Autistic imagination sub-traits are unrelated to visual temporal integration or inference

Abstract

Autistic individuals experience atypical temporal integration, and previous studies have suggested that this atypicality is related to difficulties in imagination. Temporal integration in vision was assessed using slit-viewing tasks that require unifying successive partial views of a moving target into a coherent percept. This task also requires inference ability when the targets are complex, as in line drawings. We hypothesized that imagination difficulties would be more closely related to inference difficulties than to temporal integration difficulties. To test this, we conducted an experiment with 67 Japanese university students using slit-viewing tasks with line drawings, characters, and meaningless shapes, as well as a fill-in-the-blank task involving word inference. The results showed that neither imagination difficulties nor overall autistic traits were associated with performance in any of these tasks. These findings suggest that imagination difficulties and autistic traits may not be related to visual temporal integration or inference abilities.

Keywords

temporal integration inference visual processing slit viewing imagination autism spectrum disorder

How to Cite Article

Tsuji, Y., Nishiguchi, Y., Noda, A., & Imaizumi, S. (2026). Autistic imagination sub-traits are unrelated to visual temporal integration or inference. i-Perception, 17(4), 1-19. https://doi.org/10.1177/20416695261465782

Introduction

Atypical sensory processing is common in autistic individuals. Some studies have suggested that autistic individuals experience difficulties in temporal integration, which is a component of atypical sensory processing. Temporal integration refers to the unification of successive sensory inputs into a single percept.

Many studies in this field have focused on temporal processing in multisensory integration. Previous studies that used simultaneity judgement and temporal order judgement tasks have reported atypical responses to bimodal stimuli in autism spectrum disorder (ASD) (Casassus et al., 2019; Meilleur et al., 2020). In particular, evidence of atypical temporal processing in ASD has been demonstrated using the simultaneity judgement task with social and complex stimuli (Regener et al., 2024; Stevenson, Siemann, Schneider et al., 2014). This suggests that autistic individuals may have difficulty in temporal processing at the global level, as processing more complex cues with stronger semantic correspondence likely involves stronger involvement of global processes (Regener et al., 2024). Stevenson, Siemann, Schneider et al. (2014) also found that performance on the simultaneity judgement task with complex stimuli was related to performance on the McGurk effect, an illusion in which visual facial information influences auditory speech perception. In this effect, listeners integrate temporally coincident visual and auditory stimuli. Autistic individuals may exhibit a weaker McGurk effect than those with typical development (TD) (e.g., Bebko et al., 2014; Irwin et al., 2011; Mongillo et al., 2008; Stevenson, Siemann, Schneider et al., 2014). Individuals in the general population with high autistic traits also showed weaker effects (Ujiie & Wakabayashi, 2022). While many studies investigating the McGurk effect in autistic individuals have used only synchronous audiovisual conditions, Woynaroski et al. (2013) examined it under both synchronous and asynchronous conditions. Although they found no significant difference between the ASD and TD groups, they reported that autistic children demonstrated an expanded temporal binding window of audiovisual stimuli in the McGurk task (Woynaroski et al., 2013). This suggests that autistic individuals exhibit atypical temporal integration of audiovisual information. The double-flash and stream-bounce illusions also require temporal integration of audiovisual information. It has also been suggested that autistic individuals exhibit a weaker effect of these illusions (Stevenson, Siemann, Woynaroski et al., 2014) and a broader temporal binding window for the double-flash illusion (Foss-Feig et al., 2010). Furthermore, some autistic sub-traits are associated with the stream-bounce illusion effect (Yaguchi & Hidaka, 2021).

Atypical temporal integration has also been observed during unisensory processing in autistic individuals. In the auditory domain, difficulties in the temporal integration of speech can lead to difficulties in perceiving speech in noise with temporal dips (Alcántara et al., 2004; Dunlop et al., 2016; Groen et al., 2009). Alcántara et al. (2004) and Groen et al. (2009) indicated that autistic individuals have such difficulty perceiving speech in noise with temporal dips, which are moments when the overall level of the noise is low, but not in noise without temporal dips, such as speech-shaped noise or pink noise. Participants must temporally integrate fragments of the target speech to perceive it as a complete word. Additionally, difficulties in speech perception in noise were associated with autistic traits among university students (Tsuji & Imaizumi, 2024; Tsuji et al., 2025).

In the visual domain, performance in a slit-viewing task (Parks, 1965) is lower in autistic individuals (Nakano et al., 2010; Peiker et al., 2015). In this task, participants perceive images of objects passing behind a slit. Thus, the task requires temporal integration of visual stimuli to perceive the complete object because only a small part of the image is presented at any given moment and participants cannot recognize the object from that part alone.

Imagination difficulties, which are characteristic of ASD, have been suggested to be associated with atypical temporal integration. The imagination subscale of the Autism-Spectrum Quotient (AQ; Baron-Cohen et al., 2001) is used to measure imagination difficulties in the general population. For example, the strengths of the McGurk effect and the stream-bounce effect, both of which require multisensory temporal integration, are related to imagination difficulties as measured by the AQ. A weaker McGurk effect is associated with greater difficulties in imagination (Van Laarhoven et al., 2019). Individuals with greater imagination difficulties require more time to integrate stimuli into the stream-bounce effect among the general population (Yaguchi & Hidaka, 2021). Furthermore, imagination difficulties are associated with difficulties in speech perception in noise with temporal dips, which requires temporal integration of fragmented speech (Tsuji et al., 2025). Van Laarhoven et al. (2019) explained that the integration of sensory stimuli, within and across modalities, is based on the prior expectation that stimuli presented in close spatial and temporal proximity should be processed as a single unified percept. According to their explanation, individuals with reduced imagination abilities may have a more literal perception of the world, and their perception may rely more on sensory input than on prior experiences (Van Laarhoven et al., 2019). Therefore, imagination difficulties may be associated with atypical temporal integration.

However, in the previous study (Tsuji et al., 2025), neither imagination difficulties nor overall autistic traits in the general population, as measured by the AQ, were associated with performance on the slit-viewing task using words written in katakana, a Japanese writing system. These findings are inconsistent with those of two previous studies reporting reduced performance on slit-viewing tasks in autistic individuals (Nakano et al., 2010; Peiker et al., 2015). The abilities required to perceive stimuli differed between these experiments, likely because Nakano et al. (2010) and Peiker et al. (2015) used line drawings of everyday objects, whereas Tsuji et al. (2025) used written words or characters. Regardless of the type of stimuli, slit-viewing tasks require temporal integration. Hence, if imagination difficulties are related to atypical temporal integration, they should be related to performance on all slit-viewing tasks. The discrepancy between the findings from slit-viewing tasks using line drawings (Nakano et al., 2010; Peiker et al., 2015) and written words (Tsuji et al., 2025) suggests that factors other than temporal integration contribute to differences in performance across tasks.

