comparing thompson’s thatcher effect with faces and non-face objects elyssa twedt 1, david...
TRANSCRIPT
Comparing Thompson’s Thatcher effect with faces and non-face objectsElyssa Twedt1, David Sheinberg2 & Isabel Gauthier1
Vanderbilt University1, Brown University2
Poster Number #
Introduction
Method
Results
References
Thompson’s Thatcher Illusion
Locally inverted the eyes and mouth of a face to create a “Thatcherized” face
We collected images from 12 categories: •Faces: Adult, grimacing, baby, animal•Objects/Scenes: Buildings, cars, close-up scenes, large scenes•Letter-strings: HF/LF words, HF/LF non-words
We chose categories that would help test the role of familiarity and bizarre expression in experiencing the Thatcher effect
We created 2 levels of Thatcherized images by locally inverting 1 and 2 internal features (e.g., eyes and mouth) 180°
We created non-words by transposing letters from actual words
We quantified the Thatcher effect using sensitivity, comparing performance on upright and inverted trials (d’ upright - d’ inverted).
Bartlett, J. C., & Searcy, J. (1993). Inversion and configuration of faces. Cognitive Psychology, 25, 281-316.Boutsen, L., & Humphreys, G. W. (2003). The effect of inversion on the encoding of normal and “Thatcherized” faces. Quarterly Journal of Experimental Psychology Section A – Human Experimental Psychology, 56, 955-975.Parks, T. E. (1983). Letter to the Editor. Perception, 12, 88.Thompson, P. (1980). Margaret Thatcher: A new illusion. Perception, 9, 483-484.Valentine, T., & Bruce, V. (1985). What’s up? The Margaret Thatcher illusion revisited. Perception, 14, 515-516.
Global inversion makes local changes difficult to detect
Upright Thatcherized face looks grotesque
Same-Different Task
Image pairs (always from the same identity and in the same orientation)presented sequentially both upright and inverted
Trial Types:•Same: Normal, Thatcherized 1, Thatcherized 2•Different: Normal vs. Thatcherized 1, Normal vs. Thatcherized 2
Sample trial:
750 ms 300 ms 750 ms 2250 ms
Overall:
•Significant Thatcher effect for all categories relative to zero
•Largest Thatcher effect for HF and LF words
•Thatcher effect is not face specific or largest for adult faces
Face Categories:
•If familiarity was a determining factor for the Thatcher effect, adult faces should show a larger Thatcher effect than animal faces, which was NOT the case.
•If perception of bizarreness were an important factor of the Thatcher effect, we would predict grimacing faces, already bizarre without Thatcherization, to show a smaller Thatcher effect - but adult and grimacing faces show similar Thatcher effects.
Objects/Scenes:
•Smaller Thatcher effects than faces or letter-strings. Why? Upright and inverted d’ values suggest this is due to smaller inversion effect, rather than greater difficulty detecting changes.
Letter-Strings:
•Main effect for word type and word frequency - Words showed larger Thatcher effects than non-words and LF letter-strings showed larger Thatcher effects than HF. Interestingly, although inverted words show no advantage over inverted non-words, we find a word frequency advantage for both types of inverted strings.
•Is the Thatcher effect face specific?
•Is the Thatcher effect stronger for adult faces?
Stimuli Examples Question: What influences the Thatcher effect?Hypothesis 1: Perceived bizarreness influences size of Thatcher effect
Method: •Subjects rated images on bizarreness - Scale of 1 (normal) to 7 (very bizarre)
•Correlated ratings with size of Thatcher effect
Hypothesis 2: Familiarity with an object at a given orientation
Method: •Defined orientation familiarity: Speed at which an observer can determine an object’s orientation•Recorded RT and correlated with size of Thatcher effect
Results: •Insignificant correlation (r = -0.113)•Perhaps we need a better measure of object familiarity
Our results suggest the TE is not exclusive to faces - it does not appear to uniquely depend on factors such as expertise or the bizarre appearance of the transformation.
Faces and letter-strings showed larger TEs than objects/scenes, although all categories showed significant Thatcher effects.
•Subgroup differences may be explained by experience with faces and letters, which transfers broadly to similar objects.
ConclusionsQuestions
Empirical research limited to faces - many assume the effect is face specific
Demonstration of TE in words (Parks, 1983):
Correlation suggests that as an image appears more bizarre, the size of the Thatcher effect increases. However, correlations for face categories and objects/scenes are opposite - suggests that bizarreness is NOT a major predictor of the Thatcher effect
Response
Animal
Car
Close-up Scene
HF Non-word
+
Letter-Strings
0
0.5
1
1.5
2
2.5
3
3.5
HF UprightHF Inverted
HF Thatcher Effect
LF UprightLF Inverted
LF Thatcher Effect
d' (or delta d' for TE)
Non-WordsWords
xx x
X = N.S.
N=21