Knowledge mediates attitude in autism spectrum disorders? Data from Malaysia

Abstract

Limited efforts have been made in middle- and low-income countries to raise public awareness of autism spectrum disorders (ASD) over the past ten years. We report data from a general population survey exploring the relationship between awareness, knowledge, and attitudes regarding autism in Malaysia. A questionnaire consisting of items related to awareness, knowledge, and attitudes regarding ASD was distributed. The scores of knowledge of ASD and attitude towards ASD were 6.87 (SD = 4.17; with a total possible score of 12) and 6.39 (SD = 2.77; with a total possible score of 9), respectively, indicating that, on average, the participants had moderate levels of knowledge and attitude. Structural equation modeling showed that improving awareness and knowledge about ASD will enhance more positive attitudes towards ASD. Efforts should be geared towards disseminating information about ASD, in order to increase acceptance of the disorder.

Introduction

Autism spectrum disorder (ASD) is a complex congenital disorder characterised by neurological deficits that manifest prior to the age of three years, with core symptoms of impairment in social interaction and communication, as well as restricted and repetitive behaviours of varying severity (APA 2013). The prevalence of autism spectrum disorder (ASD) has risen dramatically in children lately, from one in 110 (Centers for Disease Control 2009) to one in 88 (Centers for Disease Control 2012) and most recently, to one in 59 (Centers for Disease Control 2018). Global epidemiological data indicate that about 1-2% of the world’s population has ASD, which corresponds to about 50 million individuals with the disorder (Baxter et al. 2015, Hahler and Elsabbagh 2015, Sun and Allison 2010). Nevertheless, uncertainty exists in the incidence of ASD, particularly as it pertains to its causes and significance, with no definite answer available as to why the incidence of autism increased rapidly during the 1990s and has continued to do so in the 2000s (Neggers 2014). Some possible reasons for why the incidence of ASD has increased rapidly include more awareness and understanding of ASD, resulting in better diagnostic tools and more parents seeking ASD evaluations for their children (Rice et al. 2012). In developing countries, reliable epidemiologic data on ASDs are limited due to a lack of research (Elsabbagh et al. 2012). Whereas educating the public and raising awareness regarding ASD has been done across the world, particularly in the United Kingdom (UK) and the United States (US), less effort has been made in middle- and low-income countries (Scherzer et al. 2012).

In many parts of the world, children with ASD are not being diagnosed until three or four years old (Rhoades et al. 2007). This late diagnosis, which frequently occurs in middle- and low-income countries, causes delays in treatment, resulting in increased medical costs, long term special education needs, and a greater reliance on continual living support (Bello-Mojeed et al. 2017, Montiel-Nava et al. 2017). An early diagnosis of ASD facilitates early intervention, resulting in improved behavioural outcomes in the social, cognitive, and communication deficits faced by children with ASD ( Koegel et al. 2014). Applied behaviour analysis (ABA) has been found to be an effective method for treating children with ASD to improve their communication, language, social attention, and daily living skills (Eldevik et al. 2009, Rafiee and Khanjani 2020). This approach uses external reinforcement to modify or diminish behaviours based on the principles of respondent and operant conditioning.

Late diagnosis, and the pervasive consequences of this, has the potential to reinforce negative societal attitudes and stigma towards people with ASD. A conceptual approach to stigma, introduced by Thornicroft et al. (2007) in the context of mental illness, is useful here. They suggested that stigma comprises three ‘problem’ elements: problems of knowledge (leading to ignorance), attitudes (leading to prejudice), and behaviour (leading to discrimination). According to this approach, the knowledge problem is linked to the elicitation of negative attitudes and discrimination, which can result in rejecting and avoidant behaviours. Thus, increased knowledge and awareness of a specific disorder (i.e. mental illness) may positively affect people's attitudes and behaviours, reducing fear, avoidance and discrimination. In this context, people are more likely to offer support to those who are suffering from mental illness or to accept treatment themselves. This work by Thornicroft et al. (2007) helps explain the notion of stigma and demonstrates how knowledge, attitudes and behaviours can be linked. Moreover, it provides a framework to support how societies can prioritise better knowledge about conditions, including how and when to diagnose and treat them, in order to decrease discrimination and, ultimately, uphold human rights. It is also useful to consider these ideas within the specific cultural contexts of particular societies. Cultural beliefs can influence the stigma associated with disabilities and mental health conditions (Angermeyer et al. 2004). Studies conducted in Europe, North America, Canada, Australia, New Zealand, and Japan comparing public knowledge and attitudes toward depression, schizophrenia, and substance misuse disorders have demonstrated that more anger is directed towards individuals with substance misuse, due to beliefs that these individuals are ‘irresponsible’ and ‘blameworthy’ (Link et al. 1999). It is unclear whether the general public is consistent in their attitudes, emotional responses, and behaviours toward childhood disorders, such as autism, compared with mental disorders that begin in adulthood. Given that familiarity with mental illness may lead to a decrease in discriminatory behaviours (Corrigan et al. 2003, Lee and Seo 2018), understanding the knowledge and attitudes of lay people toward ASD may help explain how people with autism spectrum are treated in society.