To clarify the rationale of the present study, we distinguish among three cognitive processes that may contribute to performance in these tasks: temporal integration, imagination, and inference. In the present study, temporal integration refers to the perceptual process by which sensory inputs that are distributed over time are combined into a single coherent percept. This process is required when the relevant information is not available at any single moment and must be accumulated across successive sensory samples. Imagination refers to the generation or reconstruction of information that is not directly specified by the sensory input. In the present context, imagination is particularly relevant when participants must internally reconstruct the overall appearance of a familiar but visually complex object from limited visual fragments. Inference refers to the process of using partial or incomplete information to identify a meaningful whole by relying on stored knowledge, such as lexical or object knowledge. Unlike temporal integration, inference does not necessarily require information to be accumulated over time; it can also occur when incomplete information is presented simultaneously.

These processes are conceptually separable but not mutually exclusive. Slit-viewing tasks generally require temporal integration because visual information is revealed sequentially through a narrow aperture, but their demands on imagination and inference depend on the stimulus type. A slit-viewing task using line drawings of familiar objects is assumed to require not only temporal integration but also imagination of the overall appearance of a complex object from fragmentary visual information. In contrast, a slit-viewing task using characters requires temporal integration of simpler and familiar visual forms, but the number of possible representations is limited and the need to imagine a complex object is reduced. A slit-viewing task using meaningless shapes also requires temporal integration when local features alone are insufficient for perceiving the whole shape, but it should require little imagination based on prior object knowledge because the shapes do not correspond to familiar objects. Finally, a fill-in-the-blank task requires inference from partial information, because participants must retrieve a meaningful word from incomplete character information; however, it does not require temporal integration because all information needed to perceive each character is presented simultaneously. Thus, the tasks were selected not as measures that selectively reflect a single process, but as paradigms that differentially emphasize temporal integration, imagination, and inference.

The main purpose of the present study was therefore to examine which of these cognitive processes are associated with imagination difficulties. Rather than assuming that all slit-viewing tasks reflect the same underlying ability, we compared tasks that share a requirement for temporal integration but differ in their demands on imagination and inference. It is difficult to identify specific autistic traits that contribute to perceptual difficulties by comparing ASD and TD groups, because autistic individuals tend to display elevated levels across multiple autistic traits. Therefore, studies of the general population are useful for examining how specific autistic traits are associated with specific cognitive processes.

One possibility is that the ability to imagine the overall appearance of objects plays a role in integrating fragments of visual stimuli and identifying objects in a slit-viewing task using line drawings. Imagination in this context refers to the process of generating or reconstructing the overall appearance of a familiar object, rather than simply piecing together perceived visual fragments. This ability is similar to visual imagery, which is a form of mental imagery. Mental imagery produces the experience of perception in the absence of the corresponding sensory input (Kosslyn, 2005). Conversely, participants use parts of objects to produce a unified visual percept in slit-viewing tasks. Imagination difficulties in autistic individuals may be related to difficulties in imagining the appearance of complex visual stimuli rather than difficulties in integrating simple visual stimuli. In a slit-viewing task using line drawings, it may be difficult to imagine the appearance of an object from its fragments because participants must select an object from countless options.

In contrast, in slit-viewing tasks using characters, participants are required to integrate much simpler stimuli than line drawings and recall the character by matching these fragments to their stored knowledge of 46 types of katakana character representations. Consequently, a lower level of imaginative ability may be required. The ability to recall characters from their parts may also be unrelated to imagination difficulties and overall autistic traits, as autistic individuals exhibit a comparable ability to form mental representations of alphanumeric characters from their parts (Soulieres et al., 2011). Therefore, imagination difficulties are expected to be associated with lower performance in the slit-viewing task when using line drawings that require the ability to imagine the appearance of complex visual stimuli. No such relationship is anticipated when using characters.

In addition, a slit-viewing task using meaningless shapes also requires temporal integration of simple visual stimuli when the slit is sufficiently narrow and the viewed shape cannot be recognized by local features in the image (Orlov et al., 2021). However, this task does not require participants to imagine the overall appearance of the shape, because they do not know what the shape looks like and therefore cannot imagine it, unlike line drawings of familiar objects or characters. Therefore, performance in the slit-viewing task using meaningless shapes is also not presumed to be associated with imagination difficulties.

Taken together, comparing slit-viewing tasks using different types of stimuli may help clarify which abilities involved in these tasks are related to autistic traits. If imagination is related only to performance in the line-drawing slit-viewing task, then it would be related to the ability to imagine the appearance of complex visual stimuli. However, contrary to our predictions, if imagination is also related to the performance in the meaningless-shape slit-viewing task, it would be related to temporal integration.

In addition to temporal integration, we focused on inference processes in slit-viewing tasks and their relationship with imagination difficulties. In a slit-viewing task using written words (Tsuji et al., 2025), inference of an entire word from several characters is required in addition to perceiving each character. This process involves retrieving the target word by accessing stored lexical information on the basis of partial word information. Although this process does not involve imagining visual stimuli, imagination difficulties may be associated with impairments in inferring words from partial character information because both processes require the ability to infer a meaningful whole from its parts. Tsuji et al. (2025) reported that imagination difficulties were associated with difficulty perceiving spoken words in noise with temporal dips. The ability to perceive speech in noise likely requires the ability to infer words from fragmentary sounds. A similar inference process may operate in the visual domain and, if so, it can be assessed using a fill-in-the-blank task in which participants complete a word by providing missing characters. Because the entire character is presented at once, temporal integration is not required in this task. Therefore, performance on this task should be associated with imagination difficulties rather than with temporal integration.

The present study tested the following four hypotheses (Table 1). First, imagination difficulties would be associated with lower performance in the slit-viewing task using line drawings, because this task was expected to place relatively strong demands on imagining the overall appearance of familiar but complex objects from fragmentary visual information. Second, imagination difficulties would be associated with lower performance in the fill-in-the-blank task, because this task requires inference of a word from incomplete character information, although it does not require temporal integration. Third, imagination difficulties would not be associated with performance in the slit-viewing task using characters, because this task requires temporal integration of simple and highly familiar visual forms but only limited imagination of a complex visual object. Fourth, imagination difficulties would not be associated with performance in the slit-viewing task using meaningless shapes, because this task requires temporal integration but does not involve familiar object representations that can guide imagination. Thus, by comparing these tasks, we aimed to clarify whether imagination difficulties are related primarily to temporal integration itself, imagination of complex visual forms, or inference from partial information.

Table 1.

Abilities that each of the four tasks is intended to measure and the hypotheses regarding the relationships between these abilities and imagination autistic sub-trait.

Task	Required Abilities	Hypothesized Relationship Between Task Performance and Imagination Autistic Sub-Trait
Line-drawing slit-viewing task	Temporal integration; Imagination of complex appearance	Yes
Character slit-viewing task	Temporal integration; Imagination of simple appearance	No
Meaningless-shape slit-viewing task	Temporal integration	No
Fill-in-the-blank task	Inference of a word from characters	Yes

Material and Methods

Participants

The sample size was determined according to a stopping rule using an open-ended sequential Bayes factor (BF) design (Schönbrodt & Wagenmakers, 2018). We tested the correlation between the AQ imagination score and performance in each task until moderate evidence for H1 or H0 was obtained using BF₁₀ > 3 or BF₁₀ < 1/3 as the critical boundaries, respectively. Neither the minimum nor the maximum sample size was pre-specified. The increment in sample size was not fixed and was approximately 10. The BF₁₀ for correlations between the imagination score and performance on the four tasks had exceeded the boundary and converged by the time the sample size reached 41. However, at that point, the BF₁₀ for the correlations between the AQ total score and performance on the four tasks had not exceeded boundaries. Therefore, we decided to continue data collection to obtain evidence for these correlations. Although the BF₁₀ for the correlation between the AQ total score and performance on the fill-in-the-blank task did not exceed boundaries, data collection was stopped due to constraints on our attainable sample size.

