Binding of Features and Responses in Inhibition of Return: The Effects of Task Demand

Hsuan-Fu Chao; Fei-Shan Hsiao; Shih-Ching Huang

doi:10.5334/joc.247

Inhibition of return (IOR) is the phenomenon in which it takes longer for people to respond to a target presented at a previously cued location than to one at an uncued location (Posner & Cohen 1984; Posner et al. 1985; see Klein 2000, for a review). Several hypotheses have been proposed to explain this phenomenon. According to the attentional inhibition hypothesis, IOR reflects the aftereffect of attentional inhibition of previously attended spatial locations (Posner & Cohen 1984; Posner et al. 1985). As the attended location is initially activated and then inhibited, time is required to overcome the lingering inhibitory effect to respond to a target at this cued and attended location. According to the constructive retrieval hypothesis, when a target is presented at the cued location, this perceptual event is integrated with the episodic representation of the previous cue event due to their spatiotemporal proximity. A new episodic representation is established when a target is presented at an uncued location. Therefore, the cuing effect depends on the relative efficiency of these two processes: integration with old representations and establishment of new representations (Milliken et al. 2000). Moreover, according to the habituation hypothesis, IOR reflects the habituation of spatial orientation (Dukewich 2009). Finally, according to the detection cost hypothesis, IOR demonstrates the difficulty of detecting a target that is similar to the episodic representations of the previous cue (Lupiáñez 2010; Lupiáñez et al. 2007; Lupiáñez, Martín-Arévalo & Chica 2013; Martín-Arévalo, Kingstone & Lupiáñez 2013). The constructive retrieval, habituation, and detection cost hypotheses suggest that memory representations of cue events, rather than attentional inhibition, play an essential role in IOR.

Episodic representations, or object files (Kahneman, Treisman & Gibbs 1992), of cue events are essential in both the constructive retrieval and detection cost hypotheses of IOR (Lupiáñez, Martín-Arévalo & Chica 2013; Milliken et al. 2000). The cuing effect is related to the integration of the target into the object files of the cue event. Additionally, when the repetition of responses is taken into account, such as in the target-target paradigm of IOR (Maylor & Hockey 1985), the operation of event files (Hommel 1998, 2004) should be further considered (Lupiáñez, Martín-Arévalo & Chica 2013). In the target-target paradigm of IOR (Maylor & Hockey 1985), participants respond to the first target, followed by a second target. These two targets can be presented at the same or different locations. Slower responses in the same-location condition than in the different-location condition revealed an IOR effect. In this paradigm, responses to the target may also be integrated into the episodic representations of the first target, hence influencing the processing of the second target. Thus, the idea of event files (Hommel 1998, 2004), which further proposes that actions and stimulus features are bound together, can account for the underlying processes in the target-target paradigm of IOR.

Following the idea of event files (Hommel 1998, 2004), it is assumed that the responses to the target event are integrated into episodic representations of the first target in the target-target paradigm of IOR. When some or all of the features associated with the first target (e.g., location, form, and response) were repeated in the second trial, the episodic representation of the first trial was retrieved. Retrieved event files or episodic representations would affect the processing of the second target. Therefore, it was predicted that response repetition would affect the processing of the second target. According to the framework of transfer-inappropriate and transfer-appropriate processing (Neill 2007; Neill & Mathis 1998), a repetition benefit is expected when the retrieved response is compatible with the current response. Furthermore, a repetition cost is predicted when the retrieved response is incompatible with the current response. To examine these predictions, Chao and Hsiao (2021) combined Hommel’s (1998) response-cuing procedure with a target-detection task, demonstrating the effect of response repetition in the target-target paradigm of IOR. When the response to the first target was not repeated as a response to the second target, the IOR surfaced in the location repetition condition. When the responses were repeated, facilitation of return (FOR: benefit of location repetition) was observed. The interaction between location repetition and response repetition suggests that, in the target-target paradigm of the IOR, the responses to the target are integrated into the object file of the first target. Therefore, the effect of location repetition on performance is contingent on whether the response is repeated. These findings are consistent with predictions based on the concept of event files (Hommel 1998, 2004) and the framework of transfer-inappropriate and transfer-appropriate processing (Neill 2007; Neill & Mathis 1998).

