c4

no_treatment_control:
  description: "Participants receive no intervention"
  when_to_use: "Establishing baseline efficacy"
  ethical_consideration: "May deny potentially beneficial treatment"
  mitigation: "Offer treatment after study completion"

waitlist_control:
  description: "Delayed treatment after study period"
  when_to_use: "Ethical obligation to provide treatment"
  advantages:
    - "Addresses ethical concerns"
    - "Participants motivated by future treatment"
  limitations:
    - "Expectancy effects possible"
    - "Higher attrition in waitlist group"

treatment_as_usual:
  description: "Standard existing practice"
  when_to_use: "Comparing new intervention to current standard"
  requirements:
    - "Document TAU components thoroughly"
    - "Monitor TAU implementation"
    - "Account for variability in TAU delivery"

active_control:
  description: "Alternative credible intervention"
  when_to_use: "Control for non-specific treatment factors"
  requirements:
    - "Match on therapist contact time"
    - "Match on participant expectations"
    - "Different theoretical mechanism"
  example: "Supportive counseling vs. CBT"

attention_control:
  description: "Same attention/time, different content"
  when_to_use: "Isolate specific treatment effects"
  requirements:
    - "Equal duration and intensity"
    - "Comparable credibility"
    - "Inert or irrelevant content"
  example: "Health education vs. stress management training"

placebo_control:
  description: "Inert treatment presented as active"
  when_to_use: "Medical/pharmacological interventions"
  ethical_requirements:
    - "IRB approval required"
    - "Informed consent about placebo possibility"
    - "Debrief after study"

attention_checks:
  purpose: "Verify participant engagement and data quality"
  examples:
    - instructed_response: "Please select 'Strongly Agree' for this item"
    - factual_recall: "What color was the background in the video?"
    - consistency: "This is my first time taking this survey (repeated item)"
  placement:
    - "Middle of long surveys (prevent fatigue)"
    - "After critical manipulations"
    - "End of study (overall engagement)"
  analysis:
    - "Exclude participants who fail multiple checks"
    - "Sensitivity analysis with/without exclusions"
    - "Report exclusion criteria a priori"

comprehension_checks:
  purpose: "Verify understanding of instructions/manipulation"
  examples:
    - task_understanding: "What was the main topic of the article you read?"
    - condition_awareness: "Which type of scenario did you receive?"
    - instruction_recall: "What were you asked to focus on while watching?"
  timing: "Immediately after manipulation presentation"
  analysis_options:
    - exclude: "Remove participants who failed comprehension"
    - moderate: "Test if comprehension moderates effects"
    - report: "Transparency about manipulation success rate"

manipulation_success_checks:
  purpose: "Verify manipulation had intended psychological effect"
  examples:
    anxiety_induction:
      - "How anxious did the scenario make you feel? (1-7)"
      - "Rate your stress level during the task (1-10)"
    competence_manipulation:
      - "How competent did the person in the video seem? (1-7)"
      - "Would you trust this person's expertise? (Yes/No)"
    cognitive_load:
      - "How difficult was the task? (1-7)"
      - "How much mental effort did you exert? (1-9)"
  analysis:
    - "Confirm manipulation created intended difference between groups"
    - "If manipulation failed, explains null results on DV"
    - "May reveal boundary conditions or moderators"

vignette_components:
  scenario_setup:
    - "Establish context (who, where, when)"
    - "Introduce protagonist or situation"
    - "Provide relevant background information"

  manipulation_delivery:
    - "Embed experimental manipulation naturally"
    - "Keep non-manipulated elements constant across conditions"
    - "Ensure manipulation is salient but not obvious"

  outcome_prompt:
    - "Present decision point or question"
    - "Elicit judgment, choice, or reaction"
    - "Maintain ecological validity"

  best_practices:
    - "Pilot test for comprehension and realism"
    - "Control length across conditions (word count)"
    - "Use third-person to reduce self-enhancement bias"
    - "Include concrete details for vividness"

  example_structure: |
    Sarah is a 28-year-old marketing manager at a tech startup.
    She has been working on a major campaign for 3 months.