A total of 68 Japanese university students participated in this experiment. The participants were native Japanese speakers with normal or corrected-to-normal visual acuity. One participant was excluded from the analysis because they did not respond to the AQ. Hence, 67 participants (24 men and 43 women; mean age = 20.9 years, SD = 3.0) were included in the analysis. Three participants’ performances in the fill-in-the-blank task (see below) were excluded from the analysis because they did not follow the experimenter's instructions. One participant did not provide any four-letter words in the fill-in-the-blank task, and two participants did not respond in katakana during the typing speed measurement. Therefore, 64 participants were included in the fill-in-the-blank task analysis.

This study was approved by the Ethics Committee of the Faculty of Education, Chiba University (approval no. 156).

Apparatus

The experiment was run on PsychoPy 2021.2.3 (Peirce et al., 2019) on macOS 12.6.2. The visual stimuli and typed responses were presented on a liquid-crystal display monitor (LCD-AH271XDB, I-O DATA) with a refresh rate of 60 Hz, spatial resolution of 1920 × 1080 pixels, and display size of 597.9 × 336.3 mm. Participants used a chin rest, and the viewing distance was fixed at approximately 57 cm. The participants responded using a standard QWERTY keyboard. Physiological measures were not collected.

Slit-Viewing Task

Participants completed three types of slit-viewing tasks using line drawings, characters, and meaningless shapes (Figure 1A–C). The four tasks, comprising three slit-viewing tasks and one fill-in-the-blank task (see below), were completed in one of eight predetermined orders. A Latin square design was used to create four orders so that each task appeared once in each serial position. Furthermore, the reverse orders of these four orders were also used to control for order effects. The order of the tasks was assigned sequentially according to the order in which participants took part. In all three slit-viewing tasks, the target stimulus appeared while moving behind the narrow horizontal slit at 3.5°/s after a 1-s presentation of a white slit. We used a horizontal slit, as Peiker et al. (2015) suggested that performance on a slit-viewing task was reduced in autistic individuals when a horizontal slit was used but not when a vertical slit was used. All stimuli subtended to approximately 3° and the slit length was 6.0° in all three slit-viewing tasks. Participants were asked to identify the stimulus that passed behind the slit. During a trial, they fixated on a red dot centered on the slit. Each of the three task types consisted of 40 trials. Behavioral performance was indexed using an error rate.

Figure 1.

Slit-viewing task using (A) line drawings (Snodgrass & Vanderwart, 1980), (B) characters, and (C) meaningless shapes (Orlov et al., 2021). (D) Fill-in-the-blank task. The gray rectangles were opaque in the actual experiment. The targets were observed only through the slits or windows.

In the line-drawing slit-viewing task, we used 119 black line drawings on a white background from a standardized set of 260 pictures (Snodgrass & Vanderwart, 1980), depicting objects, animals, and humans as target stimuli. The target stimuli were selected based on a preliminary study using a full-viewing task involving 11 participants. They were asked to observe 260 pictures (Snodgrass & Vanderwart, 1980) moving from top to bottom at 3.5°/s without a slit. After each presentation, the participants typed the object name. Line drawings that all participants in the pilot study answered correctly were used as targets in the main experiment. In the slit-viewing task in the main experiment, the stimuli were moved behind a white slit with a width of 0.17° from top to bottom. The stimulus movement speed, motion direction, and slit size were identical to those used by Peiker et al. (2015), who conducted a line-drawing slit-viewing task. Participants were asked to type the names of the presented objects immediately after stimulus presentation. Typed letters were presented synchronously on a monitor in katakana.

In the character slit-viewing task, 71 characters written in katakana using the Hiragino Kaku Gothic ProN font were used as target stimuli. Japanese is a mora-counting language. One mora is a sound unit that corresponds to one katakana character. Therefore, each katakana character typically represents one sound unit. The stimuli were moved behind a white slit with a width of 0.07° from bottom to top. This was a natural direction because native readers of Japanese are accustomed to reading vertically written text. The task requires temporal integration under both motion directions because only a small part of the image is presented at any given moment in either case. The slit width was determined during a preliminary experiment. Since characters are simpler than line drawings, there was a possibility that a 0.17°-wide slit would make character slit viewing too easy. In another preliminary experiment, 10 participants completed a character slit-viewing task using slits with widths of 0.07° and 0.12°, respectively. Peiker et al. (2015) reported that the average correct rate in a line-drawing slit-viewing task using a 0.17°-wide slit ranged between 60 and 70%. We used a slit with a width of 0.07° in the character slit-viewing task because the mean correct rate in our preliminary experiment was 63.5%. If we used a 0.17°-wide slit, a ceiling effect would likely have occurred, and temporal integration would likely not have been measurable because the character could have been recognized from local features visible at a given moment. Participants were asked to type the character via Romaji input immediately after stimulus presentation. Each katakana character can consist of a single vowel or a sequence of a consonant and a vowel. Japanese language has a core system consisting of five vowels and 14 consonants. A chart that places vowels in columns and consonants in rows is commonly used. The same consonant-vowel combinations are used for Romaji input. Romaji input was used in the present study because it reflects the core system of Japanese, and most Japanese people use Romaji input to type Japanese on a hardware keyboard. The participants in the present study were Japanese university students, and most of them were assumed to be accustomed to using Romaji input. Typed letters were presented synchronously on a monitor in katakana.

In the meaningless-shape slit-viewing task, 24 pairs consisting of 48 meaningless shapes (Orlov et al., 2021) were used as target stimuli. Local contour curvatures were highly similar between the two shapes for each pair (Figure 2). One stimulus from each pair was moved behind a white slit with a width of 0.17° from top to bottom, as in the line-drawing slit-viewing task. The slit was sufficiently narrow to require temporal integration (Orlov et al., 2021). Immediately after the motion of the shape ended, the two shapes of the pair appeared side by side: one on the left and the other on the right. Participants were asked to report which shape moved behind the slit using either the left or right key.

Figure 2.

A pair of meaningless shape stimuli for the slit-viewing task (Orlov et al., 2021).

Fill-in-the-Blank Task

The stimuli used for the fill-in-the-blank task were four-letter words. One of the letters was occluded by a gray square (Figure 1D). All words were written vertically in katakana using Hiragino Kaku Gothic ProN font, according to the traditional Japanese style of writing. The size of the four-letter words was approximately 3.0 × 11.6°. The target word and the square occluding one of its letters were moved from bottom to top at a speed of 3.5°/s behind a white window (6.0 × 5.0°). The window was sized to fully reveal a single letter; however, it was too narrow to display two letters simultaneously.

