Exploring social innovation (SI) within the research contexts of higher education, healthcare, and welfare services – a scoping review

Introduction: Nordic countries face societal challenges for which social innovation may represent solutions. The aim of this scoping review is to explore the concept of social innovation within the research contexts of higher education, ealthcare, and welfare services.
Method: A scoping review methodology was used, including a literature search and the identification of eligible studies published between 2007 and 2019, in addition to data extraction and synthesis. Forty-three studies were included in this review.
Results: Across the research contexts, social innovation is conceptualized as a set of novel, creative, human-centred, and value-driven processes aiming to bring about change. Qualitative research methods dominate social innovation research. In welfare services, social innovation concerns the relationship between policy and praxis, new forms of leadership and management, and the promotion of societal inclusion and cohesion. Social innovation in healthcare comprises the use of technology to digitalize service, enhance patients’ well-being, and improve service quality. In higher education, social innovation research focuses on educational reforms involving non-profit stakeholders.
Discussion: Social innovation is a multifaceted concept related to change at the organizational or societal level, often with various stakeholders working together to create improvements. The lack of a common definition and framework of social innovation makes this concept difficult to measure or quantify, reflecting the dominance of qualitative research methods in the selected research contexts. Across these research contexts, social innovation can be defined and used for various research purposes, which are often political and value-based, with the latter connected to the common good and people’s well-being. Moreover, few social innovation studies have been performed in Nordic countries.


Introduction
This scoping review explores the concept of social innovation (SI) as it appears in three specific contexts: higher education, healthcare, and welfare services. In this study, welfare services are understood as social services provided to citizens to meet societal needs. Exploring the concept of SI is useful because SI has become a significant concept in policy and management rhetoric, often put forward as a crucial element in strategies aimed at making better use of scarce resources. In addition, the expectation that SI can drive societal change and maintain welfare services under economic constraints has become evident in public and academic discourses.
In fact, SIs seem to be gaining importance over technical innovations in dealing with societal challenges (Howaldt & Kopp, 2012). While innovation carries technological and economical connotations, SI emphasises valuecreation beyond fiscal growth, including better health, wellbeing, and safety of the SI target group. Some analytical definitions use the term 'social' to distinguish SIs from technical innovations (Borzaga & Bodini, 2014).
Moreover, this term often refers to the ways in which this value can be achieved through new ways of working (together) as well as empowering actors (i.e., professionals and citizens) to take part in the production of welfare (Avelino et al., 2019). A popular definition of this concept came from Murray et al. (2010), who defined SI as 'new ideas (products, services and models) that simultaneously meet social needs and create new social relationships or collaborations. In other words, they are innovations that are both good for society and enhance society's capacity to act' (p. 3).
As this definition implies, the concept of SI is often understood as processes that spring from real-life problem-solving, but it simultaneously addresses a basic level of capacity-building in groups, organizations, or society (Murray et.al, 2010). In addition, SI is deemed relevant when addressing challenges concerning societal changes (macro level), social needs, and people's wellbeing (micro level) (Mulgan, 2015;Nicholls et al., 2015). Such innovations are not only beneficial for society because they solve problems in practice, but they also enhance society's capacity to act. This means that they generate newer, better infrastructures and competencies for innovation beyond the demands of the actual situation (Caulier-Grice et al., 2012). SI is thus associated with social change for 'the common good', indicating that SI is meant to create social value and must be experienced as useful in a given field of practice (Mulgan, 2015).

