Speeded reasoning moderates the inverse relationship between autistic traits and emotion recognition

Abstract

People with diagnosed autism or being high in autistic traits have been found to have difficulties with recognizing emotions from nonverbal expressions. In this study, we investigated whether speeded reasoning (reasoning performance under time pressure) moderates the inverse relationship between autistic traits and emotion recognition performance. We expected the negative correlation between autistic traits and emotion recognition to be less strong when speeded reasoning was high. The underlying assumption is that people high in autistic traits can compensate for their low intuition in recognizing emotions through quick analytical information processing. A paid online sample (N = 217) completed the 10-item version of the Autism Spectrum Quotient, two emotion recognition tests using videos with sound (Geneva Emotion Recognition Test) and pictures (Reading the Mind in the Eyes Test), and Baddeley’s Grammatical Reasoning Test to measure speeded reasoning. As expected, the inverse relationship between autistic traits and emotion recognition performance was less pronounced for individuals with high compared to low speeded reasoning ability. These results suggest that a high ability in making quick mental inferences may (partly) compensate for difficulties with intuitive emotion recognition related to autistic traits.

Lay abstract

Autistic people typically have difficulty recognizing other people’s emotions and to process nonverbal cues in an automatic, intuitive fashion. This usually also applies to people who—regardless of an official diagnosis of autism—achieve high values in autism questionnaires. However, some autistic people do not seem to have any problems with emotion recognition. One explanation may be that these individuals are able to compensate for their lack of intuitive or automatic processing through a quick conscious and deliberate analysis of the emotional cues in faces, voices, and body movements. On these grounds, we assumed that the higher autistic people’s ability to reason quickly (i.e. to make quick logical inferences), the fewer problems they should have with determining other people’s emotions. In our study, we asked workers on the crowdsourcing marketplace MTurk to complete a questionnaire about their autistic traits, to perform emotion recognition tests, and to complete a test of the ability to reason under time constraints. In our sample of 217 people, we found the expected pattern. Overall, those who had higher values in the autism questionnaire scored lower in the emotion recognition tests. However, when reasoning ability was taken into account, a more nuanced picture emerged: participants with high values both on the autism questionnaire and on the reasoning test recognized emotions as well as individuals with low autistic traits. Our results suggest that fast analytic information processing may help autistic people to compensate problems in recognizing others’ emotions.

Keywords

autism autistic traits emotion recognition intuition reasoning

Compared to neurotypical people, autistic individuals have difficulties in recognizing others’ emotions from nonverbal signals such as facial expressions (for a meta-analysis, see Uljarevic & Hamilton, 2013). Lower levels of emotion recognition have also been found for people with high levels of self-reported autistic traits, independent of an autism diagnosis (e.g. Poljac et al., 2012). However, some studies did not find such differences (e.g. Tang et al., 2019). One reason for this inconsistency may be that autism frequently co-occurs with alexithymia, but not all autistic people are equally affected (Cook et al., 2013).

Some autistic people might perform as well as neurotypical people on emotion recognition tasks due to compensatory information processing (Livingston & Happé, 2017). That is, they may be able to compensate for a lack of automated or intuitive emotion recognition ability by more effortful, logical processing of the visual or auditory cues in emotional expressions. Such compensation should be facilitated by high levels of general mental ability or intelligence (Livingston, Colvert, et al., 2019; Livingston & Happé, 2017).

Here, we propose that a high ability to process information quickly in order to make logical inferences under time constraints (referred to as speeded reasoning; e.g. Wilhelm & Schulze, 2002) may buffer the association between high autistic traits and lower emotion recognition performance. First, high information processing speed may facilitate extracting and encoding relevant nonverbal cues (such as changes in eyebrow positions, voice pitch, or drooping shoulders), especially when emotional expressions are dynamic and multimodal (i.e. presented as videos with facial, vocal, and bodily cues instead of static pictures). Second, high-speeded reasoning may facilitate the integration of these cues with existing representations of emotional displays and semantic knowledge when choosing the correct emotion label (Schlegel et al., 2020).