Participants were asked to infer the four-letter words from the three presented letters and type them via Romaji input immediately after the stimulus presentation. Typed letters were presented synchronously on a monitor in katakana. Participants were instructed to infer the word as accurately as possible and to respond as quickly as possible. Responses were considered correct if participants typed a four-letter word that included the three presented letters in the correct order, was listed in the Digital Daijisen Japanese dictionary, and was not a proper noun. Forty trials were conducted in the fill-in-the-blank task. Behavioral performance was indexed using the error rate. However, the fill-in-the-blank task without time limits was extremely easy. Participants could try different characters in the blanks one after another and eventually identify the correct answer. To exclude responses that may have been obtained in this way, we calculated the error rate by treating extremely slow correct responses as errors. Because response latency in the task was likely to be affected by word-typing latency, we used each participant's word-typing latency to identify slow responses. This latency was measured after the three slit-viewing tasks and the fill-in-the-blank task. Participants were asked to type the fully presented four-letter words as quickly as possible. Out of the 15 trials for each participant, the within-participant mean plus 3SD response time of the first ten correct responses was used as the criterion for long word-typing latency. If the latency taken to type a four-letter word during the fill-in-the-blank task exceeded the criterion, the response was considered an error. For descriptive purposes, the analyses using the error rate calculated without treating extremely slow correct responses as errors were also reported in the Supplementary Material.

Measures of Autistic Traits

The participants responded to the Japanese version of the AQ (Baron-Cohen et al., 2001; Wakabayashi et al., 2004) after the experiment. The AQ measures the degree of autistic traits in adults with typical intelligence using a self-administered questionnaire (Baron-Cohen et al., 2001). It comprises 50 questions that assess five different areas (10 questions per area): social skill, attention switching, attention to detail, communication, and imagination. The items are rated on a four-point Likert scale (definitely agree, slightly agree, slightly disagree, or definitely disagree). “Definitely agree” and “slightly agree” responses score 1 point on items that reflect behaviors typically associated with ASD, while “slightly disagree” and “definitely disagree” responses score 1 point on reverse-scored items. Higher scores indicate stronger autistic traits.

Statistical Analysis

We investigated the distribution of autistic traits by visual inspection and the Shapiro-Wilk test for testing normality.

A Bayesian Pearson correlation analysis was conducted between the AQ imagination scores and performance on the slit-viewing and fill-in-the-blank tasks to test the hypothesis that greater difficulties with imagination reduce the ability to temporally integrate complex visual stimuli and infer the appearance of objects and words. We used BF₁₀ as an index of evidence and interpreted it between 0.33 and 3.00 as anecdotal evidence; that is, BF₁₀ larger than 3.00 and smaller than 0.33 provides moderate evidence for the alternative and null hypotheses, respectively (Lee & Wagenmakers, 2014). We also reported p-values adjusted using the false discovery rate correction (Benjamini & Hochberg, 1995) for descriptive purposes. Partial correlation analyses examining the relationship between the AQ imagination score and performance in each task were performed to rule out confounding effects of abilities assessed in the other three tasks. In addition, Bayesian generalized linear mixed models with the standardized AQ score as a fixed effect and participants and stimuli as random effects were fitted to the trial-level responses (correct/incorrect) to confirm whether the results of the correlation analyses were supported even after accounting for participant- and stimulus-level variability.

Moreover, we conducted a Bayesian Pearson correlational analysis of the total AQ score, AQ subscale scores other than those for imagination, and performance on slit-viewing and fill-in-the-blank tasks as an exploratory analysis.

Data analysis was performed using JASP version 0.19.3 (JASP Team, 2025).

Results

Distribution of Autistic Traits

Descriptive statistics for each measure are presented in Table 2. The average AQ score in the present study was similar to that of the general population reported in the study that developed the Japanese version (Wakabayashi et al., 2004). The AQ score was widely distributed from low to high (range = 7–37), and the cut-off score of 33 for the Japanese version of the AQ (Wakabayashi et al., 2004) was included within this range. The Shapiro-Wilk test indicated that the AQ scores were normally distributed (p = .174, Figure 3). Although Baron-Cohen et al. (2001) and Wakabayashi et al. (2004) reported higher AQ scores in men than in women, no such difference was observed in the present study (t(65) = −0.41, p = .684, Cohen's d = −0.10).

Figure 3.

Distribution of the total autism-spectrum quotient (AQ) scores of Japanese university students who participated in the experiment (n = 67).

Table 2.

Descriptive statistics for task performance and AQ (n = 67, but n = 64 for the Fill-in-the-Blank Task).

	Mean	SD	Minimum	Maximum	Cronbach's α
Task performance (error rate)
Line-drawing slit-viewing	.27	.14	.10	.83
Character slit-viewing	.42	.19	.03	.93
Meaningless-shape slit-viewing	.25	.09	.05	.45
Fill-in-the-blank	.35	.12	.10	.63
AQ
Imagination (0–10)	3.28	1.69	0	8	.37
Attention switching (0–10)	4.51	2.05	0	8	.57
Attention to detail (0–10)	4.94	2.19	0	10	.59
Social skill (0–10)	3.60	2.44	0	10	.71
Communication (0–10)	3.55	2.10	0	8	.60
Total (0–50)	19.88	6.72	7	37	.79

AQ: Autism-Spectrum Quotient.

Cronbach's α for the social skills subscale and the total AQ score were within an acceptable range. However, the values for the other subscales were insufficient and the imagination subscale fell within an unacceptable range.

Relationships Between Autistic Traits, Visual Temporal Integration, and Inference

We performed a Bayesian correlation analysis between the AQ scores and the error rates of the slit-viewing and fill-in-the-blank tasks (Table 3). The results showed moderate evidence that the AQ imagination score was not correlated with performance on the three slit-viewing tasks or with performance on the fill-in-the-blank task (r = −.12– .13, BF₁₀ = 0.16–0.27, Figure 4; for scatterplots of other correlations, see Supplementary Material). Correlations involving the error rate for the fill-in-the-blank task, calculated without treating extremely slow correct responses as errors, also showed moderate evidence in favor of the null hypothesis (see Supplementary Material).

Figure 4.

Scatterplots showing the relationships between AQ imagination scores and task performance (error rate). N = 67; n = 64 for the fill-in-the-blank task.

Table 3.

Correlations between task performances and AQ scores (N = 67; for correlations with the fill-in-the-blank task, n = 64).