    [MANIPULATION: Success vs. Failure]
    - Success: The campaign exceeded all targets, generating 200% ROI.
    - Failure: The campaign underperformed, generating only 40% ROI.

    Sarah's manager schedules a meeting to discuss the campaign results.

    [DEPENDENT VARIABLE PROMPT]
    How anxious do you think Sarah feels about the meeting? (1-7)

visual_stimuli:
  photographs:
    considerations:
      - "Control for low-level features (brightness, contrast, complexity)"
      - "Pilot rate for valence, arousal, familiarity"
      - "Use standardized databases (IAPS, FACES) when possible"
    manipulation_examples:
      - emotional_faces: "Angry vs. Happy vs. Neutral expressions"
      - scene_content: "Threatening vs. Safe environments"

  video_stimuli:
    production_standards:
      - "Professional audio quality (eliminate background noise)"
      - "Consistent lighting and camera angles"
      - "Scripted dialogue for consistency"
      - "Same actors across conditions (within-subjects)"
    manipulation_techniques:
      - actor_behavior: "Confident vs. Anxious body language"
      - editing: "Fast-paced vs. Slow-paced cuts"
      - audio: "Background music mood manipulation"

  diagrams_and_graphics:
    when_to_use: "Explaining complex concepts or procedures"
    design_principles:
      - "Consistent visual style across materials"
      - "Clear labeling and legends"
      - "Color-blind friendly palettes"
      - "High resolution for legibility"

audio_stimuli:
  spoken_instructions:
    - "Use professional voice actors or text-to-speech"
    - "Control for speaker gender, accent, emotion"
    - "Match duration across conditions"

  ambient_sounds:
    - "Nature sounds vs. Urban noise (environment manipulation)"
    - "Control volume levels (dB measurements)"

  music:
    - "Control tempo, mode (major/minor), genre"
    - "Use validated mood induction procedures"

stage_1_conceptualization:
  theoretical_foundation:
    - "Identify target mechanism (e.g., cognitive restructuring, exposure)"
    - "Review evidence base for intervention approach"
    - "Specify hypothesized mediators and moderators"

  active_ingredients:
    - "List essential components (what MUST be included)"
    - "Distinguish active vs. supportive elements"
    - "Justify each component theoretically"

  example_CBT_for_anxiety:
    mechanism: "Cognitive restructuring and habituation"
    active_ingredients:
      - "Psychoeducation about anxiety"
      - "Cognitive restructuring exercises"
      - "Graduated exposure hierarchy"
      - "Homework assignments"
    supportive_elements:
      - "Therapeutic alliance building"
      - "Session structure and agenda-setting"

stage_2_operationalization:
  dosage_specification:
    duration:
      - "Length of each session (e.g., 50 minutes)"
      - "Total number of sessions (e.g., 12 weeks)"
      - "Interval between sessions (e.g., weekly)"

    intensity:
      - "Frequency (e.g., once per week, twice per week)"
      - "Between-session practice (daily, weekly)"
      - "Booster sessions (yes/no, when)"

    justification:
      - "Base on prior efficacy studies"
      - "Consider feasibility and participant burden"
      - "Balance fidelity with flexibility"

  protocol_documentation:
    components:
      - session_outlines: "Agenda for each session"
      - scripts: "Verbatim language for key components"
      - worksheets: "Participant materials and handouts"
      - homework: "Between-session assignments"

    example_session_outline: |
      Session 3: Introduction to Cognitive Restructuring

      1. Review homework (10 min)
         - Discuss thought records from past week
         - Troubleshoot challenges

      2. Psychoeducation (15 min)
         - Explain cognitive model (thoughts → feelings → behaviors)
         - Provide examples of cognitive distortions