In this study, participants from a general population sample reported their autistic traits and completed two standard emotion recognition tests along with a test of speeded reasoning in an online setting. In line with previous studies, we expected autistic traits to correlate negatively with emotion recognition, both when measured using static pictures of faces and dynamic, multimodal stimuli. We further predicted an interaction between autistic traits and speeded reasoning such that the association between autistic traits and emotion recognition would be less strong when speeded reasoning is high.

Method

Sample and procedure

To calculate the required sample size, we considered that interaction effects are often small and that we might have to take several covariates into account. An a priori power analysis using G*Power 3.1 (Faul et al., 2009) indicated that a sample size of at least 170 participants would be required to detect a small effect (power analysis: linear multiple regression, random model, a priori; input parameters: two-tailed, H1 ρ² = 0.10, H0 ρ² = 0, α = 0.05, 1 − β = 0.80, number of predictors = 8). We recruited our sample on Amazon’s Mechanical Turk platform. The final sample (i.e. after participants were excluded according to the criteria reported below) consisted of 217 individuals from the United States (54% male, 46% female; range of age: 21–72 years). Eleven participants (5% of the sample) reported having an autism diagnosis and four participants (2%) were not sure; 14 participants (6%) scored over the suggested clinical cut-off on the Autism Spectrum Quotient-10 (AQ-10; that is, ⩾6; Allison et al., 2012). Ethnic composition was as follows: White (n = 166; 77%), Black/African American (n = 25; 12%), Hispanic/Latino (n = 16; 7%), Asian/Pacific Islander (n = 13; 6%). Fifty-one (24%) participants had a Master’s degree or higher, 91 (42%) had a college degree, 45 (21%) completed some college, 24 (11%) had a high school degree, 1 (0.5%) had a general education diploma, and 1 (0.5%) did not complete high school. Most participants were employed by a private company or organization (n = 142; 65%) or self-employed (n = 45; 21%).

Participants provided written informed consent and received US$2.50 for participation. On average, the study took about 35 min to complete.

Measures

Attention check (Oppenheimer et al., 2009): the question “Who was the first president of the United States of America?” was followed by the instruction not to check any of the three response options, but to click “continue” instead. Participants correctly following this instruction received one point; those who ticked a response received zero points.

Autism Spectrum Quotient (AQ-10; Allison et al., 2012): The AQ-10 consists of 10 items tapping different diagnostic features of autism (e.g. “I often notice small sounds when others do not”) that are rated on a four-point scale from definitely agree to definitely disagree. Definitely/slightly agree and definitely/slightly disagree are awarded one and zero points, respectively (or zero/one points when items are reversed), which are summed to form a total score. Higher scores reflect higher autistic traits. Booth et al. (2013) found that the AQ-10 measures autistic traits similarly well as the original 50-item AQ (but see Taylor et al., 2020 for recent critical findings on the psychometric properties of the AQ-10). The AQ-10 was randomly presented either before or after the following three tests.

Geneva Emotion Recognition Test (brief version, GERT-S; Schlegel & Scherer, 2016): this test consists of 42 brief video clips with sound, each showing the upper body and head of an actor expressing one of 14 emotions such as joy, amusement, pleasure, anxiety, despair, or irritation while saying a sentence in a pseudo-language without meaning. After each clip, participants select, out of 14 emotion words, the one that best describes the expressed emotion. The number of correct responses served as our measure of multimodal emotion recognition performance.

Baddeley’s Grammatical Reasoning Test (BGRT; Baddeley, 1968): participants are presented with items consisting of a statement that describes the order of two letters (A and B) using the verbs “precede” or “follow” in active or passive voice and positively or negatively (e.g. “A does not follow B,” “B is preceded by A”), followed by a letter pair (“BA” or “AB”). For each item, participants decide as quickly as possible whether the statement is true or false with respect to the letter pair. Participants have 3 min to solve as many items out of 64 as possible. The number of correctly solved items was our indicator of speeded reasoning. In previous research, BGRT scores were highly correlated with scores obtained from other standard measures of reasoning and processing speed (i.e. r = 0.53 or higher; Furnham et al., 2005; Vernon & Kantor, 1986).

Revised Reading the Mind in the Eyes Test (RMET; Baron-Cohen et al., 2001): participants are presented with 36 pictures of an individual’s eye region during an emotional experience, each accompanied by four words (e.g. playful, comforting, irritated, bored), and chose the word that best describes the individual’s emotional state for each picture. The sum score of correct answers represented our measure of facial emotion recognition.

