cont ontext xt in af affectiv fective e or or non on-affec - - PowerPoint PPT Presentation

Neur urological

• logical rol
• le

e of

• f

cont

• ntext

xt in af affectiv fective e or

• r

non

• n-affec

affectiv tive e pr primacy imacy

• Asmita Bhattacharya

• OVERA

RALL: L: To establish a neurological support for the Ad hoc cognition model.

• DONE SO FAR: To duplicate experiments as described in the paper

discussed earlier, to establish context’s role in processing.

• TO DO NEXT

T WEEK: EK: To image the brains of the subjects by functional magnetic resonance imaging (fMRI) as the experiment is being done on them.

AIM :

Ad hoc cognition :

• All categorizations, concepts and semantics are constructed ad hoc,

whenever we use them

• Affective : emotional aspect, and non-affective: semantic aspect
• Relative speed with which affective/non-affective processing occurs in

response to stimuli is context-dependent HYPO POTHE THESIS SIS : In affective context, the affective processing centres of the brain should show more activity, whereas in non-affective context, the non-affective processing centres of the brain should show more activity

Duplication of non-FMRI experiment :

• 8 subjects selected from among English speakers
• For each expt, 2 subjects were assigned to affective context group, and 2

to non-affective context group

• Task-Set-Inertia paradigm is used for context – alternating target and

filler items

• Targets (Filler for Expt 2): 32 words of 4 categories : positive human

(princess), negative human (murderer), positive animal (rabbit), negative animal (cockroach)

• Fillers (Target for Expt 2): 32 pictures of 4 categories : pleasant
• utdoor, pleasant indoor, unpleasant outdoor and unpleasant indoor
• Pre-test done to normalize the valence of the items used in the test,

which were shown for 0.5 seconds

• The reaction time (RT) (verbal affective/non-affective judgment) was

recorded for the targets

RESUL ULTS TS -I :

Experiment 1 My result:

20 40 60 80 100 120 140 160 180

Affective judgment Non affective judgment

Affective context Non-affective context

Experiment 1

RT in centi-seconds

Paper’s result:

RESUL ULTS TS -II :

Experiment 2 My result: Paper’s result:

20 40 60 80 100 120 140 160 180

Affective judgment Non affective judgment

Affective context Non-affective context

Experiment 2

RT in centi-seconds

INFERENCES ERENCES :

• In experiment 1, as predicted, with affective contexts, affective

judgments were made faster than non-affective ones; and with non- affective contexts, non-affective judgments were made faster than affective ones.

• In experiment 2, as predicted, with affective contexts, affective

judgments were made faster than non-affective ones; however, with non-affective contexts, non-affective judgments were made faster than affective ones.

• The biasing towards affective judgment, predicted by authors of the

paper while using picture stimuli, was not seen here, probably because of less sharp nature of images used.

• The overall idea was successfully reproduced in these

experiments

PRO PROBLEMS LEMS :

• Error bars and average RT were much larger in my data than

previous work, probably because of poorer measurement techniques

• The number of subjects used by me was only 8 whereas the authors

had used around 40 subjects

• The sample set for target stimuli had 32 items in my data, whereas

the authors had used 96 items

• The quality of images used by me was not very good (e.g., mainly

pictures of pollution spills for “unpleasant outdoors” and hotel brochure photos for “pleasant indoors”)

• My own reaction time in using the stopwatch is also a confusing

factor here, since reaction times of the average of ~0.70 seconds was being measured

REMA MAINING INING WORK :

• The final step involves repeating the experiment using the same

target-filler set up while getting FMRI images of the subject’s brains

• To be done on 4-8 people on 20th November at SGPGI, Lucknow
• The scan time is to be synchronized with the time lapse between the

stimuli (i.e. effective reaction time)

• Full brain imaging is to be done, as areas to focus on are not clear
• Response to stimuli is to be a motor signal, and not speech, as

speech entails greater processing (i.e. more noise)

• Occipital cortex noise is to be excluded from further analysis

Th Thank ank Y You

• u