Keywords

Next one, [statement] 28, uh, “Believes in some way, but does not view him or herself as religious” [FQS28]. Uh, well, it’s partly true about me; that is, I feel, uh, well, that something exists, but I’m not sure that Orthodox Christianity, for instance, is something I should identify myself with. I just kind of, um, I believe there’s something, but, uh, I’m not sure it can be covered by a single religion, uh, right. This is why I don’t consider myself to be religious. Interviewee from Russia (YRUPV043) commenting on a statement from FQS

1.1 Introduction

The citation above illustrates how the views which young people hold today on religion and spirituality, including non-religious worldviews, can be diverse and multifaceted. Voices like these raise questions such as, what does the religiosity of young adults look like today? How is it formed by specific cultural and national contexts? How can we successfully investigate questions like these from a multinational perspective? The citation originates from interview data collected in the Centre of Excellence in research (CoE) and research project Young Adults and Religion in a Global Perspective (YARG). It is a cross-cultural, comparative and mixed-method study of religious subjectivities and values in their context. More precisely, the preliminary research questions were:

  • What are the characteristics of the religious subjectivities and values among young adults globally in terms of the configurations of religious, spiritual, and secular assumptions regarding beliefs, attitudes, practices, and experiences?

  • What are the main discourses that constitute and shape the above subjectivities in terms of institutional, social, cultural and other related influences?

  • What methodological and theoretical implications follow from our results with regard to how contemporary religion is conceived?

As these questions suggest, the YARG project has been ambitious, and it has also been successful in many ways. We set out to gather material from 13 countries – Canada, China, Finland, Ghana, India, Israel, Japan, Sweden, Peru, Poland, Russia, Turkey, USA – and we also succeeded with this, creating an active research network in the process. In Japan, however, we only implemented one part of our study, namely the survey. The ambition of this volume is not to provide definite answers to the questions posted above. First, the mere idea of providing definite answers to questions of this kind is disputable. On the one hand, the ‘global’ research horizon of YARG is broad and complex: it includes many culturally, linguistically, politically and historically different contexts that in themselves would require more thorough research efforts than what the YARG study was aiming for. On the other hand, we can also critically ask if our multinational approach is comprehensive enough with only 12 countries (or 13 with Japan). In some respect, YARG has a much too limited focus to fully justify the ambitious questions above. This concerns many different aspects, ranging from the cross-cultural design of the project and issues relating to sampling, to representativeness, to potential ways of analysing data, and to the dissemination of results.

There is also a second reason as to why this volume does not aim to provide definite answers to the questions presented above, and this is of a more pragmatic nature. This volume is part of a series of publications that all stem from YARG data. A special issue on religion and socialization was already published by the journal Religion (Klingenberg & Sjö, 2019) and a volume on religion and media, Digital Media, Young Adults and Religion. An International Perspective (Moberg & Sjö, 2020), by Routledge. Additional books and articles will be published over the coming years.

The aim of this volume could better be described as explorative (see Stebbins, 2001), i.e. to present some relevant snapshots into the research interest defined by the initial questions. The broad-ranging explorative approach fits well with our investigation into a landscape that is not yet clearly mapped and where we also have reasons not to trust current maps. It helps generate new ideas and perspectives. This is connected to two important aspects of the project. First, YARG is on the one hand a study of young people worldwide. This aspect is critically discussed in Chap. 2 of this volume. This focus means that our findings should be discussed in relation to other important contributions addressing how young people today negotiate and form religious identities, views and meanings (e.g. Gareau et al., 2019; Arweck & Shipley, 2019; McNamara & Abo-Zena, 2014). Still, in our case the choice of young adults was instrumental in regards of our interest in how religion is currently changing. Although we did not collect trend data to observe and document changes over time, we have comprehended young people as harbingers of religious and cultural change . Our findings can be seen as trajectories of how religion, spirituality and secularity are diversified and reconfigured.