Common stigmas are faced by parents of children with ASD in high-income countries and low- and middle- income countries (Green et al. 2005, Shyu et al. 2010, Ambikile and Outwater 2012, Divan et al. 2012, DePape and Lindsay 2015, Tilahun et al. 2016). For example, parents will worry about their child experiencing social exclusion and discrimination from society. They may feel guilty, depressed, and ashamed about their child, and may make efforts to either not let others know about their child’s condition or to assure others that their child’s condition is not due to mental illness or ‘madness’, as Tilahun et al. (2016) found in Ethiopia. Meanwhile, several studies have found that there is no relationship between knowledge and attitudes towards ASD (Gardiner and Iarocci 2014, Mac Cárthaigh and López 2020). Nevertheless, public awareness of disability may affect people’s attitudes towards ASD, although such a relationship is yet to be studied in this area.

Research related to perceptions of mental disorders suggests that people who are part of a collectivist culture, which is more common in many non-Western countries (Hofstede 1980), show stronger stigma and a more negative attitude towards people with disabilities (Mak and Kwok 2010, Durand-Zaleski et al. 2012). In a collectivist culture, people tend to be more accepting of the common value and try to preserve harmony within a group. Considering that the characteristics of ASD include communication difficulties and atypical behaviours, such as stereotypic movements, rigidity, and limited social interest, these individuals are likely to be spurned for not adhering to the norms that uphold social connection and harmony in a collectivist community (Hofstede 1986). Therefore, this puts them in a position of social exclusion and stigmatisation, especially with the lack of public awareness of ASD in collectivist societies, and low acceptance of the uniqueness in individuals with ASD. As a result, people with disabilities are more likely to be hidden from public view by their caregivers in order to avoid stigmatisation, discrimination, and social exclusion (Wong et al. 2015). High value is placed on social norms with less permissiveness towards social deviation (Papadopoulos et al. 2013). Studies have shown that people from Asia reported more negative attitudes towards help-seeking behaviours (Chen and Mak 2008, Jimenez et al. 2012) and use mental health services less frequently than Westerners (Leong et al. 2007, Wang et al. 2007).

As a result of the rising prevalence rates of ASD, autism awareness has become one of the foci of government and non-government initiatives in several countries. In Northern Ireland, high levels of awareness and knowledge about ASD exist among adults and children (Dillenburger et al. 2013, Dillenburger et al. 2017). Only two studies on public awareness, knowledge, and attitudes regarding ASD have been conducted in Malaysia (Nur Adli et al. 2017, Suhaily and Syuhada 2014). Whereas Malaysia consists of three main ethnic groups (i.e. Malay, Chinese, and Indian) living in urban and rural areas, the research conducted by Nur Adli et al. (2017) and Suhaily and Syuhada (2014) was based in a specific urban region and focused only on Malay ethnicity. In addition, there are currently no epidemiological data on the prevalence of autism in Malaysia.

The gap in the current state of knowledge about ASD implies a need to increase it through systematic awareness campaigns. The lack of information on the prevalence of autism is a concern for health service providers who work with individuals with autism (Kadar et al. 2015, Nor et al. 2019). Although raising autism awareness has been the key focus of international campaigns, such as Autism Awareness Month in April, Autism Awareness Day on April 2nd (United Nations 2019), and the Light It Up Blue campaign (Autism Speaks 2019), it is not clear to what extent the general public is informed about this disorder. An increase in the awareness of the prevalence of autism has been associated with changes in the perception of autism, leading to the development of a broader spectrum of symptomatology, diagnostic criteria, and identification of biological or environmental factors that contribute to the diagnosis (Boyd and Shaw 2010, White 2012).