Data cleaning

In order to identify participants who showed little motivation or ability to follow the instructions, we used the data from the BGRT. Participants were excluded if one of the following conditions was met: (1) less than 1/3 of the 64 items was completed (n = 28), (2) no further response was provided after the first of the 3 min in this task (additional n = 20), (3) the last response was provided more than 10 s before the end of the 3-min task while at least four items remained unanswered, meaning that the person stopped ahead of time (additional n = 16). Performance on the emotion recognition tests was not used for data cleaning since participants high in autistic traits may perform low in these tasks despite high motivation and cognitive ability.

Results

Descriptive statistics are displayed in Table 1. As expected, higher autistic traits were associated with lower emotion recognition performance in both the GERT-S and the RMET, and speeded reasoning correlated positively with both emotion recognition tests; see Table 1.

Table 1.

Descriptive statistics and bivariate correlations (N = 217).

Measure	α	Maximum possible score	M (SD)	Correlations
Measure	α	Maximum possible score	M (SD)	1	2	3	4	5	6
1. Emotion recognition (GERT-S)	0.88	42.00	20.62 (8.02)	–
2. Emotion recognition (RMET)	0.90	36.00	24.21 (7.80)	0.83***	–
3. Autistic traits (AQ-10)	0.65	10.00	2.89 (1.79)	–0.39***	–0.36***	–
4. Speeded reasoning (BGRT)	0.95	64.00	34.88 (13.29)	0.39***	0.36***	–0.16*	–
5. Age	–	–	37.96 (11.07)	0.27***	0.31***	–0.16*	–0.01	–
6. Gender^a	–	–	–	0.09	0.11	–0.02	–0.04	0.15*	–
7. Attention to study materials^b	–	–	–	0.28***	0.23***	–0.17*	0.09	0.00	–0.01

SD: standard deviation; GERT-S: Geneva Emotion Recognition Test–short form; RMET: Reading the Mind in the Eyes Test; AQ-10: Autism Spectrum Quotient–10-item form; BGRT: Baddeley’s Grammatical Reasoning Test.

Coding: 1 = male, 2 = female.

Coding: 0 = low, 1 = high.

p < 0.05. **p < 0.01. ***p < 0.001.

To investigate whether speeded reasoning moderates the relationship between autistic traits and emotion recognition, we conducted hierarchical multiple regression analyses. GERT-S and RMET scores were regressed on AQ-10 scores and speeded reasoning scores (both predictors were centered) in the first block and on the product of these two centered predictors in the second block. All predictors, and most important for our hypothesis, also the interaction between autistic traits and speeded reasoning, were significantly related to both dynamic multimodal emotion recognition (GERT-S) and static facial emotion recognition (RMET); see Table 2.

Table 2.

Hierarchical multiple regression analysis regressing emotion recognition (GERT-S or RMET) on autistic traits (AQ-10), speeded reasoning (BGRT), and their interaction (AQ-10 × BGRT) (N = 217).

Block and predictor	Full model with interaction (GERT-S)	Full model with interaction (RMET)
Block 1: main effects
Autistic traits	B = –1.52, SE B = 0.26, 95% CI B = [–2.04, –1.00], β = –0.34, p < 0.001	B = –1.38, SE B = 0.27, 95% CI B = [–1.91, –0.88], β = –0.32, p < 0.001
Speeded reasoning	B = 0.23, SE B = 0.04, 95% CI B = [0.16, 0.30], β = 0.38, p < 0.001	B = 0.18, SE B = 0.04, 95% CI B = [0.14, 0.28], β = 0.31, p < 0.001
Block 2: interaction
Autistic traits × speeded reasoning	B = 0.06, SE B = 0.02, 95% CI B = [0.02, 0.10], β = 0.18, p = 0.003	B = 0.06, SE B = 0.02, 95% CI B = [0.02, 0.10], β = 0.18, p = 0.003
Overall model R²	0.29	0.26
Adjusted R²	0.28	0.25
Overall F	29.31 (p < 0.001)	24.68 (p < 0.001)
df for overall F	3, 213	3, 213

GERT-S: Geneva Emotion Recognition Test–short form; RMET: Reading the Mind in the Eyes Test; AQ-10: Autism Spectrum Quotient–10-item form; BGRT: Baddeley’s Grammatical Reasoning Test; SE: standard error; CI: confidence interval.

