Abstract
Anecdotal experiences show that the human perception of time is subjective, and changes with one’s emotional state. Over the past 25 years, increasing empirical evidence has demonstrated that emotions distort time perception and usually result in overestimation. Yet, some inconsistencies deserve clarification. Specifically, it remains controversial how valence (positive/negative), arousal (high/low), stimulus type (scenic picture/facial expression/word/sound), and temporal paradigm (reproduction/estimation/discrimination) modulate the effect of emotion on time perception. Thus, the current study aimed to conduct a meta-analysis to quantify evidence for these moderators. After searching the Web of Science, SpiScholar, and Google Scholar, 95 effect sizes from 31 empirical studies were calculated using Hedges’g. The included studies involved 3,776 participants. The results a highlighted significant moderating effect of valence, arousal, stimulus type, and temporal paradigm. Specifically, negative valence tends to result in overestimation relative to positive valence; the increasing arousal leads to increasing temporal dilating; scenic picture, facial picture, and sound are more effective in inducing distortions than word; the overestimation can be better observed by discrimination and estimation paradigms relative to reproduction paradigms, and estimation paradigm is likely to be the most effective. These results suggest that the effect of emotion on time perception is influenced by valence, arousal, stimulus type, and temporal paradigm. These mitigating factors should be considered by scientists when studying time perception.
Similar content being viewed by others
Introduction
We frequently experience distortion of time when we encounter emotional stimuli or events in our daily lives. This phenomenon is called emotional temporal distortion (Lake et al., 2016). As early as 1890, James (1890) noted that our perception of time changed with different mental moods. Later, some studies found that emotions affected time perception (e.g., Falk & Bindra, 1954; Gulliksen, 1927; Hare, 1963; Langer et al., 1961; Rosenzweig & Koht, 1933; Thayer & Schiff, 1975). However, these studies suffered from methodological limitations, making their findings inconsistent and difficult to interpret (Lake et al., 2016). For example, some studies made inappropriate comparisons between emotional and non-emotional conditions (Hare, 1963; Langer et al., 1961; Thayer & Schiff, 1975). Other studies failed to induce the targeted emotion (Rosenzweig & Koht, 1933) or induced other confounding psychological processes (Gulliksen, 1927; Langer et al., 1961). Due to those limitations, the above studies only found that emotions affected time perception but failed to explain how. With the development of emotion research methods, Angrilli et al. (1997) started to use standardized emotional materials to explore this phenomenon to make up for the above limitations and to further understand the causal link between emotion and time perception. Since then, much evidence has demonstrated that emotions distort time perception (Gil et al., 2009; Mioni et al., 2020; Ogden et al., 2021; Yin et al., 2021b, 2022; Yuan et al., 2020).
As the number of studies exploded, many moderators were discovered. For example, based on emotion dimension theory (Lang et al., 1998; Russell, 1980), valence and arousal may modulate emotional temporal distortion (Angrilli et al., 1997; Noulhiane et al., 2007). In addition, there are many kinds of stimuli to induce emotions, such as word, picture, sound, and video. They have different effectiveness and validity in eliciting emotion and physiological reactions (Ellard et al., 2012; Siedlecka & Denson, 2019). Therefore, stimulus type may modulate the effect of emotion on time perception. Thirdly, there are many paradigms for measuring time perception, and they differ in response (Thoenes & Oberfeld, 2017). Therefore, temporal paradigm may modulate the emotional temporal distortion (Gil & Droit-Volet, 2011).
Though previous studies found that valence, arousal, stimulus type, and temporal paradigm modulated emotional temporal distortion, there was some evidence to the contrary. Meta-analysis is an effective method of exploring moderators, but no researcher has yet used meta-analysis to examine these moderators. Therefore, the current meta-analysis aimed to systematically clarify the effect of these moderators.
Valence
Emotional valence refers to the degree of the pleasantness of an emotion, ranging from pleasant to unpleasant (Bradley et al., 2001; Cacioppo & Gardner, 1999). In the simplest of empirical studies, valence is divided into three levels: positive, neutral, and negative. As such, there are many studies on valence-related temporal distortion, but the results are mixed. Some studies found that positive stimuli and negative stimuli led to temporal overestimation. Specifically, compared with neutral stimuli, both positive and negative stimuli extended the perceived duration of time passing (Droit-Volet et al., 2004, 2016; Grommet et al., 2011; Jones et al., 2017; Li & Tian, 2020; Smith et al., 2011). However, others found inconsistent results. For example, Lui et al. (2011) observed that both negative and positive stimuli shortened time perception; Tipples (2008) found that happy expressions did not lengthen time perception relative to neutral expressions; Eberhardt et al. (2020) found that angry expressions did not cause time perception to be overestimated.
The comparison between negative and positive stimuli is also mixed. Some studies found that time perception of negative stimuli was longer than that of positive stimuli (Buetti & Lleras, 2012; Mereu & Lleras, 2013; Noulhiane et al., 2007; Yamada & Kawabe, 2011). This is consistent with everyday experience: when you are happy, time flies; in sorrow, days seem years. However, other studies found the opposite (e.g., Van Volkinburg & Balsam, 2014). Therefore, it is necessary to clarify whether valence has a moderating effect on emotional temporal distortion.
Arousal
Arousal refers to the degree of physiological activation, ranging from calm to excitement (Bradley et al., 2001; Lang et al., 1998; Russell, 1980). It is an essential dimension of emotion and plays a key role in emotional temporal distortion. In empirical study, arousal is usually manipulated by different emotional stimuli (Clark, 1983; Gross & Levenson, 1995; Lang et al., 1993). How emotional arousal affects time perception has been explored in many ways. For example, participants stated that they feel aroused (Yin et al., 2021b); experimenters chose stimuli they believe to be arousing (Gil et al., 2007); experimenters measured physiological arousal directly (Mella et al., 2011). Correspondingly, the phenomenon of temporal overestimation with the increase of arousal has been verified by various experimental manipulations (Campbell & Bryant, 2007; Dirnberger et al., 2012; Droit-Volet et al., 2020; Zhou et al., 2021).
Although some studies observed that arousal modulated time perception, a few studies found inconsistent results (e.g., Noulhiane et al., 2007). Therefore, there were reasons for caution. Firstly, there were exceptions to the usual pattern: high arousal was not perceived as lasting longer than low arousal (Noulhiane et al., 2007). Secondly, studies using pictures found that the relationship was modulated by valence and not always in the same way (Angrilli et al., 1997; Smith et al., 2011; Van Volkinburg & Balsam, 2014). The interaction between valence and arousal has also been found in auditory stimuli (Noulhiane et al., 2007). To summarize, although some research suggested that arousal prolonged temporal estimations, other studies showed that an increase of emotional arousal did not result in the length of time perception (Noulhiane et al., 2007). Across multiple studies, there is still inconsistent evidence, which needs to be clarified.
Stimulus type
There are many mood induction procedures (MIPs). They can be divided into two categories: mood-induction situation and mood-eliciting material (Zheng et al., 2013). The former includes Imagination MIP, Velten MIP, Social Interaction MIP, Gift MIP, etc. (Zheng et al., 2013). The latter includes video, sound, word, and picture (Siedlecka & Denson, 2019).
In the field of time perception, although some researchers adopt mood-induction situations (Benau & Atchley, 2020; Matsuda et al., 2020; Piovesan et al., 2019), most of them used mood-induction material: word (Johnson & MacKay, 2019; Tipples, 2010), facial expression (Bar-Haim et al., 2010; Fayolle & Droit-Volet, 2014; Li & Yuen, 2015; Tipples, 2011; Zhang et al., 2014), scenic picture (Gable et al., 2016; Grondin et al., 2014; Tian et al., 2018; Tipples, 2019), sound (Droit-Volet et al., 2010; Noulhiane et al., 2007; Wackermann et al., 2014), and video (Özgör et al., 2018; Wöllner et al., 2018).