75
NJSR -Nordic Journal of Social Research Vol. 12, 2021 The diagnoses of societal challenges vary, yet there seems to be agreement regarding the role of SI: its potential in empowering and mobilising civic creativity in addition to problem-solving. This agreement is especially manifested in policy discourses across the EU (Avelino et al., 2019). Policy documents emphasise the social dimension of societal challenges and the immediate social challenges that shape the process of social interaction to improve individuals' well-being (Commission of the European Communities, 2008). Hence, SI is regarded as a response that 'can offer a way forward in coping with the societal challenges and the crisis that EU is facing (Bureau of European Policy Advisers [BEPA], 2011, p. 7). The somewhat optimistic assumptions that SI can 'save' welfare states, however, seem to underestimate the systemic complexities of these challenges (Rittel & Webber, 1973). In fact, there is a growing concern in both theory and practice regarding the systemic level of SI and ways in which systems are rewired, which is considered crucial for lasting change and for understanding the concept (Mulgan, 2012).
SI is a complex, contested concept, and knowledge about its nature, management, and facilitation in specific contexts is limited. Thus, further research is necessary to develop knowledge about how SI appears within different scholarly contexts, how (and if) it is conceptualized, the methods researchers apply, and the foci of relevant research.
Nordic countries are facing some of the same societal challenges affecting other European societies, including increased income inequality and a higher risk of poverty (Fritzell et al., 2012). The Nordic welfare states are experiencing cutbacks on? public spending/budgets, and innovative public service models are needed (Norwegian Institute of Public Health, 2019).
Hence, SI combined in healthcare, higher education, and welfare services may become a strategy for addressing the societal challenges facing these Nordic welfare systems.
Globally, in welfare fields such as higher education, healthcare, and welfare services, work practices are continually evolving due to changing demographic, political, and economic circumstances. The need for alternative conceptions of traditional processes and product models of innovation is relevant because the contexts of welfare innovations mainly take place as part of everyday interactions between people. The concept of innovation is often 76 NJSR -Nordic Journal of Social Research Vol. 12, 2021 associated with technological development and private enterprises associated with products and markets, and it does not easily translate to health and welfare services in the public sector (Hartley, 2005). A literature review on innovation within healthcare, education, and welfare services in Europe noted the following distinctive features of innovation in these contexts : i) the relational dimension, as the relationship between the user and the service provider is direct; ii) the processual dimension, as the process of innovating and the diffusion of the innovation are never fully accomplished; and iii) the interactional dimension, as the generating and dissemination of innovation unfold within a complex system and among different systems, contexts, and implementing environments (Crepaldi et al., 2012). Therefore, innovation in various welfare fields is often initiated in specific contexts, and its value is evaluated by the actors in each of these contexts (Tanggaard & Author, 2016).
With the growing interest in SI, there has also been increased attention toward not only the social ends (problem-solving) but also the social means (new ways of collaborating and the mobilization of new actors). These new social means often involve changes in power relations, enforced ethical considerations, and the need for situated decision-making by various professionals working with individuals in need of care, support, and/or learning (Author, 2016). Since the outcomes of innovations are not solely new and measurable, products, services, or processes accordingly become a mix of human-to-human interactions, methods of collaborating during problemsolving, and the quality of relations between the innovation stakeholders (Author, 2014). Although SI requires distinct professional competencies and a managerial strategy, it also involves sensemaking and commitment, the support of experimentation, critical dialogue, and risk-taking in expansive learning environments (Fuller & Unwin, 2004;Fuller et al., 2018).
Despite a growing interest in the social aspects of innovation, there is still little knowledge about the practice of SI (Phillips et al., 2019) nor consistency in the use of the concept across EU policy and national policies (Krlev et al., 2020).
In the same vein, there is no reliability in discussions of this concept across scholarly fields. Additionally, Mulgan et al. (2013), stated that social innovation is 'theoretically underdeveloped with few, if any, accepted concepts or causal relationships' (p. 426). Accordingly, in this paper, we will ask the following: How is SI conceptualized and studied in the contexts of higher education, healthcare, and welfare services, and what are the foci of the related 77 NJSR -Nordic Journal of Social Research Vol. 12, 2021 research? Thus, the aim of this scoping review is to contribute to a description of research on SI across these three scholarly fields.

Scoping Review Design
The study applied a scoping review method following the framework of Arksey and O'Malley (2005). This framework suited our aim of making a broad mapping of the concept of SI in an interdisciplinary context and allowed the inclusion of studies that applied various methodological approaches. It comprises five steps: 1) identifying the research question, 2) performing a literature search to identify relevant studies, 3) selecting relevant studies, 4) charting the data, and 5) synthesizing the data (Arksey & O'Malley, 2005).

Identification of Research Questions
The study's scope and research question were discussed and developed by The scope of this study included exploring and describing the concept of SI, which resulted in the following research questions: 1) Which methodological approaches were applied in the various studies analysed?
2) How has SI been conceptualized in research within the contexts of higher education, healthcare, and welfare?
3) How does SI unfold as a process within the welfare research contexts of higher education, healthcare, and welfare services?