Block 1: predictors were centered; Block 2: interaction term was built with centered predictors. The values for the interaction term regarding the GERT-S and the RMET differ only from the third decimal place onwards.

To interpret the significant interactions, we conducted simple slope analyses at low/high (i.e. ±1SD) speeded reasoning levels. For the GERT-S, at low (−1SD) speeded reasoning levels, the negative relationship between autistic traits and emotion recognition was substantial, B = −2.34, SE B = 0.38, 95% CI B = [−3.10, −1.59], β = −0.52, p < 0.001. In contrast, the same relationship was relatively small and statistically nonsignificant at high (+1SD) speeded reasoning levels, B = −0.69, SE B = 0.37, 95% CI B = [−1.43, 0.05], β = −0.15, p = 0.07. Similarly, for the RMET, the simple slope was substantial at low (−1SD) speeded reasoning levels, B = −2.21, SE B = 0.38, 95% CI B = [−2.96, −1.46], β = −0.51, p < 0.001, but nonsignificant at high (+1SD) speeded reasoning levels, B = −0.58, SE B = 0.37, 95% CI B = [−1.31, 0.16], β = −0.13, p = 0.12. Figures S1 and S2 illustrate these interactions; see Supplementary Material.

Given that age and scores on the attention check question were correlated with one or more of our main measures of interest, we ran additional regressions in which these variables were added as predictors and their influence was controlled for. We also included gender in these analyses. None of the findings described above changed; see Table S1 in Supplementary Material. Furthermore, the results remained stable when the 19 participants who failed the attention check were excluded from the analyses; see Table S2 in Supplementary Material. Finally, the results were robust when we repeated the regression analyses without the data of the 15 participants who had received an autism diagnosis or were unsure in this regard respectively; see Table S3 in Supplementary Material.

Discussion

Our results are in line with previous findings that people high in autistic traits have difficulties with recognizing others’ emotions. However, these difficulties can apparently be (partly) overcome for both static facial and dynamic multimodal expressions when speeded reasoning and information processing capacities are high. This result is compatible with earlier considerations on compensatory processes in autism (Frith, 2004) and contributes to the currently flourishing research on this topic (e.g. ‘Livingston, Colvert, et al., 2019, Livingston et al., 2020; Livingston & Happé, 2017; Livingston, Shah, & Happé, 2019). We assume that people with autism or high levels of autistic traits recognize emotions in a less intuitive or automated way than neurotypical individuals, and that high speeded reasoning as a central component of intelligence may facilitate the effortful and deliberative processing of nonverbal cues.

Our findings may also inform research into the recently introduced dual process theory of autism which posits that higher autistic traits go along with a high deliberative and low intuitive reasoning style (Brosnan et al., 2016). While this line of work has focused on differences in preferences for reasoning styles between neurotypical and autistic individuals, the present study focused on information processing differences within individuals high in autistic traits. Our results suggest that it can make a difference for autistic people how quickly and correctly they can analytically think. Therefore, future research within the framework of the dual process theory of autism may also take into account individual differences in analytical processing to explain behavior within the group of autistic people (or such high in autistic traits, respectively).

Future research should also investigate whether the need to infer others’ emotions consciously or logically can become obsolete through automatization over time (“deep compensation”; Livingston et al., 2020). It may be that some autistic people or such high in autistic traits recognize others’ emotions quasi-intuitively after a certain point in their development. Intellectual abilities, personality traits, and environmental factors may determine the extent to which emotion recognition becomes automated or quasi-intuitive.