However, the results of studies using different emotional stimuli to explore emotional temporal distortion were inconsistent. For example, Noulhiane et al. (2007) and Angrilli et al. (1997) used sounds and scenic pictures to explore the effect of emotion on time perception, respectively. Despite both studies using the same time-reproduction paradigm, and using positive and negative stimuli high and low in arousal, the two studies found different outcomes. Angrilli et al. (1997) found that low-arousal positive pictures caused overestimation as compared to low-arousal negative pictures, but high-arousal negative pictures caused overestimation as compared to high arousal positive pictures. In contrast, Noulhiane et al. (2007) found that negative sounds were judged to be longer than positive sounds, regardless of whether arousal was high or low. In addition, Zhang et al. (2017) did not observe the overestimation in word. However, many studies found emotional temporal distortion in facial picture, scenic picture, and sound (Bar-Haim et al., 2010; Fayolle & Droit-Volet, 2014; Li & Yuen, 2015). To summarize, the results of studies using different stimuli are inconsistent and need further clarification.
Temporal paradigm
Paradigms used to study time perception encompass time estimation, time reproduction, and duration discrimination (Thoenes & Oberfeld, 2017). They were used to study the effect of emotion on time perception. Therefore, the effect has been generalized to different temporal paradigms: (a) Estimation paradigms, typically incorporate rating scales, in which participants rate the perceived duration from short to long using a Likert-type scale (Noulhiane et al., 2007; Ogden et al., 2019); (b) reproduction paradigms that ask participants to reproduce a given interval (Angrilli et al., 1997; Bar-Haim et al., 2010; Doi & Shinohara, 2009; Noulhiane et al., 2007); (c) discrimination paradigms that require participants to decide whether a specific duration is longer or shorter than a standard duration (Doi & Shinohara, 2009; Gil & Droit-Volet, 2012; Grommet et al., 2019).
Some empirical findings with different temporal paradigms have revealed high levels of correlation (Wearden, 2003; Wearden & Lejeune, 2008). However, many studies about emotional temporal distortions have found that different temporal paradigms caused different results (e.g., Gan et al., 2009; Gil & Droit-Volet, 2011; Huang et al., 2018a). For example, Gil and Droit-Volet (2011) tested anger-related temporal distortion using estimation, reproduction, and different kinds of discrimination paradigms (i.e., bisection and generalization). Results showed that in the estimation and one discrimination paradigm (bisection), the time of angry faces was estimated to be longer than that of neutral faces, but not in the reproduction and another discrimination paradigm (generalization). To summarize, the temporal paradigm is a possible variable that modulates emotional temporal distortion and needs to be further clarified. Despite the claim that the influence of emotion on time perception generalizes across paradigms, results seem mixed when examined across different temporal paradigms.
The current study
Since the study by Angrilli et al. (1997), increasing studies have focused on emotional temporal distortion. Although most work has found that emotions distort time perception, results of how arousal, valence, stimulus type, and temporal paradigm modulate emotional temporal distortion are inconsistent. Given that many of the studies reviewed above had relatively small sample sizes, it is clear that some results were limited by the lack of statistical power and had an increased risk of type I and random errors. However, these results can be well suited for meta-analysis, which is a powerful statistical method that can identify trends across numerous small sample studies based on effect sizes (Borenstein et al., 2009). Therefore, the current study aimed to clarify how valence, arousal, stimulus type, and temporal paradigm modulate emotional temporal distortion through meta-analysis.
Specifically, the current study would firstly adopt meta-regression to examine the moderating effects of valence and arousal, respectively; considering previous studies have found the interaction between valence and arousal (Angrilli et al., 1997), the meta-regression would also be used to test their interaction; furthermore, since previous studies usually manipulated valence and arousal into categorical variables (positive high arousal, positive low arousal, negative high arousal, and negative low arousal), which made them challenging to satisfy the linear relationship, we would thus use subgroup categorical analysis to test their interaction, too. Besides, the subgroup categorical analysis would also be used to examine the moderating effect of stimulus type and temporal paradigm. Lastly, the analysis of publication bias would be conducted.
Method
Literature search
We conducted an exhaustive literature search using sequential strategies to locate studies that provide data on the effects of emotion on time perception. First, we searched for relevant studies in Web of Science and SpiScholar. The primary keywords were “time perception,” “time estimation,” “time judgment,” “time evaluation,” “interval,” “duration,” “temporal” in conjunction with “emotion,” “affective,” “fear,”* “disgust,”* “ang,”* “sad,”* and “surprise.”* To collect literature as much as possible, we also supplemented it with Google Scholar. In addition, we performed a search of the reference lists of all included articles and relevant review articles in the field. The time window of our literature search was from January 1997 to 31 May 2021, because the first article using standardized emotion materials to explore emotional temporal distortion scientifically was published in 1997 (Angrilli et al., 1997).
Study selection
Exclusion criteria were as follows:
-
(1)
The duration: Fraisse (1984) thought that time perception has an upper limit that hardly exceeds 5 s. Therefore, studies with a duration of more than 5 s were excluded.
-
(2)
Peer-reviewed: Studies that were not published in peer-reviewed journals according to indices SCI, EI, SSCI, CSSCI,Footnote 1 and CSCDFootnote 2.
-
(3)
Control condition: Studies that did not provide a control condition, including emotional condition and neutral condition; or studies with inconsistent variables other than emotion between the experimental and control groups.
-
(4)
Emotion type: Studies in which emotion type could not be determined for the absence of valence, or arousal information of stimulus.
-
(5)
Temporal paradigm: Studies that did not use estimation, discrimination, or reproduction paradigm, or used a retrospective paradigm.
-
(6)
Stimulus type: Studies that did not induce emotion with word, picture (facial picture and scenic picture), sound, and video.
-
(7)
Language: Studies that were not written in English or Chinese.
-
(8)
Sample: Studies that did not involve healthy human participants or participants' average age in these studies were not between 18 and 60.
-
(9)
Modality: Studies that used tactile stimuli to measure time perception.
-
(10)
Article type: Review, meta-analysis, editorial, or commentary.
-
(11)
Compute effect size: Studies reporting results with insufficient information to compute effect size.
Notably, when study results were ambiguous or insufficient for inclusion in the meta-analysis (e.g., information required to calculate effect size was not reported), we contacted the corresponding authors of the studies to request further information. After these exclusion criteria were applied to the 4,116 potentially relevant articles, 31 articles remained. In total, 95 effect sizes were included in the current meta-analytic review (Fig. 1).
Data extraction
Data were extracted independently by two candidates and cross-checked until consensus was reached. The following variables were extracted from each eligible article: study identification data (i.e., author and publication year), participants’ mean age, sample size, arousal, valence, emotion type, stimulus type, temporal paradigm, and the statistics for the calculation of effect size.
Valence
We extracted the value of the valence from the articles and converted it uniformly to a nine-point Likert-scale (1 = “extremely negative,” 9 = “extremely positive”). Specifically, we divided the values provided by the scoring scale employed and then multiplied it by nine. If the value is greater than 5, it would be assigned into positive, while less than 5 would be assigned into negative (Bradley et al., 2001; Cacioppo & Gardner, 1999). If no value of valence is provided, we would assign it by the emotion species, that is, happiness would be assigned into positive, while anger, fear, disgust, etc., would be assigned into negative.
Arousal
Similar to valence, we extracted the value of arousal from the articles and converted it uniformly to a nine-point Likert-scale (1 = “low arousal,” 9 = “high arousal”). If the value is greater than 5, it would be assigned into high, while less than 5 would be assigned into low (Bradley et al., 2001; Cacioppo & Gardner, 1999).
Stimulus type
We coded stimulus types into five categories: scenic picture, facial picture, word, sound, and video. After coding, we found that only one study used video as emotion-eliciting stimuli, providing four effect sizes (Droit-Volet et al., 2011). Therefore, we excluded videos as a stimulus type in the subsequent analysis.
Temporal paradigm
We encoded temporal paradigm into three categories: estimation, reproduction, and discrimination. We included the verbal estimation task and rating scales as types of temporal estimation paradigms. Also, we regard bisection (Droit-Volet et al., 2015; Li & Yin, 2019), generalization (Huang et al., 2018b), and S1/S2 temporal discrimination paradigm (Lui et al., 2011) as discrimination paradigm (Table 1).
