METHODS

Part 2 - Methods

For Methods, you want to know what was done, why it was done, what each technical term means, and what you personally did versus what was already done in the wider project.

The overall pipeline is:

Recruit adolescents → YPI → EEG preparation → CS2 gameplay + continuous EEG → preprocessing → extract alpha power → calculate F4/F3 and F8/F7 FAA → merge with YPI → statistical analysis

1. Study design

Your study used a:

Cross-sectional correlational design involving secondary analysis of EEG data from a broader adolescent violent-videogaming project.

Cross-sectional

Variables were examined at one period in time rather than following participants longitudinally.

Correlational

You examined whether naturally occurring individual differences in GM, CU and II were associated with FAA.

You did not manipulate psychopathic traits.

Therefore, you cannot establish causality.

Secondary analysis

The EEG/videogaming data had already been collected as part of a broader research project. Your dissertation used those existing data to answer a new question about psychopathic traits and FAA. FINAL PSYCHOPATHIC DISSERTATION

Variables

Predictors:

• GM

• CU

• II

Outcomes:

• F4/F3 FAA - primary

• F8/F7 FAA - exploratory.

2. What did YOU actually do?

This is one of the most important distinctions to remember.

You did not personally preprocess the raw EEG and calculate the FAA values from scratch.

For your secondary analysis, you:

cleaned and merged the sociodemographic, FAA and YPI datasets, excluded participants with missing FAA data, scored the YPI subscales and calculated the dimension and total scores, and conducted the statistical analyses. FINAL PSYCHOPATHIC DISSERTATION

So if you're asked:

"Did you collect the EEG data yourself?"

"No. This was a secondary analysis of data collected as part of a broader project. I cleaned and merged the relevant datasets, scored the YPI and calculated its dimension and total scores, and conducted the statistical analyses. I therefore needed to understand the EEG acquisition and preprocessing pipeline, but I did not personally derive the FAA values from the raw EEG."

That distinction makes you sound more, not less, knowledgeable because you're accurately representing your contribution.

3. Participants - know the numbers

The broader project originally collected data from:

30 adolescents.

Then:

9 excluded for poor EEG signal quality/excessive noise

↓

N = 21

aged:

12-17 years

M age = 15.00

SD = 1.38. FINAL PSYCHOPATHIC DISSERTATION

Then during your statistical screening:

Case 17 was excluded

↓

Final analytic sample = N = 20

Keep these two types of exclusion separate:

EEG quality exclusions: 9 participants before your analysis.

Statistical exclusion: Case 17 after initial screening.

4. Original N = 21 characteristics

Before the statistical exclusion:

15 male - 71.4%

5 female - 23.8%

1 preferred not to disclose - 4.8%.

Ethnicity:

16 White - 76.2%

2 Mixed/multiple - 9.5%

2 Asian/Asian British - 9.5%

1 Black/Black British/Caribbean/African - 4.8%. FINAL PSYCHOPATHIC DISSERTATION

You probably don't need to memorise every percentage. More important:

The sample was predominantly male and White.

That matters later for generalisability.

5. Their videogaming experience

This wasn't a sample of adolescents who had never encountered videogames.

Weekly PC/console gaming at N=21:

• <5 hours: 3

• 5-<10 hours: 4

• 10-<15 hours: 5

• 15-<20 hours: 0

• ≥20 hours: 9 (42.9%)

Primary device:

PC = 9

console = 12. FINAL PSYCHOPATHIC DISSERTATION

All participants reported that at least some of their gaming time involved violent videogames.

Why could gaming experience matter?

More experienced gamers may differ in:

• familiarity with FPS controls

• performance

• cognitive load

• emotional engagement

• arousal.

The practice periods and standardised protocol help reduce some variability, but they cannot completely eliminate individual differences in gaming experience.

6. Recruitment

Participants were recruited:

directly + through social-media advertisements

targeting adolescents aged:

12-17 years.

Eligibility was assessed using a screening form before participation. FINAL PSYCHOPATHIC DISSERTATION

7. Exclusion criteria

Adolescents were excluded if they had:

neurodevelopmental difficulties

uncorrected visual/hearing difficulties

or:

long-term physical or mental health conditions that could interfere with the study tasks.