Methodology is a second important theme. YARG involved an innovative methodological aspect, and particular attention is therefore given to the Faith Q-Sort (FQS). The Faith Q-Sort is based on Q-methodology and was originally developed by David Wulff (2019) for the assessment of religion. In cooperation with Wulff, FQS was further developed within the YARG study and implemented for cross-cultural use. One of the main epistemological strengths of FQS is that it allows us to study contemporary religiosity from a bottom-up perspective and with sensitivity for emerging subjectivities. This chapter will shed light on why this is crucial to the study of religions of today.

The chapters in this volume stem from analyses of data collected in YARG, and the issues they deal with reflect the research interests of YARG as they have developed during the project and via collaborations within the research network. The introductory notes provided in this chapter are central to the chapters as a whole, and provide background information not discussed in the individual chapters. In the following, we will shed further light on four main questions:

  • What theoretical ideas were at the core of the project?

  • What did the Q-methodology and the mixed method approach involve?

  • How was the multinational research process organized?

  • What ethical issues were raised and addressed?

1.2 Understanding Contemporary Religiosity?

Two observations have from the beginning been central to the YARG study. Scholars in the study of religions have become increasingly aware that religion in the first place has been changing rapidly over the last decades and, secondly, that our understanding of religion is based not only on a limited Western gaze, but also on a narrow empirical basis. Taken together, both observations raise serious questions about how we conceive of religion; what we take it to mean. Therefore, they also imply that our means for studying, assessing, and measuring religion are not up to date. Consequently, several scholars have called for a critical discussion of the conceptual toolkit traditionally employed in the study of religions (e.g. Bowman & Valk, 2012; Droogers & van Harskamp, 2014; Lassander, 2012, 2014; McGuire, 2008; Nynäs et al., 2015).

Lately, the prominent discussion about secularization in the West has also shifted towards being more concerned with the matter of contemporary religious change , rather than with a linear and simple disappearance of religious themes, ideas, and phenomena (see e.g. Nynäs et al., 2012; Woodhead, 2012). On-going processes of religious change have been conceptualized through a range of interrelated theoretical frameworks such as de-secularization (Berger, 1999), re-sacralization (Davie, 2010), re-enchantment (Partridge, 2005), post-secularity (Habermas, 2006; Nynäs et al., 2012), un-churching (Fuller, 2001), and de-Christianization (Brown & Lynch, 2012), to name just a few. Some researchers approach religious change against the background of a general ‘subjective or expressive turn’ (Heelas & Woodhead, 2005), while others speak of an “Easternization of the West” (Campbell, 2007) or the emergence of a ‘new style religion’ that is replacing ‘reformation style religion’ (Woodhead, 2012).

For this chapter, we extract a first important observation from this discussion: even though (Western) societies continue to become increasingly secularized, we need to be more attentive to how these processes of secularization are far from incompatible with certain forms of religion and spirituality (e.g. Berger, 1999; Day et al., 2013; Turner, 2010; Nynäs et al., 2015). Rigid juxtapositions between religion and secularity produce simplistic and distorted pictures of the complexities involved in the formation of values and religiosities today. Secularization does not erase religion but comprises a change in the conditions for ‘religious belief,’ and this has further consequences for the ways in which they can be expressed (Taylor, 2007; Warner et al., 2010). Current change is not taking place on a societal level only; it is also visible in the cultural and individual spheres. For instance, a growing body of research highlights the ways in which people increasingly mix ideas, practices, and identities in novel ways, following the changing organization of religion, secularization, and increasing religious diversity (e.g. van der Braak & Kalsky, 2017; Bruce & Voas, 2007; Woodhead, 2012; Gilhus & Sutcliffe, 2013; af Burén, 2015; Nynäs et al., 2015; Nynäs, 2017).