Meta-analysis
Effect size
For each study, the effect sizes relevant to this analysis were calculated as Hedges’ g, as it shows a lower level of bias (Borenstein et al., 2009). In the current analysis, Hedges’ g was calculated as follows. If the study provided the mean and standard deviation of the emotion condition and the neutral condition, it was calculated according to the formula g = (Mean1 - Mean2)/SDpooled. If the related statistics of this formula were missing, Hedges’ g would be derived from t value and sample sizes according to the formula \(g=t \sqrt{\left({n}_{1}+{n}_{2}\right)/\left({n}_{1}\times {n}_{2}\right)}\). If the t value was also not reported in the study, the p value reported in the article was converted to a t value. Extraction of p value was referred to previous studies (Yuan et al., 2019). If results were reported as insignificant, it was conservatively assigned a one-tailed p value of 0.50, such that Hedges' g was 0. If the reported results were significant, but exact p values were not provided, p values were assumed to be 0.05, 0.01, or 0.001 (two tails), respectively. For example, Grommet et al. (2019) did not report an exact p value but instead p < 0.001. In the Comprehensive Meta-Analysis (Version 3; CMA; Biostat, Englewood, NJ, USA) software package, we calculated the effect size according to p = 0.001 (2-tails). Similarly, Mella et al. (2011) reported p < 0.05; we calculated the effect size based on p = 0.05 (two tails). The effect size was positive if duration judgments were longer for emotional stimuli than for neutral stimuli.
We used the Comprehensive Meta-Analysis (Version 3; CMA; Biostat, Englewood, NJ, USA) software package to order, calculate, and compare effect sizes.
Model selection
Most meta-analyses were based on fixed- or random-effects models. According to Borenstein et al. (2010)'s suggestion, since most included articles in our meta-analysis were inconsistent with temporal paradigm and stimulus type, and we expected the results to generalize to a broader population, a random-effects model was selected for the current meta-analysis.
Heterogeneity
The heterogeneity of the distribution of effect sizes was assessed by Q and \({I}^{2}\) tests. In the Q test, a statistically significant Q value (p < 0.1) shows heterogeneity in the distribution of effect sizes. In the \({I}^{2}\) test, \({I}^{2}\) is the proportion of total variation in the estimates of effects that is due to heterogeneity rather than to chance, and higher \({I}^{2}\) values indicate greater heterogeneity (Higgins, 2003). Furthermore, heterogeneity can be used to assess the rationality of model selection. Consistent with most meta-analyses, we regarded \({I}^{2}\) values of 25%, 50%, and 75% as low, moderate, and high heterogeneity, respectively, and \({I}^{2}\) > 25% is a necessary condition for random-effects models (Borenstein et al., 2009).
Publication bias
Publication bias was identified and assessed by funnel plots (Sterne & Egger, 2001), Egger’s regression test (Egger et al., 1997), trim-and-fill (Duval & Tweedie, 2000), and classic fail-safe N (Begg & Mazumdar, 1994; Rosenthal, 1979). If no publication bias is present, the funnel plot should appear symmetric for the distribution of effect sizes. In Egger’s regression test, the intercepts that do not differ significantly from zero (p > 0.05) indicate the absence of publication bias. The classic fail-safe N considers the question of how many new studies averaging a null result are required to bring the overall effect size to nonsignificance. If the classic fail-safe N is greater than the level of 5k + 10, the publication bias is tolerant (Rosenthal, 1979). In Duval and Tweedie's trim-and-fill, the distribution of the effect sizes in included studies is trimmed or filled on the left or right to provide a symmetrical distribution, and insignificant differences between adjusted and observed effect sizes indicate the impact of publication bias is not serious.
Results
A total of 31 papers offering 95 effect sizes were included in the primary meta-analysis of the emotional temporal distortion. The total number of participants was 3,776 (Fig. 2).
Overall effect size
The overall effect size was statistically significant, g = 0.200, 95% CI [0.134, 0.265], Z = 5.966, p < 0.001, showing that the emotional time perception was longer than the neutral one. Heterogeneity analysis showed a moderate heterogeneity across the included studies, Q(94) = 185.601, p < 0.001, \({I}^{2}\)= 49.354%, suggesting a moderate degree of variation between included studies. According to Borenstein et al. (2010) 's suggestion, when there was heterogeneity between studies, the random-effects model was appropriate.
Moderator of emotional temporal distortion
Valence
We first included the valence value into the meta-regression. The result revealed that there was no significant moderating effect of valence, k = 66, β = -0.018, se = 0.017, Z = -1.020, p = 0.309, 95 % CI [-0.051, 0.016].
Previous studies mostly manipulated valence and arousal into categorical variables (positive and high arousal, positive and low arousal, negative and high arousal, negative and low arousal), which made them challenging to satisfy the linear relationship. Therefore, we combined them into the variable named emotion type to conduct subgroup categorical analysis.
The subgroup categorical analysis showed that the moderating effect of emotion type was statistically significant, Q(3) = 17.570, p < 0.001. Further analysis showed that valence was a significant moderator. Specifically, pair comparisons revealed the overall effect size of the negative high arousal stimuli was significantly bigger than positive high arousal stimuli, Q(1) = 10.371, p = 0.001. In addition, the overall effect size of the negative low-arousal stimuli was also bigger than positive low-arousal stimuli, Q(1) = 2.733, p = 0.098 (Table 2).
Arousal
Similarly, we explored the moderating effect of arousal through meta-regression and subgroup categorical analysis. The meta-regression analysis revealed that arousal was a significant moderator, k = 95, β = 0.083, se = 0.023, Z = 3.610, p < 0.001, 95% CI [0.038, 0.129], showing that time perception was prolonged with the increase of arousal. It accounted for 17% of the heterogeneity. Similar results were found when both valence and arousal were simultaneously included in the meta-regression. Arousal was a significant moderator, k = 66, β = 0.048, se = 0.024, Z = 2.030, p = 0.043, 95% CI [0.002, 0.095]; valence was not, k = 66, β = -0.014, se = 0.017, Z = −0.800, p = 0.425, 95% CI [-0.047, 0.020]. They accounted for 7% of the heterogeneity.
The result of subgroup categorical analysis also supported the moderating effect of arousal. Though there was no significant difference between positive low-arousal stimuli and positive high-arousal stimuli Q(1) = 0.069, p = 0.792, the overall effect size of the negative high-arousal stimuli was significantly bigger than negative low-arousal stimuli Q(1) = 3.965, p = 0.046. This suggested that there was an interaction between valence and arousal.
Stimulus type
The moderating effect of stimulus type was statistically significant, Q(3) = 13.806, p = 0.003. Pair comparisons revealed that the effect size of word was significantly more negative than scenic picture Q(1) = 8.949, p = 0.003, facial expression Q(1) = 13.058, p < 0.001, and sound Q(1) = 4.026, p = 0.045. There was no significant difference between scenic picture, facial expression, and sound (ps > 0.1). These results suggested that stimulus type modulated emotional temporal distortion.
Temporal paradigm
The moderating effect of temporal paradigm was statistically significant, Q(2) = 11.188, p = 0.004. Pair comparisons revealed that the overall effect size for estimation was significantly bigger than discrimination Q(1) = 10.058, p = 0.002, and reproduction Q(1) = 7.705, p = 0.006. There was no significant difference between the overall effect size for discrimination and reproduction Q(1) = 0.579, p = 0.447. These results suggested that temporal paradigm modulated emotional temporal distortion.
Additional analyses of arousal and valence
Although the results showed the moderating effect of stimulus type and temporal paradigm, another possibility is that it is caused by the difference in valence and arousal. Therefore, we conducted a series of analyses of variance (ANOVAs) to clarify whether the moderating effect of stimulus type and temporal paradigm is independent of arousal and valence.
We performed two ANOVAs to test the difference in arousal. When valence, paradigm, and stimulus type were included in the same ANOVA, some levels lacked corresponding values. Therefore, we conducted two ANOVAs. Firstly, we analyzed arousal value using a two-way factorial ANOVA in a 2 (Valence: positive, negative) × 3 (Paradigm: discrimination, estimation, and reproduction). The main effect of valence, F (1, 89) = 0.714, p = 0.400, η2 = 0.008, and temporal paradigm, F (2, 89) = 2.239, p = 0.113, η2 = 0.048, did not reach statistical significance. In addition, the interaction of valence × temporal paradigm was not significant, F (2, 89) = 0.346, p = 0.708, η2 = 0.008. These results suggested that the arousal values were similar in the three paradigms and two kinds of valences. Consequently, the moderating effects of valence and paradigm reported above should be independent of the arousal effect. Secondly, we conducted a one-way ANOVA with arousal value on stimulus type (facial picture, scenic picture, sound, and word). The result revealed a significant main effect of stimulus type, F (3, 91) = 6.775, p < 0.001, η2 = 0.183. Post hoc comparisons with the LSD test showed that the arousal of word was lower than facial picture (p = 0.001), scenic picture (p = 0.011), and sound (p < 0.001). The arousal of sound was higher than facial picture (p = 0.062) and scenic picture (p = 0.006). The arousal of facial picture was similar to scenic picture (p = 0.139).