Parents/carers involved in the wider project were also screened for learning or psychiatric difficulties that could affect their ability to provide informed consent. FINAL PSYCHOPATHIC DISSERTATION

Why have exclusion criteria?

Primarily to reduce factors that could:

• interfere with task completion

• influence EEG measurement

• compromise data quality

• make participation inappropriate

• interfere with valid informed consent.

Don't say they were designed to make the sample "normal".

8. Ethics

Ethical approval:

UCL Research Ethics Committee

ID 27381/002. FINAL PSYCHOPATHIC DISSERTATION

Because participants were minors, ethical considerations are particularly important.

Before attending:

adolescent + parent/carer received written study information.

Electronic consent was obtained from:

adolescent + parent/carer.

And on the day:

adolescent verbal assent was obtained. FINAL PSYCHOPATHIC DISSERTATION

Consent vs assent

Consent = informed agreement to participate.

Assent = the young person's affirmative agreement to participate, particularly relevant when working with minors.

Your study actually obtained adolescent consent as well as parent/carer consent, with additional verbal assent on the day.

9. Why is EEG ethics important?

EEG is generally considered a safe, non-invasive neurophysiological technique, but researchers still need to:

• minimise discomfort

• explain procedures appropriately

• protect sensitive biological data

• maintain confidentiality/anonymisation

• monitor the participant.

Your additional EEG material emphasises that EEG data are biological and sensitive personal data, requiring appropriate anonymisation, data protection and security. 05 EEG-ERP_Data collection

10. Violent content and distress

Because adolescents were exposed to violent game content, your protocol included an additional safeguard:

Participants watched a calming video following the session to reduce potential distress associated with exposure to violent gaming content.

The principal investigator was also available if additional support was required. FINAL PSYCHOPATHIC DISSERTATION

That's a nice detail to know if you're asked specifically about ethics.

11. YPI - Youth Psychopathic Traits Inventory

You need to know this measure very well because it produced all three predictor variables.

YPI = Youth Psychopathic Traits Inventory

Developed by:

Andershed et al. (2002).

It is a:

50-item self-report questionnaire

containing:

10 subscales

with:

5 items per subscale.

Those 10 subscales form:

3 higher-order dimensions. FINAL PSYCHOPATHIC DISSERTATION

12. Exact YPI structure

GM - 4 subscales

Dishonest Charm

Grandiosity

Lying

Manipulation

So:

4 × 5 items = 20 GM items.

CU - 3 subscales

Callousness

Unemotionality

Remorselessness

So:

3 × 5 = 15 CU items.

II - 3 subscales

Impulsiveness

Irresponsibility

Thrill-Seeking

So:

3 × 5 = 15 II items. FINAL PSYCHOPATHIC DISSERTATION

Memory trick

GM = 4

CU = 3

II = 3

4 + 3 + 3 = 10 subscales

10 × 5 = 50 questions.

13. How was the YPI scored?

Every item uses a:

4-point Likert scale

from:

1 = does not apply at all

to:

4 = applies very well.

Three Callousness items are reverse-scored:

items 23, 35 and 49.

Each five-item subscale therefore ranges:

5-20.

Relevant subscales are summed to produce the higher-order:

GM, CU and II scores.

Higher score:

greater level of that psychopathic trait dimension. FINAL PSYCHOPATHIC DISSERTATION

14. Possible examiner trap - why reverse-score items?

To ensure that the direction of scoring is consistent.

For example, if an item is phrased in the opposite direction to the construct, reverse scoring means that after recoding:

higher scores consistently represent more of the trait being measured.

15. Why use the YPI?

There are several good reasons.

It was specifically developed to assess psychopathic traits in youth and allows psychopathy to be examined dimensionally rather than assigning adolescents a categorical "psychopath" label.

It also directly provides the:

GM + CU + II

structure required by your theoretical question.

The YPI has shown generally adequate:

reliability + validity

in adolescent samples. FINAL PSYCHOPATHIC DISSERTATION

16. YPI reliability - know the important numbers

Hachtel et al. (2025):

GM α = .89

CU α = .70

II α = .77.

So broadly:

GM = good

CU = adequate

II = acceptable/adequate.

However, previous work found weaker reliability for:

CU α = .60-.66

and particularly:

Callousness α = .54. FINAL PSYCHOPATHIC DISSERTATION

Why does this matter?

