Skip to main content

Young Scientist Journal

Young Scientist Journal

Evaluating the Emotional Salience of Environmental Images Using the Attentional Blink Paradigm

ABSTRACT

Attention is inextricably linked to emotion, with highly salient imagery attracting more attention at earlier time frames. Environmental issues pose an imminent threat to our planet, but the salience of images associated with the environment is unclear. The Attentional Blink Paradigm (AB) was used to compare participant (N = 63) accuracy when identifying target imagery following a disgusting, pleasant, neutral, or unpleasant environmental image at differing lag times (200, 400, 600, 800 ms). In the AB, accuracy is reduced for the identification of target imagery when it follows shortly after an emotionally salient distractor image. Thus, if unpleasant environmental imagery were perceived as emotionally salient images, we would see poorer accuracy in those conditions. We hypothesized that unpleasant environmental images would be more salient than pleasant and neutral imagery but less salient than disgust imagery, and that accuracy in all categories would improve with increased lag time. We observed a significant effect of distractor category on accuracy at different time intervals (i.e., Lag; p < .001) and individual main effects of Lag and distractor category on accuracy (ps < .001). More specifically, when the environmental imagery served as distractors at Lag 2, participants were most accurate, whereas when pleasant images served as distractors at Lag 8, participants were most accurate (ps < .001). Although participants expressed concerns in a survey about environmental attitudes, the results were not reflected as increased emotional salience toward unpleasant environmental images in the AB. Contrary to our hypothesis, unpleasant environmental images were not shown to be salient distractors.  One possible explanation is that the inclusion of facial imagery in other categories may have elicited stronger attentional engagement than the unpleasant environmental images, which lacked faces entirely. In the future, these results could help identify salient imagery for environmental educational efforts.

INTRODUCTION.

Issues like climate change, environmental degradation, and pollution present some of the largest threats to modern day society.   Identifying the emotional salience of imagery related to these topics—that is, the extent to which these images receive prioritized cognitive processing— can inform more effective environmental education efforts. By emphasizing imagery that reliably captures attention, communication efforts and messages designed for the public might be enhanced and more impactful. To develop this approach, however, requires a systematic understanding of  individuals’ responses to imagery depicting negative environmental crises like climate change, environmental degradation, and pollution.

Previous work has shown that attention is inextricably linked to emotion, with highly salient images capturing attention at greater magnitudes than neutral imagery [1]. Research in this field has often utilized a rapid serial visual presentation (RSVP) paradigm, specifically the Emotional Attentional Blink (AB) task, to assess emotional modulation of attention at different time intervals [2]. This paradigm presents a series of images in rapid succession with emotionally salient, distracting imagery causing participants to experience temporary emotion induced blindness as they fail to recognize subsequent imagery in close proximity to the emotionally distracting target. Thus, participants are less accurate at recognizing target images following an emotional distractor [2]. High accuracy in target recognition in the AB indicates that the distractor had low emotional salience, whereas low accuracy suggests the distractor had a higher valuation of emotional salience.

Research has also shown that emotion induced blindness is most noticeable at shorter time frames. At later presentations, the effects of emotion induced blindness are less pronounced, resulting in increased accuracy in the detection of targets following an emotional stimulus [3]. For example, a previous study using the Attentional Blink task to assess the effects on pleasant, neutral, and disgust stimuli showed impaired attention to emotional imagery at shorter time frames [2]. In that study, pleasant and disgust stimuli were shown to impair attention at earlier lag times with accuracy improving at later lag times. These findings are in alignment with an evolutionary perspective: stimuli with a high emotional salience, such as images relevant to survival (i.e., a threatening dog), demand immediate attention because they signal importance. However, this heightened attention would be maladaptive if attention could not be efficiently reallocated to other relevant information when necessary. Importantly, it has been shown that subsets of unpleasant emotional categories (e.g., disgust, fear) have distinct effects on attention [4-5].

