Merged study

What are the 3 core executive functions?

Working memory, inhibitory control, and cognitive flexibility.

Why is 'working memory' preferred over 'short-term memory'?

Working memory is dynamic, involving active manipulation rather than passive storage.

Is there a single brain region dedicated to working memory?

No, working memory is distributed, modality-dependent, and recruits sensory-specific regions.

What is the key takeaway from auditory and visual WM decoding studies?

WM content localizes to specific regions, with trial lengths varying by neural recording method.

What is inhibitory control, and what is its behavioral signature?

Suppressing distractors; signaled by slower, less accurate responses on incongruent trials.

Why is an incongruent trial handled faster after another incongruent trial (II vs CI)?

Prior conflict primes the brain, leading to cognitive adaptation and reduced conflict cost.

What did the Flanker EEG decoding study show?

Current-trial congruency decoded well with a medial-frontal peak; previous-trial congruency did not.

What common confound affects cognitive control decoding studies?

Decoding motor response or perceptual differences rather than the actual cognitive control process.

How does cognitive flexibility differ from inhibitory control?

Cognitive flexibility involves adapting to rule changes, which develops later and is more abstract.

How do WCST and task-switching paradigms differ structurally?

WCST requires inferring implicit rules from feedback; task-switching uses explicit, cued rules.

When do bilinguals show a performance advantage in task switching?

Specifically on switch trials, not on single-task control blocks.

How can higher language proficiency increase performance while reducing brain activation?

Neural efficiency: practiced inhibitory systems achieve better task performance with less neural effort.

Do all core executive functions mature at the same rate?

No, working memory and inhibitory control mature earlier than cognitive flexibility.

What are the 3 forms of cognitive offloading?

Offloading onto the body, world/others, or technology/AI as task difficulty increases.

What is the core critique of current AI-and-cognition research?

Studies are often correlational, lack proper controls, and conflate distinct cognitive constructs.

Why compare baseline demographics before testing EF differences across groups?

To rule out demographic confounds that could explain differences instead of target group variables.

What is the primary rule of spatial and temporal trade-offs in neuroimaging?

No method achieves both high spatial resolution and high temporal resolution simultaneously.

Why do EEG and fMRI have opposite spatial and temporal strengths?

EEG measures direct millisecond electrical signals; fMRI measures slow, spatially precise blood-oxygen responses.

Which two neuroimaging methods dominate experimental cognitive psychology?

EEG and fMRI, because they are non-invasive and practical for human research.

Why do neural compensation mechanisms undermine lesion studies?

Surrounding brain regions reorganize to take over lost functions, obscuring original local function.

What is volume conduction in EEG?

Electrical signal spreading through the skull, meaning electrodes record broad activity, not just local tissue.

How does EEG frequency relate to mental state?

Lower frequencies indicate passive states; higher frequencies reflect active task engagement and decision-making.

Why is multivariate decoding often more sensitive than traditional ERP analysis?

Decoding evaluates subtle, distributed patterns across multiple signals rather than single component amplitudes.

What is the purpose of a searchlight analysis?

It slides a local decoding window across the brain to map spatial accuracy.

Why are overlapping sliding time windows used in time-resolved EEG decoding?

To capture continuous signals without boundary loss and reduce noise artifact effects.

What was the key finding of the own-race face EEG decoding study?

Same-race faces decoded with higher accuracy and peaked later than unfamiliar other-race categories.

How do researchers mitigate participant fatigue in cognitive experiments?

By counterbalancing block orders, keeping sessions short, and using strict attention checks.

What are the two major categories of research statistics in cognitive neuroimaging?

Group difference comparisons and association/prediction analyses.

When should you use a one-sample t-test in neural decoding?

To compare classifier accuracy against a theoretical chance level baseline.

What should you do if neural data violates normality assumptions?

Use non-parametric alternative tests like the Wilcoxon signed-rank or Mann-Whitney U.

Why must effect size be reported alongside p-values?

P-values confirm statistical significance, whereas effect size quantifies the actual magnitude of effect.

Why are multiple comparisons corrections necessary in neuroimaging analyses?

Testing thousands of electrodes or voxels inflates false positive rates.

What unique insight does an ANOVA provide that multiple t-tests cannot?

ANOVA reveals interaction effects between multiple independent variables.

What is the main limitation of simple bivariate correlation?

It cannot control for third variables driving the observed relationship.

How do regression and decoding/classification differ fundamentally?

Regression predicts continuous outcome values; decoding predicts discrete categorical labels.

Why is standard linear regression rarely applied directly to neural signals?

Neural signals are complex and non-linear, violating linear regression model assumptions.

What key methodological issues does study pre-registration address?

It prevents p-hacking and HARKing by fixing hypotheses before data collection.

What does a Bayes Factor quantify?

The relative strength of evidence supporting one hypothesis over an alternative hypothesis.

Why is neural reconstruction strictly distinct from 'mind reading'?

Reconstruction recovers biased internal representations, constrained by algorithm design and training datasets.

What is the main vulnerability of simple regression-based image reconstruction?

Overfitting to training data, resulting in poor generalization to new stimuli.

What is overfitting in machine learning?

When a model fits training noise so closely that it fails on new data.

How does RSA/MDS-based reconstruction differ from direct regression reconstruction?

RSA captures relational similarity across all stimuli rather than reconstructing items in isolation.

What is the two-step approach for validating neural stimulus reconstructions?

First establish physical accuracy, then evaluate higher-level psychological or relational distortions.