Lower reliability means more measurement error.

That can attenuate associations and could potentially contribute to your null CU result.

But don't say:

"CU was non-significant because the measure was unreliable."

You don't know that.

Say:

"Previous evidence of weaker CU reliability could have reduced sensitivity to detect an association, although this cannot be established from the current study."

17. Self-report - strength and limitation

Strength

Efficient and standardised way to assess subjective/personality characteristics across adolescents.

Limitation

Psychopathic traits include characteristics such as:

lying + manipulation + limited insight

which creates an obvious potential issue for self-report.

Responses can also be affected by:

• social desirability

• response style

• self-awareness.

That's why one of your future recommendations is to supplement self-report with something like:

PCL:YV - Psychopathy Checklist: Youth Version

or collateral/interview information.

18. The CS2 task

Participants played:

Counter-Strike 2

for:

30 minutes.

This consisted of:

three × 10-minute Deathmatch rounds.

Each round used:

a different map.

And:

map order was randomised across participants. FINAL PSYCHOPATHIC DISSERTATION

Why randomise map order?

To reduce systematic order effects.

If everyone played exactly the same map sequence, differences related to fatigue, practice or adaptation could become confounded with a particular map.

Randomisation distributes those effects more evenly.

19. What did Deathmatch involve?

Participants continuously engaged opposing players with the goal of:

achieving as many kills as possible. FINAL PSYCHOPATHIC DISSERTATION

That gives you a task involving:

• continuous engagement

• goal pursuit

• rapid responding

• competition

• repeated success/failure

• behavioural adaptation.

That's why it is useful for studying FAA during a sustained motivational context.

20. Practice periods

Before each round, participants completed approximately:

2-3 minutes of practice

on the corresponding map.

This allowed them to:

• familiarise themselves with the map

• familiarise themselves with controls

• adjust mouse sensitivity

• adjust game volume where needed. FINAL PSYCHOPATHIC DISSERTATION

Why is practice important?

It reduces variation caused simply by:

not knowing how to control the game.

Without practice, novice participants might spend early gameplay learning the controls while experienced players are already actively competing.

21. Breaks

Short breaks were permitted:

between rounds

but:

kept to a minimum. FINAL PSYCHOPATHIC DISSERTATION

Why?

You need some accommodation for comfort/fatigue, particularly with adolescents and an EEG net.

But excessive or inconsistent breaks could disrupt the continuity of the task.

22. Standardising CS2

The game protocol was standardised across participants.

They played:

Deathmatch

on:

Workshop Maps

as:

Counter-Terrorists

using:

M4A4.

Keyboard settings prevented:

bonus/alternative weapon selection

and:

random weapon selection was disabled. FINAL PSYCHOPATHIC DISSERTATION

Why?

Experimental control.

If different participants use substantially different weapons or game configurations, you introduce unnecessary variation into the experience.

So standardisation helps ensure differences in FAA are less likely to reflect arbitrary differences in game setup.

23. Why was the researcher in the room?

The researcher remained present to:

monitor the participant + monitor EEG signal

and intervened only if:

assistance/adjustment was required.

But they stayed:

outside the participant's visual field where possible

to minimise interference with gameplay. FINAL PSYCHOPATHIC DISSERTATION

That's a balance between:

data/participant monitoring

and:

reducing researcher influence/distraction.

24. Overall procedure

Know this sequence:

Study information

↓

adolescent + parent/carer consent

↓

verbal assent on day

↓

YPI + sociodemographic questionnaire

↓

EEG preparation

↓

other wider-project tasks

↓

30-minute CS2 + continuous EEG

↓

debrief

↓

calming video

↓

£10 adolescent voucher + £5 family payment.

Total broader laboratory session:

approximately 90 minutes. FINAL PSYCHOPATHIC DISSERTATION

25. Now EEG - start with the absolute basics

EEG = electroencephalography.

It measures electrical potential differences at the scalp resulting primarily from summed postsynaptic activity in populations of cortical neurons.

An electrode does not independently measure "the voltage of the brain underneath it".

Voltage is a difference in electrical potential between points.

That's why EEG requires a reference. Your additional EEG material explicitly emphasises that voltage always involves two points and that an active electrode is compared with a reference. 05 EEG-ERP_Data collection

26. What is a reference electrode?

A reference provides the comparison point against which the voltage at active electrodes is measured.