To date, research has focused on the descriptive analysis of knowledge and attitudes regarding ASD in specific populations (de Vries et al. 2020, Tipton and Blacher 2014), with very few studies measuring awareness or assessing indirect effects of knowledge and attitudes. As a result, there is an absence of a model explaining the relationship between awareness, knowledge and attitude regarding ASD. In a study with Australian adults (n = 1054), 81.5% had high levels of knowledge and 81.3% had strong, positive attitudes toward ASD (Kuzminski et al. 2019). In another study conducted in Malaysia, 76.7% of the respondents were familiar with the word ‘autism’, whereas 21.5% had never heard of it (Suhaily and Syuhada 2014). However, the same study also reported that awareness of autism was not necessarily accompanied by knowledge about the appearance of specific characteristics of autism. One commonality shared by these studies was that they did not consider the indirect effect of attitude on ASD via knowledge. In this investigation, we used a mediation analysis to explore direct and indirect relationships between the general public’s awareness, knowledge, and attitudes regarding ASD. We hypothesized that a positive change in awareness of ASD leads to a positive change in attitudes toward ASD via an increase in knowledge of ASD. We addressed the following question: ‘Could an increase in the awareness of autism expand knowledge and change attitudes regarding ASD?’ The results of this study will assist us in targeting future educational intervention programs in broader communities aimed at increasing the acceptance of individuals with autism.

Method

Instrument

Based on previous studies (Dillenburger et al. 2013, Helmy 2017), a survey was created to identify the respondents’ awareness, knowledge, and attitudes regarding ASD. The questionnaire (Appendix A, Supplementary material) consisted of four sections: demographic information (Section A), awareness of ASD (Section B), knowledge about ASD (Section C), and attitudes toward ASD (Section D). The content of the survey items was reviewed by 17 certified speech-language pathologists (SLPs), 11 from the US and six from Malaysia. The questionnaire was revised based on their feedback. Work experience for the SLPs ranged from one to 23 years (M = 8.53, SD = 6.95). The American Speech-Language-Hearing Association (ASHA), the Child Neurology Society, and the American Academy of Neurology expect SLPs, among other professionals of the health care team, to play a central role in the assessment, diagnosis, and treatment of ASD (Schwartz and Drager 2008). Considering that deficits in communication are a diagnostic criterion for autism, approximately 88% of school-based ASHA-certified SLPs provide services to children with ASD (Plumb and Plexico 2013). The feedback provided by these 17 experts assisted in verifying whether the items in the survey reflected our research question, rectifying grammatical mistakes, rephrasing sentences, and amending some of the content to become more specific, in order to avoid confusion about the questions and to make the questionnaire more understandable to the public. The revised version of the questionnaire was administered to 12 participants of various educational levels to ensure that the content of the questionnaire was easy to comprehend. Feedback from the pilot study guided the research team in selecting and producing the final version of the questionnaire for data collection. The final version of the questionnaire consisted of 42 items and the estimated time for completion was approximately 20 min.

The awareness section consisted of four items relevant to having heard the term ‘Autism Spectrum Disorder,’ the source of knowing the term, and the perception of the prevalence of ASD in Malaysia. Participants were given one point if they had heard of ASD and one point if they answered correctly regarding its prevalence. No points were deducted for incorrect responses. However, for the SEM analysis, only three of these items were used, which included Q1: Have you heard of Autism Spectrum Disorders (or Asperger's syndrome)?, Q3: How many people in Malaysia do you think have ASD?, and Q4: Boys are more likely to be diagnosed with ASD than girls. The second item (Q2: From where did you hear the term Autism Spectrum Disorder?) had six response options that were nominal in nature and, therefore, could not be included in the analysis.

The knowledge section included 20 items pertaining to the basic understanding of the definition, characteristics, cause(s), and warning signs and symptoms of ASD. Each correct item received a score of one point and one point was deducted for a wrong response. A neutral response did not receive any points. Higher scores implied that the respondents had more knowledge about ASD. In the attitude section (10 items), a five-point Likert scale was used to obtain the public’s perception and acceptance of ASD in terms of individuals and community. Each positive response was scored as one point, and one point was deducted for a negative response. A neutral response was scored as zero points. Higher scores implied more positive attitudes toward ASD.