The process of religious change is fueled by global processes and sociocultural shifts in societies. In particular, the role of media (e.g. Granholm et al., 2015; Moberg & Sjö, 2020), consumerism (e.g. Gauthier & Martikainen, 2013; Gauthier, 2020) and social movements (e.g. Nynäs & Lassander, 2015) are vital to these changes. Several accounts (e.g. Brown & Lynch, 2012) of contemporary religious change direct particular focus at the ways in which these developments have entailed cultural change, with subsequent profound implications for traditional understandings of religious authority and mechanisms of religious socialization. In the light of this, Ulrich Beck’s (2010, p. 42) claim makes sense when he observes that instead of previous perceptions of a fusing of nation and religion “we see the formation of a new, religiously determined, global sociality in which increased significance is attached to transnational, religious imagined communities which complement, and enter into competition and conflict with the institutionalized forms of national societies and national institutions.” Religion is both reconfigured and relocated.

From giving relevance to global sociocultural shifts follows a growing need to engage with religion outside the Western frame, both in geographical and conceptual terms. Fenggang Yang (2018) discusses what he calls the Global East, namely not only the East Asian societies and cultures but also diasporic communities of East Asians and the more general impact of East Asian culture and religion on the West, and claims that this “presents theoretical and methodological challenges for the social scientific study of religion” (Yang, 2018, p. 7). This takes us to the second main observation of the YARG study. There is a strong agreement that religion has often been one-sidedly conceptualized and assessed as a transhistorical universal essence, while religion as a concept has often in practice been provincial (e.g., Asad, 1993, 2003; Balagangadhara, 2005; Chakrabarty, 2000; Masuzawa, 2005; Winzeler, 2008). This is a complex issue. It implies that scholars have often been forced to presuppose a biased conceptual similarity between various religious traditions. This follows from how both general views of religion as well as restricted measurement or assessment tools tend to be predominantly Christian centered. Research on religion has further been predominantly conducted by Western scholars and on religion in the West. Also, research on young adults and religion has presented a Christian focus and bias, and has to a large extent emerged from the US (Shipley & Arweck, 2019).

Of direct further relevance for this aspect, is the debate on universality vs. particularism: the universality assumption clashes with the increasingly prevailing notion of cross-cultural incommensurability (e.g. Balagangadhara, 2014a, b). For example, as Balagangadhara (2014b, p. 41) states concerning the application of “Western” understandings of religion on the study of religion in India, scholars tend to assume “that religion is a cultural universal and that the difference between Indian and western culture (among other things) lies in the difference between their ‘religions’”. A critical view of assumptions of universality is essential to aspirations regarding cross-cultural studies on religion and spirituality. Nevertheless, we also need to stress the risks of becoming the victim of an opposite methodological and conceptual trap that is based on assumptions of, for instance, ‘the totally different Western culture’ or the ‘totally different Asian or African culture’ or similar notions attributed to national and cultural geographies. This is equally problematic and challenging since it might become dependent on the process of essentializing differences and historicities in terms of incommensurable particularities.

The complex problem we address here does not mean blindness or bias in relation to non-Western cultures only. As Woodhead (2010) claims with regard to the concept of spirituality, it is often understood to be socially precarious when we think of institutions in terms of established churches and hierarchical structures. “Here again”, Woodhead (2010, p. 42) writes, “we see the distorting effect of identifying ‘real’ religion with historic western churches”. Bruce and Voas (2007) maintain that the processes of secularization and increasing religious plurality have altered contemporary religiosity in a fundamental way. For this reason, the typologies we use in order to describe – for example – religious organizations, such as church, sect, denomination and cult, have lost much of their earlier heuristic and explanatory value (Bruce & Voas, 2007), and the usefulness of taken-for-granted concepts in the study of religions such as public and private have been questioned (Woodhead, 2013) .