Your recording used:

Cz as the online reference.

So during acquisition, scalp channels were initially measured relative to Cz.

But importantly:

The data were later re-referenced to the average of retained channels during preprocessing.

These are not contradictory.

Online reference = Cz during recording

Offline reference = average reference after preprocessing.

27. Why does reference choice matter?

Because EEG voltage is relative.

Changing the reference can alter the apparent waveform/amplitude across channels. Your supplementary EEG material explicitly notes that waveforms change depending on the reference. 05 EEG-ERP_Data collection

This is why referencing isn't just a meaningless technical detail.

28. Why use average reference?

In high-density EEG, you have electrodes distributed widely across the scalp.

Average referencing uses the average activity across retained electrodes as the reference.

Your supplementary material notes that average referencing is commonly relevant to high-density systems, although its effectiveness depends on electrode density and distribution. 05 EEG-ERP_Data collection

With a dense 256-channel system, this is a reasonable approach.

29. What EEG system was used?

Hardware

256-channel NetAmps system

from:

Electrical Geodesics Inc. - EGI.

Recording software

NetStation v4.5.2.

Sampling rate

250 Hz.

Online reference

Cz.

Impedance

maintained:

<50 kΩ. FINAL PSYCHOPATHIC DISSERTATION

Know those four numbers/details.

30. Why 256 channels?

This is a high-density EEG system.

Compared with sparse electrode systems, high-density recording gives greater:

spatial sampling across the scalp.

It also facilitates:

• electrode clustering

• interpolation of noisy channels using neighbours

• potential source-localisation approaches.

Your supplementary material specifically notes that high-density EGI systems facilitate cluster extraction, interpolation and source localisation. 05 EEG-ERP_Data collection

But more electrodes aren't automatically perfect

Potential issues include:

• more opportunities for noisy channels

• inter-electrode bridging

• increased complexity

• cluster "dilution" if noisy/irrelevant electrodes are included. 05 EEG-ERP_Data collection

31. What type of electrodes?

The EGI system used a:

semi-dry electrode net.

Before application, it was soaked in a:

water-based solution

for approximately:

15-17 minutes.

It was then positioned on the participant's scalp. FINAL PSYCHOPATHIC DISSERTATION

Your supplementary material describes semi-dry systems as offering signal quality comparable with wet electrodes while allowing relatively fast/convenient setup, although they can dry relatively quickly. 05 EEG-ERP_Data collection

That explains the next step...

32. Why were electrodes re-wet?

Signal quality was:

visually inspected before recording

and:

monitored throughout recording.

Electrodes were:

re-wet when necessary

to maintain adequate signal quality. FINAL PSYCHOPATHIC DISSERTATION

This is especially relevant during a long developmental protocol because the electrode interface can deteriorate over time.

33. What is impedance?

Think of impedance as:

opposition to the flow of the EEG electrical signal at the electrode-skin interface.

Poor electrode contact can increase impedance and degrade signal quality.

Your protocol kept electrode impedance:

below 50 kΩ.

Don't panic if 50 kΩ sounds high compared with traditional EEG systems. High-input-impedance modern systems such as EGI can tolerate higher electrode impedances than older systems.

34. Why minimise movement?

Participants were positioned at a consistent distance from the screen and asked to minimise unnecessary movement.

Because EEG is extremely sensitive to artefacts from:

• muscle activity

• head movement

• facial movement

• electrode movement.

Gameplay creates a particular challenge because participants naturally move while playing.

So you're balancing:

ecological engagement

against:

EEG signal quality.

35. What does a sampling rate of 250 Hz mean?

It means the EEG signal was sampled:

250 times per second.

So one sample every:

4 milliseconds.

Why is this sufficient for alpha?

Alpha is only:

8-13 Hz.

According to the Nyquist principle, the sampling rate must exceed twice the highest frequency you want to represent.

250 Hz is therefore easily sufficient for analysing 8-13 Hz alpha.

36. EEG preprocessing software

Preprocessing was conducted using:

MATLAB 2021b

with:

EEGLAB v13.5.6b

and:

ERPLAB v12.0. FINAL PSYCHOPATHIC DISSERTATION

What are EEGLAB and ERPLAB?