Task Performance (error rate)		Imagination	Attention Switching	Attention to Detail	Social Skill	Communication	Total
Line-drawingslit-viewing	r	.02	.11	.19	−.08	−.03	.07
	BF₁₀	0.16	0.22	0.51	0.18	0.16	0.18
	p	.879	.855	.691	.855	.879	.855
Characterslit-viewing	r	.06	.06	.04	−.08	−.06	.00
	BF₁₀	0.17	0.17	0.16	0.19	0.17	0.15
	p	.855	.855	.879	.855	.855	.969
Meaningless-shape slit-viewing	r	.13	.14	−.10	.10	−.05	.06
	BF₁₀	0.27	0.28	0.21	0.21	0.17	0.17
	p	.855	.855	.855	.855	.861	.855
Fill-in-the-blank	r	−.12	−.04	−.10	−.21	−.20	−.22
	BF₁₀	0.24	0.16	0.21	0.59	0.53	0.69
	p	.855	.879	.855	.691	.691	.691

Note. AQ = Autism-Spectrum Quotient; BF: Bayes factor. p-values were adjusted by calculating the false discovery rate (Benjamini & Hochberg, 1995).

To rule out confounding effects of abilities assessed in the other three tasks, we performed partial correlation analyses examining the relationship between imagination and performance in each task. The results indicated that AQ imagination was not associated with performance in any task (r = −.14–.10, p = .289–.719). Moderate evidence for the null hypotheses (BF₁₀ = 0.17–0.28) was obtained in the Bayesian partial correlation analyses (Mulder et al., 2025).

Bayesian generalized linear mixed models with the standardized AQ imagination score as a fixed effect and participants and stimuli as random effects were fitted. The response variable was binary (correct/incorrect) and was modeled using a binomial error distribution with a logit link function. The 95% credible interval for the AQ imagination score included zero, indicating that the results did not support associations between the AQ imagination score and trial-level responses in the line-drawing slit-viewing task (b = −0.04, 95% CI [−0.28, 0.21]), character slit-viewing task (b = −0.05, 95% CI [−0.31, 0.22]), meaningless-shape slit-viewing task (b = −0.07, 95% CI [−0.21, 0.07]), or fill-in-the-blank task (b = 0.07, 95% CI [−0.10, 0.24]).

There was also moderate evidence that the total AQ score is not related to the error rate of the slit-viewing tasks (r = .00– .07, BF₁₀ = 0.15–0.18). There was anecdotal evidence for a null relationship between the total AQ score and error rate of the fill-in-the-blank task (r = −.22, BF₁₀ = 0.69). In Bayesian generalized linear mixed models with the standardized AQ total score instead of the imagination score as a fixed effect, the 95% credible intervals for the AQ total score included zero in the analyses of trial-level responses in the line-drawing slit-viewing task (b = −0.09, 95% CI [−0.34, 0.16]), character slit-viewing task (b = 0.01, 95% CI [−0.25, 0.26]), meaningless-shape slit-viewing task (b = −0.04, 95% CI [−0.18, 0.11]), or fill-in-the-blank task (b = 0.16, 95% CI [−0.02, 0.33]). These results did not support associations between the AQ total score and trial-level responses in any of the tasks.

Furthermore, anecdotal to moderate evidence was found that AQ subscale scores, other than imagination, were not related to the error rates in the four tasks (r = −.21– .19, BF₁₀ = 0.16–0.59). These results suggest that imagination and other autistic traits are not associated with the temporal integration of visual stimuli or the inference of object appearance and words.

Discussion

In the present study, we investigated the following four hypotheses: Imagination difficulties measured by the AQ are associated with lower performance on (1) slit-viewing tasks using line drawings and (2) the fill-in-the-blank task, but are not associated with performance on slit-viewing tasks using (3) characters or (4) meaningless shapes. The results suggest that imagination difficulties were not associated with performance in any of the four tasks, contrary to Hypotheses (1) and (2), but consistent with Hypotheses (3) and (4).

Imagination difficulties were not associated with performance on slit-viewing tasks using characters or meaningless shapes, as hypothesized. These difficulties were not associated with trial-level responses in these tasks after accounting for participant- and stimulus-level variability. These findings are consistent with those of a previous study, which indicated that these difficulties were not associated with performance on a slit-viewing task using written words (Tsuji et al., 2025). Slit-viewing tasks using characters or meaningless shapes require the ability to temporally integrate visual stimuli, which may not be related to imagination. However, the internal consistency of the imagination subscale scores in the AQ was markedly low in the present study (Cronbach's α = .37). The alpha coefficient for the imagination subscale of the original English version of the AQ was .65 (Baron-Cohen et al., 2001). Wakabayashi et al. (2004) developed the Japanese version of the AQ and reported an alpha of .51 for the imagination subscale. Other studies using the Japanese version of the AQ reported alpha coefficients for the imagination subscale ranging from .39 to .49 across four studies (Kitazoe et al., 2015; Nagase, 2019; Tada et al., 2022; Yamamoto, 2009), .53 in Kimura (2019), and .60 in Tsuji et al. (2025). Although the alpha coefficient of the imagination subscale in the present study was particularly low, the imagination subscale of the Japanese version of the AQ may generally show low internal consistency, and its internal consistency in the English version was also not high (Baron-Cohen et al., 2001). Because imagination can involve many related concepts (Crespi et al., 2016), the AQ imagination subscale may include multiple dimensions. In the present study, a confirmatory factor analysis did not support a one-factor structure for the imagination subscale [χ²(35) = 53.96, p = .021, CFI = .44, TLI = .28, RMSEA = .09, and SRMR = .22]. However, the 10 items in this subscale are too few to measure specific aspects of imagination. A more reliable and multidimensional scale or task than the AQ should be used to measure imagination. Hence, although the results suggested that imagination difficulties were not associated with performance on the slit-viewing tasks using characters or meaningless shapes, it would be hasty to conclude that there was no association between them. The present findings remain tentative, and further studies are needed to confirm the relationship between imagination and performance on the tasks used in the present study.

Performance on the slit-viewing task using line drawings was also not associated with imagination. Although the findings remain tentative because of the extremely low internal consistency of the AQ imagination subscale, the present study did not support our expectation that imagination difficulties would be associated with performance on the slit-viewing task using line drawings, given that this task requires the ability to infer the appearance of an object from its fragments. Imagination difficulties were also not associated with performance on the fill-in-the-blank task, which requires the ability to infer words. Therefore, imagination may not be related to the ability to infer the appearance of objects, infer what a word is, or temporally integrate visual stimuli. However, it is necessary to re-examine this relationship using a different measure of imagination, as should be the relationships involving the other tasks.

In addition to limitations of the AQ imagination subscale, one possible explanation for the absence of correlations with task performance is a restricted range of performance. However, the mean ± 1 SD of the error rates for all four tasks fell within the scale range (Table 2), and visual inspection revealed no extreme clustering of performance (Figure 4). Therefore, it is unlikely that ceiling or floor effects, or extremely skewed performance distributions, interfered with the detection of correlations between task performance and the imagination subscale or AQ scale scores.