De Roover claims that, “the contemporary study of religion has a unique opportunity to settle the debate on the cultural universality of religion” (de Roover, 2014, p. 2017). What does this mean? In this study, we have taken this to mean the need to establish a third option beyond dichotomous academic positions on universality versus particularism. This means, we argue, that we need to be cautious with regard to essentialist, limited, generic understandings of religion which are based on theistic, doctrinal, institutionally-based faith (Beckford, 2003; Day, 2010, 2011). As Lassander (2012) suggests, we must abandon dysfunctional categories and models and instead approach religion as a hybrid, or as Taira (2006) writes, account for the liquid character of religion with reference to Bauman’s (2000) distinction between a solid culture that existed in earlier times and the fluid or shifting character possessed by the present. Religion needs to be conceptualized as being part of human and social interpretations and negotiations. Concepts such as religion, spirituality and belief are created within various forms of academic enterprises; we need to be more attentive to the everyday, individual uses of the term ‘religion’, and be open to new understandings of that term.

In other words, with our study we attempt to be attentive to the social location of religion and “its role in bringing into being forms of identity that actors strategically create in order to adapt to and integrate themselves into various social situations” (Day, 2010, p. 10). In terms of method and on a pragmatic level, this theoretical outset by necessity equals a bottom-up approach to the category of religion that is characterized by a systematic empirical cross-cultural approach. This might have the potential to realize what Lambek (2014, p. 147) calls a “moving balance between distinct epistemological positions” and acknowledge that the ideal to obtain a neutral ‘view from nowhere’ is an illusion (Nagel, 1986) or, as Haraway (1988) aptly puts it, a ‘God trick’.

In this volume, we often choose to call our respondents’ outlooks worldviews. The precedent of worldview, Weltanschauung, was first connected to German idealism. The English version ‘worldview’ has been used and theorized in various ways: for interpersonally shared ideologies, for privately held outlooks, for explicitly formulated philosophies and for implicit and intuitive structures (Holm, 1996; Naugle, 2002; Nilsson, 2013; Kontala, 2016) . We prefer to use it as a parent term that can incorporate religious, spiritual, and non-religious outlooks and views on life. Furthermore, we also emphasize that even though worldviews are primarily expressed verbally these are not necessarily systematic or rational only. As cognitive representations they involve both emotional and conative dimensions (Holm, 1996). This approach helps us accommodate all kinds of emerging viewpoints that could reject, accommodate, approve, or doubt religion or spirituality including simultaneities and ambiguities. It also matches the design of the Faith Q-method we have used.

1.3 Q-methodology and the Faith Q-Sort

In light of the cross-cultural scope of the YARG study, we found that our ambitious criteria were best met by Prof. David Wulff’s work on the Faith Q-Sort (FQS). FQS was designed by David Wulff (2019) in order to overcome some of the shortcomings of earlier approaches to surveys on religiosity and simultaneously to find a nuanced and less biased way to assess religious subjectivities. Q-methodology is rather unknown and usually not discussed in volumes on methodology, with some exceptions (e.g., Newman & Ramlo, 2010; Tashakkori & Teddlie, 1998). It has been used in a variety of fields, ranging from studies of political opinions and marketing research to studies of educational settings and personality psychology, including studies that assess worldviews (Block, 1978, 2008; Brown, 1980; Gabor, 2013; Watts & Stenner, 2012; van Exel & de Graaf, 2005; McKeown & Thomas, 2013; Nilsson, 2013).

Q-methodology was developed in the 1930s by the British researcher William Stephenson (1993/1994) for assessing subjective viewpoints on a specific topic, or subjectivities. Our opinion is that this focus on subjectivity is important to the study of religions today, since it helps us to shift from limited views on religion, as e.g. belief, to a richer perspective. Subjectivity refers to the range of individual experiences that serve as a platform for agencies, identities, and social identifications, such as variations regarding preferences, emotions, values, desires, interests, practices, views, and beliefs. Subjectivities are also relational, and emerge as parts of interpretative communities at play in various contexts, and they are also fluid to some extent (Biehl et al., 2007). From the perspective we develop rooted in Q-methodology, worldviews therefore always involve an affective dimension in addition to ways of thinking, viewing, and doing things. This aspect of Q-methodology is central to the purpose of studying contemporary religion where lately the relevance of e.g. emotions and practices has been underlined (e.g. Bowman & Valk, 2012; McGuire, 2008; Riis & Woodhead, 2010) .