Literature Search
The scoping study method requires a comprehensive approach. Therefore, all kinds of literature and studies within SI that intersected the contexts of interest were initially determined to be relevant (Arksey & O'Malley, 2005). To obtain an overview of the three research fields, a literature search was performed using Google Scholar, resulting in a high number of retrieved titles (n = 103,300). This led to a more restricted, manageable approach to our search strategy, which involved a second, identical search performed during March 78 NJSR -Nordic Journal of Social Research Vol. 12, 2021 2017 in the databases of Scopus, the Web of Science, Cumulative Index of Nursing and Allied Health Literature (CINAHL), and Political Science by a university librarian together with two co-authors (ALP, MBØ). Broad search terms such as 'education', 'health', and 'welfare' were applied in combinations with 'SI' in all of the searches (Table 1). Subsequently, we decided to narrow 'education' to 'higher education' and to include only studies that explored and/or reported on higher education in the health and welfare sectors. The first author (AMLH) conducted a second, updated search in the same databases in August 2019, which yielded 173 new records. All of these searches were limited to the documents' titles or abstracts. The publication period was set to January 2007-August 2019.

Table 1 Search history
The searches identified 828 records. After screening the titles and removing duplicates, 290 abstracts were included in the study selection stage (Table 1).

Study Selection
This stage began with two of the authors sorting the identified records from the initial search (n = 655) into different files in Google Docs (ALP, MBØ), with each file representing one of the three research contexts. In each file, titles and abstracts were sorted under the predefined headings of 'Education',

79
NJSR -Nordic Journal of Social Research Vol. 12, 2021 'Health', and 'Welfare' to aid further review. The records identified in the updated literature search (n = 173) were subsequently added to the bibliographical files from the first search. During study selection, the authors worked in pairs according to their research context of interest and professional expertise: higher education (CW, MKAA), healthcare (AMLH, MS), and welfare (EW, AØ). First, each pair independently read the identified records, which was followed by a joint review of abstracts eligible for full-text examination.
Disagreement regarding eligibility was resolved through dialogue. Across the three research contexts, the texts of 290 abstracts were selected to be read in full. Next, full-text versions of eligible abstracts were obtained and independently read by each pair to determine which articles met the inclusion criteria.

Inclusion and Exclusion Criteria
The criteria for the inclusion and exclusion of articles were made post hoc based on our increasing familiarity with the literature following a comprehensive reading of abstracts. In reading the abstracts, we developed an overall idea of the focus of each paper. During this process, the documents were sorted into two categories: 'Research about SI' and 'Documents that define SI'. We then developed the following refined criteria for inclusion: 1. SI on an individual, group, or organizational level within the context of higher education, health, or welfare; 2. Education innovations within social work, healthcare, or higher education (bachelor's degree); 3. Welfare innovations, including social services, the public sector, and civil society; 4. health innovations within the health care system; 5. peer-reviewed articles; 6. empirical studies; and 7. case studies.
The final and more detailed criteria for inclusion further guided the abstract evaluation process, thus studies that reported on urban environmental innovations, workplace development innovations, or innovations in the development of medicines were excluded. In addition, scientific papers that described innovation without the social context or described SI as limited to policy were excluded. Beyond this, we excluded review articles, book chapters, study protocols, and non-empirical studies.

Data extraction and charting
In Step 4, the authors extracted and charted data from each research context in a pairwise manner. Data was extracted from each of the identified articles and coded according to the following descriptive information: author(s), country of origin, research context, study population, and research methods.
This information is charted in Table 2.

Data Synthesis
In Step 5, each pair of authors synthesised the results from each article in text paragraphs. In a synthesis workshop, all authors read the text paragraphs from each research context (higher education, healthcare, welfare services) to elaborate on and identify common themes across these contexts (to be discussed in the final section).  (Bjerregaard, et al., 2018;Casanova et al., 2019;Giraud et al., 2014). None of the studies were performed in African countries. For detailed information regarding the study origins and other characteristics, see Table 2. Across the studies included, numerous methods, definitions, and theoretical frameworks were employed, and a variety of study aims were presented. In the following section, the findings of this review will be presented according to the three research questions.

Applied Methodology
The characteristics of the methodological approaches used in the studies are presented in Table 2. More detail on the research methods is provided below.

Higher Education
Four of the included studies were based on qualitative methods-either interviews or case studies (Kickul et al., 2012;Nandan et al., 2015;Pitt-Catsouphes & Berzin, 2015;Traube et al., 2017). One study reported on an educational intervention (Cavalcante et al., 2016), and one study adopted a four-step methodological approach to supporting and enabling a theoryinformed learning design (Alden Rivers et al., 2015). Another study was based on a quantitative survey of teachers situated in vocational and professional schools (Tafel-Viia et al., 2012), and one was a think piece on social work education (Traube et al., 2017).

