Labs

Lab 1

Literature Search skills

  • Planning (be selective):

    • Is it relevant? Understand if it’s:

      • Empirical (experiment)

      • Theoretical (predictive, no data)

      • Literature review (meta-analysis)


  • Different ways of reading:

    • Scanning

      • Brief to understand if it’s relevant

    • Skimming

      • Get good overview of text intention and meaning

    • Close reading

      • Complete understanding of the text


  • Literature search types:

    • Backwards (look through references)

    • Forwards (look through “cited by”)

    • Sideways (look through related articles)


Lab 2

Coming up with new research

  • Look for method issues or research field gaps

    • Based on what’s done previously


AI

  • Academic misconduct: plagiarism, unauthorised collaboration, impersonation, falsification, use of unauthorised materials, assisting others to commit misconduct

  • Prompt literacy: ability to prompt AI to achieve desired results

    • Be mindful of mis/disinformation

    • Be specific and detailed

    • Use iterative refinement: multiple prompts to refine its answer


Lab 3

Memory

  • A memory: specific pattern (memory trace) of synaptic strengths in a population of neurons

    • Memory trace represents events and learned info


  • Short term memory: enables brain to remember small amount of info for short period of time

  • Working memory: shortest memory type (few seconds-minutes) used to hold info while doing other cognitive processes


  • Long term memory: continuous info storage

    • Declarative memory: things you know and can tell others

      • Episodic: personal events

      • Semantic: facts

    • Nondeclarative (procedural) memory: things you know that you can show by doing

      • Skill learning

      • Priming: e.g. Being more likely to use a word you recently heard

      • Conditioning


  • Encoding: getting info into memory through attention and rehearsal

  • Recall: accessing stored info (reactivates og synaptic strength pattern)


Memory Experiment

What we did in class:

  • Experiment 1: study 6 lists of 15 words→recall words→2 min break→recognition confidence rating

    • Record mean confidence ratings for studied and non-studied (unrelated and related) words

  • Experiment 2: presented with list of names→labelled if we thought they were famous or not

    • Record probability of judgment for F(new), NF(new), and NF(old)


  • Kouwenhoven & Machado (2024):

    • Aims:

      • Measure age-related differences in inhibition and working memory

      • Is inhibitory control dependent on working memory performance?

      • How are these relationships affected by age?

    • Tasks measured:

      • Reflexive inhibition: various saccade tasks

      • Volitional inhibition: selectively ignoring distractors (anti-saccade, stroop, flanker)

      • Verbal working memory: digit-span task

      • Spatial working memory: spatial-span task

    • Conclusions:

      • Inhibition and working memory function largely independently in both age groups

      • Age-related working memory difficulties cannot account for age-related declines in inhibitory control

    • Limitations:

      • Should add more age groups

      • Measure neural activity


Lab 4

  • False memories: remembering events that never happened or quite different from the way they happened

    • Some researchers (1990s) thought the recovery of memories during therapy actually was creating false memories due to problematic therapeutic practices and constructive memory processes


  • Brewer (1997): sentence paradigm

    • The paratrooper leaped out the door implies the paratrooper jumped out of the plane.

      • Participants more likely to report remembering information that was implied not explicitly stated


  • Loftus, Miller, and Burns (1978): slide sequences

    • Participants exposed to misleading information about a (fake) car crash they saw were less accurate in recognition trials


  • Loftus and Palmer (1974): videotapes

    • Participants watched car accident videos

    • Question wording of hit/bumped/crashed etc influenced whether they remembered seeing broken glass (when there was none) a week later


  • Deese (1959):

    • 36 12-word lists where words in each list were associated with one, non-present word

      • During free recall, participants reported remembering the non-present word


  • Roediger and McDermott (1995): extended Deese’s study to include recognition tests

    • The study we did in class


Recall - What we did in class:

  • Experiment 1: study 6 lists of 15 words→recall words→2 min break→recognition confidence rating

    • Record mean confidence ratings for studied and non-studied (unrelated and related) words

    • Results: Paired samples t-test showed confidence ratings were lure (new) words than non-studied (new) words

      • Implicit associative response: during encoding participants think of associated words in the list (the lure) that when later asked in trials, they think they saw it

      • Failure of source monitoring: claim to remember seeing lure words because weren’t aware of the assoiative response when studying the lists

        • Didn’t remember where it came from?