Although the internal consistency of the total AQ score was acceptable, the autistic traits measured by the total AQ score were not associated with performance on any slit-viewing task. The total AQ score was not associated with the trial-level responses in the slit-viewing tasks after accounting for participant- and stimulus-level variability. This finding is inconsistent with previous studies suggesting that performance on a slit-viewing task using line drawings was lower in autistic individuals than in TD individuals (Nakano et al., 2010; Peiker et al., 2015). However, whereas Nakano et al. (2010) reported reduced performance among autistic individuals in a slit-viewing task using a vertical slit, Peiker et al. (2015) found no difference between autistic and TD individuals when a vertical slit was used, but observed reduced performance in autistic individuals when a horizontal slit was used. In other words, the findings reported by Nakano et al. (2010) were not replicated, even within the clinical group examined by Peiker et al. (2015). One difference between them was the number of stimuli. Forty line-drawings were used by Nakano et al. (2010) and 244 were used by Peiker et al. (2015), both of which were selected from the same set of 260 pictures (Snodgrass & Vanderwart, 1980). Their selection criteria were not reported. Therefore, the finding of reduced performance on the slit-viewing task using a vertical slit in autistic individuals, as reported by Nakano et al. (2010), may only be observed in tasks using specific stimuli and may have low reproducibility. Indeed, Nakano et al. (2010) reported that autistic individuals showed comparable accuracy rates for stimuli that included locally salient features, but lower accuracy rates for those without such features. They suggested that autistic individuals rely on local features rather than the entire picture. Lower performance to identify stimuli without locally salient features may be related to difficulties in global-level temporal processing in autistic individuals (Regener et al., 2024; Stevenson, Siemann, Schneider et al., 2014). Hence, the performance of autistic individuals on the slit-viewing task with a vertical slit may be influenced by whether and to what extent stimuli are characterized by local or global features. The presence or absence of local features of stimulus may also explain the discrepancy between the results of the present study and those of Peiker et al. (2015), both of which used a horizontal slit. To investigate the influences of different local and global features of stimuli, one can compare performances for each line drawing. However, in the present study, 40 were randomly selected from 119 stimuli for each participant, making it difficult to conduct systematic analyses comparing stimuli with different local features.

Another difference of stimuli besides local features may also have contributed to the discrepancy between the findings between Peiker et al. (2015) and the present study, which did not find a relationship between autistic traits and performance in the slit-viewing task. Although we excluded stimuli that were difficult to perceive in the full-viewing task, these stimuli may have been particularly difficult for individuals with high autistic traits to perceive in the slit-viewing task. Further studies are required to examine the influence of differences in line drawings on slit-viewing task performance.

Moreover, the difference in slit width may also be related to the differences in the results. The slit widths were 0.14° in Nakano et al. (2010) and 0.17° in Peiker et al. (2015), suggesting that the task in Nakano et al. (2010) was slightly more difficult. Autistic individuals may tend to exhibit poor performance under more difficult conditions in the slit-viewing task using a vertical slit. If the aforementioned trend is also observed in the task using a horizontal slit, the likelihood of group discrepancies may increase. In subsequent studies, it would be worthwhile to investigate the relationship between autistic traits and performance in slit-viewing tasks using a narrower horizontal slit. Alternatively, a staircase method that adjusts slit width, as used by Orlov et al. (2021), may be useful. This method may also address the difference in slit width between the line-drawing and character slit-viewing tasks in the present study.

Another possible reason why the total AQ score was not related to performance on the slit-viewing task is a non-linear relationship between the AQ score and behavioral performance. The AQ may have a categorical structure, with six subtypes identified—including one with little probability of endorsing autistic traits, one engaging in “systemizing” behaviors, three intermediate groups endorsing multiple components of autistic traits, and a group with high autistic traits (James et al., 2016). This finding suggests that high scores on the total AQ scale may not necessarily predict the manifestation of specific autistic behavioral characteristics.

Another categorical architecture of AQ, characterized by a two-component structure, was also reported (Abu-Akel et al., 2019). It was proposed that autistic traits have not only a single dimensional structure but also a categorical structure that can be explained in terms of the absence or presence of the condition (Abu-Akel et al., 2019). This finding suggests that the distinction between the clinical and non-clinical groups, as influenced by the categorical nature of autistic traits, warrants consideration. This hypothesis is consistent with the findings of a study that reported atypical use of visual facial cues observed in autistic individuals was not observed in individuals with high autistic traits in the general population (Uono et al., 2022). The difference of results between Peiker et al. (2015) and the present study may also be caused by the distinction between clinical and non-clinical groups. Given that autistic traits have been associated with temporal integration deficits in the general population in audiovisual (Ujiie & Wakabayashi, 2022) and auditory (Tsuji & Imaizumi, 2024; Tsuji et al., 2025) tasks, it is reasonable to expect that autistic traits may also be associated with performance in slit-viewing tasks, provided that the task reflects temporal integration ability. Therefore, the slit-viewing task may not accurately reflect the temporal integration deficits, and another factor that differentiates clinical and non-clinical groups may underlie the poorer performance in autistic individuals reported by Peiker et al. (2015).

Peiker et al. (2015) argued that the poor performance of autistic individuals in the slit-viewing task was related not only to integration deficits but also to abnormal interhemispheric synchronization. They recorded magnetoencephalography during a line-drawing slit-viewing task with a horizontal slit. This task requires the integration of visual input across both visual hemifields and coordination between cerebral hemispheres. In TD individuals, gamma-band coherence between bilateral posterior superior temporal sulci increased during the task. However, autistic individuals, who showed reduced task performance, did not exhibit such an increase in bilateral coherence, suggesting atypical interhemispheric communication. If such atypical brain functions co-occur with poor performance on the slit-viewing task and are observed only in the clinical group, consistent with the categorical nature of autistic traits, then it is reasonable to assume that relatively mild autistic traits in the general population do not lead to such atypical brain functions, and thus are not correlated with lower performance on the slit-viewing task. However, the present study could not examine the interhemispheric communication associated with performance on the slit-viewing task. Therefore, future studies should investigate the neural mechanisms related to slit-viewing performance and their association with autistic traits in the general population, using not only line drawings but also other types of stimuli such as characters or meaningless shapes.

From an alternative standpoint, it cannot be wholly refuted that the reproducibility of previous studies’ outcomes (Nakano et al., 2010; Peiker et al., 2015) is low and that the findings were erroneously positive. The sample sizes in prior studies were relatively small: 17 autistic individuals and 16 TD individuals in Nakano et al. (2010), and 20 autistic individuals and 20 TD individuals in Peiker et al. (2015). If the true effect size is small, small sample sizes may have resulted in inflated effect-size estimates in prior studies. To date, only one study has reported a significant difference in performance in the slit-viewing task with a vertical slit between individuals with and without ASD (Nakano et al., 2010), while another study has not observed this difference (Peiker et al., 2015). Furthermore, only one study has reported a discrepancy in performance between groups in the task with a horizontal slit (Peiker et al., 2015), and the extant evidence base remains underdeveloped. The validity of the slit-viewing task as a measure of temporal integration and inference must be verified. Since the motion direction and response format were not standardized across the three slit-viewing tasks in the present study, further studies are also needed to investigate stimulus differences under the same conditions. In addition, fixation monitoring using eye tracking should be used to confirm that participants maintain fixation on a central point. Replication studies with a sufficient sample size and a valid task for both clinical and non-clinical individuals are needed.