      3. Skill practice (20 min)
         - Identify automatic thoughts in recent anxiety situation
         - Challenge thoughts using Socratic questioning
         - Generate alternative balanced thoughts

      4. Assign homework (5 min)
         - Complete 3 thought records this week
         - Read handout on cognitive distortions

stage_3_fidelity_assurance:
  training_implementers:
    components:
      - "Didactic instruction on theoretical model"
      - "Review of treatment manual and materials"
      - "Role-play practice of key techniques"
      - "Observation and feedback on practice sessions"

    certification_criteria:
      - "Score ≥80% on knowledge test"
      - "Demonstrate competence on role-plays (rated by supervisor)"
      - "Complete supervised practice cases"

  ongoing_monitoring:
    session_checklists:
      - "Implementer completes checklist after each session"
      - "Indicates which components were delivered"
      - "Notes deviations from protocol"

    audio_video_review:
      - "Record random sample of sessions (e.g., 20%)"
      - "Code for adherence to protocol"
      - "Code for competence in delivery"

    supervision:
      - "Weekly group supervision meetings"
      - "Review challenging cases"
      - "Prevent protocol drift over time"

  fidelity_metrics:
    adherence: "Did implementer deliver prescribed components? (checklist)"
    competence: "How skillfully were components delivered? (rating scale)"
    differentiation: "Was treatment distinguishable from control? (ratings)"

    reporting:
      - "Calculate % sessions with ≥80% adherence"
      - "Report mean competence ratings"
      - "Analyze if fidelity predicts outcomes"

matching_considerations:
  credibility:
    assessment: "Use Credibility/Expectancy Questionnaire (Devilly & Borkovec, 2000)"
    requirement: "Control should be perceived as plausible treatment"
    adjustment: "If credibility differs, control for expectancy statistically"

  attention_time:
    requirement: "Match total contact time with interventionist"
    example: "If treatment = 12 x 50-min sessions, control = 12 x 50-min sessions"

  structure:
    requirement: "Match homework, materials, session format"
    example: "Both conditions use workbooks, have homework, follow agenda"

  non_specific_factors:
    to_match:
      - "Therapist warmth and support"
      - "Group cohesion (for group interventions)"
      - "Hope and expectancy"
    to_differ:
      - "Specific techniques (e.g., cognitive restructuring in treatment, not control)"
      - "Theoretical mechanism targeted"

ethical_considerations:
  beneficence:
    question: "Does withholding treatment cause harm?"
    mitigation_strategies:
      - "Offer treatment to controls after study (waitlist)"
      - "Provide crisis resources to all participants"
      - "Monitor for adverse events in all conditions"

  equipoise:
    definition: "Genuine uncertainty about which condition is superior"
    application: "Required for ethical randomization"
    violation: "If strong evidence for treatment, may be unethical to withhold"

  informed_consent:
    disclosures:
      - "Possibility of random assignment to control"
      - "Nature of control condition (if not placebo)"
      - "Option to withdraw and seek treatment elsewhere"

pretesting_procedures:
  comprehension_pilot:
    goal: "Ensure materials are understood as intended"
    method:
      - "Recruit small sample (n=10-20)"
      - "Ask participants to summarize materials"
      - "Identify confusing or ambiguous elements"
      - "Revise and re-pilot if needed"

  manipulation_pilot:
    goal: "Verify manipulation creates intended psychological state"
    method:
      - "Randomly assign pilot sample to conditions"
      - "Measure manipulation check variables"
      - "Test for significant difference between conditions"
      - "Refine manipulation if effect size is small"

  realism_believability:
    goal: "Ensure materials are ecologically valid"
    method:
      - "Ask participants to rate realism (1-7 scale)"
      - "Identify unrealistic elements via open-ended feedback"
      - "Revise to increase mundane realism"

standardization_protocols:
  presentation_mode:
    written: "Control font, size, line spacing, paragraph structure"
    audio: "Control speaker, tone, pace, volume"
    video: "Control duration, editing, production quality"

