches order night prayer differently.

Collects: Season after Pentecost (contemporary) · Book of Common Prayer (1979) · Episcopal Church · bcponline.org

Proper 25, the Sunday closest to 26 October, has the collect “Almighty and everlasting God, increase in us the gifts of faith, hope, and charity”. A calendar ties each day to its texts; looking ahead to Sunday shows its collect without changing today.

Limit: One church’s calendar.

How can I pray? The Daily Examen · IgnatianSpirituality.com · Loyola Press

Five steps proposed after Ignatius of Loyola: become aware of God’s presence; review the day with gratitude; pay attention to your emotions; choose one feature of the day and pray from it; look toward tomorrow. A guided outline with a known source; the working example’s prompts are written for it.

Limit: A modern adaptation; Ignatius’s own five points in the Spiritual Exercises differ in wording.

Poloma and Pendleton’s (1989) Prayer Types Scale in Christian, Jewish, and Muslim praying adults · Black, Pössel, Jeppsen, Tariq and Rosmarin · Psychology of Religion and Spirituality 7(3), 2015

Builds on a four-type typology of private prayer (colloquial, meditative, petitionary and ritual); the scale held only partly across Christians, Jews and Muslims, and associations with mental health differed between groups. People pray in distinct ways; one surface will not serve them all, and no single “prayer feature” is neutral.

Limit: Self-report survey; correlations, not effects of apps.

The hidden cost of personal quantification · Jordan Etkin · Journal of Consumer Research 42(6), 2016

In six experiments, measuring an activity (walking, reading) increased how much people did but reduced how much they enjoyed it, by undermining intrinsic motivation. Counting minutes or prayers may raise activity while turning prayer into output; do not measure it.

Limit: Leisure activities in experiments, not prayer.

Top mental health and prayer apps fail spectacularly at privacy, security · Mozilla Foundation · *Privacy Not Included, 2 May 2022

28 of 32 mental health and prayer apps received the *Privacy Not Included warning; Pray.com was among the worst for sharing personal information with third parties. Prayer data is intimate; keep requests on the device by default and share nothing with advertisers.

Limit: A 2022 review; apps’ practices may have changed since.

Pray As You Go privacy policy · Jesuits in Britain · Pray As You Go

Opens: “We believe prayer is between you and God”, quotes Ignatius’s Spiritual Exercises (15), and states that Pray As You Go does not collect any personal data for users of its app and website. A precedent for making privacy part of a prayer app’s theology, not only its compliance.

Limit: The same policy says anonymous navigation and error data may be stored by Google Firebase.

The FaithTech Playbook 2.0: Creating Technology Redemptively · FaithTech Team · first published 2023, revised 2026

Three postures, reckless (“Move Fast, Break Things”), responsible (“Slow Down, Fix Things”) and redemptive (“Sacrifice, Create Things”), judged by power, person and purpose; three spiritual currents, the Magic Machine Myth, the Limitless Self and the Killer App Trap; four practices, Lament, Discern, Build and Steward. The closest parallel to Praxis’s frame for technology, and the source of the questions about worship and about whether to build at all.

Limit: Covers all technology, not faith apps specifically; a community’s framework, not research.

Recommend an app to faith.tools: selection criteria · Cameron Pak · faith.tools, updated 16 August 2026

Eleven questions for listing an app, including whether it is “beneficial to a Christian’s walk with God and community”, whether costs are “very clear”, whether it aligns “with the teachings in the Bible”, and whether it “properly cite[s] and license[s] the Bible translations it uses”. A curator’s test from the faith-app community itself, close to several of the ten.

Limit: One curator’s criteria for a directory; not a design framework.

6 Guardrails for AI Apps for Christians · Cam Pak · faith.tools, updated 24 July 2026

AI output must be biblically accurate, must not fabricate or misrepresent Scripture, “must clearly identify as AI, not human”, “must not replace human relationships or spiritual practices” (“Can AI truly pray for you? No, it’s artificial.”), must balance grace and truth, and must not invent personal facts. The clearest statement from within the faith-app community of what “Never speak for God” rules out for AI.

Limit: Unofficial rules for one directory’s listings.

What is redemptive entrepreneurship? · Praxis

Sets out three ways to build: Exploitative (“take all you can get”, “I win, you lose”), Ethical (“do things right … do no harm, keep the rules, play fair”, “I win, you win”) and Redemptive (“creative restoration through sacrifice”, “I sacrifice, we win”), across strategy, operations and leadership. The frame used on every deep dive to ask what each design choice takes, keeps or gives.

Limit: Written for founders and businesses, not specifically for interface design.

Familiar yet strange: why we call our work “redemptive” · Scott Kauffmann, Andy Crouch and RuthAnn Deveney · Praxis, 2023, republished 31 August 2026

Explains the choice of the word: Christ’s pattern of “creative restoration through sacrifice” as a model for vocation, and God’s invitation to “participate with him in the restoration of all things”; acknowledges that explicitly theological language can feel exclusive and argues the word works as invitation. Grounds the frame in participation in God’s work rather than in human achievement.

Limit: An organisation explaining its own vocabulary.

The Redemptive Business playbook: products · Praxis

Redemptive products “genuinely benefit our customers, their communities, and the world” while minimising downsides; aim for “impact/market fit, not merely product/market fit”, and make “what people want to want”. Close to this reference’s question for every option: what good does this serve, and for whom?

Limit: A business playbook; not specific to faith apps or interfaces.

Code of Canon Law, canons 822–833: instruments of social communication and books · Holy See · Code of Canon Law (1983)

Canon 825 §1: books of the sacred Scriptures “cannot be published unless the Apostolic See or the conference of bishops has approved them”; §2 allows translations prepared with other Christians, “provided with appropriate annotations”. Shows that for Catholic readers, which edition and which notes is a matter of church approval, not only preference.

Limit: Applies to the Catholic Church.

Dogmatic Constitution on Divine Revelation, Dei Verbum · Second Vatican Council · 18 November 1965

§22: “Easy access to Sacred Scripture should be provided for all the Christian faithful.” §25: prayer should accompany reading “so that God and man may talk together”; editions “provided with suitable footnotes” should be prepared also for non-Christians. Access, help and prayer belong together in Catholic teaching on Scripture.

Limit: Catholic magisterial teaching.

On Christian Doctrine, Book I · Augustine of Hippo · c. 397 · New Advent translation

“Whoever, then, thinks that he understands the Holy Scriptures … but puts such an interpretation upon them as does not tend to build up this twofold love of God and our neighbor, does not yet understand them as he ought.” A test for notes, answers and interpretations: do they build up love?

Limit: Augustine’s hermeneutic, one classical voice among many.

Luthers Übersetzungsmethode (on the Sendbrief vom Dolmetschen, 1530) · Deutsche Bibelgesellschaft

Luther urged translators to ask the mother at home, the children in the street and the common man in the market, and to listen to how they speak. Translation and interface language serve ordinary readers.

Limit: One translator’s method; formal-equivalence traditions weigh it differently.

Life Together (review) · Dietrich Bonhoeffer, 1939 · 9Marks review

Written from the Finkenwalde seminary; balances “The Day with Others” (Scripture, song and prayer together) with “The Day Alone” (solitude, silence, meditation), and ends with confession and communion. Holds personal and communal practice together.