They are MATLAB-based toolboxes used for EEG/ERP:

• processing

• visualisation

• analysis.

You don't need to pretend you personally executed every preprocessing command because you didn't.

37. Preprocessing - entire pipeline

Memorise this order:

256 channels

↓

remove 83 peripheral electrodes

↓

173 retained channels

↓

0.1-30 Hz band-pass filter

↓

visually identify/remove bad channels

↓

interpolate rejected channels

↓

average re-reference

↓

divide continuous EEG into 1-second segments

↓

visually reject noisy/artifact-contaminated segments

↓

FAA processing / spectral analysis. FINAL PSYCHOPATHIC DISSERTATION

Now understand every step.

38. Why remove 83 peripheral electrodes?

Before preprocessing:

83 peripheral electrodes were permanently excluded

leaving:

173 channels. FINAL PSYCHOPATHIC DISSERTATION

Peripheral electrodes can be particularly susceptible to:

• facial/neck muscle activity

• movement

• poor contact

• other non-neural artefacts.

Your dissertation's electrode map visually shows the permanently excluded peripheral channels in grey and the frontal FAA clusters in colour. FINAL PSYCHOPATHIC DISSERTATION

39. What does band-pass 0.1-30 Hz mean?

The continuous EEG was:

band-pass filtered between 0.1 and 30 Hz.

Meaning frequencies:

below 0.1 Hz

and:

above 30 Hz

were attenuated/removed. FINAL PSYCHOPATHIC DISSERTATION

Why?

Very slow fluctuations can reflect:

drift.

High-frequency activity can contain substantial:

muscle/electrical noise.

And your frequency of interest:

alpha = 8-13 Hz

lies comfortably inside the retained range.

40. Bad-channel rejection

Channels showing:

excessive noise

or:

signal drift

were identified through:

visual inspection

and removed. FINAL PSYCHOPATHIC DISSERTATION

In this developmental sample:

approximately 10-15% of retained channels could be rejected and subsequently interpolated.

But importantly, researchers also considered:

where those channels were located.

Why?

Because interpolation requires useful neighbouring electrodes.

Losing several adjacent electrodes is more problematic than losing the same number scattered across the scalp.

41. What is interpolation?

Suppose one EEG electrode is unusably noisy.

Rather than leaving a missing channel, its signal can be estimated using neighbouring electrodes.

That's interpolation.

High-density EEG is useful here because there are many neighbouring channels from which to estimate the missing signal. Your additional material specifically identifies reliable interpolation based on neighbouring electrodes as an advantage of high-density EGI systems. 05 EEG-ERP_Data collection

Important limitation

Interpolation doesn't magically recover the true signal.

It provides an estimate based on spatial neighbours.

That's why too many bad channels, particularly adjacent ones, make reliable interpolation impossible.

Participants with insufficient usable EEG for reliable interpolation were therefore excluded. FINAL PSYCHOPATHIC DISSERTATION

42. Average re-reference

After bad channels were interpolated:

EEG was re-referenced to the average of the retained channels. FINAL PSYCHOPATHIC DISSERTATION

Again:

Recording reference = Cz

↓

Preprocessed reference = average reference.

Definitely memorise that distinction.

43. Segmentation

The continuous EEG was divided into:

1-second segments.

Those segments were visually inspected.

Segments containing substantial:

noise

signal drift

or:

movement-related artefacts

were excluded. FINAL PSYCHOPATHIC DISSERTATION

Why segment continuous data?

It allows contaminated portions of the recording to be rejected without necessarily throwing away the entire 30-minute recording.

44. Why is artefact rejection especially important here?

Because this was:

active videogame play.

Participants were:

• moving hands

• responding rapidly

• emotionally engaged

• potentially changing posture/facial expression.

EEG measures tiny electrical potentials and is highly susceptible to non-neural electrical activity.

So rigorous artefact control is particularly important in a naturalistic task like gaming.

45. How did you go from EEG → FAA?

After preprocessing, FAA was derived from the:

30-minute gameplay period

using the:

FAA plugin within EEGLAB.

Signals were averaged within predefined electrode clusters corresponding to:

F3

F4

F7

F8. FINAL PSYCHOPATHIC DISSERTATION

46. Why clusters instead of one electrode?

Rather than using one electrode to represent F3, for example, several nearby EGI electrodes were averaged.