Healthcare
Qualitative research designs were applied in the majority of the studies regarding healthcare. These studies combined the qualitative methods of 89 NJSR -Nordic Journal of Social Research Vol. 12, 2021 document analysis, interviews, observation, and the use of smart technology and social media platforms to create communication data. Five studies applied a case study design (Farmer et al. 2018;Huq, 2019;Norman & Yip, 2012;Okada et al., 2014;Windrum et al., 2017). Seven studies (Ballard et al. 2017;Cheema & Mehmood, 2019;de Freitas et al., 2017;Grindell et al., 2018;Vechakul et al., 2019;Shrimali & Sandu, 2015;Xiao et al., 2015) were qualitative interview studies. Two studies used a quantitative research design, involving methods such as surveys and the quantification of technology users' behaviour (Gravili, 2013;Kitamura et al., 2016).

Welfare services
Most of the studies applied a qualitative methodology and, more specifically, case studies comprising methods like participatory observations, semi- Three of these studies also included findings from previous survey data (Grohs, Scneiders & Heinze, 2015;2017;Meričkova et al., 2015)

Higher Education
In the research context of higher education, SI was defined by novelty and improvement, a definition requiring every new practice to be more effective than its pre-existing alternatives (Cavalcante et al., 2016;Jiang & Thagard, 2014;Kickul et al., 2012;Nandan et al., 2015;Pitt-Catsouphes & Berzin, 2015;Rivers et al., 2015;Tafel-Viia et al., 2012;Traube et al., 2018). Also found was an interest in distinguishing SI from related terms, such as 'social entrepreneurship' (Kickul et al., 2012), as well as in distinguishing 'social intrapreneurship' (i.e., an entrepreneurial behaviour exhibited by employees within an organization) from 'social entrepreneurship' (i.e., innovative, community-based approaches to social work that have been nationally replicated) (Nandan et al., 2015). Nandan et al. (2015) also provided a more elaborate definition, with SI described as 'a novel mechanism that increases 90 NJSR -Nordic Journal of Social Research Vol. 12, 2021 the welfare of the individuals who adopt it compared with the status quo' (p.6) .
Moreover, social innovations were defined as creative products and changes that are motivated by social needs and bring value to society by meeting those needs (Jiang & Thagard, 2014).

Healthcare
Within the healthcare context, most of the studies provided a definition of SI or placed themselves within an SI theoretical framework. SI was explained as a creative process involving community members or users (Cheema & Mehmood, 2019;Farmer et al., 2018;Grindell et al., 2017;Vechakul et al., 2015) and as creating opportunities to explore SI in existing service delivery programs, professional work, and values, directions, and insights (Gravili, (2019) considered SI to refer to new responses to pressing social needs. SIs were found to create new social relationships; they were 'social' in both their ends and their means and could include products, production processes, or technologies. In addition, SI could be a principle, idea, piece of legislation, or social movement in addition to being community-based, which includes microentrepreneurship, or serve as an evaluation approach, an intervention or a combination thereof (Cheema & Mehmood, 2019;Grindell et al., 2018;Huq, 2019;Shier & Turpin, 2019). A common understanding among such studies was that SI is a set of technical innovations in which innovators, users, and communities collaborate using digital technologies to co-create knowledge and solutions to a wide range of social needs. Moreover, SI was considered in relation to initiatives and how involved participants become and interact collaboratively to leverage power through, for example, the use of digital technologies.
Despite building on SI processes and presenting their study as a SI study, three studies did not provide an explicit definition of SI or discuss their research within an SI theoretical framework (Kitamura et al., 2016;Norman & Yip, 2012;Okada et al., 2014).

Welfare services
In the welfare services research context, almost all of the included studies regarding SI, such as co-creation, social entrepreneurship and intrapreneurship, and social enterprise. Furthermore, these concepts have contributed to the dynamic character of SI (Gawell, 2014;Grohs et al., 2015Grohs et al., , 2017Meričkova et al., 2015;Sabato et al., 2017).

