Common measures in Cog Sci 7.2

Question 1

Self report

Answer 1

Examples

•Surveys, questionnaires, interviews

Question 2

Self report pros

Answer 2

• Direct, fast, easy

* Generally high face validity

Question 3

Self report cons

Answer 3

• Dishonesty, lack of self awareness, different interpretations of same scales
• Socially-desirable responding
• Demand characteristics

Question 4

Behavioral measurements

Answer 4

Examples
•Reaction Time, accuracy, signal detection (d’)
•Observed behaviors, e.g., hand gestures, body language, speech disfluencies
•Analysis of verbal or written responses

Question 5

Behavioral measurements pros

Answer 5

• Generally inexpensive and easy to administer

* Many have a long history of use

Question 6

Behavioral measurements cons

Answer 6

• Some behaviors can be difficult to elicit

* Coding complex behaviors

Question 7

Response/reaction time

Answer 7

Ex: mental chronometry
•Assumes longer RT corresponds to
additional information processing

Question 8

Accuracy & errors

Answer 8

• Advantages
• Simple and intuitive
• Concerns
• Ceiling and floor effects
• Response biases

Question 9

d’ (sensitivity or discriminability index)

Answer 9

• Considers hit rate & false alarms
• Separates signal from noise
• Controls for response bias

Question 10

Accuracy and serial position effects in Memory

Answer 10

First and last items you remember more, middle lowest recollection

Question 11

Accuracy as a Manipulation Check

Answer 11

• Narratives and vignettes
• Are the participants actually reading the stimuli?
• Short, fairly easy comprehension questions
• Did the mother buy some clothing?
• Visual stimuli
• Intermittent attention and memory tests
• Was there a house in that last photo?
• AFC tasks and MC surveys
• Select ”A” as the answer for this response

Question 12

Observed behaviors

Answer 12

•There are many behaviors that 
can be measured
•Often coded via ratings
ex: •Imitation learning of aggression 
in preschool children

Question 13

Physiological measurements

Answer 13

• Examples

* single-unit recordings, EEG, ERP, fMRI, eye-tracking…

Question 14

Physiological measurement pros

Answer 14

•Objective measurements

Question 15

Physiological measurements cons

Answer 15

• Equipment might be expensive

* Equipment might be disruptive to natural behavior

Question 16

Single unit recordings

Answer 16

Great spatial resolution
Great temporal resolution
•Invasive
•Can (mostly) only be done with animals

Question 17

Electroencephalography (EEG)

Answer 17

•Electrical activity at the scalp (brain
waves)
•Summed, synchronous postsynaptic potentials
of cortical neurons
•High temporal resolution
•Poor spatial resolution
•Different brainwaves are associated with
different processes
•Delta (<4Hz) slow wave sleep
•Beta (16-31 Hz) thinking
•Mu (8-12 Hz) suppression – mirror neuron
activity

Question 18

Event Related Brain Potentials

Answer 18

Averaged EEG, time-locked to stimulus
onset
•Negative voltage typically plotted up
•Different ERP components associated
with different cognitive processes
•Good for examining real-time information
processing in human brains (not locations)

Question 19

Functional Magnetic Resonance Imaging

fMRI

Answer 19

•Indirect measure of brain activity
•Assumes changes in blood flow index changes in
neural activity
•BOLD (Blood Oxygen Level Dependent contrast)

•High spatial resolution
•Poor temporal resolution
•Good for addressing questions about 
wherebrain activity differs but not 
timing issues
•Images reflect contrasts between 
conditions

Question 20

Eye-tracking

Answer 20

•Logic – we tend to look at 
things we are thinking about
•High temporal resolution
•Different features can be 
measured (e.g., looking time, 
number of saccades)

Question 21

Different measures differ in their reliability

Answer 21

•EEG records brain activity at the scalp, and the brain is doing much more than simply
processing your stimuli
•Response times are a product of multiple processes, each of which contributes to the
variability of an RT
•The mean of many trials is a more reliable indicator than the mean of a few