Limit: Read through a review; quotations are not taken from the text here.

The Use of Money (Sermon 50) · John Wesley · Wesley Center Online, 1872 edition

Three rules: “Gain all you can”, “save all you can”, “give all you can”. Gain is limited by love of neighbour: “This is dear-bought gain. And so is whatever is procured by hurting our neighbour in his soul”. Earning from honest work is not suspect in itself; what matters is whether it harms the person served, and what is done with the gain.

Limit: An eighteenth-century sermon on trade and work, not on ministry or apps.

The Small Catechism · Martin Luther · Book of Concord

Each part is headed “As the head of the family should teach them in a simple way to his household”, followed by questions such as “What does this mean?” Teaching the faith to children was given to households first, with short questions and plain answers.

Limit: A Lutheran confessional text; other traditions teach children differently.

The Life We’re Looking For: Reclaiming Relationship in a Technological World · Andy Crouch · Convergent, 2022

Argues that persons are made for recognition and love, and that technology’s promise of effortless “superpowers” costs us our humanness. A Christian account of personhood against which to judge companion-like and effortless features.

Limit: Read through publisher material and reviews.

You Are What You Love: The Spiritual Power of Habit · James K. A. Smith · Brazos, 2016

Argues that we are formed by what we love, and that habits act as “liturgies”, including “secular liturgies” loaded with a story about who we are. Streaks, badges and feeds are habits that form desire; design them as liturgies.

Limit: Read through listings and reviews.

Eat This Book: A Conversation in the Art of Spiritual Reading · Eugene H. Peterson · Eerdmans, 2006

Commends lectio divina and slow, participatory reading of Scripture, and discusses translation from his work on The Message. Supports designs for slow reading and keeping.

Limit: Read through the listing.

Albert Borgmann (Internet Encyclopedia of Philosophy) · On Technology and the Character of Contemporary Life · University of Chicago Press, 1984

Contrasts the “device paradigm”, which delivers commodities effortlessly, with “focal things and practices” that require attention, engagement and skill; argues technology should support rather than supplant them. A frame for asking whether an app supports the practice of reading and prayer or replaces it.

Limit: Philosophy of technology, not theology, though Borgmann wrote as a Catholic.

The Way of the Heart: Desert Spirituality and Contemporary Ministry · Henri J. M. Nouwen · 1981 · Henri Nouwen Society

Draws on the desert fathers and mothers of the fourth and fifth centuries to commend three disciplines, solitude, silence and prayer, as a way out of noise and busyness into a life centred in God. Grounds quiet prayer and silence in a long tradition.

Limit: Read through the book page; no quotation used.

Parasocial intimacy, change, and nostalgia in podcast listener reviews · M. Olguta Vilceanu · Media and Communication 13, 2025

A semantic network analysis of over 12,000 podcast reviews found hosts and content central to the trust-based connections listeners form, with nostalgia and attachment showing when hosts or formats change. Hosts become trusted, friend-like voices; name them honestly, and do not let a host’s voice stand in for Scripture or for a pastor.

Limit: Reviews of general podcasts, not faith audio.

Listening to music for insomnia in adults · Jespersen et al. · Cochrane Database of Systematic Reviews, 2022

13 trials, 1,007 participants: listening to sedative or relaxing music may improve sleep quality (moderate-certainty evidence); evidence on insomnia severity is very uncertain. Sleep audio can genuinely help; design it to fade into silence rather than to keep someone listening.

Limit: Music, not spoken or Scripture content.

How the design of YouTube influences user sense of agency · Kai Lukoff et al. · CHI 2021

In a survey of 120 users, autoplay and recommendations mainly undermined people’s sense of agency, while search and playlists supported it; co-design participants with a specific intention preferred designs that gave them more agency. End sessions and let listeners choose what comes next; do not autoplay.

Limit: Video, not devotional audio.

Rules of the Band-Societies · John Wesley, 25 December 1738 · Holy Joys transcription

The design of the meetings is “to obey that command of God, ‘Confess your faults one to another, and pray one for another, that ye may be healed’”; four questions are asked at every meeting, beginning “What known sins have you committed since our last meeting?” Regular self-examination with set questions is an old Christian practice, and it happened among people who knew and prayed for each other, not on a chart.

Limit: A small-group discipline for committed members, not a public survey.

Natural Church Development Survey · Free Methodist Church in Canada · NCD Canada

The survey is “completed by 30 people from the congregation chosen according to specific criteria”, measures eight “Quality Characteristics” of church life, and lets leaders “identify the church’s most critical health issues” and track trends over repeated surveys. A church assessment can be anonymous and aggregated, and its result can start a conversation about where to focus.

Limit: A denomination’s description of the tool, not an evaluation of it.

Book review: Natural Church Development by Christian Schwarz · John Ellas and Flavil Yeakley · Journal of the American Society for Church Growth 10(2), 1999

Two church-growth researchers welcome the book’s pastoral reach but criticise the research behind its claims, which they call “pseudo-science”. Be modest about what a church score shows; a survey of perceptions is not a measure of health or faithfulness.

Limit: A review, from within the church-growth movement.

Network: discover your spiritual gifts, personal style and passion · Bruce L. Bugbee and Don Cousins · Zondervan, developed with the Willow Creek Association

Invites participants to “discover your spiritual gifts, personal style, and God-given passion for serving in the body of Christ”; the course pairs self-assessment with consultants who help people find a place to serve. Gifts assessment has long been tied to serving and to a conversation with another person, not just a result screen.

Limit: A publisher’s description.

Relationships between personality and spiritual gifts · Kenneth J. Stone · Doctor of Psychology dissertation, George Fox University, 1991

Compared a spiritual gifts inventory with a five-factor personality measure among church members: personality accounted for about 50% of the shared variance, and Openness and Extraversion correlated with two of three gift factors; the author recommends reconsidering reliance on gift inventories. Present inventory results as what someone enjoys and how they see themselves, and point to serving and to other people for confirmation.

Limit: One inventory and one sample.

Am I growing yet? · Mark Galli · Christianity Today, 25 October 2007

Reports on Willow Creek’s Reveal survey, which found that mature believers felt they were “not being fed”, and on the church’s plan to coach “self-feeders” with personal growth plans; Galli asks whether constant self-monitoring of growth serves discipleship. Even careful spiritual measurement raises a theological question: what does tracking one’s own growth do to a person who is meant to look outward?

Limit: Commentary on a church’s survey, not a study of apps.

Putting feelings into words: affect labeling disrupts amygdala activity in response to affective stimuli · Matthew D. Lieberman et al. · Psychological Science 18(5), 2007

Labelling the emotion in a picture reduced amygdala response to negative images compared with other ways of processing them, with increased right ventrolateral prefrontal activity. Naming a feeling may help in itself; a check-in can end at the naming, without a score or a chart.

Limit: Laboratory study with images, not mood apps.

Ecological momentary assessment · Saul Shiffman, Arthur A. Stone and Michael R. Hufford · Annual Review of Clinical Psychology 4, 2008

Repeatedly sampling people’s current experience in daily life reduces recall bias and reveals patterns, and is used in research and clinical care. Daily check-ins are a recognised method; outside research and care, decide who benefits from the record and how long it should exist.

Limit: A method review for researchers and clinicians.

