Belief bias in Conditional Reasoning. - Plymouth University

Within Within--Trial Repetitive Transcranial Magnetic Stimulation Affects Belief Bias in Reasoning M. E. Roser ., J. S. T. Evans ., L. S. Carroll ., N. A. McNair ., G. Fuggetta ., and A. Y. Kharko 1

1

1

2

3

1

82.13

School of Psychology, Plymouth University, UK. University of Sydney, AUS University of Leicester, UK.

1

2

3

[1] OVERVIEW

[5] PROCEDURE:

Aim: To investigate the effect of rTMS on belief bias in conditional reasoning. Specifically, whether selectively impairing brain regions implicated in reasoning and belief may modulate belief bias.

Three-term conditional statements. Participants were instructed to assume all information in the premises was true and to determine whether the conclusion followed according to logic. Pared-down statements and word-by-word presentation was used to force reading speed and provide the temporal specificity needed for time locking rTMS stimulation to immediately after the premises. Six blocks of 24 trials each (half un/believable). Logical forms (4) MP, MT, AC, DA, balanced across believability within each block. One TMS location per block (order randomised). TMS 10Hz for 500ms at 80% of Active Motor Threshold (AMT) on each trial immediately after the presentation of the major OR minor premise.

[3] OUR PREVIOUS fMRI CONDITIONAL REASONING STUDY: We investigated the brain areas involved in conditional reasoning and belief bias. Our previous fMRI study found activation in the right inferior frontal cortex (IFG; implicated in inhibitory processing) at the major premise stage and the addition of activation in the left dorsolateral prefrontal cortex (DLPFC; involved in working memory) when the whole argument was presented (minor premise) which correlated with degree of belief bias. P1 > Control task

PAll > Control

[4] CURRENT rTMS STUDY: Based on our fMRI results, we targeted 6 locations for rTMS: Left IFG, Left MFG, Right IFG, Right MFG, Vertex stimulation (Cz), Vertex sham (Cz*; note: these included lateral homologues of fMRI findings and two control locations). We applied within-trial rTMS to these areas, manipulating the stage of processing targeted between subjects. Functionally active regions from our fMRI were de-normalised into each individual’s brain space as TMS target locations.

Middle: Brainvoyager TMS real-time neuronavigation guide Right: The experimental set-up

[6] THE TASK: (Unbelievable modus ponens example) IF DEAD THEN BREATHING [Major premise] DEAD [Minor premise] BREATHING [Conclusion] Does the conclusion logically follow?

[7] RESULTS Overall results showed little belief bias or evidence for experimental rTMS effects. However, this may be as our study differed from orthodox reasoning paradigms due to methodological changes necessitated to run the study with TMS (e.g., pared-down structure, TMS distraction, much greater number of trials). The extensive number of trials and practise effects resulted in little or no belief bias past block 4. Therefore, we focused on data from the first 3 blocks only as this more closely replicated traditional reasoning studies.

Belief bias by block 8 6 4 2 0 -2

Block 1 Block 2 Block 3 Block 4 Block 5 Block 6

-4 -6

Block order

[10] REFERENCES: [1] Evans, J. S. B. (2008). Dual-Processing accounts of Reasoning, Judgement, and Social Cognition. Annu. Rev. Psychol., 59, 255-278 [2] Handley, S., Capon, A., Beveridge, M., Dennis, I., & Evans, J. S. B. (2004). Working Memory, Inhibitory Control and the Development of Children’s Reasoning. Thinking and Reasoning, 10, 175-195

Left: Individual extracted brain image with TMS target locations mapped

[8] DATA FROM BLOCKS 1-3: Exclusions:  Blocks 1-3 only  Individual trials with RTs>3SDs from each participant’s individual overall RT removed  Data for participants for whom we did not get a reliable AMT  Block/location data if TMS intensity was dropped below 40%

Belief bias by TMS location 16.00 14.00 12.00 10.00 Belief Bias

Decision making is influenced by cognitive biases which may impair reasoning ability. We studied Belief Bias in conditional reasoning. Belief bias is the tendency to endorse more inferences derived from conditional statements if they are believable, regardless of logical validity. Dual process theories posit two distinct cognitive systems: Type 1 processes occur quickly and are automatic, whereas Type 2 processes are slower and require effort [1]. Default interventionist theories state that in order to perform well in a reasoning task in which there is conflict between logic and belief, responders must inhibit and overcome initial intuitive belief-based judgements (type 1) in favour of more systematic logical processing (type 2). This is effortful and requires cognitive resources and working memory (WM) [2]. Individuals with high IQ and WMC typically perform better at these tasks. The roles of inhibitory and WM processes in determining individual differences remain unresolved.

Belief bias

[2] BACKGROUND:

8.00 Major Premise (T1)

6.00

Minor Premise (T2)

4.00 2.00 0.00 Left IFG

Left MFG

-2.00

Right MFG

Mean Sham

TMS Location

[9] CONCLUSION: Results differed from predictions, particularly the lack of Right IFG and Left DLPFC rTMS effect. rTMS to the left IFG increased susceptibility to belief bias, suggesting that the left IFG in particular is involved in the moderation of belief bias and plays a central part in reasoning. rTMS to the left IFG may have disrupted the analytical (type 2) processing of conditionals at the major and minor premise stage of the task thus increasing the likelihood of participants reverting to automatic (type 1) processes and giving a default heuristic response which may have been biased by the believability of the statements. These results are consistent with a model of concurrent processes of inhibition and integration rather than discreet stages.

WEBSITE/MORE INFO: This work was supported by the Economic and Social Research Council Grant: RES- 062-23-3285

Right IFG

www.psy.plymouth.ac.uk/research/mattroser