The fill-in-the-blank task required the ability to infer the words. This ability also appears to be involved in the perception of speech during temporal dips in noise, where performance has been shown to be reduced in autistic individuals (Alcántara et al., 2004; Dunlop et al., 2016; Groen et al., 2009) and associated with autistic traits (Tsuji & Imaizumi, 2024; Tsuji et al., 2025), as participants must infer words from fragmented speech in noise. However, we only obtained anecdotal evidence regarding the relationship between autistic traits and performance on the fill-in-the-blank task. The inconsistency between previous studies on speech perception in noise and the present study on the fill-in-the-blank task may be due to differences across modalities. Individuals with high autistic traits may find it easier to infer words from visual stimuli than auditory stimuli. Additional research is required to investigate the relationship between autistic traits and inference difficulties using stimuli from both modalities.

Conclusions

The present study showed that imagination difficulties related to ASD were not associated with performance on slit-viewing tasks using line drawings, characters, or meaningless shapes, as well as on the fill-in-the-blank task. Overall autistic traits were also unrelated to these tasks. These findings suggest that imagination difficulties and overall autistic traits may not be associated with the ability to temporally integrate visual stimuli, infer the appearance of objects, or infer a word from incomplete letter strings. However, because the reliability and validity of the measures of autistic traits, particularly imagination, were insufficient, our findings should be considered tentative. The effects of the methodological differences between the slit-viewing tasks should be investigated in future studies. In addition, the structure of autistic traits should also be examined to clarify discrepancies between studies targeting clinical and non-clinical groups.

Supplemental Material

sj-pdf-1-ipe-10.1177_20416695261465782 - Supplemental material for Autistic imagination sub-traits are unrelated to visual temporal integration or inference

Supplemental material, sj-pdf-1-ipe-10.1177_20416695261465782 for Autistic imagination sub-traits are unrelated to visual temporal integration or inference by Yurika Tsuji, Yuki Nishiguchi, Akari Noda and Shu Imaizumi in i-Perception

Footnotes

ORCID iDs

Yurika Tsuji

Shu Imaizumi

Ethical Considerations

This study was approved by the Ethics Committee of the Faculty of Education, Chiba University (approval no. 156) on September 22, 2023.

Consent to Participate

All participants provided written informed consent prior to enrolment in the study.

Author Contribution(s)

Yurika Tsuji: Conceptualization; Data curation; Formal analysis; Formal analysis; Investigation; Writing – original draft.

Yuki Nishiguchi: Conceptualization; Formal analysis; Supervision; Writing – review & editing.

Akari Noda: Conceptualization; Data curation; Investigation; Writing – review & editing.

Shu Imaizumi: Conceptualization; Formal analysis; Formal analysis; Supervision; Writing – review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: YT was supported by JSPS KAKENHI (22J11083 and 24KJ1123) and the Institutional Research Fund of Tokyo Metropolitan University. YN was supported by JSPS KAKENIHI (23K02933). SI was supported by JSPS KAKENHI (23K11785).

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The raw data is publicly available at .

Supplemental Material

Supplemental material for this article is available online.

References

Abu-Akel

Allison

Baron-Cohen

Heinke

(2019). The distribution of autistic traits across the autism spectrum: Evidence for discontinuous dimensional subpopulations underlying the autism continuum. Molecular Autism, 10, 24. https://doi.org/10.1186/s13229-019-0275-3

Alcántara

J. I.

Weisblatt

E. J. L.

Moore

B. C. J.

Bolton

P. F.

(2004). Speech-in-noise perception in high-functioning individuals with autism or Asperger’s syndrome. Journal of Child Psychology and Psychiatry, 45, 1107–1114. https://doi.org/10.1111/j.1469-7610.2004.t01-1-00303.x

Baron-Cohen

Wheelwright

Skinner

Martin

Clubley

(2001). The autism-spectrum quotient (AQ): Evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. Journal of Autism and Developmental Disorders, 31, 5–17. https://doi.org/10.1023/a:1005653411471

Bebko

J. M.

Schroeder

J. H.

Weiss

J. A.

(2014). The McGurk effect in children with autism and Asperger syndrome. Autism Research, 7, 50–59. https://doi.org/10.1002/aur.1343

Benjamini

Hochberg

(1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society: Series B, 57, 289–300. https://doi.org/10.2307/2346101

Casassus

Poliakoff

Gowen

Poole

Jones

L. A.

(2019). Time perception and autistic spectrum condition: A systematic review. Autism Research, 12, 1440–1462. https://doi.org/10.1002/aur.2170

Crespi

Leach

Dinsdale

Mokkonen

Hurd

(2016). Imagination in human social cognition, autism, and psychotic-affective conditions. Cognition, 150, 181–199. https://doi.org/10.1016/j.cognition.2016.02.001

Dunlop

W. A.

Enticott

P. G.

Rajan

(2016). Speech discrimination difficulties in high-functioning autism spectrum disorder are likely independent of auditory hypersensitivity. Frontiers in Human Neuroscience, 10, 401. https://doi.org/10.3389/fnhum.2016.00401

Foss-Feig

J. H.

Kwakye

L. D.

Cascio

C. J.

Burnette

C. P.

Kadivar

Stone

W. L.

Wallace

M. T.

(2010). An extended multisensory temporal binding window in autism spectrum disorders. Experimental Brain Research, 203, 381–389. https://doi.org/10.1007/s00221-010-2240-4

10.

Groen

W. B.

Van Orsouw

Huurne

N. T.

Swinkels

Van Der Gaag

R. J.

Buitelaar

J. K.

Zwiers

M. P.

(2009). Intact spectral but abnormal temporal processing of auditory stimuli in autism. Journal of Autism and Developmental Disorders, 39, 742–750. https://doi.org/10.1007/s10803-008-0682-3

11.

Irwin

J. R.

Tornatore

L. A.

Brancazio

Whalen

D. H.

(2011). Can children with autism spectrum disorders “hear” a speaking face? Child Development, 82, 1397–1403. https://doi.org/10.1111/j.1467-8624.2011.01619.x

12.

James

R. J.

Dubey

Smith

Ropar

Tunney

R. J.

(2016). The latent structure of autistic traits: A taxometric, latent class and latent profile analysis of the adult autism spectrum quotient. Journal of Autism and Developmental Disorders, 46, 3712–3728. https://doi.org/10.1007/s10803-016-2897-z

13.

JASP Team. (2025). JASP 0.19.3. https://jasp-stats.org/.

14.

Kimura

(2019). Characteristics of social anxiety of university students with high autistic traits: Links with self-esteem and public self-consciousness. The Japanese Journal of Personality, 28, 97–107. https://doi.org/10.2132/personality.28.2.2

15.

Kitazoe

Inoue

Izumoto

Kumagai

Iwasaki

(2015). The Autism-Spectrum Quotient in university students: Pattern of changes in its scores and associated factors. Asia-Pacific Psychiatry, 7, 105–112. https://doi.org/10.1111/appy.12094

16.

Kosslyn

S. M.

(2005). Mental images and the brain. Cognitive Neuropsychology, 22, 333–347. https://doi.org/10.1080/02643290442000130

17.

Lee

M. D.

Wagenmakers

E. J.

(2014). Bayesian cognitive modeling: A practical course. Cambridge University Press.