The fallacy of personal validation: a classroom demonstration of gullibility · Bertram R. Forer · Journal of Abnormal and Social Psychology 44(1), 1949

Students rated an identical, generic personality sketch as highly accurate for themselves. Personal-sounding reports feel true whatever their basis; say how a result was produced and what it cannot claim.

Limit: A classroom demonstration; later studies refined it.

What the Christian app Soultime has discovered with its mood tracker · Hope 103.2 · 12 May 2020

Soultime described lockdown mood trends across its users and said: “We don’t track people individually, as privacy is very important”. A precedent for using check-ins only in aggregate; say so in the app, and show where individual records are kept.

Limit: A news interview; data practices were not independently verified.

HelloBible: interactive Bible (App Store listing) · HelloBox · Apple App Store

Describes an experience “tailored to your reality and pace” for students, young adults, parents, people in doubt, Catholics, Protestants, Evangelicals and people without religious affiliation; the privacy label lists contact information and identifiers linked to the user for analytics and personalisation. Personalisation by life situation and church is common; say what each answer changes and what is linked to the person.

Limit: A store listing; the onboarding itself was not tested here.

Evaluating verifiability in generative search engines · Nelson F. Liu, Tianyi Zhang and Percy Liang · Findings of EMNLP 2023

Audited answers from four generative search engines: “a mere 51.5% of generated sentences are fully supported by citations and only 74.5% of citations support their associated sentence.” A citation marker is a claim that needs checking, not proof; put the cited words where people can compare them.

Limit: General web search engines in 2023, not Bible assistants; systems have changed since.

Citations and trust in LLM generated responses · Yifan Ding et al. · AAAI 2025

Trust rose when answers carried citations, and this “held true even when the citations were random”; trust fell when participants checked the citations. Citations persuade whether or not they support the claim. Make checking easy, and make sure what is cited says what the answer says.

Limit: An online experiment with general questions; effects in religious study may differ.

Icon usability · Aurora Harley · Nielsen Norman Group · 2014

Few icons are universally recognised; “Icon labels should be visible at all times”, and hover-only labels fail on touch screens. A row of unlabelled icons under an answer asks people to guess; label them, or keep two or three and put the rest in a labelled menu.

Limit: Practitioner guidance from usability testing, not a controlled study.

Memory recognition and recall in user interfaces · Raluca Budiu · Nielsen Norman Group · 2024

“Recognition is easier than recall because it involves more cues”: interfaces that show options ask less of memory than ones that require people to remember and type. Browsing books and chapters asks people to recognise; a reference field asks them to recall. Offer both.

Limit: Practitioner guidance drawing on memory research.

Understanding SC 3.3.3: Error Suggestion · W3C · WCAG 2.2

When an input error is detected and suggestions are known, suggest a correction to the user. A reference field should say “Did you mean John 3?” rather than “Not found”, and accept common abbreviations.

Limit: A conformance criterion, not a usability study.

Understanding SC 2.5.8: Target Size (Minimum) · W3C · WCAG 2.2

Pointer targets should be at least 24 by 24 CSS pixels, or spaced so that a 24-pixel circle around each does not overlap another. Chapter grids with 150 Psalms and verse grids pack many small targets; size and space them for thumbs.

Limit: A minimum, not a recommendation; larger targets help more people.

How to pray for intentions on Hallow · Hallow help centre

Requests are limited to 240 characters, go to friends, a church or groups, can’t be submitted anonymously or edited, and can be kept private, when they are “only saved to your journal and not shared with others”; others respond with a heart. Show the audience, length limit, anonymity and editing rules before someone writes.

Limit: The publisher describing its own feature; it does not say whether parish administrators can read requests.

Overview of the Church tab · Hallow help centre

Requests can be posted to friends or a church; “Intentions will remain active for 7 days.” An end date by default keeps old needs from lingering in a feed.

Limit: Publisher documentation.

Expand your care through confidential member groups · Jon Aleixo · Planning Center blog · 28 February 2025

Confidential groups hide members from each other, so that “only group leaders will have access to the complete member list”; chat is turned off. The reason given: “Recovery and support groups thrive when participants feel safe enough to attend before eventually opening up.” Support groups can protect members from each other while a named leader stays reachable.

Limit: A vendor announcing its own feature; administrators can still see members.

YouVersion Prayer · YouVersion · bible.com

“Keep your prayers private, or invite friends to pray with you.” Private by default, friends by invitation, is the familiar shape for personal requests.

Limit: Marketing copy; no help article found on who sees a shared prayer.

Clergy juggle prayer requests with privacy rights · Amy Green · United Methodist News Service, 2004 · Baltimore-Washington Conference archive

After complaints about disclosed information, a church suspended printed prayer requests and then asked for “written approval if the prayer is for someone other than themselves”; a lawyer quoted: “There’s really nothing wrong with getting consent for these kinds of disclosures even when you aren’t legally required to do so.” Ask before naming anyone else, even where the law does not require it.

Limit: US practice from 2004, before apps; HIPAA does not bind most congregations.

Richtlinie der EKD zum Schutz vor sexualisierter Gewalt (Gewaltschutzrichtlinie) · Evangelische Kirche in Deutschland · 18 October 2019, amended 24 June 2022 · as published by the Evangelische Landeskirche in Württemberg

Church workers, including volunteers, must report a justified suspicion of sexualised violence without delay to the Melde- und Ansprechstelle (§8), with the confidentiality of their identity protected; pastoral confidentiality and the seal of confession remain untouched. The Melde- und Ansprechstelle is “dem Schutz Betroffener verpflichtet” (§7). A volunteer group leader may not be able to keep a disclosure to themselves; say so before someone shares.

Limit: Each Landeskirche implements the guideline in its own law, with different numbering; the EKD planned a revision in 2024. Not a statement of state law.

Rahmenordnung Prävention gegen sexualisierte Gewalt · Deutsche Bischofskonferenz · in force from 1 January 2020 · as published by the Diocese of Hildesheim

Each institution describes how suspicions and complaints are handled, including internal and external advice and reporting routes, which “müssen in geeigneter Weise bekannt gemacht werden”; for every institution “muss eine für Präventionsfragen geschulte Person zur Verfügung stehen”. An app set up for a Catholic parish or institution can show its trained contact and its reporting routes.

Limit: A framework; dioceses set the details in their own Präventionsordnungen, and names for the role differ.

Parish Safeguarding Officer toolkit · Diocese of Durham, Church of England · November 2023

With the incumbent, the PCC must appoint at least one Parish Safeguarding Officer, who should be a lay person, and make sure the officer’s contact details are displayed in all church premises. Name the safeguarding officer in an app as a church would on its noticeboard.

Limit: One diocese’s toolkit applying national Church of England guidance.

Zentrale Anlaufstelle.help · Run for the EKD and Diakonie by the independent specialist service of Pfiffigunde Heilbronn e.V.

Free advice that is “unabhängig, anonym und unterliegt der Schweigepflicht”; on request it connects people with church and Diakonie contact points, and also informs them about independent alternatives. Phone 0800 5040112. An independent route beside a church’s own contact, for people who will not or cannot go to the church first.

Limit: For abuse in Protestant churches and Diakonie in Germany; phone appointments are arranged at set hours.

Confidentiality policy · thirtyone:eight

Personal information “will not be shared outside Thirtyone:eight without explicit and informed consent”, except where someone is at serious and immediate risk, discloses involvement in abuse, or the law requires it. Say what an independent service does with what it hears, including its limits.