Similarly, we conducted two ANOVAs to test the difference in valence. Firstly, we analyzed valence values using a two-way factorial ANOVA in a 2 (Arousal: high, low) × 3 (Paradigm: discrimination, estimation, and reproduction). The main effect of arousal, F (1, 61) = 0.041, p = 0.841, η2 = 0.001, and temporal paradigm, F (2, 61) = 0.442, p = 0.645, η2 = 0.014, did not reach statistical significance. The interaction of arousal × temporal paradigm was not significant, F (1, 61) = 1.413, p = 0.239, η2 = 0.023. These results suggested the valence value was similar in three paradigms and two levels of arousal. Consequently, the moderating effects of arousal and paradigm should be independent of the valence effect. Secondly, we conducted a one-way ANOVA with valence on stimulus type (facial picture, scenic picture, sound, and word). The main effect of stimulus type was not significant, F (3,62) = 0.590, p = 0.624, η2 = 0.028. The result suggested that the valence value was similar across all four categories of stimuli. Consequently, the moderating effect of stimulus type should be independent of the valence effect.
Publication bias
The publication bias was identified and assessed via funnel plots, Egger’s regression test, classic fail-safe N, and trim-and-fill. The funnel plot was asymmetrical, see Fig. 3. In addition, Egger’s regression test indicated a possible publication bias, t (93) = 5.364, p < 0.001. Due to publication bias, we further assessed its impact by classic fail-safe N and trim-and-fill. According to classic fail-safe N, the number of missing studies that would bring the overall effect to nonsignificance was 1,708. The classic fail-safe N (1,708) was greater than a tolerance level of 5k + 10 (485, k = 95). The trim-and-fill showed that 20 effect sizes were missing on the left of the overall effect size. When the 20 effect sizes were filled, the overall effect size reduced to g = 0.084, 95% CI [0.009, 0.158]. However, the adjusted effect size was not significantly different from the observed overall effect size, g = 0.200, 95% CI [0.134, 0.265]. The results showed that although there was a publication bias, it did not affect the conclusions.
Discussion
As a subjective feeling, time perception is flexible and affected by many factors. For the past 25 years, a growing body of empirical research has increased our knowledge of how emotion affects time perception. Although increasing empirical evidence has proved that emotions distort time perception and usually result in overestimation, it is controversial how valence (positive/negative), arousal (high/low), stimulus type (scenic picture/facial expression/word/sound), and temporal paradigm (reproduction/estimation/discrimination) modulate the effect of emotion on time perception. Therefore, the current study used meta-analysis to quantify existing evidence, aiming to clarify the effects of these moderators on emotional temporal distortion.
Valence
The current meta-analysis suggests that valence is a moderator of emotional temporal distortion. The subgroup categorical analysis showed that the effect size of negative valence was greater than that of positive valence, both under high- and low-arousal conditions. The meta-regression did not detect this trend, possibly because previous studies have generally treated valence as a categorical variable, making the valence not satisfy the linear relationship.
Though an authoritative classification is to bisect emotional stimuli symmetrically into positive and negative categories (Lang et al., 1998; Russell, 1980), the effects of positive and negative stimuli on us are rarely symmetrical (Yuan et al., 2019). The current finding of valence on time perception could be considered as negativity bias, a phenomenon in which the response to negative stimuli is more intense than to positive stimuli, and it is also found in a variety of cognitive processes such as memory, emotional response, and decision-making (Kress & Aue, 2017; Lam et al., 2020). A common explanation is that negativity bias may reflect an evolutionarily based activation of the aversive motivational system. The brain allocates more resources to negative emotional processing, which helps to detect environmental danger and mobilize defensive behavior, such as escaping from danger and maintaining vigilance. Thus, it is conducive to survival and environmental adaptation (Yuan et al., 2019).
In addition, subgroup categorical analyses found that there was an interaction between valence and arousal. Therefore, the moderating effect of valence on emotional temporal distortion should take into account arousal. This is discussed in more detail in the next section.
Arousal
The moderating effect of arousal was observed by both meta-regression and subgroup categorical analyses, that is, higher arousal leads to greater temporal overestimation.
Within the time perception literature, arousal has been considered the key mechanism for determining the length of the perceiving time. Particularly for clock-like models (i.e., internal clock model, attention gate model, and scalar timing model), arousal is conceptualized as any manipulation that changes the speed of the clock (Gibbon et al., 1984; Treisman, 1963; Zakay & Block, 1997), with an increase in arousal equivalent to an increase in clock speed. The evidence from physiological and pharmacological manipulations in both animals and humans has produced changes in arousal and observed covariation in time perception (Cheng et al., 2006; Meck, 1983; Mella et al., 2011). Therefore, increased arousal generally results in increasing temporal distortion.
However, as mentioned above, the current meta-analysis found that there may be an interaction between valence and arousal, that is, the negative valence boosts the moderating effect of arousal on emotional temporal distortion. Specifically, under positive valence, there is no significant difference between the effect sizes of high and low arousal; however, under negative valence, the effect size of high arousal is greater than that of low arousal. Additional ANOVA on arousal showed that the arousal degree between negative and positive stimuli was matched. These results suggest that the moderating effect of arousal on emotional temporal distortion is affected by valence. In other words, valence and arousal, as basic dimensions of emotion, jointly modulate emotional temporal distortion.
Recently, an adaptive perspective has emerged in the time perception. Emotional temporal distortion has been thought to allow individuals to adaptively respond to changes in the environment (e.g., Droit-Volet & Gil, 2009; Harrington et al., 2011; Lake et al., 2016; Matthews & Meck, 2014). Specifically, temporal distortion may allow individuals to have more subjective time to approach, attack, or flee. Although the bipolar structure theory of emotion posits that both positive and negative valence have essential associations with adaptive survival (Lang et al., 1998; Russell, 1980), the former is generally related to reward pursuit, and the latter is usually associated with threat avoidance (Cacioppo & Berntson, 1994; Cacioppo & Gardner, 1999). However, the increased arousal is always induced by a more intense situation. Consequently, it evolutionarily boosts human’s adaptive response more significantly in defensive than appetitive motivational systems, as it is more important to avoid a threatening event than to approach a rewarding target (Peeters & Czapinski, 1990; Taylor, 1991). Therefore, elevated arousal is linked with the prioritized processing of negative over positive stimuli (Schupp et al., 2007), and eventually leads to a larger emotional temporal distortion.
Stimulus type
The moderating effect of the stimulus type is significant, indicating emotional temporal distortion varies with stimulus type. Specifically, the subgroup categorical analysis showed that the facial expression, scenic picture, and sound led to significant temporal overestimation, while the word did not. These results suggest that word may be weaker in inducing emotional temporal distortion relative to facial expression, scenic picture, and sound.
An outstanding difference between facial expression, scenic picture, sound, and word, the most common stimulus types that people receive emotional information, is that emotional word is generally associated with a lower level of emotional arousal than other emotional material (Bayer & Schacht, 2014; Hinojosa et al., 2009; Liu et al., 2010), which has also been demonstrated by the additional ANOVA on arousal-value of the current study. Since the increase in arousal has been considered the key to emotional temporal distortion (Gibbon et al., 1984; Lake et al., 2016; Zakay & Block, 1997), it is reasonable to observe that word is not an effective stimulus to induce emotional temporal distortion.
However, the difference in arousal between stimulus types could not fully explain the moderating effect of stimulus type, because the additional ANOVA on arousal value showed that the arousal level of sound was significantly higher than other stimuli, but the emotional temporal distortion of sound is not significantly different from that of facial expression and scenic picture. One possible explanation can be attributed to evolution. Based on sensory channels, facial expression and scenic picture can be classified as visual stimuli, while sound can be classified as auditory stimuli. Although both vision and hearing are the two main channels for humans to receive emotional information (Royet et al., 2000), humans have evolved into diurnal animals and therefore rely more on vision (Paulmann & Pell, 2011). Due to the adaptability shaped by evolution, the functional mobilization of physiological response (i.e., arousal) to visual stimuli would be stronger (Delaney-Busch et al., 2016). Therefore, it is reasonable to observe that although the sound has a higher degree of arousal, the emotional temporal distortion of the sound is not significantly greater than that of facial expression and scenic picture.