  counterbalancing:
    when_needed: "Multiple stimuli per condition (e.g., 4 vignettes)"
    method:
      - "Create multiple versions (ABCD, BCDA, CDAB, DABC)"
      - "Randomly assign participants to versions"
      - "Ensures order effects don't confound condition effects"

  stimulus_sets:
    requirement: "Use multiple exemplars of each condition"
    rationale: "Generalize beyond single stimulus"
    example: "Don't use 1 sad photo vs. 1 happy photo; use 10 of each"

quality_assurance:
  expert_review:
    - "Have subject matter experts review for accuracy"
    - "Have naive raters categorize stimuli (should match intended condition)"

  technical_checks:
    - "Test materials on multiple devices/browsers"
    - "Verify audio/video files play correctly"
    - "Check for broken links or missing images"

  accessibility:
    - "Provide alt text for images"
    - "Use readable fonts (minimum 12pt)"
    - "Ensure color contrast for visually impaired"

treatment_protocol:
  theoretical_rationale: |
    [Explain theoretical foundation and active mechanisms]

  session_structure:
    - session_1:
        title: "Session title"
        duration: "50 minutes"
        objectives:
          - "Objective 1"
          - "Objective 2"
        activities:
          - activity: "Activity description"
            duration: "15 min"
        homework: "Homework assignment"

  dosage_justification: |
    [Explain why this duration/frequency/intensity]

  implementer_training: |
    [Describe training procedures]

  fidelity_monitoring: |
    [Describe how adherence/competence will be assessed]

control_condition:
  type: "waitlist / attention control / active control / etc."
  rationale: |
    [Explain why this type of control is appropriate]

  matching_elements:
    - "Element 1 matched to treatment"
    - "Element 2 matched to treatment"

  differentiating_elements:
    - "What control condition does NOT include"

  ethical_justification: |
    [Address equipoise, beneficence concerns]

manipulation_checks:
  attention_checks:
    - item: "Attention check item"
      placement: "After manipulation / End of survey"
      pass_criterion: "Correct response"

  comprehension_checks:
    - item: "What was the main topic of the article?"
      placement: "Immediately after manipulation"
      analysis_plan: "Exclude if incorrect / Analyze separately"

  manipulation_success:
    - construct: "Anxiety"
      item: "How anxious did this make you feel? (1-7)"
      expected_difference: "High anxiety condition > Low anxiety condition"

stimulus_materials:
  vignettes:
    - condition: "Condition A"
      vignette_text: |
        [Full text of vignette]
      manipulation_element: "[What varies across conditions]"
      word_count: "250"

    - condition: "Condition B"
      vignette_text: |
        [Full text of vignette]
      manipulation_element: "[What varies]"
      word_count: "250"

  visual_materials:
    - type: "Photograph / Video / Diagram"
      source: "IAPS image #1234 / Original creation"
      manipulation: "Valence (positive vs. negative)"
      pilot_results: "Valence ratings: M=6.2 vs. M=2.1, t(18)=12.5, p<.001"

  pilot_testing_results:
    comprehension: "95% correctly identified manipulation"
    manipulation_check: "d = 1.2 on intended construct"
    realism: "M = 5.8/7 on believability"

model_tier: MEDIUM
model: sonnet
temperature: 0.3
thinking_mode: enabled

reasoning:
  - "Sonnet provides good balance for structured protocol design"
  - "Temperature 0.3 ensures consistency while allowing creativity in stimulus design"
  - "Extended thinking helps navigate ethical tradeoffs in control condition selection"

keywords:
  korean:
    - "처치"
    - "중재"
    - "통제 조건"
    - "개입"
    - "조작 점검"
    - "자극 자료"
    - "비네트"
    - "실험 자극"

  english:
    - "treatment"
    - "intervention"
    - "control condition"
    - "manipulation check"
    - "stimulus materials"
    - "vignette"
    - "experimental manipulation"
    - "attention check"
    - "fidelity"

context_patterns:
  - "User is designing an experiment or RCT"
  - "Discussion of intervention development"
  - "Questions about control group design"
  - "Requests for manipulation check examples"
  - "Need for stimulus material creation"

treatment_protocol:
  theoretical_rationale: |
    Mindfulness interventions reduce stress through:
    1. Attention regulation (sustained focus on present)
    2. Emotion regulation (non-judgmental awareness)
    3. Decentering (observing thoughts without identification)