Attentional processing is tied to the valence and arousal of a given stimulus. The Self-Assessment Manikin (SAM) is a self-report measure used to categorize the valence and arousal associated with an object or event [6]. Valence describes how pleasant or unpleasant an image is, whereas arousal reflects the intensity of an emotional state [5]. Lower arousal and higher valence usually point to a calmer image that is still pleasant (e.g., a baby or a sunset; see Figure 1). Conversely, an image of a person throwing up at a party is highly arousing, but unpleasant. Research has shown that high arousing images attract attention and are typically viewed longer than low-arousing images [7].

Figure 1. Figure was adapted from [6] to depict representative images from the International Affective Picture System (IAPS; [8]) in a 2-dimensional affective space, as defined by SAM pleasure and arousal ratings.
In the present study, we employ the AB paradigm to assess emotional modulation of attention at different time intervals, focusing specifically on whether unpleasant environmental imagery (e.g., climate change, environmental degradation, and pollution) is perceived as emotionally salient. To assess whether unpleasant environmental stimuli, like pleasant and disgust stimuli in previous work [2], similarly impairs attention at shorter lag times but improves it at longer lag times [9], we varied the distractor and target images interval (lag) at 200 ms, 400 ms, 600 ms, and 800 ms [2]. We predicted that unpleasant environmental images would be more distracting and lead to lower accuracy than neutral and pleasant images but be less distracting than disgust images. We also expected accuracy in detecting unpleasant environmental images to improve as lag time increased. These findings will contribute to our understanding of the emotional salience of imagery related to climate change and pollution and may offer insight into how society perceives these issues.

MATERIALS AND METHODS.

Participants.

This study was conducted at Vanderbilt University in Nashville, TN with a total of 63 participants. Participants ranged from high school students to adults with a mean age of 24.5 years (SD = 13; range = 14 – 66). Participants were 62% female and 57% Caucasian with a majority (62%) being in high school (Table 1).

Table 1. Demographics Table of Participants
N=63 Ethnicity                % (N) Education      %(N)
Average Age (SD) African American 10.8 (7) High School 62.3 (41)
24.5 (13) Caucasian               57 (37) Post-Secondary 1.5 (1)
Sex                     %(N) Asian                    16.9 (11) Bachelors            9 (6)
Female       62.3 (41) Prefer not to say      4.6 (3) Masters            7.5 (5)
Male           37.7 (27) Other                      12.3 (8) Doctoral       19.7 (13)

Pre-Test Survey.

Participants filled out a demographics questionnaire prior to completion of the AB Task. Participants were asked for their age, ethnicity, and gender (Table 1). A second component of the questionnaire asked participants about their experience with the Attentional Blink, their involvement with climate activism (e.g., recycling, composting, volunteering), and their exposure to extreme weather events caused by climate change (i.e., flooding). This part of the survey was used to determine if participants could have any bias for the images used.

Stimuli.

The visual stimuli consisted of images formatted to a 4:3 ratio in four different categories: 48 unpleasant environmental images, 48 disgust images, 48 positive images, and 48 neutral images. Additionally, there were 252 upright landscape or architectural images that served as background stimuli, and 80 target images. Target images were landscape or architectural images rotated to the right 90 degrees or to the left 90 degrees [2]. Disgust images were sourced from the International Affective Picture System (IAPS) [8], the Emotional Picture Set (EmoPicS) [10], the Open Affective Standardized Image Set (OASIS) [11], and similar pictures from publicly available sources, contained images of contaminated or diseased items such as feces, roaches, and diseased human flesh, and were selected on their low valence (Avg: 2.46) and high arousal (Avg: 5.12) ratings. Pleasant images were sourced from both the EmoPicS and IAPS datasets and contained images of happy or excited faces, adventure sports, and skydiving, and were selected because of their high valence (Avg: 7.06) and arousal scores (Avg: 5.44). Neutral images were solely sourced from the EmoPicS dataset, consisted of 24 ordinary objects such as chairs or clocks and 24 people with neutral expressions and body language, and were selected due to their medium valence (Avg: 4.93) and low arousal (Avg: 3.15). Unpleasant environmental images were found online and focused on issues facing the planet such as pollution, habitat loss, and global warming. Stimulus size and selection was based on the work of Ciesielski and colleagues [2].

