Introduction
Individuals with autism spectrum disorders (ASD) usually have a difficulty relating socially to others. The social difficulties they face are among the major symptoms of ASD. Persons with ASD who do not show considerable language or cognitive delays e.g. individuals diagnosed with high functioning autism or Asperger syndrome, usually make significant progress when put through certain interventions. Significant progress to the extent that they improve in terms of communication. However, even with interventions, social difficulties usually remain and may sometimes end up impacting the vocational or academic success of such individuals later in life. Even individuals with autism who have higher or average intelligence usually find it difficult to overcome social difficulties and to communicate in a manner that observes all the unwritten rules of conversation and social interactions (Radley, O’Handley & Sabey, 2017). For example, an individual diagnosed with autism spectrum disorders may not always say the right thing or the most appropriate thing for a specific situation. They may also not know how to observe turn-taking in conversations, how to be brief and clear, and what information to leave out in some instances. Sometimes individuals with ASD also experience difficulties finding what topics to share with whom and how to maintain such topics like other people without the condition normally would.

Background

Social skills educators are at times surprised by how learners can sometimes perfectly show or demonstrate a skill in class or during a practice class and then fail to apply the same skill to a real-life scenario. In education circles, the term generalization is often used. It is usually used to show when a person’s behavior or response is produced in scenarios where it is actually not directly instructed/ or taught. This is a very common experience among social skill educators. It is an experience that also seems to negatively affect social skills instruction (Radley et al., 2017). Addressing this requires a thorough look at how social skills are taught and adjusting the curriculum to make generalization much more likely.

It is usually difficult for young people diagnosed with autism to develop and utilize social skills. According to Shindorf (2016), social skills include all elements of understanding, behavior, and communication that are important for effective social interactions. It is crucial to take cognizance of the complexity of social skills and how they are difficult even for individuals without ASD to master. For individuals with ASD, especially those who have high functioning ASD can master some social skills but usually find it difficult to interact in social situations. They usually also have difficulties with social comprehension. Persons with autism spectrum disorders usually have difficulties interacting in social situations resulting in fewer friends, minimal interactions with others, poor satisfaction with friends, and loneliness (Stichter et. al., 2010). Therefore, it is often quite difficult for them to maintain relationships with their teachers, family members, and peers.

A number of social skills interventions exist that can help to address the social difficulties experienced by persons with ASD. Considering the complexity of social skills, most social skills interventions have specific components that address various social challenges. A good example of a social skill intervention for people with ASD is the SCIP (Social Communication Intervention Project). SCIP has been successfully utilized to improve learning skills, social communication, pragmatic functioning, and conversational competence over the years (Adams et al., 2012). The EBSST (Emotional-Based Social Skills Training) intervention is also a good intervention. It has been shown to boost emotional competence significantly. The SCI (Social Competence Intervention) is also great social skills intervention for people with ASD. It has been proven to boost executive functioning, to improve social abilities, and to enhance problem solving in young people with autism spectrum disorders.

There are also interventions that are framework-based e.g. the SST (Social Skills Training) intervention (Shindorf, 2016). Such framework-based interventions take into account theoretical strategies and perspectives to provide even more effective treatment. Garcia Winner’s Social Thinking intervention is also a good example of a framework-based intervention (Winner & Crooke, 2009), and so is the YETI (Youth Engagement Through Intervention) program.

Study objective

The objective of this present study is to develop a group intervention for social skills training for youths with autism in a high school setting.

Literature review

Social Skills

Social skills include all elements of understanding, behavior, and communication that are important for effective social interactions. Young people with autism spectrum disorders usually face social difficulties because of poor social skills (Shindorf, 2016). However, it is difficult to improve the social skills of persons with autism because of their condition and because of how complex social skills and abilities are. The diversity of social skills also make them difficult to teach and to master for young people with ASD.

Evidenced-Based Practices in Social Skills Interventions

The American Psychological Association (APA) has for many years now been advocating for the use of evidence-based practices. Evidence-based practices are those that require professionals to seek and take into account the latest and most authoritative research and combine it with their special expertise to offer patients quality treatment that also considers the patient’s preferences, culture, characteristic, and circumstances (APA, 2006).

The theoretical basis for utilizing evidence-based practices was initially meant for research in health service (Ubbink, Guyatt & Vermeulen, 2013). Thus, to take into account a theory or evidence to utilize for an evidence-based practice, there must be the consideration of various research evidence to make sure that the theory or evidence or practice is really regarded to be evidence-based. The types of research that are considered include meta-analyses, efficacy research studies, effectiveness research concerning interventions, process-outcome studies, ethnographic studies, public health studies, single-case experimental studies, systematic case studies, qualitative research studies, and clinical observations (APA, 2006).

In short, the APA requires the consideration of rigorous research studies to establish evidence as well as the use of professional knowledge when providing treatment to patients. This is evidence-based practice according to APA. Thus, this paper also regards a similar use of evidence to find research-supported...…group (2) they involve a social skill intervention by professionals for a multi-modal group with at least two children (3) they include participants who are between 14 and 21 years old (4) they entail the examination of social skills utilizing SSRS and/ or SRS. To improve internal validity and reduce heterogeneity, only the randomized control trials that make use of a delayed treatment control group are going to be retained.

Published works will not be eligible for this study if they: (1) are conducted another language which is not English (2) include young people with other intellectual disabilities (3) are protocols, theses, abstracts, conference proceedings, or reviews. Studies not in English will be excluded to reduce possible translation errors or lost cultural meanings, while studies including children with intellectual disabilities will be excluded to ensure low sample heterogeneity.

Quality Assessment: Risk of Bias

Two independent reviewers will separately conduct the quality assessment of the selected studies utilizing the Cochrane Collaboration Risk of Bias (RoB) v2 tool (Higgins & Green, 2011). The two reviewers will assess for bias in selective reporting, incomplete outcomes, blinding of outcome assessments, blinding of personnel and participants, baseline measurements, allocation concealment, and sequence generation (Higgins & Green, 2011). Solutions to disagreements between the two independent reviewers will be arrived at via discussions.

Data Extraction

Two independent reviewers will separately help with the data extraction using a customized data extraction spreadsheet. The data that will be extracted will include the characteristics of the interventions, the characteristics of the patients, the outcome measures utilized, and the outcome scores. In case any additional information is required, the authors of the published works will be contacted. They authors will also be contacted to provide sub-scale and total scores of SSRS and SRS in cases where they have not published them. The co-variates will include the type of intervention, the duration of the intervention, the intensity, the teaching strategy utilized, and whether there is direct parental involvement.

Data Analysis

Meta-analysis

A statistical analysis utilizing STATA 14 will be done. The standardized mean difference and 95 percent CI (confidence interval) for every outcome will be utilized as the summary statistic. Post treatment outcomes for the treatment and for the delayed groups shall be compared for every study. The standardized mean difference shall be interpreted as small, moderate, and large for the following values respectively, 0.20-0.50, 0.50-0.80, and 0.80-1.30 (Cohen, 2013). They shall be interpreted as very large for the values more than 1.30 (Cohen, 2013). If heterogeneity is suspected, the random-effects model shall be employed. And heterogeneity shall be examined utilizing the Higgins heterogeneity I2 statistic. The extent of heterogeneity shall be interpreted as low, moderate, and high for the following values respectively 25-49 percent, 50-74 percent, and 75 percent or more (Higgins & Green, 2011). Only when p < 0.05 will heterogeneity be considered to be statistically significant.

Sensitivity Analyses

The trim and fill approach plus funnel plots with Egger’s test will be utilized to assess publication bias (Egger et al., 1997).…