  • Experiment 2: pre-exposure to 25 names (in previous lab) →1 week delay→presented with list of names→labelled if we thought they were famous or not

    • Record probability of judgment for F(new), NF(new), and NF(old)

    • Results: ANOVA showed there was a significant main effect of name type

      • Post-hoc tests showed real famous names were judged as famous more than new and pre-exposed not famous names (and didn’t differ btn pre-exposed and new)

    • Hyps not supported


Lab 5

  • Developing a research proposal argument:

    • Addressing a gap


Data

Types of quantitative data:

  • Binary/dichoromous: two non-numerical categories

  • Nominal categories: assigned a number which doesn’t indicated value or rank

  • Ordinal categories: assigned a number that represents sequence order but no meaning beyond this (e.g. Likert scales)

  • Continuous: numerical (e.g. Height, time etc)


Quantitative designs:

  • Correlational: relationship between 2 numerical variables and whether one predicts the other

    • Can be positive or negative

    • Pearson correlation r = linear correlation

      • For continuous variables

    • Spearman correlation rho = rank order correlation

      • Ordinal or continuous variables

  • Comparing groups or conditions:

    • Within-subjects: all receive same treatment

    • Between-subjects: allocated or natural groups


Which tests to use, when:

  • One IV, two levels:

    • If DV is continuous → t-test

      • Independent samples t-test = between subjects

      • Paired samples t-test = within-subjects

    • If DV is categorical or ordinal → non-parametric tests

      • Chi-square = binary variables

      • Mann-Whitney U = ordinal variables, btn subjects

      • Wilcoxon signed-rank test = ordinal variables, within subjects

  • One IV, multiple levels:

    • If DV is continuous → ANOVA

      • One-way ANOVA = btn subjects

      • Repeated-measures ANOVA = within subjects

    • If DV is non-continuous → non-parametric tests

      • Kruskal-Wallis test = one-way ANOVA for ordinal data

      • Friedman test = similar to repeated measures ANOVA for non-parametric data

  • Multiple IVs with multiple levels

    • Two or more categorical IVs → factorial ANOVA

      • All within-subjects

      • All between-subjects

      • Mixed ANOVA (within and between variables)

    • One continuous IV and/or DVs are not continuous → regression


  • Main effects: when have only one IV (looking at relationship btn two variables of 1IV)

  • Interactions: when have multiple IVs with multiple levels


Lab 6

They said in lab these concepts will all be on exam:

  • Allocentric space: uses landmark position relative to other objects

    • Doesn’t change with observer’s position

  • Egocentric space: observer’s spatial relationship to environmental features

    • Changes depending on observer position


Types of spatial info:

  • Location based: fixed point references

    • Uses identifiable features of environment for reference

  • Movement based: generated by own movements

    • Gather info as we move through space (integrates it from all senses perceiving the environment to build cognitive map)


Types of movement cues:

  • Optic flow: visual information movement in our visual fields

    • How brain processes objects in focal and peripheral view as we move

    • Estimates movement direction and speed

  • Sensory flow: similar to optic flow but with non-vision senses

    • E.g. Changes in air temperature and air movement speed

  • Vestibular information: inner ear mechanisms keeping track of head’s position in 3D space

    • Detects head acceleration/deceleration and rotation

  • Motor efferent copy: when brain sends movement generation signal, a copy is sent to rest of NS to monitor movement (to note you’ve done the movement)

  • Kinaesthetic information: actual movement feedback (muscles, receptors, sensory signals, joints etc) that tells us we’re moving

    • To estimate how far we’ve travelled and incline level


Navigation strategy:

  • Piloting: using multiple fixed reference location points to determine position and navigate

    • Allocentric processing, landmarks in relation to each other, not observer

    • Constructs cognitive map

  • Beacon homing: walking towards one landmark

    • Requires little cognitive mapping or spatial memory

    • Can’t use if object is obscured so will switch to piloting

  • Path Integration: using movement cues to continually update place in space and create cognitive spatial maps

    • Vulnerable to NS noise so is more accurate for small distances

      • Humans mostly rely on landmark-based navigation


Lab Experiment

  • Experiment 1: the effect of terrain

    • Estimate distance from start of street A (up a hill) and of street B (flat)

      • Counterbalanced for order effects (half class did A first and other half did B)

    • Paired samples t-test

    • Ppl estimate distance as further away when it’s up a slope (affected by required effort to move across that distance)

  • Experiment 2: finding way through Barcelona

    • Complex navigational instructions and making heading and distance estimations at certain points along route

    • When using path integration in real world - NS integrates all movement cues to create a cognitive map of surroundings

      • Drawback is that errors accumulate as distance moved increases

    • Repeated measures ANOVA and post hoc tests

    • Can’t show this effect bc we did it virtually


Lab 7

  • Brain as a black box:

    • Reaction time is a common measure in many behavioural tasks but provide little insight into underlying mental processes

    • Donders (1868) proposed subtraction method to time the different mental processes that occur during different cognitive tasks


Donders (1868): used subtraction method for reaction time

  • Measured RT on 3 tasks (know these)