Reliability and validity

Cronbach’s alpha was 0.5960 for knowledge of ASD and 0.68 for attitude towards ASD (standardized α = 0.64 and 0.69, respectively). Items having a standardized factor loading of <0.30 (8 knowledge and 1 attitude items) were excluded from the analysis. After removing those unreliable items, Cronbach’s alpha improved to 0.77 for knowledge of ASD and 0.70 for attitude towards ASD (standardized α = 0.78 and 0.70, respectively). These values indicated acceptable internal consistency and reliability of the scales. Content validity was initially performed by the research team and experienced SLPs, and this was followed by face validity conducted among the SLPs on the clarity and understanding of the items.

Data collection procedures

Participants were recruited using purposive, convenience, and snowball sampling methods (Biernacki and Waldorf 1981). These included social media platforms like Facebook and Whatsapp, newsletters, and personal contact via public venues, such as book clubs. The questionnaire was distributed via Google form and paper-pencil format among the general public to people ranging in age from 20 to 61 years old across all four regions in Peninsular Malaysia and two regions in East Malaysia. Approval to conduct the survey was obtained from the university’s ethics board (NN-2018-166) and all participants signed a consent form electronically prior to completing the questionnaire. The survey was administered between February 2019 and July 2019 with a completion of 400 responses (353 via Google form, 47 via paper-pencil format). Of the 400 responses obtained, ten were excluded due to incomplete surveys or multiple submissions.

Data analysis

Demographic characteristics of the participants and their responses on the survey were summarized using descriptive statistics (i.e. frequencies and proportions for categorical variables and means and standard deviations for continuous variables). Item analysis and confirmatory factor analysis (CFA) were utilized to explore and refine the scale-level (i.e. Cronbach’s alpha) and item-level (i.e. factor loading) reliability of the instrument.

Structural equation modeling (SEM; i.e. path analysis) was conducted to examine the hypothesized predictive and mediation relationships among the study variables while controlling for demographic characteristics (i.e. covariates: age, gender, and education level). The SEM included consideration of these demographic characteristics to better understand the details of existing knowledge and their link to attitudes. The greater prevalence of ASD in boys than girls is a key example, where, for example, girls with ASD may be at risk of late diagnosis because of an assumption that this is a problem for boys (Giarelli et al. 2010, Horovitz et al. 2012). Specifically, mediation was assessed by decomposing the total effect on an outcome into direct and indirect effects (Brown 1997). In Figure 1, for example, the direct effects include the following regression paths: awareness of ASD (3 variables/domains) → knowledge of ASD [denoted by a₁-a₃], knowledge of ASD → attitude toward ASD [denoted by b], and awareness of ASD → attitude toward ASD [denoted by ${c^{\prime }}_1$-${c^{\prime }}_3$]. The indirect effects [denoted by a₁b-a₃b, product of a₁-a₃ and b] represent the impact of awareness of ASD (predictor) on attitude toward ASD (outcome), mediated by knowledge of ASD (mediator). If an ab is significant, a significant $c^{\prime }$ suggests partial mediation while a non-significant $c^{\prime }$ reveals full mediation. Given that the distributional property of a product term is never known, we derived an empirical distribution of each ab from 2,000 bootstrap samples and then constructed a 99% bias-corrected confidence interval (BC CI) to determine significance of the indirect effects (MacKinnon et al. 2004). As an alternative model, we also tested the mediation in the other direction: attitude toward ASD (mediator) mediates the impact of awareness of ASD (predictor) on knowledge of ASD (outcome). All analyses were performed using SAS 9.4 (SAS Institute 2002–2012) and Mplus 8.0 (Muthén and Muthén 1998–2007).

Figure 1.

Hypothesized SEM model with standardized estimates of the direct effects.

*p < .05, **p < .01, ***p < .001.

Note. The covariances among the three awareness variables and the control variables (age, gender, and education level) are omitted for simplicity in this figure.

Participants

The participants were recruited from all six regions of Malaysia, with most participants residing in the central region (see Table 1). The participants included adults in the following age ranges: 20–30 years old (67.3%), 31–40 years old (18.7%), and 41 years or older (14.1%). About half of the participants were Chinese (49.5%) while the other half included Malays (42.3%) and Indians (6.2%). In addition, 78.5% of the participants were female, 91.9% had levels of education higher than secondary school, 46.5% were employed, and 51% knew someone with an ASD diagnosis.

Table 1.

Demographic characteristics of the study sample (N = 390).

