Effects of hunger on model-based and model-free decision-making

van Swieten MMH
Manohar SG
Bogacz R
Description

In this study, healthy participants were tested on the effects of hunger on model-free and model-based decision-making. Each first-stage choice rocket was predominantly associated with one of the second-stage planets (common transition: 70% of the trials) and sometimes with the other second-stage planet (rare transition: 30% of the trials). Each second-stage alien probabilistically lead to a reward. The reward probabilities for each second-stage alien fluctuated across trials between 25% and 75% according to a Gaussian random walk.
Data Two Step (excel file) contains the following fields:

Sheet 1: Behavioural data 
Subject:                Number of participant
Condition:            Hunger (1) or Sated (2)
Trial:                     Trial number
Block:                   Block number 1-4
Choice stage 1:    Left (1) or right (2)
RT stage 1:           reaction time at stage 1 in ms
Stage 2 state:        indicated the second stage planet
Transition type:      common (1) or rare (0)
Choice stage 2:     Left (1) or right (2)
RT stage 2:           reaction time at stage 2 in ms
Reward:                rewarded (1) or unrewarded (0)
Total Reward:            Running total number of points
Stim order stage 1:        [1,2] or [2,1] indicates whether the stimulus 1 or stimulus 2 was presented on the left or right.
Stim order stage 2:        1,2] or [2,1] indicates whether the stimulus 1 or stimulus 2 was presented on the left or right.
Reward probabilities:        Reward probabilities of the second stage planet (S1 or S2) and alien (A1 or A2)

Sheet 2: Study details
Subject:             number of participant
Session 1:         hunger (1) or sated (2)
Session 2:        hunger (1) or sated (2)
Age:                 age of participant
Gender:            male (1) or female (2)
Height:             height in m
Weight:             weights in kg
Highest level of education
BMI:                 BMI of participant in kg/m2

Sheet 3: Processed data
Subject:            Number of participant
Condition:         Hunger (1) or Sated (2)
CR:                  Stay probability for Common-rewarded trials
RR:                  Stay probability for Rare-rewarded trials
CU:                  Stay probability for Common-unrewarded trials
RU:                  Stay probability for Rare-Unrewarded trials
MB_index:        Index for model-based behaviour
MF_index:        Index for model-free behaviour
Missed trials:    Number of missed trials per session
TotalPoints:       Total number of points earned
MedianRT stage 1:         Median reaction time for stage 1 (ms)
MedianRT stage 2:         Median reaction time for stage 2 (ms)

Sheet 4: Parameters Gaussian
Subject:            Number of participant
Condition:            Hunger (1) or Sated (2)
bMB:                Gaussian distributed model-based weighting parameter
bMF:                Gaussian distributed model-free weighting parameter
b_2:                Gaussian distributed softmax parameter for stage 2
a_1:                Gaussian distributed learning rate parameter for stage 1
a_2:                Gaussian distributed learning rate parameter for stage 2
a_21:                Gaussian distributed learning rate parameter for eligibility traces
p:                Gaussian distributed perseverance parameter    

Sheet 5: Parameters transformed
Subject:            Number of participant
Condition:            Hunger (1) or Sated (2)
beta_MB:            Transformed model-based weighting parameter
beta_MF:            Transformed model-free weighting parameter
Beta_2:                Transformed softmax parameter for stage 2
alpha_1:            Transformed learning rate parameter for stage 1
alpha_2:            Transformed learning rate parameter for stage 2
alpha_21:            Transformed learning rate parameter for eligibility traces
pi:                Transformed perseverance parameter    

PlotFigures (matlab file):
Plots the results figures used in the paper

Stay-switch behaviour at first stage. A) Food deprived participants are more likely to stay after receiving a reward than after reward omission. B) Sated individuals show less model-free behaviour. C) Change in stay probability with food deprivation. Food deprivation increased the tendency to stay after receiving a reward and increased the tendency to switch after reward omission. Error bars represent SEM.
Year Published
2021
DOI
10.5287/bodleian:R5Emr4pov
Funders & Grant Numbers
Medical Research Council, UKRI (MC_ST_U16043)
Medical Research Council, UKRI (MC_UU_12024/5)
Medical Research Council, UKRI (MC_UU_00003/1)
Medical Research Council, UKRI (MR/P00878/X)
Biotechnology and Biological Sciences Research Council, UKRI (BB/S006338/1 )
Publisher
University of Oxford
Group
Bogacz Group
Terms

Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA 4.0)

This is a human-readable summary of (and not a substitute for) the licence.

You are free to:

Share — copy and redistribute the material in any medium or format

Adapt — remix, transform, and build upon the material for any purpose, even commercially.

This licence is acceptable for Free Cultural Works. The licensor cannot revoke these freedoms as long as you follow the license terms. Under the following terms:

Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.

ShareAlike — If you remix, transform, or build upon the material, you must distribute your contributions under the same licence as the original.

No additional restrictions — You may not apply legal terms or technological measures that legally restrict others from doing anything the licence permits.