    Evidence: Meta-analysis (Khoury et al., 2015) shows d=0.55 for stress reduction.

  session_structure:
    session_1:
      title: "Introduction to Mindfulness"
      duration: "45 minutes (self-paced online module)"
      objectives:
        - "Understand mindfulness definition and benefits"
        - "Learn basic breath awareness technique"
      activities:
        - activity: "Video: What is Mindfulness? (10 min)"
        - activity: "Guided practice: Breath awareness (10 min audio)"
        - activity: "Reflection journal: Notice one thing in present moment"
      homework: "Practice breath awareness 5 min/day, log practice"

    session_2:
      title: "Body Scan Meditation"
      # [Additional sessions...]

  dosage_justification: |
    - 6 weeks chosen based on Carmody & Baer (2009): 8 weeks is gold standard,
      but 6 weeks shows significant effects with lower attrition in students
    - 45-min modules: balance depth with student time constraints
    - Daily 5-15 min practice: sufficient for skill acquisition

  fidelity_monitoring:
    platform_tracking:
      - "LMS tracks module completion"
      - "Audio players track time listened"
    self_report:
      - "Weekly practice logs (minutes practiced per day)"
      - "Technique used (breath, body scan, etc.)"

recommended_control: "Active Attention Control"

rationale: |
  Given that mindfulness is an active skill-training intervention,
  a no-treatment or waitlist control would not match for:
  - Expectation of benefit (demand characteristics)
  - Time and attention from program
  - Engagement with online content

  An active attention control is preferred over waitlist.

control_condition_design:
  name: "Academic Success Skills Training"

  description: |
    6-week online program on study skills and time management.

  matched_elements:
    - duration: "6 weeks, weekly modules"
    - format: "Self-paced online modules with videos"
    - engagement: "Weekly homework assignments"
    - attention: "Weekly check-in emails (same as mindfulness group)"

  content_per_week:
    week_1: "Goal-setting and SMART objectives"
    week_2: "Time management and scheduling"
    week_3: "Note-taking strategies"
    week_4: "Reading comprehension techniques"
    week_5: "Test preparation and study habits"
    week_6: "Overcoming procrastination"

  differentiating_elements:
    what_control_lacks:
      - "No meditation or mindfulness practices"
      - "No focus on present-moment awareness"
      - "No emotion regulation training"

    what_control_includes:
      - "Cognitive strategies for academic performance"
      - "Plausible benefit (students value study skills)"

  credibility_assurance:
    - "Both programs presented as evidence-based"
    - "Both receive same amount of email support"
    - "Pilot test credibility/expectancy questionnaire to ensure equivalence"

ethical_note: |
  Offer mindfulness program to control group after study completion.
  Both programs have potential benefit, maintaining equipoise.

manipulation_checks:

  # 1. Comprehension check (did they pay attention?)
  attention_check:
    item: "What position was the candidate applying for?"
    options:
      - "Marketing Manager"  # Correct
      - "Software Engineer"
      - "Financial Analyst"
      - "I don't remember"
    placement: "After reading resume but before DV"
    analysis_plan: "Exclude participants who select incorrect option"

  # 2. Manipulation success check (did manipulation work?)
  perceived_competence:
    items:
      - "How competent does this candidate seem? (1=Not at all, 7=Extremely)"
      - "How qualified is this candidate for the position? (1-7)"
      - "How likely is this candidate to succeed in this role? (1-7)"

    composite_score: "Average of 3 items (α should be >.80)"

    expected_result: |
      High competence condition should score significantly higher than
      Low competence condition. If not, manipulation failed.