Variable	n a	%
Age
20 years old and below	8	2.1%
21–30 years old	254	65.1%
31–40 years old	73	18.7%
41–60 years old	53	13.6%
61 years old and above	2	0.5%
Gender
Male	84	21.5%
Female	306	78.5%
Race b
Malays	165	42.3%
Chinese	193	49.5%
Indian	24	6.2%
Other	8	2.1%
State
Northern region	58	14.9%
East Coast region	34	8.7%
Central region	260	66.7%
Southern region	48	12.3%
East Malaysia	20	5.1%
Highest education
No formal education	0	0.0%
Primary school	1	0.3%
Secondary school	14	3.6%
Pre-university (diploma, vocational training, etc.)	124	31.8%
Bachelor/Degree	225	57.7%
M.D./Ph.D.	26	6.7%
Occupation
Professional	125	32.1%
Sales	56	14.4%
Student	181	46.4%
Other/No response	28	7.2%
Know someone with ASD
Yes	200	51.3%
No	190	48.7%

Note.

^a There are 32.4 million Malaysians in 2018 (Department of Statistics Malaysia, 2021).

^b Ethnics’ distribution corresponds to the Malaysia Demographics Profile 2018, wherein Malaysia consists of 69.6% Bumiputera (Malays and indigenous peoples), 22.6% Chinese, 6.9% Indian, and 1.0% others.

Results

In terms of awareness, 87.7% (n = 342) of the participants had heard of ASD either through social media (37.1%), family or friends (26.9%), school (25.7%), radio or television (11.4%), newspaper (4.7%), or other (4.1%). The majority (84.1%) correctly estimated the ASD population in Malaysia and 54.6% correctly indicated the higher likelihood of diagnosis in boys than girls. The scores of knowledge of ASD and attitude towards ASD were 6.87 (SD = 4.17; with a total possible score of 12) and 6.39 (SD = 2.77; with a total possible score of 9), respectively, indicating that, on average, the participants had moderate levels of knowledge and attitude.

The correlations between awareness of ASD, knowledge about ASD, and attitude towards ASD are shown in Table 2. All correlations were significant except for the ones with awareness of the incidence of ASD. Table 3 presents the unstandardized estimates of the direct effects and indirect effects among awareness of ASD, knowledge about ASD, and attitude toward ASD, and Figure 1 shows the standardized estimates of the direct effects. The direct effects of awareness of ASD on knowledge about ASD (a₁ and a₃ paths) were significant (p < .001), except for the path from the variable representing awareness of ASD incidence in Malaysia (correct/incorrect = 1/0). The direct effect of knowledge about ASD on attitude toward ASD (b path) was also significant (p < .001). These results indicate that greater awareness of ASD was significantly associated with higher knowledge about ASD, which was also significantly associated with a more positive attitude toward ASD. For the two variables representing general awareness (yes/no = 1/0) and awareness of gender disparity in ASD diagnosis (correct/incorrect = 1/0), their indirect effects on attitude toward ASD via knowledge about ASD were significant at .01 alpha level (i.e. their 99% BC CIs did not include 0; see Table 3). Given that their direct effects on attitude toward ASD were not significant, the significant indirect effects suggested full mediation. That is, the effects of awareness of ASD were fully mediated by knowledge about ASD highlighting the importance of one’s knowledge in improving her/his attitude toward ASD. In contrast, mediation in the other direction was not supported—i.e. none of the indirect effects of awareness of ASD on knowledge about ASD via attitude toward ASD were significant.

Table 2.

Correlations among study variables.

Variable	1	2	3	4	5
1. Awareness of ASD: General	1.00
2. Awareness of ASD: Incidence	.10	1.00
3. Awareness of ASD: Gender-disparity	.30***	.06	1.00
4. Knowledge of ASD	.30***	.02	.43***	1.00
5. Attitude toward ASD	.17***	.05	.15**	.29***	1.00

^*p < .05.

^**p < .01.

^***p < .001.

Table 3.

Unstandardized estimates of the direct effects and indirect effects.

	Direct effect on:		Indirect effects on Attitude toward ASD via Knowledge of ASD
Variable	Knowledge of ASD (R2 = 0.24***)	Attitude toward ASD (R2 = 0.12**)
Awareness of ASD
General	2.27 (0.58)***	0.67 (0.50)	0.36 [0.10; 0.81]
Incidence	–0.01 (0.54)	0.45 (0.41)	0.00 [–0.25; 0.24]
Gender-disparity	3.00 (0.42)***	0.04 (0.29)	0.47 [0.16; 0.90]
Knowledge of ASD		0.16 (0.04)***

^*p < .05.