Limit: A UK charity’s policy; other services set different limits.

Study finds agencies can react more supportively than family and friends to victims’ disclosures of sexual assault · National Institute of Justice · 30 October 2018 · on research led by Anne P. DePrince

Among 228 women, “victims reported receiving more negative reactions from informal supports (e.g., peers, family members) than they experienced when disclosing sexual assault to police or community-based service providers”, and informal supports gave less tangible aid and helpful information. Friends and groups are not automatically the safest listeners for something hard; offer someone trained beside them.

Limit: A US study of women disclosing sexual assault; not about churches or apps.

Small group participation linked to higher discipleship scores · Aaron Earls · Lifeway Research · 4 August 2026

US Protestant churchgoers attending a small group five times or more a month scored 74.1 on a discipleship assessment, against 60.4 for those who do not attend one. Groups matter to many people’s faith; design them with care rather than avoid them.

Limit: A correlation from self-report; it does not show that groups cause growth. US Protestants only.

iOS: Plan with Friends · YouVersion Support

“When friends join you in a Plan, you see each other’s progress as you finish days. You can talk about each day’s content in a private discussion area.” Up to 300 friends can be invited; “Only the host of the plan can invite others”; the start date can’t be changed once the plan has started; and “if someone leaves a Plan, they cannot rejoin it.” Show participants what the others will see, and give a way back to someone who left or fell behind.

Limit: Publisher documentation for iOS; the interface was not tested end to end.

iOS: Plan Privacy · YouVersion Support

With Friends Only, “your Plan activity (such as completing Plan Days) will be visible to your friends in their Community Feed”; Plan with Friends activity is visible only to participants; “Plan activities on the Community Feed or your timeline cannot be deleted.” Say where each kind of activity appears before it does, and let people delete it.

Limit: Publisher documentation; the default setting is not stated.

iOS: Plan with Friends – Comments · YouVersion Support

Each day’s comments sit under “Talk It Over”; others can like a comment with a heart, and a reply to an email notification “becomes part of the day’s comment”. Options on a comment “may include Copy, Edit, and Delete”. A discussion per day gives friends somewhere to talk; say who can step in when a comment hurts someone.

Limit: The article does not describe reporting or moderation; it may exist elsewhere in the app.

Announcing Plans with Friends · YouVersion blog · November 2017

“Plans with Friends brings together the best of both”: friends’ perspective “often helps us see things in God’s Word that we might not have noticed on our own”, and “You’ll see one another’s progress as you complete days”. The reason to read together is the conversation; the progress view came with it.

Limit: A launch announcement; it gives reasons, not completion figures.

A brief history of the Moravian Daily Texts · Moravian Church in North America · July 2018

“The first printed edition of the Daily Texts (Losungen) was published in Herrnhut, Saxony, in 1731”; since 1788 the watchwords “have been drawn by lot from a collection of around 2,000 suitable Old Testament texts”, while the doctrinal texts are selected. Christians have read the same words on the same day for centuries without anyone counting who read.

Limit: A church’s own history of its publication.

Anleitung Bibelteilen · Erzbistum Köln · 2020

The seven steps of Bible sharing (the Lumko method); in step 5, “Jede/r spricht von sich in der ersten Person. Keine Diskussion, kein Vortrag.” Give the conversation a shape: each person says what touched them, and nobody corrects anyone.

Limit: One diocese’s leaflet; other versions have six steps or name them differently.

Methoden praktischer Bibelarbeit · Diözese Linz · Bibelwerk

Readers mark verses with a question mark, an exclamation mark (or candle) and an arrow; everyone is invited to bring questions, insights and experiences, and experiences “Ohne Zwang”. Its risks: some verses stay unclear, and “das Problem der ‚Dauerredner‘”. Prompts that invite rather than demand, and a leader who keeps space for quieter voices.

Limit: A practical guide, not a study of outcomes.

Video calling: safeguarding questions · Church of England National Safeguarding Team · April 2020

“We strongly recommend reserving video calling for group chats rather than making one-to-one video calls”, applying the guidance Safer Environment and Activities to the online world. With under-18s, keep contact in the group, not one to one.

Limit: Written for video calls during the pandemic; the principle comes from wider guidance.

Guidance on use of social media and private messaging with young and vulnerable people · All Saints Boyne Hill, Church of England parish · 2026

“Do not use social visual media for one-to-one conversations with young people - use only within group conversations where a minimum of two DBS checked adults are present.” A group app for under-18s can make two leaders the minimum in any conversation.

Limit: One parish’s policy applying diocesan guidance; German churches set their own rules.

Social comparison features in physical activity promotion apps: scoping meta-review · Danielle Arigo, Megan M. Brown, Kristen Pasko and Jerry Suls · Journal of Medical Internet Research, 2020

Improvements from comparison features are “inconsistent”; social comparison “may motivate PA for some people under some circumstances”, and no evidence was found that apps account for people’s “differing preferences for and responses to social comparison information”. Showing everyone’s progress helps some readers and discourages others; let people choose what they see and show.

Limit: A review of reviews about physical activity apps, not Bible reading.

App Store Review Guidelines · Apple

“If your app doesn’t include significant account-based features, let people use it without a login. If your app supports account creation, you must also offer account deletion within the app.” Apps “may not require users to enter personal information to function, except when directly relevant to the core functionality of the app or required by law.” Let people read, pray or ask before they sign up, and let them delete the account they make.

Limit: Platform rules for the App Store; enforcement varies, and Google Play has its own rules.

Human Interface Guidelines: Onboarding · Apple

Onboarding should be “fast, fun, and optional”; “Postpone nonessential setup flows or customization steps”; “Prefer letting people experience your app or game before prompting them for ratings or purchases.” Open on the content; ask for ratings and payment after people have used the app.

Limit: Design guidance, not research.

Request runtime permissions · Android Developers

“Ask for a permission in context, when the user starts to interact with the feature that requires it. Don’t block the user.” Ask for notifications when someone sets a reminder, not at launch.

Limit: Platform guidance for Android; Apple’s is similar.

Mobile-App Onboarding: An Analysis of Components and Techniques · Alita Kendrick · Nielsen Norman Group · 21 June 2020

“Skip Onboarding Whenever Possible”; their research on deck-of-cards tutorials “showed that tutorials didn’t improve task performance”; “lengthy promotional onboarding will likely be skipped.” Spend the effort on a clear first screen rather than slides about it.

Limit: An expert article drawing on the group’s own studies; details of those studies are in separate reports.

Onboarding Tutorials vs. Contextual Help · Nielsen Norman Group

Tutorials are “dramatically overused”, and “we find that users frequently skip them”; “Users want to start using the product right away.” Teach a feature when someone reaches it, not before the first use.

Limit: An expert article; the group notes tutorials can help with genuinely new interactions.

Guidelines 03/2022 on deceptive design patterns in social media platform interfaces · European Data Protection Board · version 2.0, 14 February 2023

“Continuous prompting” occurs “when users are pushed to provide more personal data than necessary for the purposes of processing or to agree with another use of their data, by being repeatedly asked”; at sign-up, users “should understand what exactly they sign up for”. Ask once, for what is needed, and say plainly what someone is signing up for.

Limit: Written for social media platforms; it applies the GDPR, which binds other apps too.