Procedure.

Our procedure closely resembles that of Ciesielski and colleagues. Similar to their study, participant accuracy in the attentional blink task was calculated on participants’ identification of  directional rotation of landscape images following an emotional or neutral distractor [2]. On each trial of the AB task, 17 images were presented on a white background in rapid succession using the computer software Testable (testable.org). Each image was presented for 100 ms and each trial lasted a total of 1700 ms. Each trial included one distractor image (disgust, pleasant, unpleasant environmental, neutral) and one rotated target image (landscape), presented at variable intervals (Figure 2). Stimulus presentation was randomized, and each trial consisted of either a pleasant, neutral, unpleasant environmental, or disgust distractor image that occurred either 400, 600, or 800 ms after trial onset. The distractor image appeared 200 ms (Lag 2), 400 ms (Lag 4), 600 ms (Lag 6), or 800 ms (Lag 8) before the rotated target image (Figure 3). Participants completed a total of 196 trials in one sitting with a 30 second break halfway through the testing period. The experiment lasted approximately 6 minutes. After each trial, participants were asked to respond in which direction the target image was presented: rotated to the left or to the right using the L and R keys on a keyboard. The accuracy of these responses was recorded for each of the 196 total trials, with each distractor type being presented 48 times with an additional 1 trial per distractor type without a subsequent target. The 4 trials without a target image were used to ensure participants were not simply guessing at random ‘yes’ or ‘no’ for seeing a target. Thus, if they are answering that they did see a target when none was present then they would be simply guessing, however accuracy for all categories proved well above chance for these trials. The 4 lags were equally distributed among trials. Following each trial, participants were instructed to indicate if they saw a rotated image (yes or no; detection) and then asked what direction the rotated image was turned (right or left; accuracy).

Figure 2. Schematic of the trial procedure for the attentional-blink paradigm.
Figure 3. Schematic depicting the variable timing of the distractor and target images by lag time. P4/6/8: Indicates location position within the temporal stream. Lag position indicates the temporal delay (lag) after the distractor image (e.g., Lag 2 indicates a 200 ms delay after the distractor image), when the target image can first appear.

Statistical Analysis.

RStudio was used to conduct a 4×4 ANOVA on the Emotion (disgust, pleasant, unpleasant environmental, neutral) x Lag (200, 400, 600, 800 ms) interaction as well as a summary of the participant questionnaire responses and two one-way ANOVAs on the Emotion (disgust, pleasant, unpleasant environmental, neutral) x Accuracy interaction and the Lag (200, 400, 600, 800 ms) x Accuracy interaction.

RESULTS.

Survey.

Participants reported a high level of concern about climate change and environmental issues, rating their concern at an average of 8.8 out of 10, with 10 representing the highest level of concern. Additionally, participants rated their involvement in environmental protection efforts as moderately high, reporting an average of 5.6 out of 10.

Emotion Content, Time Course, and Target Detection.

The 4 (Emotion; disgust, pleasant, neutral, and unpleasant environmental) x 4 (Lag; 2, 4, 6, 8) ANOVA on accuracy revealed a significant Emotion x Lag interaction (p<.001). To examine this interaction, a subsequent ANOVA assessing accuracy by emotional category at each lag time was conducted. At each lag, a main effect of emotion was observed (all ps<.001). Figure 4 shows that relative to all other trials of the same lag, participants were significantly more accurate in identifying target imagery with unpleasant environmental distractors at 200 ms (ps<.001). Furthermore, participants were significantly more accurate in target imagery when the distractor image was positive at 800ms (ps<.001; see Table 2).

