Scene Gist Narrative Priming: Sequential Expectations Influence Scene Gist Recognition Performance Maverick E. Smith and Lester C. Loschky Kansas State University DISCUSSION – EXP1
RESULTS – EXP1
The Scene Perception and Event Cognition Theory (SPECT), (Loschky et al., 2014; 2018; Magliano et al., 2013) distinguishes front-end (i.e., information extraction & attentional selection) from back-end processes (i.e., event understanding): • SPECT proposes event knowledge influences information extraction
Research Question: Do back-end processes influence front-end processing? • Can sequential expectations of scene categories prime scene gist recognition?
Alternative Competing Hypotheses Feed-forward: Expectations do not influence scene gist recognition Coherent Sequence: Expectations improve gist recognition
METHOD – EXP1 N = 48 Procedure: • 48 trials • 9 briefly flashed primes, 1 masked target (10 images per trial, 1 target) • Task: Identify target scene category from 8 AFC labels
HOWEVER Follow-up Question: Do sequential expectations influence what people “see” (i.e., sensitivity) vs. what they “say” (i.e., response bias)?
100%
On-Campus (Observed)
90%
Off-Campus (Observed)
80%
On-Campus (Fitted) Off-Campus (Fitted)
70% 60% 50% 40% 30%
• 24 narratives; 5 categories of scenes per narrative; 10 images per narrative shown
Trial Schematic
..
• Object Consistency Advantage: • Objects recognized more accurately when in consistent rather than inconsistent scenes (Davenport & Potter, 2004)
1.25
12.5% Chance
0.25 -0.25 -0.75
Narrative Sequence
Mixed Effects Logistic Regression Random effects: Image, Subject • Narrative Sequence [Coherent = 1], β = 0.47, z = 9.57, p < .0001 • Location [Off-Campus = 1], β = -0.14, z = -2.25, p = .02 • Interaction, β = 0.008, z = 0.16, p = 0.87 Answer: Coherent Sequence hypothesis supported
N = 31 *Data collection is ongoing Target: N = 96 Procedure:
(DeCarlo, 1998)
Random effects: Image, Subject
24 trials All primes in coherent sequence Targets were coherent with sequence for 50% of trials Task: Does post-scene category label match target: [Yes] or [No]?
• Number of primes from immediately preceding category • Consistency of target scene with expectation [Con vs. Inc] • Consistency of post-scene category label with expectation [Con vs. Inc]
Term Intercept Correct Response [Yes = +.5, No = -.5] Consistency of Target [Con = +1, Inc = -1] Correct Response*Consistency of Image 2
no
Fitted
Hallway
or
12.5% Chance
1
2 Number of Primes
or
2
or
3
Number of primes from category prior to target
Con
Parking Inc Lot
Target Image Scene Category Label
Trial Schematic
3
Fixation point (until button press) Scene category label [Consistent vs. Inconsistent]
Answer: Accuracy improved with target predictability
yes
-1.5 -2
Inconsistent Consistent Consistency of category label
Term Intercept Correct Response [Yes = +.5, No = -.5] Consistency of label [Con = +1, Inc = -1] Correct Response*Consistency of Label
Estimate Std. Error Z value -0.32 0.06 -5.6 0.81 0.10 7.73 0.08 0.05 1.67 0.17 0.10 1.59
Pr(>|z|) Randomized
Probability Correct
Observers rapidly identify gist of briefly flashed and masked scenes (Greene & Oliva, 2009; Loschky & Larson, 2010)
RESULTS – EXP2
d prime
INTRODUCTION
VSS 2018
48 ms mask Target image (24 ms) [Consistent vs. Inconsistent] Neutral gray screen followed each prime (276 ms)
24 ms primes
Fixation point (until button press) Fixation point (until button press)
*Catch trial example
Data collection is ongoing (N = 31 of 96) Scene gist narrative priming may influence both what people “see” and what they “say”
REFERENCES Davenport, J. L., & Potter, M. C. (2004). Scene consistency in object and background perception. Psychological Science, 15(8), 559-564. DeCarlo, L. T. (1998). Signal detection theory and generalized linear models. Psychological Methods, 3(2), 186. Greene, M. R., & Oliva, A. (2009). The briefest of glances: The time course of natural scene understanding. Psychological Science, 20(4), 464-472. Hollingworth, A. (1998). Does consistent scene context facilitate object perception?. Journal of Experimental Psychology: General, 127(4), 398. Loschky, L. C., & Larson, A. M. (2010). The natural/man-made distinction is made before basic-level distinctions in scene gist processing. Visual Cognition, 18(4), 513-536. Loschky, L. C., Hutson, J. P., Smith, M. E., Smith, T. J., & Magliano, J. P. (2018). Viewing Static Visual Narratives Through the Lens of the Scene Perception and Event Comprehension Theory (SPECT). J. Laubrock, J. Wildfeuer, & A. Dunst (Eds.), The Empirical Study of Comics, Rutledge. Loschky, L. C., Hutson, J., Magliano, J. P., Larson, A. M., & Smith, T. (2014, June). Explaining the film comprehension/attention relationship with the scene perception and event comprehension theory (SPECT). Paper presented at the annual conference of the Society for Cognitive Studies of the Moving Image, Lancaster, PA. Magliano, J. P., Loschky, L. C., Clinton, J. A., & Larson, A. M. (2013). Is Reading the Same as Viewing?. B. Miller, LC and McCardle, P., editors, Unraveling the Behavioral, Neurobiological and Genetic Components of Reading Comprehension, 78-90. Corresponding author email:
[email protected] VanRullen, R., & Thorpe, S. J. (2002). Surfing a spike wave down the ventral stream. Vision Research, 42(23), 2593-2615.