    • Simple RT

      • Press a specific button for any stimulus

      • Detect stimulus→identify stimulus→select response→execute response

    • Go/no-go

      • Push a specific button for a specific stimulus (no response if it’s a different stimulus)

        • Detect stimulus→identify stimulus→select response→execute response

    • Choice RT

      • If you see stimulus A press specific button (different buttons for other stimuli)

      • Detect stimulus→identify stimulus→select response→execute response


  • So:

    • Choice RT - go/no-go RT = response selection time

    • Go/no-go RT - simple RT = time to identify stimulus


Assumptions of subtraction method:

  • Assumption of seriality: no overlap btn tasks

    • Logically there will be brain processing overlap but assume there isn’t

  • Assumption of independence: removal of one task doesn’t affect timing of others


  • We also use subtraction methods in fMRI to isolate active regions from background noise


Lab 8

  • Theory of signal detection: how ppl detect and make decisions on certain things and how we deal with noise/conditions of uncertainty

    • Useful for medical diagnostics, forensics, security screening etc


  • Decision-making example: airport security luggage scan

    • Quickly inspect luggage and identify potentially dangerous items

      • Difficult task and requires a lot of training and experiencec

    • Signal (target): dangerous item in scan

    • Noise: other packed items

    • Decision: dangerous item? Yes/no


  • Outcomes:

    • Hit: Decision=yes, target=present

    • False alarm: decision=yes, target=not present

    • Miss: decision=no, target=present

    • Correct rejection: decision=no, target=not present


  • Information acquisition and training increases likelihood of correct decision


Lab experiment

  • Clear or search bags coming in when inspecting for weapons in two sets of trials:

    • Condition A: no extra info

    • Condition B: security alert of higher change of weapons being present

  • Sensitivity (d’ (said as d prime)): quantifies observer's ability to distinguish between a signal (target stimulus) and noise (background or no stimulus)

    • Larger d’ = person is more sensitive to difference btn signal present and absent

      • Little to no uncertainty

      • No response bias (c=0)

    • d’ near 0 = chance performance

      • High uncertainty

      • Response bias occurs (-c or +c)


  • Two noise factors:

    • External noise: poor image quality, number/type of suitcase items

      • Not much to do to reduce external noise

    • Internal noise: hypothetical ‘weapon detector’ neurons (give noisy responses)

      • No evidence to suggest these exist but there is some internal response (reflected by neural activity) that determines officer’s decision


  • Response bias:

    • Yes: high-stakes scenarios - expect this when there are lots of other luggage items and when additional info (of risk) has been provided

    • No: in low-stakes situations - could indicate fatigue, time pressure etc

    • Response criterion c: use signal detection theory to determine response bias

      • C=0 (absence of bias, equal yes/no)

      • C>0 (“No” bias)

      • C<0 (“Yes” bias)

      • Influencing factors on response criterion:

        • Context: low/high pressure environment

        • Motivation: e.g. Fatigue, high risk

        • Consequences of wrong decision (miss or false alarm)


Lab 11

  • EEG waveform trace attributes:

    • Frequency: number of waves/second (Hz)

    • Amplitude: height of waveform

    • Latency: timing of electrical activity

    • Direction: up/peaks or down/troughs


  • Stimulus time-locking:

    • EEG is recorded at same time for each stimulus presentation and averaged to show ERP (averages out noise)


  • Localisaiton of function:

    • Occipital lobe: visual processing

    • Temporal lobe: auditory processing

    • Parietal lobe (sensory cortex): somatosensory processing

    • Frontal lobe: planning, initiation of behaviour

      • Frontal lobe (motor cortex): motor control


Lab Experiment

  • Present visual or auditory stimulus and measure EEG

    • Visual: Look at a chequerboard pattern on a screen

      • If measuring from the occipital lobe we should be able to see a visual ERP after the data has been processed

      • If measuring from the temporal lobe there should be no ERP after the data has been processed

    • Auditory: Listen to a tone with eyes closed

      • If measuring from the occipital lobe there should be no ERP after the data has been processed

      • If measuring from the temporal lobe we should be able to see an auditory ERP after the data has been processed

        • It’s difficult to get an auditory ERP due to issues with electrode placement


  • Electrode placement:

    • Reference electrode: behind right ear

      • The standard against which active electrodes are assessed

    • Ground electrode: behind left ear

      • Reduces electrical noise

    • Visual cortex electrode: same side as reference electrode (right)

      • Active electrode to record visual cortex activity

    • Auditory cortex electrode: same side as reference electrode (right)

      • Active electrode to record auditory cortex activity


  • EEG is better for looking at timing of events whereas fMRI and PET scans are better for localisation of function (area of activity)