Temporal paradigm
A significant moderating effect has been observed between temporal paradigms, revealing that both estimation and discrimination lead to significant emotional temporal distortion; further pair comparisons revealed that the emotional temporal distortion measured by estimation is significantly larger than by discrimination and reproduction, suggesting that estimation is likely to be the most sensitive paradigm.
The response differences between paradigms may have important contributions to the moderating effect of the temporal paradigm, although the estimation, discrimination, and reproduction all need participants to encode time information. The estimation needs participants to verbally report their estimate of the duration of a stimulus (usually in ms), while the discrimination requires participants to just indicate comparisons of duration between stimuli and standards (i.e., longer or shorter). In the estimation, participants report specific values. In discrimination, however, participants convert numerical values into "long" or "short" reactions. For example, for 400 ms and 800 ms, in the estimation, both would be reported as 400 and 800; in the discrimination, both would be converted to less than 1,000. This conversion results in the loss of information in the discrimination compared to the estimation. Correspondingly, this loss of information leads to a smaller effect of the discrimination than the estimation. In the reproduction, participants were required to reproduce the duration of the emotional stimulus through a neutral stimulus, but this makes emotion-induced arousal gradually decrease during reproduction (e.g., pressing a key until it is subjectively equal to the duration of the emotional stimulus). Since increased arousal is associated with increased temporal distortion (Droit-Volet & Meck, 2007), the reproduction is likely to weaken the emotional effect.
Nevertheless, in previous empirical studies, the emotional temporal distortion has been observed by using estimation (Noulhiane et al., 2007; Ogden et al., 2021), discrimination (Doi & Shinohara, 2009; Droit-Volet, 2016; Effron et al., 2006; Yuan et al., 2020), and even reproduction (Angrilli et al., 1997; Noulhiane et al., 2007; Yin et al., 2021a). Since meta-analysis can only identify trends across numerous small sample studies based on effect sizes, the moderating effect of the temporal paradigm found in the current meta-analysis reflects a trend to some extent, rather than a final conclusion about the efficacy of temporal paradigms.
Limitations
Several important issues warrant consideration in the interpretation of current results. Firstly, several effect sizes in the current meta-analysis were derived from p value in combination with the sample size (e.g., Huang et al., 2018b; Nicol et al., 2013). However, since some of them only provided significance (i.e., p < 0.05, 0.01, or 0.001), to avoid overestimation, their values were assumed to be 0.05, 0.01, or 0.001, respectively. This may slightly underestimate the effect size. Secondly, the included studies used stimuli from different material systems. It needs to be noted that these material systems used different Likert-scales (e.g., five-point, seven-point, or nine-point) which are distinct in validity to represent raters’ power of discrimination (Matell & Jacoby, 1972). In this regard, the approach of converting all the rating data uniformly to the nine-point Likert-scale should be considered tentative, and caution should be taken with this approach. Thirdly, because the small number of studies may increase the risk of Type I and random errors, the current meta-analysis did not include video studies due to insufficient eligible studies; similarly, the current meta-analysis did not include the studies using the time production paradigm because of the small number of studies and its particularity. Specifically, its particularity is mainly reflected in the production, which first presents the neutral stimulus and then uses the emotional stimulus to produce the time interval. This results in the reverse of the other paradigms: the produced interval is short, which means the time perception is overestimated. Thus, future studies may pay more attention to both video and time production until there are enough studies for revelation. Lastly, although meta-analysis uses statistical methods to identify trends across numerous small sample studies based on effect sizes, it could not replace the large sample study that directly provides empirical evidence. Therefore, it will still be valuable to use a large sample to verify current findings.
Conclusion
The current study used meta-analysis to clarify the moderating effects of valence (positive/negative), arousal (high/low), stimulus type (scenic picture/facial picture/word/sound), and temporal paradigm (discrimination/ estimation/ reproduction) on emotional temporal distortion. The results revealed that negative valence tends to result in overestimation relative to positive valence; the increasing arousal leads to increasing temporal dilating; scenic picture, facial expression, and sound are more effective in inducing overestimation than word; both discrimination and estimation are effective in measuring emotional temporal distortion relative to reproduction, and estimation is likely to be the best.
Transparency and openness
We reported how we searched, excluded, and coded literature. The complete set of data will be publicly stored in APA's Repository upon publication. The Comprehensive Meta-Analysis (Version 3; CMA; Biostat, Englewood, NJ, USA) software package was used to order, calculate, and compare effect sizes. The meta-analysis was not pre-registered.
Data availability
The complete set of data will be publicly stored in APA's Repository upon publication.
Code availability
Not applicable.
Notes
Chinese Social Sciences Citation Index, CSSCI.
Chinese Science Citation Database, CSCD.
References
References marked with an asterisk indicate studies included in the meta-analysis. The in-text citations to studies selected for meta-analysis are not preceded by asterisks.
Angrilli, A., Cherubini, P., Pavese, A., & Manfredini, S. (1997). The influence of affective factors on time perception. Perception & Psychophysics, 59(6), 972–982. https://doi.org/10.3758/BF03205512
Bar-Haim, Y., Kerem, A., Lamy, D., & Zakay, D. (2010). When time slows down: The influence of threat on time perception in anxiety. Cognition & Emotion, 24(2), 255–263. https://doi.org/10.1080/02699930903387603
Bayer, M., & Schacht, A. (2014). Event-related brain responses to emotional words, pictures, and faces – a cross-domain comparison. Frontiers in Psychology, 5(OCT), 1–10. https://doi.org/10.3389/fpsyg.2014.01106
Begg, C. B., & Mazumdar, M. (1994). Operating characteristics of a rank correlation test for publication bias. Biometrics, 50(4), 1088. https://doi.org/10.2307/2533446
Benau, E. M., & Atchley, R. A. (2020). Time flies faster when you’re feeling blue: Sad mood induction accelerates the perception of time in a temporal judgment task. Cognitive Processing, 21(3), 479–491. https://doi.org/10.1007/s10339-020-00966-8
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2009). Meta-analysis methods based on direction and p-values. In Introduction to meta-analysis (pp. 325–330). John Wiley & Sons, Ltd. https://doi.org/10.1002/9780470743386.ch36
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2010). A basic introduction to fixed-effect and random-effects models for meta-analysis. Research Synthesis Methods, 1(2), 97–111. https://doi.org/10.1002/jrsm.12
Bradley, M. M., Codispoti, M., Cuthbert, B. N., & Lang, P. J. (2001). Emotion and motivation I: Defensive and appetitive reactions in picture processing. Emotion, 1(3), 276–298. https://doi.org/10.1037/1528-3542.1.3.276
Buetti, S., & Lleras, A. (2012). Perceiving control over aversive and fearful events can alter how we experience those events: An investigation of time perception in spider-fearful individuals. Frontiers in Psychology, 3(SEP), 1–17. https://doi.org/10.3389/fpsyg.2012.00337
Cacioppo, J. T., & Berntson, G. G. (1994). Relationship between attitudes and evaluative space: A critical review, with emphasis on the separability of positive and negative substrates. Psychological Bulletin, 115(3), 401–423. https://doi.org/10.1037/0033-2909.115.3.401
Cacioppo, J. T., & Gardner, W. L. (1999). Emotion. Annual Review of Psychology, 50(1), 191–214. https://doi.org/10.1146/annurev.psych.50.1.191
Campbell, L. A., & Bryant, R. A. (2007). How time flies: A study of novice skydivers. Behaviour Research and Therapy, 45(6), 1389–1392. https://doi.org/10.1016/j.brat.2006.05.011
Cheng, R.-K., MacDonald, C. J., & Meck, W. H. (2006). Differential effects of cocaine and ketamine on time estimation: Implications for neurobiological models of interval timing. Pharmacology Biochemistry and Behavior, 85(1), 114–122. https://doi.org/10.1016/j.pbb.2006.07.019
Clark, D. M. (1983). On the induction of depressed mood in the laboratory: Evaluation and comparison of the velten and musical procedures. Advances in Behaviour Research and Therapy, 5(1), 27–49. https://doi.org/10.1016/0146-6402(83)90014-0
Delaney-Busch, N., Wilkie, G., & Kuperberg, G. (2016). Vivid: How valence and arousal influence word processing under different task demands. Cognitive, Affective, & Behavioral Neuroscience, 16(3), 415–432. https://doi.org/10.3758/s13415-016-0402-y
Dirnberger, G., Hesselmann, G., Roiser, J. P., Preminger, S., Jahanshahi, M., & Paz, R. (2012). Give it time: Neural evidence for distorted time perception and enhanced memory encoding in emotional situations. NeuroImage, 63(1), 591–599. https://doi.org/10.1016/j.neuroimage.2012.06.041
Doi, H., & Shinohara, K. (2009). The perceived duration of emotional face is influenced by the gaze direction. Neuroscience Letters, 457(2), 97–100. https://doi.org/10.1016/j.neulet.2009.04.004
Droit-Volet, S., & Meck, W. H. (2007). How emotions colour our perception of time. Trends in Cognitive Sciences, 11(12), 504–513. https://doi.org/10.1016/j.tics.2007.09.008.