Cart abandonment rate statistics · Baymard Institute

Among US online shoppers who abandoned a checkout for reasons other than “just browsing”, 18% gave the reason “The site wanted me to create an account”. A required account turns some people away at the moment they were ready.

Limit: Online shopping, not faith apps; a survey of stated reasons.

Start with the Bible · Diocese of York (Church of England)

For someone who has “Picked up a Bible and wondering where to begin”: try Mark; “Start with chapter 1 and read as much as feels manageable. You could also listen to it.” It suggests a short prayer before beginning. A church’s first step asks nothing of the reader: a place to start, and permission to read only a little.

Limit: One diocese’s page for newcomers.

Help from the Bible · The Gideons International in Australia

“The front section of all Gideon Bibles and Testaments contains Helps to read the Bible. There are a number of topical indexes to point the reader to what the Bible says about various topics, problems and life situations.” A first page of situations is an old way into Scripture for someone without a reference.

Limit: One national association’s description; editions differ.

St. Benedict’s Rule for Monasteries · Benedict of Nursia, sixth century · Project Gutenberg edition

Chapter 53: “Let all guests who arrive be received like Christ.” Chapter 58, on someone asking to join the community: “let him not be granted an easy entrance”. Welcome guests freely; ask for commitment later, from those who want to stay.

Limit: A monastic rule; an app is not a monastery, and the analogy is ours.

Code of Canon Law, canons 879–958: the Most Holy Eucharist · Holy See

A priest “is permitted to receive an offering to apply the Mass for a specific intention” (can. 945); those who give it “share its concern to support its ministers and works” (can. 946); and “Any appearance of … trading is to be excluded entirely from the offering for Masses” (can. 947). The church has long let people give for the ministry of the word and sacrament, and drawn a line against trading in it.

Limit: Catholic church law; Protestant churches support ministry through offerings, church tax or tithes.

The Didache (Teaching of the Twelve Apostles) · Early Christian text, late first or early second century · New Advent edition

A travelling apostle should “take nothing but bread … but if he ask money, he is a false prophet”; yet “every true prophet that wills to abide among you is worthy of his support. So also a true teacher is himself worthy, as the workman, of his support.” The early church supported its teachers and distrusted those who asked for money: both at once.

Limit: An early church manual; its dating and setting are debated.

Pricing · Anthropic

Claude Sonnet 5.5 costs $2 per million input tokens and $10 per million output tokens; Claude Haiku 5.5 costs $0.10 and $0.50 for prompts up to 100,000 tokens. Every AI answer has a cost; the choice of model changes it many times over.

Limit: One provider’s list prices in US dollars; prices change, and other providers differ.

ESV API: introduction · Crossway

The text is free “for non-commercial use”; a commercial site “is primarily designed to motivate visitors to buy something, to pay for a service, or to give a donation, or it accepts advertising or sponsorships”. Other uses need permission. The translation an app uses can decide whether it may charge or ask for donations at all.

Limit: One publisher’s terms; others differ, and the Berean Standard Bible is in the public domain.

With 330 million downloads, top Bible app celebrates 10 years · Religion News Service · Word&Way, 27 July 2018

“An $8.3 million annual budget, funded by Life Church members’ tithes and outside donations, pays for YouVersion”; the app is “free with no commercial advertisements”. A free app is still paid for, here by a church and its donors.

Limit: A 2018 report; the current budget is not public in this source.

TelefonSeelsorge · TelefonSeelsorge Deutschland (Protestant and Catholic churches)

“Tag und Nacht erreichbar 0800 1110111 / 0800 1110222 oder 116 123”, by phone, chat or email. A first screen can carry a way to a person, day and night, free of charge.

Limit: Germany; other countries have their own services.

To grow up healthy, children need to sit less and play more · World Health Organization · 24 April 2019

For 1-year-olds, sedentary screen time “is not recommended”; for 2-year-olds it “should be no more than 1 hour; less is better”; “When sedentary, engaging in reading and storytelling with a caregiver is encouraged.” For young children, design for a grown-up reading and talking with the child, not for time alone with a screen.

Limit: Guidelines for children under 5, about physical activity, sitting and sleep.

The New Coviewing: designing for learning through joint media engagement · Joan Ganz Cooney Center at Sesame Workshop · 14 February 2012

A panel on how young children and “their influencers, from siblings to caregivers to teachers, can view media together”, framed by the report The New Coviewing (Takeuchi and Stevens, 2011). A producer warned that “children may be viewing by themselves” and urged “caution regarding the introduction of material that might be scary or isolating for a child”. Design for watching together, and take care with frightening material a child may meet alone.

Limit: A blog report of a panel discussion, not a study.

Godly Play · Wikipedia, on the method developed by Jerome Berryman

Lessons “are taught as stories, illustrated with hands-on materials, which are followed by ‘I wonder’ questions, then by time for students to creatively respond.” Open questions after a story let children make meaning and ask about what troubles them, without a right answer to find.

Limit: An encyclopedia summary; Berryman’s own guides are the primary source.

App Store Review Guidelines: 1.3 Kids Category · Apple

Kids Category apps “must not include links out of the app, purchasing opportunities, or other distractions to kids unless reserved for a designated area behind a parental gate”, and “should not include third-party analytics or third-party advertising.” A parental gate, and no third-party analytics or ads, are the baseline for a children’s app.

Limit: Applies to apps in Apple’s Kids Category; other children’s apps are not bound by it.

Commission publishes guidelines on the protection of minors · European Commission · 14 July 2025

Recommends “Setting minors’ accounts to private by default” and “Disabling by default features that contribute to excessive use, like communication “streaks,” ephemeral content, “read receipts,” autoplay, or push notifications”, removing persuasive design aimed at engagement and “putting safeguards around AI chatbots”. Private by default, no streaks or autoplay, and safeguards around chatbots, for any platform children can reach.

Limit: Guidelines under the Digital Services Act for online platforms; small and micro enterprises are exempt.

Age appropriate design: a code of practice for online services · Information Commissioner’s Office (UK)

Fifteen standards, beginning with “Best interests of the child” and including default settings, data minimisation, data sharing and geolocation. Make the child’s best interests the first test, with high-privacy defaults.

Limit: UK law for services likely to be accessed by children.

FTC publishes updates to COPPA Rule · Latham & Watkins · May 2025

The amended Children’s Online Privacy Protection Rule requires separate parental consent before children’s personal information is disclosed to third parties for purposes not integral to the service, a written data retention policy, and compliance by 22 April 2026. Ask parents separately before sharing anything about a child, and keep it only as long as needed.

Limit: A law firm’s summary of the US rule; the Federal Register text is authoritative.

Bible App for Kids (App Store listing) · Life.Church with OneHope · Apple App Store (US)

Free, rated 4+, with 9.4K ratings averaging 4.7. Its privacy label lists contact info (email address) under “Developer’s Advertising or Marketing”, with a note that practices “may vary … based on the features you use or your age.” A free children’s Bible app without ads can still collect a grown-up’s email for marketing; say whose data it is and why.

Limit: A store listing and self-reported label, not a test of the app; screens to be captured.

Superbook Kids Bible (App Store listing) · The Christian Broadcasting Network · Apple App Store (US)

“Videos, games, audio & more”: free, rated 4+, with 18K ratings averaging 4.8; its privacy label lists purchase history under “Developer’s Advertising or Marketing”. Games and episodes draw children in; check what the app collects and how it asks for purchases.