Higher Education
Within the educational context, the focus is was twofold: the first was on SI for networking, educational change, and solving social issues, while the second was on SI in academic programmes (i.e., to prepare students for SI, to spearhead societal change in the community, and to stimulate creativity and problem-solving). A majority of the studies focused on educational reforms and the conditions of their successful or not-so-successful implementation (Kickul et al., 2012;Pitt-Catsouphes & Berzin, 2015;Tafel-Viia et al., 2012;Traube et al., 2017).
The practice of SI was also seen as offering promising approaches to social issues. Pitt-Catsouphes and Berzin (2015) employed insights and tactics that have emerged in the SI field and used them to strengthen current macro practices, Alden Rivers et al. (2015) proposed a theoretical framework to support the embedding of SI education in existing academic programmes.
Moreover, one study provided implications for the programmatic challenges of integrating social business concepts and initiatives into curricula and pedagogy (Kickul et al., 2012).

92
NJSR -Nordic Journal of Social Research Vol. 12, 2021 In addition, some studies focused on recommendations for SI actors, who were often seen as educators who can prepare students to recognize and implement new SIs in their communities (Kickul et al., 2012), and on recommendations for social workers, who were seen as capable of leading and facilitating social changes that could have lasting impacts on communities (Nandan et al., 2015). Cavalcante et al. (2016) reported on a study aimed to promote problemsolving and creative thinking among students as a means of SI. This study also aimed to enhance students' knowledge and their willingness to help build society and shape key societal values. Another study took an explicit learning approach asking how SI education can be defined, which learning theories best support SI education, how such learning theories relate to existing models of learning in higher education, and what implications an SI pedagogy may have in learning design (Alden Rivers et al., 2015). Stavreva-Kostadinova (2018) found that the encouragement of self-inclusion, volunteering, and active implementation in educational group sessions helps students develop opportunities to participate actively and meaningfully in social innovation teams.

Healthcare
The foci of the SI studies regarding healthcare were found to involve digital technology, healthcare policy driving shifts in healthcare practices, and the use of SI to meet societal needs related to aging populations, family and reproductive health, youth mental health, and substance and addiction services. The purpose of these studies was also to assess and evaluate SI within service delivery and interprofessional work.  (2019), SI was connected to policy, system and institutional change, and the implementation of programmes in mental health, healthcare, and social services organizations.
In addition, SI appeared as processes engaging public and private actors, family caregivers, and service users to meet health and societal needs. Okada and Igarashi (2014) aimed to deepen the societal understanding of social issues related to dementia using an innovative design process. In a study by Vechakul et al. (2015), innovation appeared in the context of health and social care organizations and sectors, with the aim of applying human-centred design (HCD) to create innovative products, services, and strategies. Xiao et al. (2015) combined smartphone technology with social digital activity to engage community residents in current health challenges and social problems.
De Freitas et al. (2017) identified social innovation support strategies for patients and families to cope with rare diseases. Farmer et al. (2018) described a grassroots initiative aimed at improving the oral health of community members using SI theory to implement innovative ideas from community-based processes. Elsewhere, Cheema and Mehmood (2019) described how challenges to maternal and reproductive health services in remote and rural communities in developing countries could be approached using an SI model.

Welfare Services
The studies included in the welfare services context appeared to mainly explore the interplay between policy intentions and implementation related to SI. For instance, Giraud et al. (2015)  'disruptive technology'. Visentin (2018) studied the emerging logic of welfaresharing, particularly in terms of service design and the social processes involved. Lindsay et al (2018) focused on how co-governance and comanagement as a model could improve services for single parents. Beyond this, Cools and Oosterlynck (2019) showed how to implement a socially innovative service developed by local actors of work trajectories for foreign language newcomers in Belgium. Andersen and Bilfeldt (2017) presented how action research may be an interesting approach to the study of SI and empowerment within the welfare system to create possibilities for positive change. Table 3 provides an overview of the results. To summarise, qualitative research methods dominated the field of SI research. SI as a novel and value-based concept was found as prominent and related to addressing social needs and values. In addition, SI processes comprised creativity, engagement, social inclusion, and cohesion, focusing on change and innovation at both the micro and macro level for the benefit of society.

Discussion
The aim of this study was to investigate how SI unfolds within three research contexts. Our study results have identified an array of methodological approaches, definitions, and conceptual understandings of SI as well as complex SI emphases. Later in this article, the results will be discussed in relation to common methodological approaches used in SI research, how we can understand SI as a social concept across interprofessional contexts, and how SI unfolds within welfare research.