^**p < .01.

^***p < .001.

Note. The numbers in the parentheses indicate standard errors, and the numbers in the brackets indicate lower and upper limits of 99% bias-corrected confidence intervals.

Discussion

The aim of this investigation was to determine whether increases in awareness of ASD would lead to expanded knowledge and a change of attitude towards the disorder. In general, the study participants were knowledgeable about ASD and had positive attitudes towards ASD. However, caution is needed when interpreting the findings from this study because the sample, which included more females and young adults, was not representative of the majority Malay adult population. Considering that purposive, convenience, and snowball sampling methods were used in this study, generalization to the Malaysian population is limited. However, the composition of the participants and the areas from where the samples were recruited represent the urban population of the country in the central regions (namely the capital of Malaysia, Kuala Lumpur, and Selangor, which are geographically identified as Klang Valley, the most developed and urbanised region in Malaysia). Thus, the data obtained of the profiles of knowledge and attitude regarding ASD are reflective of the urban community of Malaysia, as well as those in developing countries. The majority of our sample was in the younger age group, which falls within the Malaysian median age (28.9) group (Department of Statistics Malaysia, 2021). The majority of the sample was also female, indicating this group has more interest in the topic of ASD and is more willing to participate in studies related to ASD. In addition, considering that females still play a major role in child rearing in Malaysia compared to their male counterparts, their awareness and knowledge will have a significant impact on how ASD is noticed and handled.

Lower awareness levels of autism were found among people from an ethnic minority background, lower educational level, and from a younger age group (i.e. 18-24 years old) (Dillenburger et al. 2013). For example, in the United States, Ratto et al. (2016) found that Latino mothers showed lower knowledge of ASD compared to White mothers, resulting in a late diagnosis of autism among children of Latino mothers. Given that ethnic differences in autism causal beliefs and treatment choices exist among mothers of children with ASD (Chaidez et al. 2018), further research is warranted in examining ethnic differences in beliefs and perceptions toward ASD. Gender was found to significantly influence attitudes towards autism, with some studies showing that females tend to demonstrate more positive attitudes towards autism (Jensen et al. 2016, Kuzminski et al. 2019) and others finding that female respondents demonstrate negative attitudes towards people with autism (Jones et al. 2021). These findings were supported by another study revealing that gender can influence attitudes in disability (Labelle et al. 2013, Pace et al. 2010). A study in France reported that women showed greater awareness and better understanding of specific characteristics of autism, schizophrenia and bipolar disorders when compared to men (Durand-Zaleski et al. 2012). These findings suggest that the focus of interventions addressing knowledge of and attitudes towards autism may be most beneficial for males. However, with an imbalanced sample, our study was not able to identify whether intervention should focus on males only. Thus, we recommend that future interventions addressing awareness, knowledge, and attitudes regarding autism continue to target both males and females. Higher educational levels were also associated with increased levels of knowledge about autism (Liu et al. 2016) and more positive attitudes towards communication disorders (Chu et al. 2019, Jones et al. 2021). However, Kuzminski et al. (2019) found that attitude towards autism was not affected by age and educational level. This warrants further study.

Our findings indicated that an increase in awareness of ASD yields a more positive attitude towards ASD by enhancing knowledge about ASD. In this study, 87% of the participants had heard of autism, which is consistent with that reported in the current literature on studies conducted in Western Northern Ireland and Saudi Arabia (Alsehemi et al. 2017, Dillenburger et al. 2013, Stewart 2008), but higher compared to the studies conducted in Malaysia (78.7% in Nur Adli et al. 2017; 72.7% in Suhaily and Syuhada 2014). The participants were knowledgeable about the number of people in Malaysia with ASD and the gender more likely to be diagnosed with the disorder. Most of the respondents in this study were computer and smartphone literate, whereby internet usage was common for them. Not surprisingly, they also reported that social media was the primary source from which they had heard about ASD. As a result, it appears that social media is a powerful tool for not only gaining accurate knowledge but translating that knowledge into positive attitudes towards ASD. Generally, social media and television are the primary sources to obtain information about autism, which, in turn, influence attitudes, depending on how the disability is portrayed (Durand-Zaleski et al. 2012, Happer and Philo 2013).