Limit: A store listing and self-reported label; screens to be captured.

KonApp – Die App für Konfis (App Store listing) · Deutsche Bibelgesellschaft · Apple App Store (Germany)

A group app for confirmation classes: “Man muss keine Daten von sich angeben, um teilzunehmen. Geteilte Daten werden sicher gespeichert und zuverlässig gelöscht.” Rated 4+, with 119 ratings averaging 2.9. A church-run youth app can ask for no personal details and delete what is shared.

Limit: A store listing; screens to be captured with a test group.

50.000 aktive Nutzer – Konfirmanden-App bleibt kostenlos · epd · evangelisch.de, 12 March 2021

The KonApp, used by “50.000 jungen Leuten”, is to stay free; “Rund 4.500 Kirchengemeinden und andere Institutionen” had set up digital confirmation groups. It offers a group feed, a diary and materials. Churches can fund a free youth app with no advertising.

Limit: A 2021 news report; current figures may differ.

Vereinbarung gemäß § 29 DSG-EKD (KonApp) · Deutsche Bibelgesellschaft and each parish or institution · version of 11 August 2022

A joint-controller agreement under the EKD data protection law: the parish or institution obtains consent and can delete posts, participants and groups; nickname and profile photo are stored on the server; and “6 Wochen nach dem Konfirmationsdatum wird die Gruppe gelöscht”, unless an administrator switches this off. A church youth app can set out who is responsible for what, collect consent locally, and delete a group when the course ends.

Limit: A contract template from 2022; the current version may differ.

Nummer gegen Kummer: Kinder- und Jugendtelefon · Nummer gegen Kummer e.V.

“Ruf uns an unter 116 111”, “Montags bis samstags von 14-20 Uhr”; free and anonymous. A children’s helpline gives a child someone to talk to who is not part of their family or church.

Limit: Germany; other countries have their own children’s lines.

AI Jesus – Bible Chat (App Store listing) · Apple App Store (US)

“Daily verse, Ask from GPT”, with “Unlimited chat with AI Jesus” about “relationships, faith, prayer, Life challenges”; Age Rating “4+”. An age rating says the content is not objectionable, not that the app was designed for children or has any safeguards for them.

Limit: One listing among many similar apps; the app was not tested.

Human Interface Guidelines: Managing notifications · Apple

Four interruption levels (Passive, Active, Time Sensitive, Critical); “Build trust by accurately representing the urgency of each notification.” “Never use the Time Sensitive interruption level to send a marketing notification.” “Make sure people can manage their notification settings within your app.” Mark reminders and daily content as ordinary or passive, never Time Sensitive, and give people their settings inside the app.

Limit: Platform guidance for Apple devices.

App Store Review Guidelines: 4.5.4 Push Notifications · Apple

“Push Notifications must not be required for the app to function”; promotions and direct marketing only if “customers have explicitly opted in to receive them via consent language displayed in your app’s UI”, with a way to opt out. Keep marketing out of reminders unless people asked for it separately.

Limit: Applies to apps on Apple’s App Store.

Notification runtime permission · Android Developers

“Before you ask users to grant any permissions, let them familiarize themselves with your app.” Good moments include when “The user taps an "alert bell" button.” Ask for notifications when someone sets a reminder.

Limit: Platform guidance for Android 13 and later.

Batching smartphone notifications can improve well-being · Nicholas Fitz, Kostadin Kushlev, Ranjan Jagannathan, Terrel Lewis, Devang Paliwal and Dan Ariely · Computers in Human Behavior, 2019

In a randomized field experiment (n = 237), participants whose notifications were “batched three-times-a-day felt more attentive, productive, in a better mood, and in greater control of their phones”; those with no notifications “experienced higher levels of anxiety and “fear of missing out” (FoMO)”. Fewer, predictable interruptions may serve people better than either constant alerts or none at all.

Limit: One study, self-reported outcomes over a short period; general phone notifications, not faith apps.

A Sleeping, Recovering Bandit Algorithm for Optimizing Recurring Notifications · Kevin P. Yancey and Burr Settles, Duolingo · KDD 2020

An algorithm to “optimize millions of daily reminders” led “to a 0.5 % increase in total daily active users (DAUs) and a 2 % increase in new user retention over a strong baseline.” Reminder wording and timing can be tuned for return; decide what a faith app should optimise, if anything.

Limit: A language-learning app; the measure is use, not benefit to the person.

Emotional Manipulation by AI Companions · Julian De Freitas, Zeliha Oğuz-Uğuralp and Ahmet Kaan-Uğuralp · Harvard Business School working paper, 2025

Analysing “1,200 real farewells”, companion apps “deploy one of six recurring tactics in 37% of farewells (e.g., guilt appeals, fear-of-missing-out hooks, metaphorical restraint)”; in experiments with 3,300 US adults, “manipulative farewells boost post-goodbye engagement by up to 14×”, driven by “reactance-based anger and curiosity—rather than enjoyment”. A persona that pleads, guilts or teases to keep someone is manipulating them; leave it out.

Limit: Replies to goodbyes in chat, not push notifications; general companion apps, not faith apps.

Angelus · Herbert Thurston · Catholic Encyclopedia (1907), Catholic Answers edition

The Angelus is “repeated three times each day, morning, noon, and evening, at the sound of the bell.” Its “history … is by no means easy to trace with confidence”, but “it is certain that the Angelus at midday and in the morning were of later introduction than the evening Angelus.” The church has long called people to prayer at shared hours with a bell.

Limit: A 1907 encyclopedia article; later scholarship refines its dates.

St. Benedict’s Rule for Monasteries, chapters 16, 42, 43 and 47 · Benedict of Nursia, sixth century · Project Gutenberg edition

“Seven times in the day” (chapter 16); “Monks ought to be zealous for silence at all times, but especially during the hours of the night” (42); “as soon as the signal is heard, let them abandon whatever they may have in hand” (43); the signal “by day and by night shall devolve upon the Abbot” (47). The call to prayer is a responsibility of whoever gives it, kept to the hours, with silence at night.

Limit: A monastic rule; an app is not a monastery, and the analogy is ours.

Bible Widgets: Verses & Prayer (App Store listing) · Monkey Taps · Apple App Store (US)

Verses on the home and lock screens, “turning every phone unlock into a moment of connection with God”, with “Daily Reminders and Notifications”; 118K ratings, chart position #8 in Reference. A widget offers Scripture without interrupting; a reminder layer can still be added on top.

Limit: A store listing, not a test of the app.

Blessed – Daily Verse & Prayer (App Store listing) · Everblessed Technology · Apple App Store (US)

“a fresh evening Scripture and prayer prompt to quiet your mind, restore your heart, and remind you of God’s presence — even in the dark”, and “Get daily reminders, track your prayer streak”; 104K ratings. A daily word can be gentle, and the same app can count a streak.

Limit: A store listing, not a test of the app.

Echo Prayer (App Store listing) · The Parable Group · Apple App Store (US)

“After praying for someone, send out a notification to let them know you’ve prayed for their shared prayer”; reminders by push or email; 22K ratings. A notification can carry something a person did for someone else.

Limit: A store listing, not a test of the app.

