Participants
This study performed a one-way within-subject analysis of variance (ANOVA) using the difference between the three types of images as a within-participant factor. In order to detect an effect (α = 0.05, 1 − β = 0.80, f = 0.25) in a one-way within-subjects ANOVA, G*Power [18] estimated the sample size needed to be 28. Thus, 28 undergraduate and graduate students from the University of Tsukuba (17 females and 11 males; mean age = 24.6; SD = 0.6 years; range = 19–33 years) participated in this study. The average score of the 28 participants in the Japanese version of the Trypophobia Questionnaire (TQ-J) [10, 19], which measures trypophobia proneness, was 33.32 (SD = 14.52, range: 17–79). Among them, 10 participants (36%) scored higher than the cut-off value (= 31). All of the participants provided written informed consent. The experiment was conducted according to the principles of the Declaration of Helsinki and approved by the Ethics Committee of the Faculty of Art and Design, University of Tsukuba (Approval Number 29-11).
Stimuli
Eight frontal facial images (two Japanese men and two Japanese women, two Caucasian men, and two Caucasian women) were taken from the Japanese and Caucasian Neutral Faces (JACNeuF) [20]. All facial images were converted to grayscale images. Using the eight pictures, we created four sets of stimuli (Fig. 1a–d). In the first set (Fig. 1a, face-only), we used eight grayscale pictures without adding dots. The average grayscale values were set to 140 in Adobe Photoshop CC2018, and the average luminance of the pictures was 100.6 cd/m2. In the second set (Fig. 1b, black dots on the face), a cluster of 42 dots was placed on each cheek of a face within an approximately 3.9° × 3.5° notional oval area. The dots were colored black and gray with circular gradation. The sizes and positions of each dot were based on a previous study [6] in which the dots were drawn by tracing real lotus seed pods. The gradation was added to enhance discomfort [10]. The 24 black dots were 0.3° in diameter, and the remaining 18 black dots were 0.4° in diameter. The average luminance of the black dots stimuli was 95.0 cd/m2, which was darker than the face-only stimuli, because of the presence of black dots on the face. In the third set (Fig. 1c, gray dots on the face), we took the position of black dots in the black dots stimuli as the coordinate axis, and the positions of the gray dots were shifted 0.1° left and 0.3° below on the left side of the face. On the contralateral side, the positions of the gray dots shifted 0.1° to the right and 0.3° below the right side of the face. Finally, we removed all the black dots and retained only the gray dots. The number of gray dots corresponds to the number of black dots in one set with black dots, for a total of 84. Gray dots were drawn with uniform gray, keeping the same sizes and shapes as the original black dots. The average luminance of gray dots stimuli was 95.7 cd/m2. In the fourth set (Fig. 1d, black dots and gray dots on the face), both black dots in the black dots stimuli and gray dots in the gray dots stimuli were presented together on the face. A cluster of black dots was presented on top of the cluster of gray dots. A cluster of 84 dots (42 black dots and 42 gray dots) was present on each side, and 168 dots (84 black dots and 84 gray dots) were placed on the face. The luminance of the black dots and gray dots stimuli was 93.3 cd/m2. In total, there were 32 stimulus images. The average luminance of the black dots was 6.0 cd/m2. The luminance of the gray dots (uniform gray) was 39.1 cd/m2. The average luminance of the face background around the dots was approximately 99.4 cd/m2. Therefore, the luminance contrast of the black dots on a face background was (99.4–6.0)/(99.4 + 6.0) = 0.89. The luminance contrast of the gray dots on a face background was (99.4–39.1)/(99.4 + 39.1) = 0.44. All images were subtended about 17.7° horizontally and 23.1° vertically and presented centrally on a white background on a 13-inch computer screen (MacBook Air, Model A1369).
Spatial frequency spectra of the four types of stimuli analyzed with the SHINE toolbox [21] were shown in Fig. 2. Contrast energies of each stimulus showed different patterns in the lower (37–60 cycles per image) vs. higher (67–150 cycles per image) ranges of the medium spatial frequencies (37.5–150 cycles per image) which were reported to contribute to discomfort [22]. In the lower (37–60 cycles per image) range, the contrast energy of the black dots was higher than those of black and gray dots, and gray dots. On the other hand, in the higher part (67–150 cycles per image), the contrast energy of the black and gray dots was higher than those of black dots and gray dots.
Apparatus
The experiment was conducted in a lit room (approximately 1,000 lx). The participants rested on a comfortable chair, and the viewing distance was fixed at 57 cm away from the display. An Apple MacBook with a 13-inch, 1440 × 900 pixels display was used as the experimental apparatus, and the presentation of images was controlled using PsyScope X Program (available at http://psy.ck.sissa.it) [23] that ran presentation software and recorded participants’ evaluations.
Procedure
The experiment consisted of two types of evaluation tasks, a discomfort evaluation task and a depth evaluation task. In the discomfort evaluation task, each trial began with a central presentation of eye fixation (“ + ”) on a white background for three seconds (Fig. 3). The cross subtended about 1.0° horizontally and 1.0° vertically and was presented centrally on the white background of a computer screen. A facial stimulus followed, and one of the thirty-two images (consisting of eight face-only stimuli, eight black dots stimuli, eight gray dots stimuli, eight black dots and gray dots stimuli, respectively) was presented in random order for three seconds. After that, the participants were asked to evaluate the discomfort she/he felt from each image. An evaluation image was presented directly after each facial stimulus. Participants were asked to rate each image using a 9-point Likert scale, ranging from − 4 (uncomfortable) to 4 (comfortable), using the numeric keypad with unrestricted time. Each participant performed 32 trials: four image types (face-only stimuli, black dots stimuli, gray dots stimuli, black dots and gray dots stimuli) × eight facial images.
After the discomfort evaluation task, the participants underwent the depth evaluation task. The procedures were the same as the discomfort evaluation task, except that the participants were asked to judge whether they perceived depth between the foreground dots and the background face. After the facial stimulus was presented, an evaluation image with an evaluation scale was presented, and participants were asked to judge whether they felt the depth between black dots and the face (“yes” or “no”) using the numeric keypad. The viewing time and judgment time were unrestricted. Each participant performed 24 trials: three image types (black dots stimuli, gray dots stimuli, and black dots and gray dots stimuli) × eight facial images.