Advantages:

• less dependence on noise in a single electrode

• more robust representation of regional activity

• useful with high-density EEG.

The high-density EEG material likewise notes that cluster-based extraction can provide a more robust indicator than relying on a single electrode. 05 EEG-ERP_Data collection

Potential downside:

averaging can dilute focal effects or incorporate noise from additional electrodes.

So clusters increase robustness but can reduce spatial specificity.

47. Know your exact electrode clusters

F3 - left mid-frontal

E41, E40, E42, E30, E36, E49, E50

= 7 electrodes

F4 - right mid-frontal

E205, E206, E213, E214, E215, E223, E224

= 7 electrodes

F7 - left lateral-frontal

E38, E39, E46, E47, E48, E54

= 6 electrodes

F8 - right lateral-frontal

E1, E2, E3, E10, E11, E222

= 6 electrodes. FINAL PSYCHOPATHIC DISSERTATION

You probably do not need to memorise every E-number unless your supervisor is particularly technical.

You absolutely should know:

F3 = left mid-frontal

F4 = right mid-frontal

F7 = left lateral-frontal

F8 = right lateral-frontal.

And remember the standard 10-20 naming rule:

odd = left

even = right. 05 EEG-ERP_Data collection

48. Power spectral density

You weren't interested in the raw voltage waveform itself.

You needed to determine:

how much signal power occurred within the alpha frequency band.

That's where power spectral density - PSD comes in.

PSD describes how signal power is distributed across frequencies.

So conceptually:

EEG signal

↓

frequency decomposition

↓

"How much power is present at 8 Hz? 9 Hz? 10 Hz?..."

↓

extract mean power from 8-13 Hz

↓

alpha power.

49. Welch's method

Power spectral density was estimated using:

Welch's method

with:

1-second windows

and:

50% overlap. FINAL PSYCHOPATHIC DISSERTATION

You should understand this rather than memorising the name.

Welch's method essentially:

1. divides the EEG into shorter windows

2. estimates the frequency spectrum for each

3. averages those estimates.

This produces a more stable estimate of spectral power than relying on one single estimate from the whole signal.

50. What does 50% overlap mean?

With 1-second windows:

Window 1 might cover:

0.0-1.0 s

then instead of the next beginning at 1.0:

Window 2 begins at:

0.5 s

and covers:

0.5-1.5 s.

Then:

1.0-2.0

etc.

So adjacent windows overlap by half their duration.

This increases the amount of information contributing to the averaged spectral estimate and helps produce smoother/more stable PSD estimates.

51. Alpha extraction

From the PSD, mean power was extracted from:

8-13 Hz

which was defined as the alpha band. FINAL PSYCHOPATHIC DISSERTATION

So:

raw EEG → preprocessing → PSD → 8-13 Hz alpha power → FAA.

52. Why log-transform alpha power?

FAA used:

natural logarithms of alpha power.

For example:

ln(F4) - ln(F3).

Power values are commonly positively skewed.

Log transformation helps:

• reduce skew

• stabilise variance

• make relative left-right differences more suitable for analysis.

You don't need to derive the mathematics, but understand why the ln is there.

53. Exact FAA calculation

The general formula was:

FAA = mean [ln(POW-R) - ln(POW-L)]

where:

POW-R = right-frontal alpha power

POW-L = left-frontal alpha power.

Therefore:

Primary

F4/F3 FAA = ln(F4) - ln(F3)

Exploratory

F8/F7 FAA = ln(F8) - ln(F7). FINAL PSYCHOPATHIC DISSERTATION

54. Important methodological modification

This is a really good technical detail.

The standard FAA plugin calculation was modified by:

removing the absolute-value function. FINAL PSYCHOPATHIC DISSERTATION

Why?

Because absolute value removes the sign.

For your question, direction matters enormously.

For example:

+0.5 and -0.5

would both become:

0.5

if absolute values were used.

Then you'd know the magnitude of asymmetry but lose whether alpha was greater on the:

right or left.

So removing the absolute-value function:

preserved the direction of FAA.

That's exactly what you needed for directional hypotheses.