This study contributed to the growing evidence-based on awareness of ASD, mainly in Chinese, female, and young age sub-populations. As this study was exploratory in nature, it offers new information the authors hope can be used in future interventions and help guide similar future studies. While this survey was limited in scope to a single country population, it highlighted public awareness of autism and created a direction for future educational programs in the broader community. Given that ethnic differences in causal beliefs about autism and treatment choices exist among mothers of children with ASD (Chaidez et al. 2018), further research is warranted in examining the differences in beliefs and perceptions towards ASD that may exist in different cultural and ethnic groups. We suggest the designing of awareness campaigns that focus on informing society about what it could be like to have autism in order to assist in increasing knowledge about ASD among the general public (Kuzminski et al. 2019). Since several developing countries in South East Asia (e.g. Singapore, Indonesia, Thailand) follow a similar system of healthcare delivery, we believe our study findings may be generalized to other countries in the region. Considering that the research infrastructure of ASD remains underdeveloped in South East Asia due to the limited number of health care providers in the region, a systematic evaluation of the knowledge and attitudes of the public in these countries may contribute to its development.

Limitations

There are some limitations to our study, owing to the convenience sampling method utilized in this study. First, most of our participants were Chinese, although we recruited participants from various ethnic, age, and gender distributions. Malaysia is a multiethnic country, comprised of 69.6% Malays, 22.6% Chinese, and 6.8% Indians (Department of Statistics, Malaysia), which is not aligned with the current sample population (i.e. 42.3% Malays and 49.5% Chinese). Second, more females participated in our study, although there are more males than females in the Malaysia population. Third, the majority of our sample was of a young age (21-30), had completed either a pre-university or bachelor’s degree, and resided in the central region of Malaysia, which is the capital city of the nation. The fact that the data was skewed to the central urban area may not be a true representation of the Malaysian public. This sample may have had higher knowledge about ASD than the general public due to their educational level. Additionally, most of the respondents reported they had heard about autism, which could have influenced the results we obtained. However, it also indicates that ASD is rather common and many people have some idea of this disorder. The findings of the present study should be used with caution as they cannot be generalized across the population without further research. Future studies should attempt to capture a more representative sample, given that previous studies have reported that socio-demographic characteristics might influence one’s knowledge and perception regarding ASD (Kuzminski et al. 2019, Liu et al. 2016, Ratto et al. 2016, Dillenburger et al. 2013).

Due to our use of a convenient, snowball sampling method, we were unable to obtain a sample of participants that represented the current population distribution. We favored this approach over others because it was cost-effective, and it allowed us to collect the desired sample within a short period of time. Nevertheless, there was no way our team was able to ensure data integrity through online collection methods. Although we approached the local Malay community during data collection, the enrollment of Malay ethnicity participants was not very successful. This could be due to the language used for the invitation email and the questionnaire, which were only in English. The potential participants may have chosen not to enroll into the study due to the language barrier (Gayet-Ageron et al. 2011).

Some of the respondents reported the questionnaire was too long. Future studies should consider reducing the number of items on the survey, in addition to conducting a rigorous examination of the psychometric properties (i.e. reliability and criterion validity) of the scale. Furthermore, the addition of socioeconomic status could be included to determine demographic factors that may influence knowledge and attitudes regarding autism. Nonetheless, questions about socioeconomic status, such as monthly income, should be included in future studies to explore variables that may influence attitudes and knowledge towards autism.

Conclusion

A lack of knowledge among the general public about ASD suggests poor levels of support for people with ASD in the community. Interventions aimed at increasing knowledge about autism for the general public may lead to a greater impetus towards timely diagnosis of, and treatment for, autism, improved attitudes and increased acceptance, which will ultimately result in increased quality of life for people with ASD. We advocate for the development of a policy to promote both knowledge of autism and health research in Malaysia. This will not only ensure the early diagnosis and identification of ASD but will also assist in the creation of an effective autism surveillance program (Chu et al. 2020b, Ilias et al. 2018). Future studies should include a participant sample that is more representative of the Malaysian population, in order to examine the social demographic factors that may further influence awareness, knowledge, and attitudes regarding autism. With this information, we can better target future interventions in the broader community, possibly leading to an increase in quality of life among families of ASD (Chu et al. 2020a).

Declaration of interest

The authors report no conflicts of interest.