Droit-Volet, S. (2016). Emotion and implicit timing. PLOS ONE, 11(7), e0158474. https://doi.org/10.1371/journal.pone.0158474
Droit-Volet, S., & Gil, S. (2009). The time–emotion paradox. Philosophical Transactions of the Royal Society B: Biological Sciences, 364(1525), 1943–1953. https://doi.org/10.1098/rstb.2009.0013.
Droit-Volet, S., Brunot, S., & Niedenthal, P. (2004). Perception of the duration of emotional events. Cognition & Emotion, 18(6), 849–858. https://doi.org/10.1080/02699930341000194
*Droit-Volet, S., Mermillod, M., Cocenas-Silva, R., & Gil, S. (2010). The effect of expectancy of a threatening event on time perception in human adults. Emotion, 10(6), 908–914. https://doi.org/10.1037/a0020258
Droit-Volet, S., Fayolle, S. , & Gil, S. (2011). Emotion and time perception: Effects of film-induced mood. Frontiers in Integrative Neuroscience, 5, 1–9.
*Droit-Volet, S., Lamotte, M., & Izaute, M. (2015). The conscious awareness of time distortions regulates the effect of emotion on the perception of time. Consciousness and Cognition, 38, 155–164. https://doi.org/10.1016/j.concog.2015.02.021
*Droit-Volet, S., Fayolle, S., & Gil, S. (2016). Emotion and time perception in children and adults: The effect of task difficulty. Timing & Time Perception, 4(1), 7–29. https://doi.org/10.1163/22134468-03002055
Droit-Volet, S., El-Azhari, A., Haddar, S., Drago, R., & Gil, S. (2020). Similar time distortions under the effect of emotion for durations of several minutes and a few seconds. Acta Psychologica, 210(August), 103170. https://doi.org/10.1016/j.actpsy.2020.103170
Duval, S., & Tweedie, R. (2000). A nonparametric, “trim and fill” method of accounting for publication bias in meta-analysis. Journal of the American Statistical Association, 95(449), 89–98. https://doi.org/10.1080/01621459.2000.10473905
*Eberhardt, L. V., Pittino, F., Scheins, A., Huckauf, A., Kiefer, M., & Kliegl, K. M. (2020). Duration estimation of angry and neutral faces: Behavioral and electrophysiological correlates. Timing & Time Perception, 8(3–4), 254–278. https://doi.org/10.1163/22134468-bja10020
Effron, D. A., Niedenthal, P. M., Gil, S., & Droit-Volet, S. (2006). Embodied temporal perception of emotion. Emotion, 6(1), 1–9. https://doi.org/10.1037/1528-3542.6.1.1
Egger, M., Smith, G. D., Schneider, M., & Minder, C. (1997). Bias in meta-analysis detected by a simple, graphical test. BMJ, 315(7109), 629–634. https://doi.org/10.1136/bmj.315.7109.629
Ellard, K. K., Farchione, T. J., & Barlow, D. H. (2012). Relative effectiveness of emotion induction procedures and the role of personal relevance in a clinical sample: A comparison of film, images, and music. Journal of Psychopathology and Behavioral Assessment, 34(2), 232–243. https://doi.org/10.1007/s10862-011-9271-4
Falk, J. L., & Bindra, D. (1954). Judgment of time as a function of serial position and stress. Journal of Experimental Psychology, 47(4), 279–282. https://doi.org/10.1037/h0061946
*Fayolle, S. L., & Droit-Volet, S. (2014). Time perception and dynamics of facial expressions of emotions. PLoS ONE, 9(5), e97944. https://doi.org/10.1371/journal.pone.0097944
Fraisse, P. (1984). Perception and estimation of time. Annual Review of Psychology, 35(1), 1–37.
*Gable, P. A., Neal, L. B., & Poole, B. D. (2016). Sadness speeds and disgust drags: Influence of motivational direction on time perception in negative affect. Motivation Science, 2(4), 238–255. https://doi.org/10.1037/mot0000044
Gan, T., Luo, Y., & Zhang, Z. (2009). The influence of emotion on time perception. Journal of Psychological Science, 32(4), 836–839.
Gibbon, J., Church, R. M., & Meck, W. H. (1984). Scalar timing in memory. Annals of the New York Academy of Sciences, 423(1 Timing and Ti), 52–77. https://doi.org/10.1111/j.1749-6632.1984.tb23417.x
*Gil, S., & Droit-Volet, S. (2011). “Time flies in the presence of angry faces”… depending on the temporal task used! Acta Psychologica, 136(3), 354–362. https://doi.org/10.1016/j.actpsy.2010.12.010
*Gil, S., & Droit-Volet, S. (2012). Emotional time distortions: The fundamental role of arousal. Cognition & Emotion, 26(5), 847–862. https://doi.org/10.1080/02699931.2011.625401
Gil, S., Niedenthal, P. M., & Droit-Volet, S. (2007). Anger and time perception in children. Emotion, 7(1), 219–225. https://doi.org/10.1037/1528-3542.7.1.219
*Gil, S., Rousset, S., & Droit-Volet, S. (2009). How liked and disliked foods affect time perception. Emotion, 9(4), 457–463. https://doi.org/10.1037/a0015751
*Grommet, E. K., Droit-Volet, S., Gil, S., Hemmes, N. S., Baker, A. H., & Brown, B. L. (2011). Time estimation of fear cues in human observers. Behavioural Processes, 86(1), 88–93. https://doi.org/10.1016/j.beproc.2010.10.003
*Grommet, E. K., Hemmes, N. S., & Brown, B. L. (2019). The role of clock and memory processes in the timing of fear cues by humans in the temporal bisection task. Behavioural Processes, 164(January), 217–229. https://doi.org/10.1016/j.beproc.2019.05.016
Grondin, S., Laflamme, V., & Gontier, É. (2014). Effect on perceived duration and sensitivity to time when observing disgusted faces and disgusting mutilation pictures. Attention, Perception, & Psychophysics, 76(6), 1522–1534. https://doi.org/10.3758/s13414-014-0682-7
Gross, J. J., & Levenson, R. W. (1995). Emotion elicitation using films. Cognition & Emotion, 9(1), 87–108. https://doi.org/10.1080/02699939508408966
Gulliksen, H. (1927). The influence of occupation upon the perception of time. Journal of Experimental Psychology, 10(1), 52–59. https://doi.org/10.1037/h0073995
Hare, R. D. (1963). The estimation of short temporal intervals terminated by shock. Journal of Clinical Psychology, 19(3), 378–380. https://doi.org/10.1002/1097-4679(196307)19:3%3c378::AID-JCLP2270190340%3e3.0.CO;2-F
Harrington, D. L., Castillo, G. N., Fong, C. H., & Reed, J. D. (2011). Neural underpinnings of distortions in the experience of time across senses. Frontiers in Integrative Neuroscience, 5(July), 1–14. https://doi.org/10.3389/fnint.2011.00032
Higgins, J. P. T. (2003). Measuring inconsistency in meta-analyses. BMJ, 327(7414), 557–560. https://doi.org/10.1136/bmj.327.7414.557
Hinojosa, J. A., Carretié, L., Valcárcel, M. A., Méndez-Bértolo, C., & Pozo, M. A. (2009). Electrophysiological differences in the processing of affective information in words and pictures. Cognitive, Affective, & Behavioral Neuroscience, 9(2), 173–189. https://doi.org/10.3758/CABN.9.2.173
*Huang, S., Liu, P., Li, Q., Chen, Y., & Huang, X. (2018a). The influence of facial expressions of pain on subsecond and suprasecond time perception. Journal of Psychological Science, 41(2), 278–284. https://doi.org/10.16719/j.cnki.1671-6981.20180204
*Huang, S., Qiu, J., Liu, P., Li, Q., & Huang, X. (2018b). The Effects of same- and other-race facial expressions of pain on temporal perception. Frontiers in Psychology, 9(NOV), 1–9. https://doi.org/10.3389/fpsyg.2018.02366
James, W. (1890). The principles of psychology. Holt.