55. FAA interpretation AGAIN

After calculation:

Positive FAA

right alpha > left alpha

= greater relative right-frontal alpha power

= because alpha inversely relates to activity:

relatively greater left-frontal cortical activity.

Negative FAA

left alpha > right alpha

= greater relative left-frontal alpha power

=

relatively greater right-frontal cortical activity. FINAL PSYCHOPATHIC DISSERTATION

This should become automatic.

56. A subtle but important point - EEG's strengths

Excellent temporal resolution

EEG records electrical activity on a millisecond timescale.

That's a major advantage for examining dynamic neural responses.

Relatively direct measure

Compared with fMRI's haemodynamic BOLD response, EEG more directly reflects electrical neural activity.

Practical

• non-invasive

• relatively affordable

• tolerable for developmental populations

• can be recorded during an interactive task such as videogaming.

57. EEG's major weakness - spatial resolution

EEG has much poorer spatial localisation than something like fMRI.

Why?

Electrical activity generated inside the brain spreads through:

• brain tissue

• cerebrospinal fluid

• skull

• scalp

before reaching electrodes.

Therefore, scalp potentials don't map perfectly onto one underlying cortical source.

This leads to the:

inverse problem

There can be multiple possible underlying neural source configurations capable of producing the same scalp EEG pattern.

Your supplementary EEG material identifies the inverse problem as a central limitation of EEG/ERP research. 05 EEG-ERP_Data collection

This is relevant to your dissertation because you're talking about:

frontal scalp asymmetry

rather than claiming precise localisation of a particular neural generator.

58. Why could source localisation improve future work?

Because conventional scalp FAA tells you about differences recorded over frontal scalp locations.

It doesn't tell you with perfect precision:

which cortical source generated them.

Source-localisation methods attempt to estimate the underlying neural generators.

That's particularly relevant given Bontemps et al. (2022):

conventional scalp FAA was null while source analysis revealed dimension-specific patterns.

Hence your future recommendation for:

source-level FAA/source localisation.

59. Why continuous EEG rather than ERP?

This is important because your supplementary material is mainly about EEG/ERP, but your dissertation did not analyse an ERP component.

ERP

Event-related potentials examine EEG responses time-locked to specific discrete events.

You would need clearly defined events/triggers and usually average many repeated trials.

Your study

You examined spectral alpha power continuously across gameplay.

You weren't asking:

"What happens 300 ms after a player is shot?"

You were asking:

"What is the participant's frontal alpha asymmetry across sustained gameplay?"

So continuous spectral FAA fits your research question.

Future possibility

Event-related FAA could be used to examine specific:

• rewards

• threats

• deaths

• kills

• failures

• goal attainment.

That could reveal effects hidden when averaging across 30 minutes.

60. Methodological strength of continuous gameplay

The 30-minute measure captures FAA during a:

sustained, interactive and motivationally engaging context.

This improves ecological relevance compared with a very brief artificial stimulus.

But it introduces a trade-off:

More ecological

because gameplay unfolds naturally.

Less experimentally specific

because many processes happen simultaneously:

• reward

• threat

• frustration

• attention

• motor responses

• cognitive control

• emotion regulation

• approach

• avoidance.

Therefore, if FAA differs, you can't identify exactly which event/process caused it.

That's an excellent viva point.

61. Why no rest condition is important

Your design did not contain a resting FAA comparison for this question.

Therefore:

You can say:

"Higher II was associated with lower FAA during violent gameplay."

But you cannot say:

"Violent gameplay caused lower FAA in adolescents with higher II."

And you cannot establish that the association is:

specific to violent gameplay.

Maybe the same participants would show the same FAA pattern:

• at rest

• during non-violent gaming

• during another cognitively demanding task.

That's why a strong future design would include:

rest + non-violent game + violent game.

62. Why no non-violent gaming condition matters

Similarly, because everybody performed the CS2 task and you weren't experimentally comparing violent versus non-violent games:

your study cannot determine whether the II association is caused specifically by:

violent content.

The relevant feature could instead be:

• competition

• cognitive demand

• goal pursuit

• motor engagement

• arousal

• failure/reward.

So when you say "during violent videogame play", that describes the context in which FAA was measured, not proof of a violent-content effect.

This distinction is VERY important.