It should be noted that although Chao and Hsiao (2021) demonstrated the integration between responses and spatial locations in the target-target paradigm of IOR, they did not observe an integration between the visual forms of the targets and responses and spatial locations. For instance, the effects of location and response repetition were not modulated by form repetition, implying that the visual form may not be integrated with the location of and response to the target. A potential explanation for this is the effect of task demand. In that study, the visual forms of the targets were task-irrelevant features. Thus, they may be less likely to be integrated into the event files and less likely to be retrieved from the event files. According to Memelink and Hommel (2013), an intentional weighting mechanism influences perception and action. The weighting mechanism assigns task-relevant dimensions (e.g., color, orientation, or location) more weight than task-irrelevant dimension (see also Hommel et al., 2001). Hence, task demand affects the processing of event files during the retrieval stage (Hommel 2019; Hommel et al. 2014). Moreover, according to Frings et al.’s (2020) framework of binding and retrieval in action control (BRAC), top-down control, such as attentional weighting, affects encoding (binding) as well as retrieval. Hence, the task demand might affect what was encoded into and retrieved from the event files. Consistent with these accounts, empirical evidence supports the effects of task demand or task relevance on binding to or retrieval from event files (Mocke et al. 2020; Moeller & Frings 2014; Singh et al. 2018). In their study on action planning, Mocke et al. (2020) showed that task-relevant features were bound to action plans, whereas task-irrelevant features were not. Moeller and Frings (2014) used a prime-probe paradigm and showed that attended stimuli modulated the retrieval of event files. Additionally, Singh et al. (2018) instructed participants to either attend to the valence or word type of the stimuli. Their findings also suggested that attention modulates binding or retrieval. Consistent with these theoretical accounts and empirical findings, Huffman, Hilchey and Pratt’s (2018) review of the attentional-orienting literature indicated that there was no clear evidence for non-spatial feature integration in detection and localization tasks when the non-spatial features were task-irrelevant. Therefore, it was reasonable to expect that in the target-target paradigm of IOR, a non-spatial feature was more likely to be integrated into (or retrieved from) the event files when discrimination of this non-spatial feature was required. Consistent with this, Schöpper et al. (2020) showed that color and location were bound together when the task was color discrimination and did not integrate when the task was target detection.

The present study investigated the hypothesized effects of task demand or task relevance on the binding of non-spatial features to the locations and responses in the target-target paradigm of IOR. This was conducted following the findings that responses and spatial locations were bound in a detection task in IOR’s target-target paradigm. Furthermore, the forms were not integrated with the response and the location (Chao & Hsiao, 2021). Detection, localization, and discrimination tasks were used in Experiments 1, 2, and 3, respectively. We expected that in Experiment 3, in which form discrimination of the target was required, the form would more likely be integrated with the locations and responses and retrieved to modulate the effects of IOR.

Experiment 1: Detection Task

In this study, the target-target paradigm of the IOR was used. The couplets consisted of two successive trials (the first and the second trial). Whether the location and form of the target of the first trial and the response to this first target were repeated in the second trial were orthogonally manipulated.

In Experiment 1, a task similar to that of Chao and Hsiao (2021) was used to manipulate whether the response was repeated, combining Hommel’s (1998) response-cuing procedure with the target detection task. It was expected that, similar to Chao and Hsiao (2021), evidence for the integration of location and response should be observed. For instance, the IOR should surface when the response is not repeated, and FOR should manifest when the response is repeated in a couplet.

Additionally, according to Frings et al.’s (2020) BRAC framework, retrieving event files is critical for the effects of the event files to manifest. Furthermore, according to Chao and Hsiao’s (2021) analysis of the vincentized cumulative response time (RT) distribution (Ratcliff 1979), the IOR surfaced when RTs were relatively long. This finding implies that it may take time for the event files of target events to be retrieved. Hence, we analyzed the vincentized cumulative RT distribution in the present study.

