Summary Attention
(Lectures + Literature)
The student will acquire general knowledge of:
Theories of human attention;
Methods to measure attention or effects of attention on brain and behaviour; and
The usefulness of this above knowledge in applied settings.
1
,Week 1: Historic overview of research on attention
& Information processing and the study of attention
Lecture 1 – 09/02/2025:
This week topics that dominate the Multiple Choice Qs:
1. Attention research between 1860 and 1909
2. Sternberg’s additive factors method
3. The relation between arousal, task performance and attention
4. Event-related potentials
5. Signal detection theory
View this course (attention) as a re-introduction to empirical psychology.
Methods in human attention research:
Early studies of attention relied on introspection (self-reported awareness), but with the information-
processing revolution, researchers adopted behavioural measures like reaction time, accuracy, and
speed-accuracy trade-offs. These methods help quantify cognitive processes and assess attentional
mechanisms.
Methods used in human attention research:
*Reaction time (=RT); reaction times as measures of attention
o One of the most common behavioural methods involved reaction time (RT). This is
the interval between stimulus onset and response execution. Researchers analyse:
Mean reaction times across multiple trials
Variability in reaction times to infer attentional fluctuations
Error rates as a secondary measure to ensure accuracy is not compromised by
the speed.
*Accuracy; accuracy-based measures of attention
o Response accuracy is crucial in tasks where speed is not the primary concern.
Accuracy is influenced by stimulus complexity, response bias, and sensitivity to
information.
*Speed-accuracy relations, signal detection analysis
o
EEG / MEG (zie samenvatting athena)
PET, fMRI (zie samenvatting athena)
TMS (zie samenvatting athena)
Peripheral physiological measures
Pharmacology
Genetics
Patient studies
Are you paying attention?:
Did you see that? Did something happen or not?
2
, How do we model this?:
Signal detection theory (Green & Swets):
The signal detection theory distinguishes perceptual sensitivity (ability to detect stimuli) from
response bias (tendency to favour certain responses).
Simplest model: Between seeing something and your
response, there is a process of detection going on.
We also need a threshold:
In a signal detection task, observers classify stimuli as present or absent, resulting in four possible
outcomes:
1. Hits (correctly identifying a signal)
2. False alarms (mistakenly identifying noise as a signal)
3. Misses (failing to detect a signal)
4. Correct reject (accurately dismissing noise)
The d’(d-prime) metric quantifies sensitivity by comparing hit and
false alarm rates.
D' is a measure of sensitivity or discriminability. It is also referred to
as D-prime. It quantifies how well an observer can distinguish the
signal from noise.
3
, Other methods to model cognition:
Three different types of reactions:
A reaction = simple reaction time;
press the button when stimuli
appears.
B reaction = choice reaction time;
choose between two buttons, select
the matching button with the
stimuli.
C reaction = go/no-go; choice from 1 button, but two stimuli, you must press the button for
only one of the two stimuli, do nothing for the other.
Donders’ subtraction logic:
RT(a) = perception & motor time
RT(b) – RT(a) = stimulus discrimination + response selection time
RT(b) – RT(c) = response selection time
The subtractive method:
Introduces by Donders, the subtractive method estimates the time needed for different cognitive
processes. It involves comparing reaction times between tasks that differ only in one processing stage,
allowing researchers to isolate cognitive stages.
Criticism: the method assumes pure insertion (i.e., adding a process does not affect other stages, but
later studies showed that adding cognitive operations can alter multiple processing stages.
DRAW EXERCISE:
+ (fixation point) X (stimulus) press button
Over time
Last arrow is the response time.
Try draw a CRT?:
4
(Lectures + Literature)
The student will acquire general knowledge of:
Theories of human attention;
Methods to measure attention or effects of attention on brain and behaviour; and
The usefulness of this above knowledge in applied settings.
1
,Week 1: Historic overview of research on attention
& Information processing and the study of attention
Lecture 1 – 09/02/2025:
This week topics that dominate the Multiple Choice Qs:
1. Attention research between 1860 and 1909
2. Sternberg’s additive factors method
3. The relation between arousal, task performance and attention
4. Event-related potentials
5. Signal detection theory
View this course (attention) as a re-introduction to empirical psychology.
Methods in human attention research:
Early studies of attention relied on introspection (self-reported awareness), but with the information-
processing revolution, researchers adopted behavioural measures like reaction time, accuracy, and
speed-accuracy trade-offs. These methods help quantify cognitive processes and assess attentional
mechanisms.
Methods used in human attention research:
*Reaction time (=RT); reaction times as measures of attention
o One of the most common behavioural methods involved reaction time (RT). This is
the interval between stimulus onset and response execution. Researchers analyse:
Mean reaction times across multiple trials
Variability in reaction times to infer attentional fluctuations
Error rates as a secondary measure to ensure accuracy is not compromised by
the speed.
*Accuracy; accuracy-based measures of attention
o Response accuracy is crucial in tasks where speed is not the primary concern.
Accuracy is influenced by stimulus complexity, response bias, and sensitivity to
information.
*Speed-accuracy relations, signal detection analysis
o
EEG / MEG (zie samenvatting athena)
PET, fMRI (zie samenvatting athena)
TMS (zie samenvatting athena)
Peripheral physiological measures
Pharmacology
Genetics
Patient studies
Are you paying attention?:
Did you see that? Did something happen or not?
2
, How do we model this?:
Signal detection theory (Green & Swets):
The signal detection theory distinguishes perceptual sensitivity (ability to detect stimuli) from
response bias (tendency to favour certain responses).
Simplest model: Between seeing something and your
response, there is a process of detection going on.
We also need a threshold:
In a signal detection task, observers classify stimuli as present or absent, resulting in four possible
outcomes:
1. Hits (correctly identifying a signal)
2. False alarms (mistakenly identifying noise as a signal)
3. Misses (failing to detect a signal)
4. Correct reject (accurately dismissing noise)
The d’(d-prime) metric quantifies sensitivity by comparing hit and
false alarm rates.
D' is a measure of sensitivity or discriminability. It is also referred to
as D-prime. It quantifies how well an observer can distinguish the
signal from noise.
3
, Other methods to model cognition:
Three different types of reactions:
A reaction = simple reaction time;
press the button when stimuli
appears.
B reaction = choice reaction time;
choose between two buttons, select
the matching button with the
stimuli.
C reaction = go/no-go; choice from 1 button, but two stimuli, you must press the button for
only one of the two stimuli, do nothing for the other.
Donders’ subtraction logic:
RT(a) = perception & motor time
RT(b) – RT(a) = stimulus discrimination + response selection time
RT(b) – RT(c) = response selection time
The subtractive method:
Introduces by Donders, the subtractive method estimates the time needed for different cognitive
processes. It involves comparing reaction times between tasks that differ only in one processing stage,
allowing researchers to isolate cognitive stages.
Criticism: the method assumes pure insertion (i.e., adding a process does not affect other stages, but
later studies showed that adding cognitive operations can alter multiple processing stages.
DRAW EXERCISE:
+ (fixation point) X (stimulus) press button
Over time
Last arrow is the response time.
Try draw a CRT?:
4