Considering Methodological Approaches in SI Research
The complexity of SI may cause problems for researchers trying to apply research methods in its study. This review suggests that research on SI is in its early stages and that possible next steps could include the development of a broader range of research methods. To guide this development, there seems to be a need for conceptual evolution and clarification regarding the Based on the review results, it is unsurprising that attempts to measure SI are still in a very early stage, and few attempts have been made to this end (Caulier-Grice et al., 2012;Mulgan et al., 2013). On a related note, explorative and qualitative approaches dominate the research field of SI. For example, quantitative studies were few in the sample of included studies, a finding supported by other researchers (Nicolls et al., 2015). Moreover, the results here showed that the methods applied in the welfare context are more diverse than those applied in higher education and healthcare. In addition, the lack of SI research within the welfare fields of higher education and health care in Nordic countries may reflect a lesser focus on scientific, empirical investigations for the benefit of non-empirical, theoretical discourses.
As argued above, the development of SI research methodology may focus on the development of conceptual models. The qualitative approach found in much of the SI literature has likely contributed to these possibilities. As shown in this review, it should be possible to develop conceptual SI models within different contexts and at different levels, with certain core model elements revealing the inner meaning of SI.
This review partly supports the notion that SI can be defined somewhat differently across various contexts. Several conceptual models may be developed based on different theories and experiences. For example, a researcher focusing on organizational elements (e.g. resources, leadership, competencies) may include other concepts in a model in comparison to a researcher focusing on team elements (e.g. members, characteristics, processes). Hence, the development of a research instrument, such as a questionnaire, will look quite different depending on which conceptual model is developed and how it is operationalized. In summary, these results may 99 NJSR -Nordic Journal of Social Research Vol. 12, 2021 provide information about very different aspects SI. This is basically an issue of validity-that is, the question of what is being measured. Without a sound contextual understanding of the meaning of SI (e.g., content, process, value), the concept is shrouded in confusion.
How can we understand innovation as a social concept across research contexts? In this review, only four of the included SI studies were performed in Nordic countries, and all of these were within the research context of welfare.
One reason could be that the global economic crises that hit the EU in 2008 had less impact on Nordic countries, which may reflect the EU's commitment to welfare development and research in countries that faced recession and thus had a greater need for change, including in terms of social innovation.
Additionally, the economic crises may mirror the role of the Nordic welfare model, in which the state and local government play a pivotal role and the principles of universality and egalitarianism constitute ideals and, consequently, a robust financial base for the citizens' welfare (Pedersen & Kuhnle, 2017), and possibly less strain for change. However, new SI research initiatives are present in Nordic countries. A recent review by Jungsberg et al.
(2020) on community-driven SI in rural areas in Nordic countries adds to our understanding of SI from a Nordic perspective, as the authors points to important characteristics of innovation actors and suggest a building capacity for SI implementations among local innovators.
Our findings showed that SI is not a straightforward concept and that it has been used to describe policy goals related to national, regional, community, and organization change (societal level). Moreover, SI has been used to describe partnerships involving specific service user groups and stakeholders in collaborative work (individual level). In the included studies, SI was not always defined, and when it was, this was done in a variety of ways. Thus, there seems to be significant uncertainty about what the concept actually entails (Murray et al., 2010). According to Caulier-Grice et al. (2012), SI must do the following: 1) represent something new for their site, market, or users or represent a new application; 2) include implementation-not just new ideas; 3) meet a current social need; 4) show higher functionality than existing solutions and create measurable improvement; and 5) empower the recipients of services, create new roles and relationships, develop new tools, and strengthen these services' abilities to deliver accurate solutions. Although social innovation has positive connotations and is meant to protect the common good (Tidd & Bessant, 2014), it is important to be critical and ask 100 NJSR -Nordic Journal of Social Research Vol. 12, 2021 oneself who will benefit from the innovation in question as well as what disadvantages it entails and for whom. This review revealed that within all three research contexts, the concept of SI is associated with value creation although this concept is connected to a degree of uncertainty in terms of a lack of agreement on definition and delineation, outcomes, and processes of involvement. Moreover, innovation as a change strategy in higher education, healthcare, and welfare mayinvolve the recognition and application of perspectives and concepts that have not been developed in terms of this context. We believe that social innovation can embrace these social aspects and contribute to a more nuanced understanding of innovation related to interpersonal interactions. Therefore, social innovation can be an appropriate approach and supplement to traditional theories of innovation (Caulier-Grice et al., 2012).
Additionally, the meaning of SI is often blurred by its intersection with other similar concepts, such as social entrepreneurship, social intrapreneurship, and co-creation, which involve both organizations and actors (Grohs et al., 2015).
Social entrepreneurship is also central in the context of higher education, where it functions as a value-based service development to enrich student welfare while familiarizing them with SI and providing tools to conduct SI.
Within the healthcare context, the majority of the studies examined here applied SI frameworks emphasizing value-based processes, digital technologies, user involvement and collaboration between stakeholders, and creativity. In other words, SI is a multifaceted concept and can be used for various-often political-value-based and research agenda purposes, thus valuing the common good (Borzaga & Bodini, 2014). SI can be understood as central in discussions related to where and how innovation arises and who can and is allowed to contribute in the innovation, as well as those regarding democracy and public value creation (Author, 2020).