Finally, the form of the target (rectangle or oval) was repeated within a couplet and was manipulated in this experiment. Considering the intentional weighting mechanism (Memelink & Hommel 2013) and the effects of task demand on the processing of event files (Hommel 2019; Hommel et al. 2014), it was expected that repetition of the form would be less likely to affect performance in the present experiment because the form of the target was not related to the task demand (target detection).

Methods

Participants

A total of 20 undergraduate students at Chung Yuan Christian University (6 men and 14 women, aged 20–25 years old, average age = 21.1 years) participated in this experiment and received NT$320. According to a Monte Carlo simulation (Lakens & Caldwell 2019), for the critical two-way interaction between location repetition and response repetition reported by Chao and Hsiao (2021), the power was 0.99 for a sample size of N = 20 and an alpha of 0.05. Therefore, the sample size was 20 in all experiments in the present study.

Each participant had a normal or corrected-to-normal vision. The Research Ethics Committee of the National Taiwan University (201512ES045) reviewed and approved this study. All participants provided written informed consent to participate in the study.

Stimuli

In this experiment, the viewing distance was approximately 57 cm. At the beginning of each trial, three boxes were presented at the center of the screen as placeholders, aligned on the horizontal meridian. The box’s outline was presented in white, with a thickness of 0.1°. Each box had a width of 3.0° and a height of 2.0°. The center-to-center distance between the left and right boxes and the central box was 4.8°. When the central cue was presented, the width of the central outline increased to 0.3°, and a Chinese word indicating left or right was presented at the center of this central box as the response cue. The height and width of the Chinese words were 0.8°. The target was a rectangle or oval, presented in the center of the left or right box. The length and width of the rectangle and oval were 1.1° and 0.7°, respectively. The target was always presented in yellow and vertically.

Procedure

This experiment was conducted using the DMDX software (Forster & Forster 2003). Given the use of the target-target paradigm of IOR, couplets comprising two successive trials (first and second trials) were used in the present experiment.

There was a practice block of 16 test couplets and four catch couplets at the beginning of the experiment, followed by 384 test couplets and 96 catch couplets. In the test couplets, a target was present in both the first and second trials. The location required response and form of the first target may be repeated in the second trial, resulting in a location repetition (repeated/unrepeated) × form repetition (repeated/unrepeated) × response repetition (repeated/unrepeated) design. There were 48 couplets for each combination of these three variables. In the catch couplets, the target was present in the first trial but absent in the second trial (32 couplets), absent in the first trial but present in the second trial (32 couplets), or absent in both the first and second trials (32 couplets).

Figure 1 illustrates an example of a couplet. At the beginning of each trial of each couplet, the placeholders were presented for 200 ms, followed by a 1200-ms central cue. After another interval of 200 ms, only the placeholders were presented. Further, a target was presented in the left or right box for 1,000 ms or until a response was given. Participants were instructed to press the mouse button specified by the central cue at the beginning of each trial upon seeing a target, regardless of the target’s form and location. The target was absent in the first, second, or both trials in catch couplets. The participants rested after completing each block of 30 couplets.

Figure 1

An Example of a Couplet (Not to Scale).

Note: Two trials were performed for each couplet. Each trial began by displaying three boxes as placeholders for 200 ms. Subsequently, the width of the outline of the central box was increased as a fixation cue for 1200 ms (Experiments 1 and 3) or 1600 ms (Experiment 2). During the same period, a response cue indicating the response to make was presented in the central box. In Experiment 1, Cue 1/Cue 2 was a Chinese word indicating the left or right. In Experiment 2, Cue 1/Cue 2 were Chinese words indicating ‘press the left button for a target at the left location; press the right button for a target at the right location’ or ‘press the right button for a target at the left location; press the left button for a target at the right location’. In Experiment 3, Cue 1/Cue 2 were Chinese words indicating that ‘press the left button for an oval; press the right button for a rectangle’ or ‘press the right button for an oval; press the left button for a rectangle’. After an interval of 200 ms, a target was presented in either the left or right box, or the target was absent.

Results

Responses to the second trial of each couplet were included in further analyses only when the responses to the first trial of that couplet were correct. This criterion excluded 1.6% of the data (0%–4.2% per participant).

