SPCD Task Overview

Sensory Preconditioning and Devaluation

Summary

The SPCD task is a multi-day, fMRI-compatible behavioral paradigm designed to study outcome devaluation and sensory preconditioning. It includes multiple phases such as cue-cue learning, cue-odor conditioning, value updating through devaluation, and recognition memory testing. Participants complete tasks both inside and outside the MRI scanner.

Participants are asked to fast for 2–4 hours before arrival to ensure consistent hunger state for the devaluation manipulation. The full study is completed over three sessions (Day 0, Day 1, and Day 2), typically scheduled within a 2-week window, although exceptions may occur in practice.

---

Day-by-Day Procedure

Day 0 – Baseline Rating & Odor Selection

• Image Rating Task
  Participants rate character and place images, which are later used as visual cues in the experiment.
• Initial Odor Rating & Selection
  Participants rate the pleasantness and intensity of several odors. Two odors—one savory and one sweet—are selected for each participant based on these ratings.
• Follow-Up Odor Rating
  Participants re-rate the same odors to ensure consistency and confirm odor selection.
• Odor Discrimination Task
  Participants judge whether two presented odors are the same or different.

---

Day 1 – Pre-conditioning & Localizer (fMRI)

• Localizer Task (in scanner)
  Participants undergo fMRI scans to measure neural responses to individual images and odors, before any associative learning has occurred.
  • odor runs
  • character runs
  • place runs
  • anatomical scan
• Pre-conditioning Training (outside scanner)
  Participants learn associations between characters and places.

---

Day 2 – Conditioning, Probe, Devaluation

• Pre-conditioning Refresher (outside scanner)
  A brief session to reinforce previously learned character-place associations.
• Odor Prediction Task (in scanner)
  Participants predict which odor follows a given conditioned cue.
• Pre-meal Probe (in scanner)
  Tests participants’ responses to pre-conditioned and conditioned cues, without any odor delivery.
• Meal & Devaluation Manipulation
  • Pre-meal odor ratings and questions.
  • Participants consume a meal flavored with one of the two previously conditioned odors.
  • Post-meal odor ratings and questions.
• Post-meal Probe (in scanner)
  Identical to the pre-meal probe but administered after devaluation.
• Recognition Memory Task (outside scanner)
  Participants judge whether specific character-place pairs were seen during pre-conditioning.
• Post-experiment Interview
  A short debriefing or follow-up assessment to gather subjective feedback or check for awareness.

---

Task Timing and Run Structure Details

This section summarizes the trial timing, run structure, and duration for each task used in the SPCD paradigm.

Day 1 (MRI Localizer + Pre-conditioning)

Odor Scan (in scanner)

• Trial: 2s sniff cue → 2s odor → 9–13s ITI
• Structure: 3 odors × 10 repetitions = 30 trials per run
• Duration: ~7.5 min per run
• Runs: 2 (320 volumes/run)

Character Image Scan

• Trial: 2s image cue → 2s image → 5–8s ITI
• Structure: 8 images (6 task + 2 filler) × 4 repetitions = 32 trials
• Duration: ~5.6 min per run
• Runs: 4 (244 volumes/run)

Place Image Scan

• Same as above

Pre-conditioning Training (outside scanner)

• Trial: 3s first cue → 0.3s gap → 3s second cue (choice) → 3–8s ITI
• Structure: 6 cue pairs × 2 alternatives × L/R = 24 trials
• Duration: ~5 min per run
• Runs: 3–5; task may end after the 3rd or 4th run if ≥90% accuracy is achieved.

---

Day 2 (Refresher + Conditioning + Devaluation + Memory)

Pre-conditioning Refresher (outside scanner)

• Same structure as Day 1
• 1–3 runs (max 3); task may end after 1 run if ≥90% accuracy is achieved.

Odor Prediction (in scanner)

• Trial: 2s cue → 3s prediction → 2s delay → 2s odor delivery → 4–10s ITI
• Structure: 6 cues × 6 orderings = 36 trials
• Duration: ~10 min per run
• Runs: 3 (404 volumes/run)
• Notes: Each cue appears 18 times across all runs

Pre-meal Probe (in scanner)

• Trial: 3s cue1 → 4–10s gap → 3s cue2 → 2s decision → 4–10s ITI
• Structure: 6 pairs × 2 sets × L/R = 24 trials
• Duration: ~9 min per run
• Runs: 3 (372 volumes/run)
• Notes: Each cue appears 12 times across all runs

Meal Time

• ~15 min meal

Post-meal Probe (in scanner)

• Identical to pre-meal probe

Recognition Memory (outside scanner)

• Trial: 3s cue1 → 0.3s gap → 3s cue2 → 3–8s ITI
• Structure: 6 cue pairs × old/foil × 3 repetitions = 36 trials
• Duration: ~7 min
• Runs: 1

---

Odor–Food Mapping

The table below lists the specific foods that were used to devalue each odor.

Odor	Food(s)
Potato Chip	Potato chips
Garlic	Minced garlic jars, Naan
Cheddar	Mac n cheese, Goldfish crackers, Cheeze-Its
Pizza	Frozen pizzas, Tomato sauce jars
Tuna	Canned tuna, Crackers
Strawberry	Strawberry strudels
Chocolate	Clif bars
Apple cinnamon	Nutrigrain bars, Apple crisps
Peach	Fruit cups, Peach strudels
Vanilla	Vanilla cookies

---

Notes on data issue handling

This only documents issues that may affect later analysis. Some abnormal data collection events are not noted here if they have already been well handled and do not impact downstream analyses. Most issue-related corrections (e.g., scan truncation, trigger adjustments) are handled in the Do_5_process_labchart.m script.

• Sub-35 – Day 2, Odor Prediction Run 3
  The scan did not complete all expected volumes. Processed the LabChart trigger data to use only the partial scan. Also corrected a minor mismatch with the number of motion parameters.

• Sub-36 – Day 2, Pre-Meal Probe Run 3
  The scan was interrupted mid-run, resulting in a short initial recording followed by a second, longer segment. Both were excluded from analysis to avoid alignment issues, since the behavioral task was not paused.

• Sub-14 – Day 2, Meal

  • Subject was initially assigned to the savory–devaluation group, but due to a lack of pizza in storage, was fed the sweet meal instead.
  • No code changes were made to preserve consistent labeling across tasks.

• Sub-33 – Day 2, Post-Meal Probe Task, Run 3

  • Breathing signals were missing, likely due to a connection issue.
  • No correction was applied; breathing signals were extracted as usual.