63. Biggest methodological strengths

If asked "What are the strengths of your methodology?", I'd give these:

1. Task-based continuous EEG

FAA measured during actual sustained gameplay rather than only at rest.

2. Developmentally relevant sample

Directly examines adolescents rather than extrapolating from adults.

3. High-density EEG

256-channel system provides dense spatial sampling.

4. Standardised gameplay

Same game mode, team/weapon protocol and controlled setup.

5. Practice

Reduced variability caused simply by unfamiliarity with controls/maps.

6. EEG quality control

Bad channels/segments were removed, channels interpolated appropriately and participants with insufficient data excluded.

7. Dimension-level psychopathy measurement

GM, CU and II examined separately rather than collapsing everything into a total psychopathy score.

8. Regional FAA comparison

F4/F3 primary plus exploratory F8/F7.

64. Biggest methodological weaknesses

Small sample

Final:

N = 20.

Limits:

• statistical power

• precision

• generalisability

• stability of regression estimates.

Predominantly male/White sample

Limits generalisability.

Self-report YPI

Potential:

• social desirability

• limited insight

• response bias.

Some YPI reliability concerns

Especially CU/Callousness in previous studies.

Cross-sectional

No causality or developmental trajectory.

No rest/non-violent control

Can't isolate violent gaming-specific effects.

Single task

CS2 findings may not generalise to other games or emotional contexts.

Movement/artifact challenge

Naturalistic gameplay increases EEG artefact risk.

Scalp EEG localisation

Limited spatial specificity/inverse problem.

Averaging across 30 minutes

Potentially obscures event-specific fluctuations.

65. Likely Methods viva questions

These are the ones I would practise aloud:

1. Why did you use a cross-sectional correlational design?

2. What does secondary analysis mean in your study?

3. Which parts of the data collection/analysis did you personally conduct?

4. Why were nine participants excluded?

5. Why did your final sample become 20 rather than 21?

6. What were your participant exclusion criteria?

7. Why was parental consent required?

8. What's the difference between consent and assent?

9. Why did participants watch a calming video?

10. Why did you use the YPI?

11. What are the ten YPI subscales?

12. How is the YPI scored?

13. Why are some YPI items reverse-scored?

14. What are the limitations of using self-report to assess psychopathic traits?

15. What do you know about YPI reliability?

16. Why CS2?

17. Why Deathmatch?

18. Why give participants practice?

19. Why randomise map order?

20. Why standardise weapon choice?

21. Why was the researcher present during gameplay?

22. How does EEG actually measure brain activity?

23. Why does EEG need a reference?

24. What was your online reference?

25. Why was the signal later average-referenced?

26. What does 250-Hz sampling mean?

27. What is impedance and why does it matter?

28. Why was <50 kΩ acceptable?

29. Why use a 256-channel system?

30. Why were 83 electrodes removed?

31. What does a 0.1-30 Hz band-pass filter do?

32. What is channel interpolation?

33. Why can't you interpolate unlimited bad channels?

34. Why divide the EEG into 1-second segments?

35. What kinds of artefacts were rejected?

36. Why is gameplay particularly challenging for EEG recording?

37. Why use electrode clusters rather than individual electrodes?

38. Which side are F3/F7 and F4/F8?

39. What is power spectral density?

40. What is Welch's method?

41. What does 50% overlap mean?

42. Why was alpha defined as 8-13 Hz?

43. Why log-transform alpha power?

44. How exactly was F4/F3 FAA calculated?

45. Why was the absolute-value function removed?

46. What does positive FAA mean?

47. What does negative FAA mean?

48. What is the inverse problem?

49. Why did you use continuous FAA rather than ERPs?

50. What does the absence of a resting condition prevent you from concluding?

51. Can you conclude that violent content caused the FAA association?

52. What would you change about the methodology if you repeated the study?

The Methods numbers to memorise

These are worth knowing exactly:

And the Methods chain to know cold

30 recruited → 9 EEG exclusions → 21 usable → YPI + 30-min CS2 → 256-channel EEG at 250 Hz → remove 83 peripheral channels → 0.1-30 Hz filter → reject/interpolate bad channels → average reference → 1-sec artefact screening → Welch PSD → extract 8-13 Hz alpha → ln(right) - ln(left) → F4/F3 primary + F8/F7 exploratory → statistical screening → exclude Case 17 → final N=20.