Analysis of the overall data

Table 1 lists each condition’s average median correct RTs and error rates. Both RTs and error rates were analyzed using location repetition (repeated/unrepeated) × form repetition (repeated/unrepeated) × response repetition (repeated/unrepeated) repeated-measures analysis of variance (ANOVA).

Table 1

Averaged Median Correct Response Times (RT, in ms) and Error Rates (ER, %) and Standard Deviations (in parentheses) as a Function of Location, Form, and Response Repetition in Experiment 1 (Detection Task).


	LOCATION REPEATED				LOCATION UNREPEATED

	FORM REPEATED		FORM UNREPEATED		FORM REPEATED		FORM UNREPEATED

	RESPONSE REPEATED	RESPONSE UNREPEATED	RESPONSE REPEATED	RESPONSE UNREPEATED	RESPONSE REPEATED	RESPONSE UNREPEATED	RESPONSE REPEATED	RESPONSE UNREPEATED

RT	317(39)	338(47)	313(33)	331(47)	334(42)	324(42)	325(47)	317(33)

ER	0.7(1.4)	1.7(2.1)	0.9(1.9)	1.5(2.5)	1.1(1.8)	1.2(2.0)	2.8(3.0)	1.5(2.3)

Analysis of the RT data revealed a significant interaction between location repetition and response repetition, F(1, 19) = 40.69, p < 0.001, $η p 2$ M1 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.68. Figure 2 shows the raincloud plots (Allen et al., 2021) for the effect of location repetition as a function of response repetition and form repetition. A follow-up simple main effect analysis indicated that location repetition produced a significant repetition benefit when the response was repeated (19 ms), F(1, 38) = 26.97, p < 0.001, $η p 2$ M2 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.42, and a significant repetition cost when the response was not repeated (–14 ms), F(1, 38) = 14.92, p < 0.001, $η p 2$ M3 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.28. Effects involving form repetition were not significant, including the main effect of form repetition, F(1, 19) = 2.71, p = 0.113, $η p 2$ M4 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.12, the interaction between location repetition and form repetition, F(1, 19) = 0.41, p = 0.535, $η p 2$ M5 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.02, the interaction between form repetition and response repetition, F(1, 19) = 1.26, p = 0.275, $η p 2$ M6 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.06, and the three-way interaction of all three variables, F(1, 19) = 0.51, p = 0.489, $η p 2$ M7 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.03. A further Bayesian repeated-measures ANOVA (JASP Team, 2022) showed anecdotal evidence for the null model for the two-way interaction of form repetition and response repetition (BFincl = 0.624) and the three-way interaction of location repetition, form repetition, and response repetition (BFincl = 0.430). Other effects were not significant (p > 0.20).

The effect of location repetition as a function of form repetition and response repetition in a detection task (Experiment 1)

Figure 2

Raincloud Plots for the Effect of Location Repetition as a Function of Form Repetition and Response Repetition in Experiment 1 (Detection Task).

Note: Effect of location repetition = unrepeated-location RT – repeated-location RT.

An analysis of error rates revealed a significant interaction between location repetition and response repetition, F(1, 19) = 4.71, p = 0.041, $η p 2$ M8 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.20. A follow-up simple main effect analysis indicated that location repetition produced a significant repetition benefit when the response was repeated (1.2%), F(1, 38) = 6.27, p = 0.016, $η p 2$ M9 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.14, and a non-significant repetition cost when the response was not repeated (–0.2%), F(1, 38) = 0.23, p = 0.626, $η p 2$ M10 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.01. Additionally, the main effect of the form was significant, F(1, 19) = 7.65, p = 0.012, $η p 2$ M11 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.29, suggesting fewer errors when the form was repeated. Other effects involving form repetition were not significant, including the interaction between location repetition and form repetition, F(1, 19) = 2.76, p = 0.110, $η p 2$ M12 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.13, the interaction between form repetition and response repetition, F(1, 19) = 3.02, p = 0.095, $η p 2$ M13 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.14, and the three-way interaction of all three variables, F(1, 19) = 0.83, p = 0.376, $η p 2$ M14 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.04. A further Bayesian repeated measures ANOVA showed anecdotal evidence for the null model for the two-way interaction of form and response repetition (BFincl = 0.517) and the three-way interaction of location, form and response repetition (BFincl = 0.487).