There are several limitations to this study. First, due to the cross-sectional study design, we cannot draw causal conclusions about the relationship between autism, emotion recognition, and speeded reasoning. Second, our measure of speeded reasoning is verbal in nature and different types or measures of reasoning may provide alternative results (e.g. unspeeded cognitive reflection; Lewton et al., 2019). Third, we did not directly examine the postulated mechanisms related to intuitive or deliberative processing of emotional information, and future research (e.g. using physiological measures or neuroimaging) is needed to substantiate our interpretation of the present findings. Fourth, this study focused on autistic traits but not on diagnosed autism. On one hand, there are authors who point out that people with high autistic traits tend to behave similarly to diagnosed autistic people in certain domains such as visual detail perception (e.g. Poljac et al., 2012). In addition, there are autistic people who have not yet undergone diagnostic clarification, but who may report high autistic traits according to their everyday experiences (Brosnan, 2020). Therefore, findings based on the measurement of autistic traits may also be relevant for clinical autism research. On the other hand, people who seek diagnostic assessment of autism but do not ultimately receive a diagnosis may also have relatively high scores in the Autism Spectrum Quotient (Happé et al., 2016). Considering that a high score in a measure of autistic traits does not always go along with a diagnosis of autism, the present findings should not be overgeneralized, but replicated with clinically diagnosed autistic people.

Supplemental Material

Supplementary_Material – Supplemental material for Speeded reasoning moderates the inverse relationship between autistic traits and emotion recognition

Supplemental material, Supplementary_Material for Speeded reasoning moderates the inverse relationship between autistic traits and emotion recognition by Alex Bertrams and Katja Schlegel in Autism

Footnotes

Author contributions

A.B. conceived the project, designed the study, conducted the study, analyzed the data, and wrote the manuscript. K.S. conceived the project, designed the study, conducted the study, prepared the data set, and wrote the manuscript.

Community involvement statement

The first author (A.B.) was diagnosed with autism.

Declaration of conflicting interests

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

Ethical approval/patient consent

The study was approved by the Institutional Review Board of the Faculty of Human Sciences at the University of Bern (reference number: 2019-05-00004). All participants provided written informed consent.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by a grant from the Uranus Foundation, Switzerland.

ORCID iD

Alex Bertrams

Supplemental material

Supplemental material for this article is available online.

References

Allison

Auyeung

Baron-Cohen

(2012). Toward brief “red flags” for autism screening: The short autism spectrum quotient and the short quantitative checklist in 1,000 cases and 3,000 controls. Journal of the American Academy of Child & Adolescent Psychiatry, 51, 202–212.e7. https://doi.org/10.1016/j.jaac.2011.11.003

Baddeley

A. D.

(1968). A 3 min reasoning test based on grammatical transformation. Psychonomic Science, 10, 341–342. https://doi.org/10.3758/BF03331551

Baron-Cohen

Wheelwright

Hill

Raste

Plumb

(2001). The “reading the mind in the eyes” test revised version: A study with normal adults, and adults with Asperger syndrome or high-functioning autism. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 42, 241–251. https://doi.org/10.1017/S0021963001006643

Booth

Murray

A. L.

McKenzie

Kuenssberg

O’Donnell

Burnett

(2013). Brief report: An evaluation of the AQ-10 as a brief screening instrument for ASD in adults. Journal of Autism and Developmental Disorders, 43, 2997–3000. https://doi.org/10.1007/s10803-013-1844-5

Brosnan

(2020). An exploratory study of a dimensional assessment of the diagnostic criteria for autism. Journal of Autism and Developmental Disorders. Advance online publication. https://doi.org/10.1007/s10803-020-04474-8

Brosnan

Lewton

Askwin

(2016). Reasoning on the autism spectrum: A dual process theory account. Journal of Autism and Developmental Disorders, 46, 2115–2125. https://doi.org/10.1007/s10803-016-2742-4

Cook

Brewer

Shah

Bird

(2013). Alexithymia, not autism, predicts poor recognition of emotional facial expressions. Psychological Science, 24, 723–732. https://doi.org/10.1177/0956797612463582

Faul

Erdfelder

Buchner

Lang

A.-G.

(2009). Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41, 1149–1160. https://doi.org/10.3758/BRM.41.4.1149

Frith

(2004). Emanuel Miller lecture: Confusions and controversies about Asperger syndrome. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 45, 672–686. https://doi.org/10.1111/j.1469-7610.2004.00262.x

10.

Furnham

Moutafi

Chamorro-Premuzic

(2005). Personality and intelligence: Gender, the Big Five, self-estimated and psychometric intelligence. International Journal of Selection and Assessment, 13, 11–24. https://doi.org/10.1111/j.0965-075X.2005.00296.x

11.

