Response retrieval in a go/no-go priming-of-popout task

Bryan R. Burnham 0 0 I thank Adam Bidas , Susannah Bruno, Yana Kim, and Rory Pfund for assistance with data collection and analysis 1 ) Department of Psychology, University of Scranton , Scranton, PA 18510 , USA Priming of popout is the finding that singleton search is faster when features of a target and of nontargets are repeated across trials than when the features switch. Theoretical accounts suggest that intertrial repetition influences perceptual and attentional selection processes, episodic retrieval processes, or both. The present study combined a popout search task with a go/no-go task. In Experiment 1, the nontarget distractors in each display carried the go/no-go feature, and in Experiment 2, the texture of all items carried the go/no-go feature. Results showed that the go/no-go task moderated the intertrial repetition effects. In Experiment 1, the target color elicited retrieval of the preceding distractor color and associated no-go response, resulting in larger interference effects. In Experiment 2, the target color elicited retrieval of the preceding target color and no-go response, resulting in reduced facilitation effects. Additional results from both experiments showed that the colors in a search display also influenced target selection on the following trial. Taken together, the results of both experiments suggest that intertrial repetition influences both early selection and postselection retrieval processes. - example, repeating visual features, even if irrelevant, can speed subsequent target localization. Theoretical accounts suggest that intertrial repetition effects are due to facilitated selection (Becker, 2008; Chun & Nakayama, 2000; Goolsby & Suzuki, 2001; Maljkovic & Nakayama, 1994, 1996), facilitated retrieval of target features (Hillstrom, 2000; Huang, Holcombe, & Pashler, 2004; Huang & Pashler, 2005; Thomson & Milliken, 2011) or both facilitated selection and retrieval (Lamy et al., 2010; Yashar & Lamy, 2011). This study addressed whether intertrial repetition effects are due to facilitated selection and retrieval by examining whether withholding a response influences visual search. Maljkovic and Nakayama (1994) provided evidence that repetition of noncritical features facilitates visual search. In their study, subjects performed a popout search where a color singleton target appeared among two homogeneously colored distractors and identified which corner of the target was missing. The target was, randomly, red among green or green among red. Although the target color was unpredictable, repeating the colors sped responding relative to when the target and distractor colors switched across trialsa priming-of-popout (PoP) effect. Initial explanations for PoP suggested that encoding the targets color sped selection of a same-colored target (Chun & Nakayama, 2000; Goolsby & Suzuki, 2001; Maljkovic & Nakayama, 1994, 1996; Olivers & Humphreys, 2003). To support this, Maljkovic and Nakayama (1994) found that repetition of noncritical features such as the response did not influence PoP. Also, Becker (2008) found that target color repetition sped saccades toward a target, suggesting that intertrial repetition influenced processes before response execution. Others propose that PoP facilitates retrieval of target information from preceding trials (Hillstrom, 2000; Huang et al., 2004; Huang & Pashler, 2005; Thomson & Milliken, 2011). According to these accounts, when an item is selected, the visual system verifies whether this is the target by retrieving target information from trial n1. If there is a feature match, responding is facilitated, but if there is no match, responding is slowed. Support for postselection retrieval accounts comes from studies showing that response information on trial n1 influences PoP. For example, Huang et al. (2004) showed that response repetition facilitated search when the target feature repeated but slowed search when the target feature switched. Similarly, Thomson and Milliken (2011) found larger PoP effect when the task on trial n was repeated. Because such information affects performance only after a response is selected, the results suggest that PoP influences postselection retrieval. Recently, Lamy et al. (2010; Yashar & Lamy, 2011) proposed a dual-stage model, which states that PoP speeds both selection and postselection processes. To support this, Lamy et al. (2010) varied the onset of a color change to the target and distractors within trials. Consistent with selection accounts, response repetition did not affect PoP when the color change occurred soon after the displays onset but did affect PoP when the change occurred long after the displays onset, which is consistent with retrieval accounts. Support for this dual-stage model is found in studies showing that distractor color repetition contributes to PoP (Burnham, Kim, & Bruno, 2011; Kristjnsson & Driver, 2008). A strict interpretation of retrieval accounts predicts that distractor color repetition will not affect PoP, because PoP reflects only the retrieval of target features. The present study examined whether response information assigned to distractors influences PoP, since such a finding would support Lamy et al.s (2010) dual-stage model. Specifically, distractor color repetition and switching effects would be consistent with PoP facilitating selection, whereas showing that response information tied to the distractors can influence popout search would be consistent with retrieval accounts. The present study combined a go/no-go task with a popout search task. The target color and distractor color were drawn from a set of four; however, if the distractors were a certain color, subjects withheld their response (no-go trial), but responded otherwise (go trial). The question examined was whether this no-go feature assigned to the distractors would influence the intertrial transition effects. If so, this would suggest that response information can be retrieved from distractors, because the go/no-go response is a decision-level process that should not affect selection. Forty-three University of Scranton undergraduates participated (30 female; 3 left-handed). Subjects ranged from 18 to 20 years old (M= 18.44, SD= 0.70). E-Prime software (v.2.0.8.22) was used to present the experiment. Displays were generated using a Dell 755 computer with a Pentium Core 2 Duo processor (2.33 GHz; 1.96 GB) and were presented on a Dell E178Fpv monitor (60 Hz; 1,024 768). Responses were collected on a five-button response box. Subjects sat 60 cm from the monitor, which was height adjusted. Example search displays appear in Fig. 1. Displays contained three diamonds: one color singleton target and two distractors on a black background (0.16 cd/m2). The target and distractor colors were red (20.44 cd/m2), blue ( 1 9 . 0 9 c d / m 2 ) , g r e e n ( 2 0 . 6 2 c d / m 2 ) , o r y e l l o w (21.84 cd/m2). Each diamond measured 1.1 1.1 and was missing its left or right corner (0. (...truncated)