Reaction time bin analysis

Table 2 lists each condition’s average median correct RTs and error rates. Figure 3 shows the effect of location repetition (i.e., unrepeated-location RT – repeated-location RT) as a function of form repetition, response repetition, and RT bin. The RT data were analyzed using location repetition (repeated/unrepeated) × form repetition (repeated/unrepeated) × response repetition (repeated/unrepeated) × RT bin (12.5%, 25%, 37.5%, 50%, 62.5%, 75%, 87.5%, and 100%) four-way repeated-measures ANOVA.

The effect of location repetition as a function of form repetition, response repetition, and RT Bin in a detection task (Experiment 1)

Figure 3

The Effect of Location Repetition as a Function of Form Repetition, Response Repetition, and RT Bin in Experiment 1 (Detection Task).

Note: Effect of location repetition = unrepeated-location RT – repeated-location RT. Error bars represent 95% confidence intervals based on Jarmasz and Hollands (2009). The MSE was based on an ANOVA of the effect of the location repetition.

Table 2

Averaged Median Correct Response Times (RT, in ms) and Standard Deviations (in parentheses) as a Function of Location, Form, and Response Repetition and RT Bin in Experiment 1 (Detection Task).


RT BIN	LOCATION REPEATED				LOCATION UNREPEATED

	FORM REPEATED		FORM UNREPEATED		FORM REPEATED		FORM UNREPEATED

	RESPONSE REPEATED	RESPONSE UNREPEATED	RESPONSE REPEATED	RESPONSE UNREPEATED	RESPONSE REPEATED	RESPONSE UNREPEATED	RESPONSE REPEATED	RESPONSE UNREPEATED

12.5%	222(48)	219(47)	216(43)	207(59)	228(60)	227(50)	225(47)	207(54)

25%	261(31)	270(42)	263(25)	265(40)	277(36)	266(36)	276(31)	260(32)

37.5%	286(30)	297(34)	283(27)	294(37)	299(33)	293(36)	301(37)	287(27)

50%	306(35)	323(39)	305(31)	318(41)	323(40)	313(39)	323(45)	306(31)

62.5%	327(41)	349(52)	323(34)	342(52)	347(47)	336(49)	349(53)	327(38)

75%	347(46)	375(63)	349(48)	376(66)	377(51)	358(58)	376(62)	352(45)

87.5%	389(63)	423(81)	388(65)	421(76)	428(67)	399(70)	416(75)	396(68)

100%	479(100)	520(93)	469(87)	524(100)	538(106)	496(96)	497(99)	500(99)