18.

Meilleur

Foster

N. E. V.

Coll

S. M.

Brambati

S. M.

Hyde

K. L.

(2020). Unisensory and multisensory temporal processing in autism and dyslexia: A systematic review and meta-analysis. Neuroscience and Biobehavioral Reviews, 116, 44–63. https://doi.org/10.1016/j.neubiorev.2020.06.013

19.

Mongillo

E. A.

Irwin

J. R.

Whalen

D. H.

Klaiman

Carter

A. S.

Schultz

R. T.

(2008). Audiovisual processing in children with and without autism spectrum disorders. Journal of Autism and Developmental Disorders, 38, 1349–1358. https://doi.org/10.1007/s10803-007-0521-y

20.

Mulder

Pfadt

Wagenmakers

E. J.

(2025). A tutorial on Bayesian hypothesis testing of correlation coefficients using the BFpack-module in JASP. Behavior Research Methods, 57, 311. https://doi.org/10.3758/s13428-025-02846-5

21.

Nagase

(2019). The traits of autism spectrum disorder in the general population influence humor appreciation: Using the autism-spectrum quotient and HSPS-J19. Cogent Psychology, 6, 1696000. https://doi.org/10.1080/23311908.2019.1696000

22.

Nakano

Ota

Kato

Kitazawa

(2010). Deficit in visual temporal integration in autism spectrum disorders. Proceedings: Biological Sciences, 277, 1027–1030. https://doi.org/10.1098/rspb.2009.1713

23.

Orlov

Raveh

McKyton

Ben-Zion

Zohary

(2021). Learning to perceive shape from temporal integration following late emergence from blindness. Current Biology, 31, 3162–3167. https://doi.org/10.1016/j.cub.2021.04.059

24.

Parks

T. E.

(1965). Post-retinal visual storage. The American Journal of Psychology, 78, 145–147. https://doi.org/10.2307/1421101

25.

Peiker

David

Schneider

T. R.

Nolte

Schottle

Engel

A. K.

(2015). Perceptual integration deficits in autism spectrum disorders are associated with reduced interhemispheric gamma-band coherence. The Journal of Neuroscience, 35, 16352–16361. https://doi.org/10.1523/JNEUROSCI.1442-15.2015

26.

Peirce

Gray

J. R.

Simpson

MacAskill

Hochenberger

Sogo

Kastman

Lindelov

J. K.

(2019). Psychopy2: Experiments in behavior made easy. Behavior Research Methods, 51, 195–203. https://doi.org/10.3758/s13428-018-01193-y

27.

Regener

Heffer

Love

S. A.

Petrini

Pollick

(2024). Differences in audiovisual temporal processing in autistic adults are specific to simultaneity judgments. Autism Research, 17, 1041–1052. https://doi.org/10.1002/aur.3134

28.

Schönbrodt

F. D.

Wagenmakers

E. J.

(2018). Bayes factor design analysis: Planning for compelling evidence. Psychonomic Bulletin & Review, 25, 128–142. https://doi.org/10.3758/s13423-017-1230-y

29.

Snodgrass

J. G.

Vanderwart

M. L.

(1980). A standardized set of 260 pictures: Norms for name agreement, image agreement, familiarity, and visual complexity. Journal of Experimental Psychology: Human Learning and Memory, 6, 174–215. https://doi.org/10.1037/0278-7393.6.2.174

30.

Soulieres

Zeffiro

T. A.

Girard

M. L.

Mottron

(2011). Enhanced mental image mapping in autism. Neuropsychologia, 49, 848–857. https://doi.org/10.1016/j.neuropsychologia.2011.01.027

31.

Stevenson

R. A.

Siemann

J. K.

Schneider

B. C.

Eberly

H. E.

Woynaroski

T. G.

Camarata

S. M.

Wallace

M. T.

(2014). Multisensory temporal integration in autism spectrum disorders. Journal of Neuroscience, 34, 691–697. https://doi.org/10.1523/JNEUROSCI.3615-13.2014

32.

Stevenson

R. A.

Siemann

J. K.

Woynaroski

T. G.

Schneider

B. C.

Eberly

H. E.

Camarata

S. M.

Wallace

M. T.

(2014). Evidence for diminished multisensory integration in autism spectrum disorders. Journal of Autism and Developmental Disorders, 44, 3161–3167. https://doi.org/10.1007/s10803-014-2179-6

33.

Tada

Hasegawa

Kondo

H. M.

(2022). Sensitivity to everyday sounds: ASMR, misophonia, and autistic traits. The Japanese Journal of Psychology, 93, 263–269. https://doi.org/10.4992/jjpsy.93.21319

34.

Tsuji

Imaizumi

(2024). Autistic traits and speech perception in social and non-social noises. Scientific Reports, 14, 1414. https://doi.org/10.1038/s41598-024-52050-2

35.

Tsuji

Nishiguchi

Noda

Imaizumi

(2025). Autistic traits and temporal integration of auditory and visual stimuli in the general population: The role of imagination. Multisensory Research, 38, 453–483. https://doi.org/10.1163/22134808-bja10160

36.

Ujiie

Wakabayashi

(2022). Intact lip-reading but weaker McGurk effect in individuals with high autistic traits. International Journal of Developmental Disabilities, 68, 47–55. https://doi.org/10.1080/20473869.2019.1699350

37.

Uono

Egashira

Hayashi

Takada

Ukezono

Okada

(2022). No influence of emotional faces or autistic traits on gaze-cueing in general population. Frontiers in Psychology, 13, 864116. https://doi.org/10.3389/fpsyg.2022.864116

38.

Van Laarhoven

Stekelenburg

J. J.

Vroomen

(2019). Increased sub-clinical levels of autistic traits are associated with reduced multisensory integration of audiovisual speech. Scientific Reports, 9, 9535. https://doi.org/10.1038/s41598-019-46084-0

39.

Wakabayashi

Tojo

Baron-Cohen

Wheelwright

(2004). The Autism-Spectrum Quotient (AQ) Japanese version: Evidence from high-functioning clinical group and normal adults. The Japanese Journal of Psychology, 75, 78–84. https://doi.org/10.4992/jjpsy.75.78

40.

Woynaroski

T. G.

Kwakye

L. D.

Foss-Feig

J. H.

Stevenson

R. A.

Stone

W. L.

Wallace

M. T.

(2013). Multisensory speech perception in children with autism spectrum disorders. Journal of Autism and Developmental Disorders, 43, 2891–2902. https://doi.org/10.1007/s10803-013-1836-5

41.

Yaguchi

Hidaka

(2021). Autistic communication and imagination sub-traits are related to audiovisual temporal integration in the stream–bounce illusion. Multisensory Research, 34, 477–492. https://doi.org/10.1163/22134808-bja10042

42.

Yamamoto

(2009). Adolescent egocentrism–sociocentrism and autism-spectrum quotient in Japanese people. In Yoshizaki

Ohnishi

(Eds.), Contemporary issues of brain, communication and education in psychology: The science of mind (pp. 237–245). Union Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.76 MB