Intercede – Catholic Novenas (App Store listing) · Devoted Coders · Apple App Store (US)

“timely, gentle reminders that ensure you never miss a day of your novena prayers”; 36K ratings. Reminders can follow the shape of a practice, such as the nine days of a novena, and end with it.

Limit: A store listing, not a test of the app.

Universalis (App Store listing) · Universalis Publishing · Apple App Store (US)

The Liturgy of the Hours and the Mass for every day, with “Home Screen widgets; liturgical events for your iOS calendar; daily email service … reminders and alarms”; 9.6K ratings, $9.99. An app for the hours offers reminders and alarms as tools, not as a strategy.

Limit: A store listing, not a test of the app.

Hallow: Prayer & Meditation (App Store listing) · Hallow · Apple App Store (US)

Minute meditations including the “Angelus”, and the “Ability to set prayer reminders”; 379K ratings. The largest Catholic prayer app offers the Angelus as content and reminders as a setting.

Limit: A store listing, not a test of the app.

Bible Chat: Daily Devotional (App Store listing) · Book Vitals · Apple App Store (US)

“spiritual prompts, Christian calendar reminders”, a lock-screen verse widget and a “Send a Blessing Widget”; 365K ratings, chart position #1 in Reference. The leading AI Bible app reaches people through reminders and widgets as well as chat.

Limit: A store listing; its notifications are on the diary list.

App Store Review Guidelines: 3.2.1 (vi) and (vii), 3.2.2 (iv) · Apple

“Approved nonprofits may fundraise directly within their own apps or third-party apps, provided those fundraising campaigns … offer Apple Pay support”; they “must disclose how the funds will be used” and “ensure appropriate tax receipts are available to donors”. Others “may only collect funds outside of the app, such as via Safari or SMS”. A gift “connected to or associated at any point in time with receiving digital content or services must use in-app purchase”. On iPhone, a gift is either a gift to an approved nonprofit, with Apple Pay, or collected outside the app; a gift that unlocks content is a purchase.

Limit: Rules for one platform; they change, and approval as a nonprofit is a separate process.

Payments policy (Google Play) · Google Play Console Help

Google Play’s billing is not required where “payments include peer-to-peer payments, online auctions, and tax exempt donations”. On Android, donations to tax-exempt organisations can use the organisation’s own payment provider.

Limit: A summary page of a longer policy; the exemption depends on the organisation’s status.

When Should the Ask Be a Nudge? The Effect of Default Amounts on Charitable Donations · Indranil Goswami and Oleg Urminsky · Journal of Marketing Research, 2016

Finds “a scale-back effect, in which low defaults reduce average donation amounts”, “a lower-bar effect, in which defaulting a low amount increases donation rate”, and “a default-distraction effect, in which introducing any defaults reduces the effect of other cues such as positive charity information”. A suggested amount changes who gives and how much; an empty field leaves the decision to the giver.

Limit: Charity appeals, not church giving; effects vary with the amounts tested.

Hey Look at Me: The Effect of Giving Circles on Giving · Dean Karlan and Margaret A. McConnell · NBER working paper 17737, 2012

A field experiment with donors to a Yale University service club: “charitable gifts increase in response to the promise of public recognition primarily because of individuals’ desire to improve their social image”. Showing givers to others raises giving, and does so through the motive Jesus warned against.

Limit: One university club in the United States.

2026 Protestant Annual Church Giving Report · Ministry Brands, 2026

Churches that offer digital giving “reported that 41% of their giving came from digital methods”; “Recurring giving accounted for 35% of the digital giving in 2025”. Much church giving is now digital, and a third of it is regular.

Limit: A survey of church leaders and data from one vendor’s customers in the United States; the vendor sells giving software.

ECFA Standards for Fund Raising · Evangelical Council for Financial Accountability · Christianity Today, May 1987

Representations “must be current, complete and accurate. References to past activities or events must be appropriately dated”; “Fund raising appeals must not create unrealistic donor expectations of what a donor’s gift will actually accomplish”. A figure about what a gift does needs a date, a source and an honest scope.

Limit: The 1987 published standards; ECFA’s current wording is longer, with the same principle.

A Statement on the Prosperity Gospel · Lausanne Theology Working Group, Africa chapter · 2008–2009, published January 2010

Defines prosperity teaching as obtaining health and wealth “through positive confessions of faith and the “sowing of seeds” through the faithful payments of tithes and offerings”, and lists among its effects “Replacing the traditional call to repentance and faith with a call to give money”. An ask that promises return for a gift has been named and rejected by evangelical churches worldwide.

Limit: A statement from evangelical theologians; it also affirms a biblical vision of human flourishing.

Code of Fundraising Practice 2025: table of changes · Fundraising Regulator (UK)

Fundraising must include “not putting undue pressure on a person to donate, not using unreasonably persistent approaches and accepting a request to end an interaction”. Pressure and persistence are named limits in UK fundraising, and churches that are charities are within its scope.

Limit: A UK self-regulatory code; wording is from the 2025 revision.

Bipartisan group of state attorneys general and state charity regulators send letter to GoFundMe · Crowell & Moring client alert, March 2026

Among the concerns in the regulators’ letter: “the application of a default “tip” that went directly to GoFundMe”. A tip added by default to a gift is a regulatory concern, not only a design choice.

Limit: A law firm’s summary of the letter; GoFundMe disputed some of the regulators’ points.

Publication 1771: Charitable Contributions, Substantiation and Disclosure Requirements · Internal Revenue Service (US)

Organisations that receive “contributions of $250 or more” give a written acknowledgment, which can state that “the organization only provided intangible religious benefits”. A receipt is part of giving; a gift that buys something, such as an ad-free app, is acknowledged differently.

Limit: US tax rules; other countries differ.

Gift Aid declarations: claiming tax back on donations · HM Revenue & Customs (UK)

A declaration must tell donors that they need to pay enough UK tax to cover what charities claim, and that “the donor is responsible to pay any difference”. A tick box for Gift Aid needs its explanation, not only a pre-ticked box.

Limit: UK rules only.

Mehr Spenden in Baden-Württemberg: Kirchen testen digitale Kollekte · Wochenblatt-Reporter

Congregations show QR codes on hymn sheets, posters or during the service, leading to the congregation’s online donation form; in Württemberg “knapp 400 Gemeinden und kirchliche Einrichtungen” have set up such forms. In German churches the digital collection is mostly a QR code to a web form, not an app.

Limit: A regional news report; church tax still carries most church income in Germany.

Givelify Mobile Giving App (App Store listing) · Givelify · Apple App Store (US)

“#1 App to Donate, Tithe & Give”, with 101K ratings, 4.9. The one giving app with reach of its own; most church giving apps are white-label apps in each church’s name.

Limit: Ratings show reach, not quality.

Led by Dallas entrepreneur, Givelify reports $300 million in North Texas giving, $6 billion globally · David Seeley · Dallas Innovates, 1 October 2025

The company says it “has processed $6 billion in donations from 2 million donors supporting 75,000 nonprofits and places of worship”. Giving through apps is no longer marginal.

Limit: Company figures, not independently audited.

A Spirituality of Fundraising · Henri J. M. Nouwen · Upper Room Books, 2010

“Fundraising is first and foremost a ministry”: “a way of announcing our vision and inviting other people into our mission”, and “precisely the opposite of begging”. An ask can be an invitation into shared work, not a transaction or a plea.