The analysis revealed that the main effects of location repetition, F(1, 19) = 4.31, p = 0.049, $η p 2$ M15 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.19, form repetition, F(1, 19) = 4.77, p = 0.040, $η p 2$ M16 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.20, and RT bin, F(7, 133) = 181.00, p < 0.001, $η p 2$ M17 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.91, were significant. Moreover, the two-way interactions of location repetition and response repetition, F(1, 19) = 40.05, p < 0.001, $η p 2$ M18 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.68, and response repetition and RT bin, F(7, 133) = 2.23, p = 0.035, $η p 2$ M19 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.11, also reached significance. More importantly, the three-way interactions of location repetition, response repetition, and RT bin, F(7, 133) = 7.98, p < 0.001, $η p 2$ M20 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.30, and form repetition, response repetition, and RT bin, F(7, 133) = 4.28, p < 0.001, $η p 2$ M21 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.18, were significant. A follow-up simple-simple main effect for the three-way interaction of location repetition, response repetition, and RT bin showed that the benefit of location repetition when the response was repeated was significant at 25%, F(1, 304) = 5.60, p = 0.018, $η p 2$ M22 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.02, 37.5%, F(1, 304) = 6.35, p = 0.012, $η p 2$ M23 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.02, 50%, F(1, 304) = 8.89, p = 0.004, $η p 2$ M24 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.03, 62.5%, F(1, 304) = 14.52, p < 0.001, $η p 2$ M25 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.05, 75%, F(1, 304) = 21.74, p < 0.001, $η p 2$ M26 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.07, 87.5%, F(1, 304) = 29.44, p < 0.001, $η p 2$ M27 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.09, and 100%, F(1, 304) = 51.91, p < 0.001, $η p 2$ M28 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.15, RT bins. The cost of location repetition when the responses were not repeated was significant in the 62.5%, F(1, 304) = 5.39, p = 0.020, $η p 2$ M29 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.02, 75%, F(1, 304) = 11.40, p = 0.001, $η p 2$ M30 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.05, 87.5%, F(1, 304) = 15.92, p < 0.001, $η p 2$ M31 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.05, and 100%, F(1, 304) = 15.69, p < 0.001, $η p 2$ M32 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.05, RT bins. A follow-up simple-simple main effect for the three-way interaction of form repetition, response repetition, and RT bin suggested that there were significant form repetition costs when the response was repeated in the 100% RT bin, F(1, 304) = 23.29, p < 0.001, $η p 2$ M33 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.07, and when the response was not repeated in the 12.5% RT bin, F(1, 304) = 8.94, p = 0.003, $η p 2$ M34 \documentclass[10pt]{article} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \usepackage{pmc} \usepackage[Euler]{upgreek} \pagestyle{empty} \oddsidemargin -1.0in \begin{document} \[ \eta _{\rm{p}}^2 \] \end{document} = 0.03.

Discussion

The present experiment’s findings are similar to those of Chao and Hsiao (2021), suggesting the integration of location and response. Therefore, there was a FOR effect (benefit of location repetition) when the response was repeated and an IOR effect (cost of location repetition) when the response was not repeated. These findings are consistent with predictions based on the concept of event files (Hommel 1998, 2004) and the BRAC framework (Frings et al. 2020).

Additionally, similar to Chao and Hsiao (2021), the findings of the present experiment show that the retrieval of previous episodic representations requires time. The RT bin analysis showed that the effects of both FOR and IOR were significant when RTs were relatively longer. Similarly, Schöpper and Frings (2022) found that the IOR was larger when RTs became longer. These findings suggest the importance of time-dependent retrieval processing of previous episodic representations to affect current performance (Frings et al. 2020).

An unexpected finding of this experiment was the significant three-way interaction between form repetition, response repetition, and RT bin. Based on the intentional weighting mechanism (Memelink & Hommel 2013) and the effects of task demand on the processing of event files (Hommel 2019; Hommel et al. 2014), we expected that the form of the target would be less likely to modulate performance in the present experiment. A possible mechanism for these findings is distractor-response binding (Frings, Rothermund & Wentura 2007). The form of the target can be viewed as a distractor dimension because it is task irrelevant. Due to the binding between the distractor form and the response, repetition of the form could facilitate performance when the response was also repeated and delayed performance when the response was not repeated. Nevertheless, it should be noted that the cost of form repetition was observed in two conditions: when the response was repeated in the 100% RT bin and when the response was not repeated in the 12.5% RT bin. This distractor-response binding hypothesis can account for when the response was not repeated but not when it was repeated.

Notably, such a form repetition cost was observed under limited conditions. Contrastingly, the FOR effect for location repetition in the repeated response condition was observed in seven successive RT bins (from 25% to 100%), and the IOR effect for location repetition in the unrepeated response condition was observed in four following conditions (from 62.5% to 100%). Hence, we speculate that the cost of form repetition observed in the present experiment might be a less robust phenomenon. This finding is examined further in Experiment 2, in which the target form is also task irrelevant.

Experiment 2: Localization Task

In Experiment 2, target localization instead of target detection task was used. Hommel’s (1998) response-cuing procedure was also used to manipulate response repetition in this experiment. We expected that, similar to the findings in Experiment 1 and Chao and Hsiao (2021), the locations and responses would be integrated, and the effect of location repetition would be modulated by response repetition. Moreover, because the form of the target was still task-irrelevant in this experiment, we did not expect strong effects from form repetition.