How Does SI Unfold in Welfare Research?
Across research contexts, SI unfolds at different levels: at the societal level, SI is connected to policy, systemic, and institutional change and implementation; it may strengthen current macro practice and complement or progressively modify the governance structures of the welfare state (Weinzierl et al., 2015;Windrum et al., 2017). Studies of individual-level SI in this review were mainly founded within the research contexts of higher education and healthcare. The latter was reported in studies aiming to develop interventions for improved 101 NJSR -Nordic Journal of Social Research Vol. 12, 2021 student performance, participation in workplace innovations (e.g., Stavreva-Kostadinova, 2018), and patient care. It appears that SI associated with technology is more prominent within the healthcare context and is less prevalent in higher education. In the latter context, SI involves networkingthat is, creating social relationships between groups such as teachers, social workers, and students; stimulating creativity; and increasing student capacities for SI processes.
Common findings across the three research contexts included understanding SI as involving three dimensions-relational, processual, and interactionalwhich is in line with the meta-study of Crepaldi et al. (2012). The processual and interactional dimensions of SI were revealed in the studies to be significant across various contexts and levels. In the processual dimension, the SIs are ongoing interplays dependent on a context; in the interactional dimension, there is a requirement of human relations and co-creation to make SI work within and between complex systems. However, the relational SI dimension in higher education seems more bottom-up-oriented when compared with the healthcare and welfare contexts, where the drivers of SI are policy and industry. Although policy also appears to play an important role within higher education, it may remain in the background of the discourse in favour of a larger interest in the relational and processual dimensions. This suggests that certain dimensions of SI are sensitive to context.

Strengths and Limitations of the Review
The methodological strengths of this scoping review included the involvement of several researchers knowledgeable within the research fields of interest.
Moreover, the review collaboration involved two workshops and digital meetings for the authors to discuss, illuminate, and establish a sound and rigid review methodology to produce trustworthy results. However, certain limitations should be recognized. The broad approach of our literature search with a limited use of keywords and databases may have resulted in missing relevant articles, even though we included a search of reference lists. In addition, we excluded all document types other than peer-reviewed articles. In the current review, systematically identifying grey literature within three research contexts was a challenging task expected to result in an unmanageable amount of data. However, the inclusion of grey and white literature is recommended in a scoping review and may have broadened the understanding of this research field, especially in terms of SI in the Nordic context. Finally, an evaluation of a possible risk of bias in the included studies 102 NJSR -Nordic Journal of Social Research Vol. 12, 2021 was not performed due to great variation in the primary data sources and a lack of well-established quality assessment criteria (Whittemore et al., 2014).

Conclusion and Implications
In summary, this scoping review illustrates how the concept of SI is used and understood within and across the research contexts of higher education, healthcare, and welfare services. Clearly, SI as a concept has gained prominence in these research fields and in EU policy during the last ten years.
Despite this development, Nordic empirical studies, especially within higher education and health care services research, are still limited. Thus, future research may focus on the role of the Nordic welfare model in the development of SI within Nordic countries.
The findings of this review also suggest that the 'social' component of SI is a relevant but problematic suggestion as to how welfare should be understood and exercised. SI may appear to be yet another 'buzzword' serving diverse political agendas, while it cannot be meaningfully encapsulated by only one definition. We suggest that empirical and conceptual development takes place in term of the common SI dimensions identified in this review: relational, processual, and interactional. Despite its various definitions and, in some studies, deficient definitions, SI holds the potential to emphasise collaborative, creative, and value-based aspects of innovation. Thus, further studies could work toward identifying generic elements of SI as well as empirically exploring its uses and effects in various social contexts-within and beyond higher education, healthcare, and social work.