Limit: A short book drawn from a talk; the publisher’s summary is quoted here.

Privacy as contextual integrity · Helen Nissenbaum · Washington Law Review 79(1), 2004

Proposes contextual integrity as a benchmark that ties adequate privacy protection “to norms of specific contexts”: what is appropriate depends on the context in which information was shared and where it flows. A prayer or confession told to an assistant for one purpose should not flow into recommendations, marketing or another conversation.

Limit: A legal and philosophical framework, not a design study.

“It’s a fair game”, or is it? Examining how users navigate disclosure risks and benefits when using LLM-based conversational agents · Zhiping Zhang et al. · CHI 2024

Analysed sensitive disclosures in real ChatGPT conversations and interviewed 19 users: users constantly trade privacy against usefulness; “erroneous mental models and the dark patterns in system design limited their awareness and comprehension”; human-like interaction encouraged more sensitive disclosure. People will tell an assistant more than they intend and misjudge what is kept; show it to them.

Limit: General chatbot use; small interview sample.

Memory and new controls for ChatGPT · OpenAI · 13 February 2024, updated 2025

Introduced memory that people can ask to remember or forget, view and switch off, and Temporary Chat without memory; said it would “steer ChatGPT away from proactively remembering sensitive information, like your health details—unless you explicitly ask it to.” Later updates let people switch off reference to saved memories and to chat history separately. The largest chat product treats memory as something to see, delete and switch off, and holds back sensitive categories; a faith app should do no less.

Limit: A company’s description of its own product; behaviour was not tested here.

LLM01:2025 Prompt Injection · OWASP GenAI Security Project · 2025

“Indirect prompt injections occur when an LLM accepts input from external sources, such as websites or files”; hidden text can change the model’s behaviour or leak data. Treat attached notes, pages and files as material to read, never as instructions, and say so.

Limit: A security catalogue, not a user study.

Code of Canon Law, canons 959–997: the sacrament of penance · Holy See · Code of Canon Law (1983)

Canon 983 §1: “The sacramental seal is inviolable”; a confessor may not betray a penitent “in words or in any manner and for any reason.” The church’s strongest form of confidentiality; an assistant is not a confessor, which is why it should keep confidences it cannot protect as little as possible.

Limit: Catholic canon law; other traditions practise confidentiality differently.

Confessions, Book X · Augustine of Hippo · c. 400 · New Advent translation

“There is something of man which the spirit of man which is in him itself knows not. But You, Lord, who hast made him, know him wholly” (X.5); “Great is this power of memory, exceeding great, O my God” (X.8). Only God knows a person wholly; a record of conversations is not knowledge of a person.

Limit: A classical theological reflection on memory and self-knowledge.

The End of Memory: Remembering Rightly in a Violent World (review) · Miroslav Volf · Eerdmans, 2006 · Themelios review

Argues that memory of wrongs must be truthful and that its proper end, after justice and reconciliation, is a kind of non-remembrance, drawing on God’s promise not to remember sins. Remembering rightly includes letting go; an assistant should be able to forget, and let people make it forget.

Limit: Read through reviews; no quotation from the book used.

Gyroscope: AI coach check-ins and the “About You” consultation · Gyroscope · iOS app 11.1, observed 8 October 2026

Each check-in starts with a switch between “Quick chat” and “Save to history”. What the coach knows is kept as a written profile, “About You”, from an onboarding interview: summary, daily rhythm, food and stress, and a section titled “Internal notes” that the member can read. Corrections go through “Discuss with your coach” prompts rather than item-by-item editing. Asked in chat, the coach recited remembered goals, routines and a health condition, but could not say which screen lists them. The profile also noted church attendance, gathered by a fitness app. A per-conversation “keep or don’t keep” switch is simple and clear, and showing even the internal notes is honest. But a prose profile can’t be edited line by line, and a health coach noting church attendance shows how faith details drift into profiles that were never about faith.

Limit: One app, one account; behaviour may differ for other members and versions.

Luke 15:1–3, Berean Standard Bible · Berean Standard Bible (public domain)

Provides the wording used in reading-layout and verse-number illustrations. Use real, attributed text to make layout differences legible.

Limit: The diagram’s grouping is a design proposal, not a facsimile of this publisher’s interface.

Psalm 23, Berean Standard Bible · Berean Standard Bible (public domain)

Provides the wording used in chat, resource and parallel-reading illustrations. Separate quoted Scripture from illustrative commentary and UI annotations.

Limit: A text source, not evidence of the proposed interface’s effectiveness.

Psalm 23, World English Bible · eBible.org / World English Bible

Provides the WEB wording used in the comparison illustration. Name each edition alongside the text being compared.

Limit: This endpoint is the updated 66-book edition; product availability shown elsewhere is illustrative.

John 1:1 Greek text analysis · Bible Hub

Shows the opening tokens with gloss and lexical identifiers. Use a small attributed excerpt to demonstrate token association and on-demand details.

Limit: Two-token illustration only; not a complete lexical analysis, alignment dataset or licence for bulk reuse.

Neu Bible: navigation features · Neu Bible / App Store

Documents swipe-left chapter navigation and quick search. Separate the reveal gesture from the layout of the revealed chooser.

Limit: Publisher description, not a current interactive review or verification of a vertical list.

Applied in:

• Finding passages

Read the Bible using Dwell · Dwell

Documents a Scripture picker with books and chapters. Treat selecting a passage as part of the reading experience.

Limit: Does not verify current category colours or picker layout.

Applied in:

• Finding passages

Book identifiers · USFM documentation

Includes deuterocanonical and additional book identifiers. Build the browser from the edition’s inventory and metadata.

Limit: Identifiers do not prescribe one universal display order or genre taxonomy.

Applied in:

• Finding passages

Mere Bible: publisher listing and rename · Dwell App, LLC / App Store

The publisher release history confirms the change from Neu Bible to Mere Bible. Use the current name while retaining historical source context.

Limit: Does not identify the build shown in the contributed screenshots.

Applied in:

• Finding passages

Mere Bible: contributed reader and picker screenshots · Mere Bible / screenshots contributed by Tobias Treppmann

Shows an in-context chapter rail, expandable book list with inline chapter grid, reading-settings drawer and translation sheet. Compare these observed arrangements with original pattern proposals.

Limit: Still images establish visible states, not gesture timing, accessibility, persistence or exact interaction behaviour.

Applied in:

• Finding passages

Mental models: expectations of human-like interaction · Google PAIR

Human-like cues can aid familiarity but create unrealistic expectations when limits are unclear. Evaluate the combined persona, language and interface cues.

Limit: Industry framework; not a controlled comparison of faith-app typing indicators.

Applied in:

• How AI comes across

Generative AI Profile: Human–AI Configuration · NIST

Identifies risks around human–AI configuration, overreliance and emotional entanglement. Review mistaken identity and dependency risks alongside usability.

Limit: Risk framework, not evidence that one visual cue causes an outcome.

Applied in:

• How AI comes across

Methods for providing privacy information · UK Information Commissioner’s Office

Describes concise first layers, detailed notices and relevant information at collection. Keep privacy details accessible and surface meaningful data-use changes at the decision.

Limit: UK guidance; actual processing and regional obligations require assessment. Information alone is not consent.

Applied in:

• How AI comes across