    analysis_strategy:
      - "t-test comparing conditions on manipulation check"
      - "Report in results: M_high=5.8, M_low=3.2, t(198)=12.4, p<.001, d=1.76"
      - "If manipulation failed (p>.05), discuss in limitations"

  # 3. Additional construct validation
  specific_dimensions:
    items:
      - "How much relevant experience does this candidate have? (1-7)"
      - "How strong are this candidate's credentials? (1-7)"

    purpose: "Verify that competence manipulation affected intended dimensions"

  # 4. Suspicion probe (did they detect manipulation?)
  awareness_check:
    item: "What do you think this study was investigating?"
    type: "Open-ended text response"
    coding:
      - "Naive: No mention of competence manipulation"
      - "Partially aware: Mentions candidate differences but not manipulation"
      - "Fully aware: Correctly identifies competence as manipulated variable"
    analysis_plan: "Analyze separately or exclude 'Fully aware' participants"

placement_sequence:
  1_after_manipulation: "Attention check (candidate position question)"
  2_before_DV: "Manipulation check items (competence, qualifications)"
  3_end_of_study: "Suspicion probe (open-ended awareness)"

reporting_template: |
  "Manipulation checks confirmed that participants in the high competence
  condition rated the candidate as significantly more competent (M=5.82, SD=0.91)
  than those in the low competence condition (M=3.15, SD=1.02),
  t(198)=12.43, p<.001, d=1.76."

works_with:
  - agent: "C1-research-design-strategist"
    interaction: "C1 defines research question → C4 develops materials to test it"

  - agent: "C2-sampling-power-analyst"
    interaction: "C4 provides effect size from manipulation pilot → C2 calculates sample size"

  - agent: "C3-measurement-psychometrician"
    interaction: "C4 develops manipulation checks → C3 validates reliability/validity"

  - agent: "D1-statistical-methodologist"
    interaction: "C4 provides fidelity data → D1 analyzes as moderator of treatment effects"

  - agent: "D4-ethics-irb-specialist"
    interaction: "C4 proposes control condition → D4 reviews for ethical concerns"

handoff_conditions:
  - "If user asks about sample size for detecting manipulation effect → D1-statistical-methodologist"
  - "If user asks about IRB approval for control condition → D4-ethics-irb-specialist"
  - "If user needs to validate manipulation check scale → C3-measurement-psychometrician"

key_references:
  treatment_development:
    - "Czajkowski, S. M., et al. (2015). From ideas to efficacy: The ORBIT model for developing behavioral treatments. Health Psychology, 34(S), 971-982."
    - "Rounsaville, B. J., et al. (2001). A stage model of behavioral therapies research. Clinical Psychology: Science and Practice, 8(2), 133-142."

  control_conditions:
    - "Mohr, D. C., et al. (2009). The selection and design of control conditions for randomized controlled trials of psychological interventions. Psychotherapy and Psychosomatics, 78(5), 275-284."

  manipulation_checks:
    - "Hauser, D. J., & Schwarz, N. (2016). Attentive Turkers: MTurk participants perform better on online attention checks than do subject pool participants. Behavior Research Methods, 48(1), 400-407."

  fidelity_assessment:
    - "Bellg, A. J., et al. (2004). Enhancing treatment fidelity in health behavior change studies. Health Psychology, 23(5), 443-451."

stimulus_databases:
  images:
    - name: "International Affective Picture System (IAPS)"
      url: "https://csea.phhp.ufl.edu/media.html"
      content: "Validated emotional images with normative ratings"

    - name: "FACES Database"
      content: "Standardized facial expressions of emotion"

  scenarios:
    - name: "Social Psychology Network Scenarios"
      content: "Vignettes for research on stereotyping, prejudice, etc."