Happé

F. G.

Mansour

Barrett

Brown

Abbott

Charlton

R. A.

(2016). Demographic and cognitive profile of individuals seeking a diagnosis of autism spectrum disorder in adulthood. Journal of Autism and Developmental Disorders, 46, 3469–3480. https://doi.org/10.1007/s10803-016-2886-2

12.

Lewton

Ashwin

Brosnan

(2019). Syllogistic reasoning reveals reduced bias in people with higher autistic-like traits from the general population. Autism, 23, 1311–1321. https://doi.org/10.1177/1362361318808779

13.

Livingston

L. A.

Colvert

, the Social Relationships Study Team Bolton

Happé

(2019). Good social skills despite poor theory of mind: Exploring compensation in autism spectrum disorder. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 60, 102–110. https://doi.org/10.1111/jcpp.12886

14.

Livingston

L. A.

Happé

(2017). Conceptualising compensation in neurodevelopmental disorders: Reflections from autism spectrum disorder. Neuroscience & Biobehavioral Reviews, 80, 729–742. https://doi.org/10.1016/j.neubiorev.2017.06.005

15.

Livingston

L. A.

Shah

Happé

(2019). Compensatory strategies below the behavioural surface in autism: A qualitative study. The Lancet Psychiatry, 6, 766–777. http://doi.org/10.1016/S2215-0366(19)30224-X

16.

Livingston

L. A.

Shah

Milner

Happé

(2020). Quantifying compensatory strategies in adults with and without diagnosed autism. Molecular Autism, 11, Article 15. https://doi.org/10.1186/s13229-019-0308-y

17.

Oppenheimer

D. M.

Meyvis

Davidenko

(2009). Instructional manipulation checks: Detecting satisficing to increase statistical power. Journal of Experimental Social Psychology, 45, 867–872. https://doi.org/10.1016/j.jesp.2009.03.009

18.

Poljac

Wagemans

(2012). Reduced accuracy and sensitivity in the perception of emotional facial expressions in individuals with high autism spectrum traits. Autism, 17, 668–680. https://doi.org/10.1177/1362361312455703

19.

Schlegel

Palese

Mast

M. S.

Rammsayer

T. H.

Hall

J. A.

Murphy

N. A.

(2020). A meta-analysis of the relationship between emotion recognition ability and intelligence. Cognition and Emotion, 32, 329–351. https://doi.org/10.1080/02699931.2019.1632801

20.

Schlegel

Scherer

K. R.

(2016). Introducing a short version of the Geneva Emotion Recognition Test (GERT-S): Psychometric properties and construct validation. Behavior Research Methods, 48, 1383–1392. https://doi.org/10.3758/s13428-015-0646-4

21.

Tang

J. S. Y.

Chen

N. T. M.

Falkmer

Bӧlte

Girdler

(2019). Atypical visual processing but comparable levels of emotion recognition in adults with autism during the processing of social scenes. Journal of Autism and Developmental Disorders, 49, 4009–4018. https://doi.org/10.1007/s10803-019-04104-y

22.

Taylor

E. C.

Livingston

L. A.

Clutterbuck

R. A.

Shah

(2020). Psychometric concerns with the 10-item Autism-Spectrum Quotient (AQ10) as a measure of trait autism in the general population. Experimental Results, 1, e3. https://doi.org/10.1017/exp.2019.3

23.

Uljarevic

Hamilton

(2013). Recognition of emotions in autism: A formal meta-analysis. Journal of Autism and Developmental Disorders, 43, 1517–1526. https://doi.org/10.1007/s10803-012-1695-5

24.

Vernon

P. A.

Kantor

(1986). Reaction time correlations with intelligence test scores obtained under either timed or untimed conditions. Intelligence, 10, 315–330. https://doi.org/10.1016/0160-2896(86)90002-4

25.

Wilhelm

Schulze

(2002). The relation of speeded and unspeeded reasoning with mental speed. Intelligence, 30, 537–554. https://doi.org/10.1016/S0160-2896(02)00086-7

Supplementary Material

Please find the following supplemental material available below.

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

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

0.00 MB

0.41 MB