| Title: The Role of the Spotlight of Attention in the Unconscious Perception of Features. Abstract: Simple features were used in an unconscious priming paradigm, and the spotlight of attention was manipulated. Unconscious feature priming was dependent on the filter and gradient effects of the spotlight. The results cannot be explained by feature integration theory. A different neural pathway theory is favored. Problem: Recent research by Sonnerat and Klinger (2001) showed that unconscious affective priming was dependent on the spotlight of attentionÕs filter effect (priming was observed when primes appeared inside, but not outside, the spotlight) and gradient effect (priming inside the spotlight decreased as the spotlight was spread across a wider area). However, unconscious location priming was not dependent on the spotlight. Two studies using simple features were designed to investigate whether Treisman and GeladeÕs (1980) feature integration theory could account for these results. Procedure: In Experiment 1A, 18 participants judged which of two targets composed of simple features [Ò(((((Ò or Ò)))))Ó] was presented in one of two central locations. A masked prime feature was presented before the target in one of these locations. Primes and targets were congruent on half the trials. A response window assured fast responses. Following this priming task, participants completed a prime perceptibility task. In Experiment 1B, 18 participants went through the same procedures except that prime and target features were presented in one of two more peripheral locations to spread participants' spatial attention. Comparing Experiments 1A and 1B tested the gradient effect of the spotlight. In Experiment 2, 28 participants went through the same procedures except that target features always appeared in one of the two central locations, thus defining the area of the spotlight, while half the primes were presented inside the spotlight (in one of the central locations) and the other half was presented outside the spotlight (in one of the peripheral locations). The filter effect was tested by comparing the unconscious priming obtained when primes were presented inside or outside of the spotlight of attention. Results: The sensitivity measure, d', was computed for the priming and prime perceptibility tasks. The indirect measure was regressed on the direct measure (Greenwald, Klinger, & Schuh, 1995). With this method, the intercept (int.) indicates the amount of priming when prime perceptibility is near zero and the slope indicates how priming varies with prime perceptibility. Large intercepts were observed in Experiments 1A [int.=.29, t(16)=5.37, p<.001] while smaller intercepts were observed in Experiment 1B [int.=.13, t(16)=3.03, p<.01]. These two intercepts were significantly different [F(1,33)=9.72, p<.005]. Thus, unconscious feature priming was dependent on the gradient effect of the spotlight of attention. In Experiment 2, intercept effects were significant when primes were presented inside [int.=.17, t(26)=3.52, p<.005], but not outside [int.=.02, t(26)=.33, p=.74] the spotlight. These two intercepts were significantly different [F(1,25)=4.25, p<.05]. Thus, unconscious feature priming is dependent on the filter effect of the spotlight of attention. No significant slopes were observed in any conditions. Discussion: These results show that unconscious feature priming is dependent on the gradient and filter effects of the spotlight of attention. This is congruent with the pattern observed for unconscious affective priming. Thus, the feature integration theory cannot account for the present data, as the simple features used in the present experiments should not have required integration. Consequently, a theory postulating different neural pathways for location and identity information may be favored. |
| This poster was presented at the Annual Meeting of the American Psychological Society in Chicago, Illinois in May, 2004. |
 |
| Click here to see the full poster |