Figure 4. Accuracy scores for each emotional category by lag time. Error bars represent standard error of the mean. Higher accuracy scores are interpreted as reflecting lower levels of emotional salience.

 

Table 2. Average Accuracy and Standard Deviation for each Distractor and Lag Time
Disgust Unpleasant  Pleasant Neutral
M (SD) M (SD) M (SD) M (SD)
Lag 2 50.26 (19.85) 74.6 (22.17) 56.75 (20.89) 50.53 (20.89)
Lag 4 71.83 (22.47) 77.38 (23.97) 69.97 (22.02) 71.83 (24.34)
Lag 6 75.13 (23.4) 80.29 (24.1) 77.25 (24.00) 77.25 (25.05)
Lag 8 55.16 (18.05) 53.04 (15.86) 79.5 (23.18) 49.74 (15.7)

Emotion Content and Target Accuracy.

A 4 (Emotion; disgust, pleasant, neutral, unpleasant environmental) x 4 (Lag; 2, 4, 6, 8) ANOVA on accuracy revealed a significant main effect of the Emotion category on attention (p<.001) (Figure 4). Since previous studies have found that overall findings for target detection and accuracy in the AB to be essentially identical [1], we present here analyses for accuracy, rather than detection, as they reflect more precise performance on the AB. Tukey post-hoc comparisons showed that emotion significantly affected accuracy. Disgust differed from unpleasant environmental stimuli and pleasant stimuli, and neutral stimuli differed from both unpleasant environmental stimuli and pleasant stimuli (all ps<.001).

Time Course and Target Detection.

The 4 (Emotion; disgust, pleasant, neutral, and unpleasant environmental) x 4 (Lag; 2, 4, 6, 8) ANOVA on percent accuracy also revealed a significant main effect of Lag on attention (p<.001) (Figure 4). Tukey post-hoc comparisons showed participants to be significantly less accurate in identifying the direction of a target at Lag 2 in comparison to Lag 4 and Lag 6 (ps<.001), but not at Lag 8 (p = .9). Additionally, we observed a significant difference in accuracy between Lag 8 and Lag 4 and between Lag 8 and Lag 6 (ps<.001) but not between Lag 4 and Lag 6 (p = .07; see Figure 4).

DISCUSSION.

Given the increasing threat that environmental dangers like climate change pose to our planet, it is essential to gain a better understanding of how people perceive unpleasant environmental imagery like those depicting environmental degradation or pollution.  We utilized the AB paradigm to evaluate the extent to which unpleasant environmental were perceived as emotionally salient images. We predicted that unpleasant environmental stimuli would elicit a more pronounced attentional blink response than neutral and pleasant images, but a less pronounced response than disgust. Contrary to our hypothesis, unpleasant environmental images did not act as salient distractors and instead followed patterns more similar to neutral stimuli [2].

The significant Emotion x Lag interaction observed here shows that both the type of distractor and lag time before target identification affected participant accuracy. Individually, both Emotion and Lag had independent main effects on accuracy, demonstrating that the emotional category of the distractor imagery itself and the time frame of the trial both impacted participants’ ability to recognize the target. In this context, this shows that the type of stimuli presented does provoke differing emotional responses given the significant differences in target recognition accuracy.

The unpleasant environmental category had a significantly higher accuracy at earlier time frames compared to all other image categories, suggesting that it was less effective at capturing participant’s attention than disgust, pleasant, and neutral images. This trend continues until Lag 8 where the accuracy of the unpleasant environmental stimulus category then drops to the same level as neutral and disgust imagery. At Lag 8, pleasant images have the highest levels of accuracy compared to the other image categories, suggesting that while images capture attention at earlier time frames, this effect is flipped at later lag times.