*Johnson, L. W., & MacKay, D. G. (2019). Relations between emotion, memory encoding, and time perception. Cognition and Emotion, 33(2), 185–196. https://doi.org/10.1080/02699931.2018.1435506
*Jones, C. R. G., Lambrechts, A., & Gaigg, S. B. (2017). Using time perception to explore implicit sensitivity to emotional stimuli in autism spectrum disorder. Journal of Autism and Developmental Disorders, 47(7), 2054–2066. https://doi.org/10.1007/s10803-017-3120-6
Kress, L., & Aue, T. (2017). The link between optimism bias and attention bias: A neurocognitive perspective. Neuroscience & Biobehavioral Reviews, 80(October 2016), 688–702. https://doi.org/10.1016/j.neubiorev.2017.07.016
Lake, J. I., LaBar, K. S., & Meck, W. H. (2016). Emotional modulation of interval timing and time perception. Neuroscience & Biobehavioral Reviews, 64, 403–420. https://doi.org/10.1016/j.neubiorev.2016.03.003
Lam, C. L. M., Leung, C. J., Yiend, J., & Lee, T. M. C. (2020). The implication of cognitive processes in emotional bias. Neuroscience & Biobehavioral Reviews, 114(March), 156–157. https://doi.org/10.1016/j.neubiorev.2020.04.022
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1998). Emotion, motivation, and anxiety: Brain mechanisms and psychophysiology. Biological Psychiatry, 44(12), 1248–1263. https://doi.org/10.1016/S0006-3223(98)00275-3
Lang, P. J., Greenwald, M. K., Bradley, M. M., & Hamm, A. O. (1993). Looking at pictures: Affective, facial, visceral, and behavioral reactions. Psychophysiology, 30(3), 261–273. https://doi.org/10.1111/j.1469-8986.1993.tb03352.x
Langer, J., Wapner, S., & Werner, H. (1961). The Effect of Danger upon the experience of time. The American Journal of Psychology, 74(1), 94. https://doi.org/10.2307/1419830
Li, L., & Tian, Y. (2020). Aesthetic preference and time: Preferred painting dilates time perception. SAGE Open, 10(3), 215824402093990. https://doi.org/10.1177/2158244020939905
Li, W. O., & Yuen, K. S. L. (2015). The perception of time while perceiving dynamic emotional faces. Frontiers in Psychology, 6(August), 1–11. https://doi.org/10.3389/fpsyg.2015.01248
*Li, D., & Yin, H. (2019). The influence of fear faces on the perception of duration among different age groups. Journal of Psychological Science, 42(5), 1061–1068. https://doi.org/10.16719/j.cnki.1671-6981.20190506
Liu, B., Jin, Z., Wang, Z., & Hu, Y. (2010). The interaction between pictures and words: Evidence from positivity offset and negativity bias. Experimental Brain Research, 201(2), 141–153. https://doi.org/10.1007/s00221-009-2018-8
Lui, M. A., Penney, T. B., & Schirmer, A. (2011). Emotion effects on timing: Attention versus pacemaker accounts. PLoS ONE, 6(7), e21829. https://doi.org/10.1371/journal.pone.0021829
Matell, M. S., & Jacoby, J. (1972). Is there an optimal number of alternatives for Likert-scale items? Effects of testing time and scale properties. Journal of Applied Psychology, 56(6), 506–509. https://doi.org/10.1037/h0033601
Matsuda, I., Matsumoto, A., & Nittono, H. (2020). Time passes slowly when you are concealing something. Biological Psychology, 155(May), 107932. https://doi.org/10.1016/j.biopsycho.2020.107932
Matthews, W. J., & Meck, W. H. (2014). Time perception: The bad news and the good. Wiley Interdisciplinary Reviews: Cognitive Science, 5(4), 429–446. https://doi.org/10.1002/wcs.1298
Meck, W. H. (1983). Selective adjustment of the speed of internal clock and memory processes. Journal of Experimental Psychology: Animal Behavior Processes, 9(2), 171–201. https://doi.org/10.1037/0097-7403.9.2.171
*Mella, N., Conty, L., & Pouthas, V. (2011). The role of physiological arousal in time perception: Psychophysiological evidence from an emotion regulation paradigm. Brain and Cognition, 75(2), 182–187. https://doi.org/10.1016/j.bandc.2010.11.012
Mereu, S., & Lleras, A. (2013). Feelings of control restore distorted time perception of emotionally charged events. Consciousness and Cognition, 22(1), 306–314. https://doi.org/10.1016/j.concog.2012.08.004
*Mioni, G., Grondin, S., & Stablum, F. (2020). Do I dislike what you dislike? Investigating the effect of disgust on time processing. Psychological Research, 0123456789.https://doi.org/10.1007/s00426-020-01425-x
*Nicol, J. R., Tanner, J., & Clarke, K. (2013). Perceived duration of emotional events: Evidence for a positivity effect in older adults. Experimental Aging Research, 39(5), 565–578. https://doi.org/10.1080/0361073X.2013.839307
Noulhiane, M., Mella, N., Samson, S., Ragot, R., & Pouthas, V. (2007). How emotional auditory stimuli modulate time perception. Emotion, 7(4), 697–704. https://doi.org/10.1037/1528-3542.7.4.697
*Ogden, Ruth S., Turner, F., & Pawling, R. (2021). An Absence of a Relationship between overt attention and emotional distortions to time: An eye movement study. Timing & Time Perception, 9(2), 127–149. https://doi.org/10.1163/22134468-bja10021
*Ogden, Ruth Sarah, Henderson, J., McGlone, F., & Richter, M. (2019). Time distortion under threat: Sympathetic arousal predicts time distortion only in the context of negative, highly arousing stimuli. PLOS ONE, 14(5), e0216704. https://doi.org/10.1371/journal.pone.0216704
Özgör, C., Şenyer Özgör, S., Duru, A. D., & Işoğlu-Alkaç, Ü. (2018). How visual stimulus effects the time perception? The evidence from time perception of emotional videos. Cognitive Neurodynamics, 12(4), 357–363. https://doi.org/10.1007/s11571-018-9480-6
Paulmann, S., & Pell, M. D. (2011). Is there an advantage for recognizing multi-modal emotional stimuli? Motivation and Emotion, 35(2), 192–201. https://doi.org/10.1007/s11031-011-9206-0
Peeters, G., & Czapinski, J. (1990). Positive-negative asymmetry in evaluations: The distinction between affective and informational negativity effects. European Review of Social Psychology, 1(1), 33–60. https://doi.org/10.1080/14792779108401856
Piovesan, A., Mirams, L., Poole, H., Moore, D., & Ogden, R. (2019). The relationship between pain-induced autonomic arousal and perceived duration. Emotion, 19(7), 1148–1161. https://doi.org/10.1037/emo0000512
Rosenthal, R. (1979). The file drawer problem and tolerance for null results. Psychological Bulletin, 86(3), 638–641. https://doi.org/10.1037/0033-2909.86.3.638
Rosenzweig, S., & Koht, A. G. (1933). The experience of duration as affected by need-tension. Journal of Experimental Psychology, 16(6), 745–774. https://doi.org/10.1037/h0074980
Royet, J.-P., Zald, D., Versace, R., Costes, N., Lavenne, F., Koenig, O., & Gervais, R. (2000). Emotional responses to pleasant and unpleasant olfactory, visual, and auditory stimuli: A positron emission tomography study. The Journal of Neuroscience, 20(20), 7752–7759. https://doi.org/10.1523/JNEUROSCI.20-20-07752.2000
Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39(6), 1161–1178. https://doi.org/10.1037/h0077714
Schupp, H. T., Stockburger, J., Codispoti, M., Junghofer, M., Weike, A. I., & Hamm, A. O. (2007). Selective visual attention to emotion. Journal of Neuroscience, 27(5), 1082–1089. https://doi.org/10.1523/JNEUROSCI.3223-06.2007
Siedlecka, E., & Denson, T. F. (2019). Experimental methods for inducing basic emotions: A qualitative review. Emotion Review, 11(1), 87–97. https://doi.org/10.1177/1754073917749016
