Intraparietal stimulation disrupts negative distractor effects in human multi-alternative decision-making

Kohl, C.^1,2, Wong, M. X. M. ¹, Rushworth, M. F. S. ³ & Chau, B. K. H. ^{1, 4}

^{1 Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Hong Kong
2 Department of Neuroscience, Carney Institute for Brain Sciences, Brown University, Providence, United States
3 Department of Experimental Psychology, University of Oxford, Oxford OX1 3UD, UK
4 University Research Facility in Behavioral and Systems Neuroscience, The Hong Kong Polytechnic University, Hong Kong}

---

This repository contains behavioural data as well as code to replicate all main behavioural findings associated with the manuscript “Intraparietal stimulation disrupts negative distractor effects in human multi-alternative decision-making”.

---

Code

This repository contains three .m files which run all main behavioural analyses, using the data provided in the ‘Data’ directory.

• GLM1.m

  • This script fits GLM1 to all NonTMS trials
  • GLM1: β0 + β1 z(HV-LV) + β2 z(HV+LV) + β3 z(D-HV) + β4 z(HV-LV) z(D-HV) + ε
  • One-sample t-tests on the resulting beta weights show a negative (HV-LV)(D-HV) effect on accuracy
  • This script plots Figure 2c

• GLM2.m

  • This script fits GLM2 to all NonTMS trials, after applying a HV-LV median split
  • GLM2: Step 1, β0 + β1 z(HV-LV) + β2 z(HV+LV) + ε1 - Step 2, β3 + β4 z(D-HV) + ε2
  • One-sample t-tests on the resulting beta weights show a positive distractor effect on hard trials and a negative distractor effect on easy trials
  • This script plots Figure 2d

• GLM3.m

  • This script fits GLM3 separately to each condition ( MIP/MT x TMS/NonTMS x Contralateral D/Ipsilateral D)
  • GLM3: β0 + β1 z(HV-LV) + β2 z(HV+LV) + β3 z(D-HV) + ε
  • The resulting beta weights of each regressor are entered in a Session x Stimulation x Distractor Location ANOVA
  • This script plots Figure 3a-c

---

Data

The ‘Data’ directory contains one .mat file per participant (01-31). Each file contains a structure called ‘data’ with three fields:

• data.MIP contains data collected in the MIP session (270 x 18, rows = trials)

• data.MT contains data collected in the MT session (270 x 18, rows = trials)

• data.Key contains a cell with 18 strings, labelling each column in data.MT/MIP

  _{1: Trial_Nr Trial number (1-270)
  2: HV_value Value of the high-value option (value = magnitude x probability)
  3: LV_value Value of the low-value option (value = magnitude x probability)
  4: D_value Value of the distractor (value = magnitude x probability)
  5: HV_magnitude Magnitude of the high-value option (1-6)
  6: LV_magnitude Magnitude of the low-value option (1-6)
  7: D_magnitude Magnitude of the distractor (1-6)
  8: HV_probability Probability of reward associated with the high-value option (12.5-87.5)
  9: LV_probability Probability of reward associated with the low-value option (12.5-87.5)
  10: D_probability Probability of reward associated with the distractor (12.5-87.5)
  11: HV_position Position of the high-value option (1 = top left, 2 = top right, 3 = bottom left, 4 = bottom right)
  12: LV_position Position of the low-value option (1 = top left, 2 = top right, 3 = bottom left, 4 = bottom right)
  13: D_position Position of the distractor (1 = top left, 2 = top right, 3 = bottom left, 4 = bottom right)
  14: TMS TMS applied (1 = TMS, 0 = NonTMS)
  15: Decision Choice made (1 = HV chosen, 2 = LV chosen, nan = D/empty quadrant chosen)
  16: RT Reaction time in ms (max. 1500)
  17: Reward Reward achieved
  18: Accuracy Choice accuracy (1 = HV chosen, 0 = LV chosen, nan= D/empty quadrant chosen)}

---

Further information, code, and data may be available upon request. Please refer to the manuscript or contact: kohl.carmen.1@gmail.com or boltonchau@gmail.com