As observed in previous studies, unpleasant environmental stimuli evoke responses similar to those of neutral images [2]. One possible explanation for this phenomenon is the uneven distribution of faces in each of our image categories. Previous research has shown that in the AB task, faces capture participant attention at higher levels than imagery that depicting landscape scenery [7]. In our study, the pleasant category had the highest ratio of pictures with faces to total pictures (31:48), followed by neutral (24:48), and disgust (7:48). Unpleasant environmental imagery had the lowest ratio of faces to total images (5:48) and mainly consisted of landscape pictures, which we know to be most well categorized as neutral stimuli. Based on this information, we hypothesize that the lack of attention-capturing facial imagery in the unpleasant environmental imagery was one of the contributing factors to the to the results reported here.

Future studies should ensure that facial imagery is evenly distributed across all stimulus categories [12]. Additionally, the inclusion of more high-arousing images could be studied in an adult population to further tease apart the category distinctions with respect to distractor imagery. Since participants did not provide valence and arousal ratings of the unpleasant environmental stimuli, we would suggest having future studies collect SAM ratings on these images to provide additional information on where these images fall within affective space. Our hope is that this research can be used in the future to optimize educational efforts on environmental issues by identifying images that capture attention most effectively. Such images could be utilized in educational settings and public campaigns to raise public awareness and encourage engagement in climate and environmental activism.

ACKNOWLEDGMENTS.

Special thanks to Dr. Deweese for her advice, Dr. Stanton, and The School of Science and Math at Vanderbilt for their support, and to Dr. Olatunji for providing the imagery from his original study.

REFERENCES

  1. J. G. Taylor, N. F. Fragopanagos, The interaction of attention and emotion. Neural Netw. 18, 353–369 (2005).
  2. B. G. Ciesielski, T. Armstrong, D. H. Zald, B. O. Olatunji, Emotion modulation of visual attention: categorical and temporal characteristics. PLoS ONE 5, e13860 (2010).
  3. M. W. Becker, Panic search: fear produces efficient visual search for nonthreatening objects. Psychol. Sci. 20, 435–437 (2009).
  4. J. M. Cisler, B. O. Olatunji, J. M. Lohr, N. L. Williams, Attentional bias differences between fear and disgust: implications for the role of disgust in disgust-related anxiety disorders. Cogn. Emot. 23, 675–687 (2009).
  5. A. Öhman, Of snakes and faces: an evolutionary perspective on the psychology of fear. Scand. J. Psychol. 50, 543–552 (2009).
  6. M. Bradley, Lang, Peter, Measuring Emotion: The self assessment manikin and the semantic differential. J. Behav. Ther. Exp. Psychiatry 25, 49–59 (1994).
  7. P. Lang, M. Greenwald, M. Bradley, A. Hamm, Looking at pictures: affective, facial, visceral and behavioral reactions. Psychophysiology 30, 261–273 (1993).
  8. P. Lang, M. Bradley, The International Affective Picture System (IAPS) in the study of emotion and attention. Handb. Emot. Elicitation Assess. 29, 29–46 (2007).
  9. B. R. Bocanegra, R. Zeelenberg, Dissociating emotion-induced blindness and hypervision. Emotion 9, 865–873 (2009).
  10. M. Wessa, P. Kanske, P. Neumeister, K. Bode, J. Heissler, S. Schönfelder, EmoPicS: subjective and psychophysiological evaluation of new imagery for clinical biopsychological research. Z. Klin. Psychol. Psychother. Suppl. 1, 11–77 (2010).
  11. B. Kurdi, S. Lozano, M. R. Banaji, Introducing the Open Affective Standardized Image Set (OASIS). Behav. Res. Methods 49, 457–470 (2017).
  12. S. Schindler, F. Bublatzky, Attention and emotion: An integrative review of emotional face processing as a function of attention. Cortex 130, 362–386 (2020).

Posted by on Monday, May 18, 2026 in May 2026.

Tags: , , , ,