*Smith, S. D., McIver, T. A., Di Nella, M. S. J., & Crease, M. L. (2011). The effects of valence and arousal on the emotional modulation of time perception: Evidence for multiple stages of processing. Emotion, 11(6), 1305–1313. https://doi.org/10.1037/a0026145
Sterne, J. A., & Egger, M. (2001). Funnel plots for detecting bias in meta-analysis. Journal of Clinical Epidemiology, 54(10), 1046–1055. https://doi.org/10.1016/S0895-4356(01)00377-8
Taylor, S. E. (1991). Asymmetrical effects of positive and negative events: The mobilization-minimization hypothesis. Psychological Bulletin, 110(1), 67–85. https://doi.org/10.1037/0033-2909.110.1.67
Thayer, S., & Schiff, W. (1975). Eye-contact, facial expression, and the experience of time. The Journal of Social Psychology, 95(1), 117–124. https://doi.org/10.1080/00224545.1975.9923242
Thoenes, S., & Oberfeld, D. (2017). Meta-analysis of time perception and temporal processing in schizophrenia: Differential effects on precision and accuracy. Clinical Psychology Review, 54, 44–64. https://doi.org/10.1016/j.cpr.2017.03.007
*Tian, Y., Liu, P., & Huang, X. (2018). The role of emotion regulation in reducing emotional distortions of duration perception. Frontiers in Psychology, 9(MAR), 1–10. https://doi.org/10.3389/fpsyg.2018.00347
*Tipples, J. (2008). Negative emotionality influences the effects of emotion on time perception. Emotion, 8(1), 127–131. https://doi.org/10.1037/1528-3542.8.1.127
*Tipples, J. (2010). Time flies when we read taboo words. Psychonomic Bulletin & Review, 17(4), 563–568. https://doi.org/10.3758/PBR.17.4.563
*Tipples, J. (2011). When time stands still: Fear-specific modulation of temporal bias due to threat. Emotion, 11(1), 74–80. https://doi.org/10.1037/a0022015
*Tipples, J. (2019). Increased temporal sensitivity for threat: A Bayesian generalized linear mixed modeling approach. Attention, Perception, & Psychophysics, 81(3), 707–715. https://doi.org/10.3758/s13414-018-01637-9
Treisman, M. (1963). Temporal discrimination and the indifference interval: Implications for a model of the “internal clock.” Psychological Monographs: General and Applied, 77(13), 1–31. https://doi.org/10.1037/h0093864
Van Volkinburg, H., & Balsam, P. (2014). Effects of emotional valence and arousal on time perception. Timing & Time Perception, 2(3), 360–378. https://doi.org/10.1163/22134468-00002034
*Wackermann, J., Meissner, K., Tankersley, D., & Wittmann, M. (2014). Effects of emotional valence and arousal on acoustic duration reproduction assessed via the “dual klepsydra model.” Frontiers in Neurorobotics, 8(FEB), 1–10. https://doi.org/10.3389/fnbot.2014.00011
Wearden, J. H., & Lejeune, H. (2008). Scalar properties in human timing: Conformity and violations. Quarterly Journal of Experimental Psychology, 61(4), 569–587. https://doi.org/10.1080/17470210701282576
Wearden, J. H. (2003). Applying the scalar timing model to human time psychology: Progress and challenges. Time and Mind II: Information Processing Perspectives, 21–39.
Wöllner, C., Hammerschmidt, D., & Albrecht, H. (2018). Slow motion in films and video clips: Music influences perceived duration and emotion, autonomic physiological activation and pupillary responses. PLOS ONE, 13(6), e0199161. https://doi.org/10.1371/journal.pone.0199161
Yamada, Y., & Kawabe, T. (2011). Emotion colors time perception unconsciously. Consciousness and Cognition, 20(4), 1835–1841. https://doi.org/10.1016/j.concog.2011.06.016
Yin, H., Bai, Y., Liu, S., & Li, D. (2021a). The influence of motivation direction and intensity on time perception in positive and negative emotions. Journal of Psychological Science, 44(6), 1313–1321. https://doi.org/10.16719/j.cnki.1671-6981.20210605
Yin, H., Cui, X., Bai, Y., Cao, G., Zhang, L., Ou, Y., Li, D., & Liu, J. (2021b). The effects of angry expressions and fearful expressions on duration perception: An ERP study. Frontiers in Psychology, 12(June), 1–10. https://doi.org/10.3389/fpsyg.2021.570497
Yin, H., Zhang, L., & Li, D. (2022). The influence of emotion on time perception:the perspective of non-embodied emotion view and embodied emotion view. Journal of Psychological Science, in press.
Yuan, J., Tian, Y., Huang, X., Fan, H., & Wei, X. (2019). Emotional bias varies with stimulus type, arousal and task setting: Meta-analytic evidences. Neuroscience & Biobehavioral Reviews, 107(August), 461–472. https://doi.org/10.1016/j.neubiorev.2019.09.035
*Yuan, J., Li, L., & Tian, Y. (2020). Automatic suppression reduces anxiety-related overestimation of time perception. Frontiers in Physiology, 11(October), 1–12. https://doi.org/10.3389/fphys.2020.537778
Zakay, D., & Block, R. A. (1997). Temporal cognition. Current Directions in Psychological Science, 6(1), 12–16. https://doi.org/10.1111/1467-8721.ep11512604
*Zhang, D., Liu, Y., Wang, X., Chen, Y., & Luo, Y. (2014). The duration of disgusted and fearful faces is judged longer and shorter than that of neutral faces: the attention-related time distortions as revealed by behavioral and electrophysiological measurements. Frontiers in Behavioral Neuroscience, 8(AUG), 1–9. https://doi.org/10.3389/fnbeh.2014.00293
*Zhang, M., Zhang, L., Yu, Y., Liu, T., & Luo, W. (2017). Women overestimate temporal duration: Evidence from Chinese emotional words. Frontiers in Psychology, 8(10), S352. https://doi.org/10.3389/fpsyg.2017.00004
Zheng, P., Liu, C.-H., & Yu, G.-L. (2013). An overview of mood-induction methods. Advances in Psychological Science, 20(1), 45–55. https://doi.org/10.3724/SP.J.1042.2012.00045
Zhou, S., Li, L., Wang, F., & Tian, Y. (2021). How facial attractiveness affects time perception: Increased arousal results in temporal dilation of attractive faces. Frontiers in Psychology, 12, 784099. https://doi.org/10.3389/fpsyg.2021.784099
Acknowledgements
This work was supported by the National Natural Science Foundation of China (31671125) and the Social Science Planning Project of Sichuan Province (SC21C054). The funding organizations had no role in the development of the study design or collection, analysis, and interpretation of the data.
Funding
This work was supported by the National Natural Science Foundation of China (31671125) and the Social Science Planning Project of Sichuan Province (SC21C054). The funding organizations had no role in the development of the study design or collection, analysis, and interpretation of the data.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Ethics approval
This is a meta-analysis. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement was used to guide the review process.
Consent to participate
Not applicable.
Consent for publication
Not applicable.
Conflicts of interest/competing interests
The authors have no conflicts of interest to declare that are relevant to the content of this article.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Xiaobing Cui and Yu Tian are shared first authorship.
Rights and permissions
About this article
Cite this article
Cui, X., Tian, Y., Zhang, L. et al. The role of valence, arousal, stimulus type, and temporal paradigm in the effect of emotion on time perception: A meta-analysis. Psychon Bull Rev 30, 1–21 (2023). https://doi.org/10.3758/s13423-022-02148-3
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.3758/s13423-022-02148-3