Although McKeown (2001) developed a Q-set for Christian Orthodoxy, Q-methodology is relatively new in religious studies and FQS is currently the only tool for assessment of religiosity based on Q-methodology. The FQS was developed to meet the challenge of how to assess individual religiosity and spirituality, and by using Q-methodology, Wulff (2019) designed an instrument that differs substantially from most other instruments in the field, such as the well-known Allport-Ross Religious Orientation Scale, ROS (Allport, 1950; Allport & Ross, 1967).

When a researcher conducts a study with Q-methodology, he or she confronts the respondents with statements that in a significant way reflect a broad array of viewpoints on a subject matter (e.g., the domains of personality or religion). The respondents are required to rank-order these statements and, unlike ratings in quantitative scales, how the statements are ranked is dependent of how other statements are perceived; the ranking of any particular statement constricts the placement of the other items (see Fig. 1.1 below). Providing respondents with a set of items that they have to sort into categories allows for individual expression of nuances and complexity, and consequently for a variety of expected and unexpected configurations to emerge, but the pool of statements (Q-set) also limits the study (Table 1.1).

Fig. 1.1
figure 1

FQS Record sheet and layout. The 101 statements of FQS are printed on cards and respondent ranks these statements by placing them in different categories on a layout. The placement should reflect the extent to which the respondent identifies with a certain statement in comparison with other statements in the Q-set of FQS

Full size image
Table 1.1 Examples of statements from the Faith Q-set
Full size table

The Q-set for a study should be broad enough to reflect potential subjectivities. It is typically derived from a range of sources that reflect views that both academic and non-academic actors might have on the topic being assessed (e.g. religiosity). Interviews, observations, and popular and academic literature on the topic are important sources for generating a specific Q-set, i.e., all sources that reflect relevant discourses and constitute the potential concourse (van Exel & de Graaf, 2005). There is of course not an endless option of significant positions that individuals take in relation to a certain subject, and the assumption behind Q-methodology is that only a limited number of distinct viewpoints exist on any topic (Brown, 1980). The validity of any given Q-set rests on its items being representative of the entire domain (concourse) of the field or discourse being studied.

As a research instrument, the FQS is a qualitative procedure that involves and is assisted by quantitative analyses. This makes it an inherently mixed-methods tool (Newman & Ramlo, 2010). The individual sorts from a Q-study can be used to paint the picture of an individual. More commonly sorts made by a number of individuals can be combined and analyzed together resulting in what we call prototypes. Prototypes represent significant shared and unique patterns in our Q-data that have been extracted through an analysis of intercorrelations among Q-Sorts, which are then factor-analyzed. In our YARG study we have used the PQMethod based on the Principal Component Analysis for the statistical part. PQMethod is a widely-used software program maintained by Schmolck (2017) and is available online. We also used the online software Ken-Q Analysis (Banasick, 2019) that builds on PQMethod.

The Q analysis produces data for the final definition of the prototypes; a separate part of the process that requires reflection and judgment on the part of the researchers. The outcome of the statistical analysis includes tables with, for instance, factor loadings, item factor scores, and distinguishing statements for each of the factors (prototypes). The factors are distinguished by particular characteristics, but they may also share characteristics with some of the other prototypes. Because of this, the researchers have to determine both which items define a particular prototype, and which items distinguish one prototype from the other. To put it simply, behind every prototype we find individuals that more or less resemble the general description of a subjectivity that a prototype is. Sometimes a prototype is constituted by a very small number of participants, but it still remains relevant because it is distinct compared to other prototypes and represents a unique point of view (Watts & Stenner, 2012).

The final prototypes in a sample can be described with more or less nuance, but they are always an informed interpretation of the preliminary factor analyses. The so-called “commentary style interpretation” is a narrative based on statements ranked high and low in combination with distinguishing statements. The final reflective part done by the researcher often also includes the labeling of prototypes (e.g. Kontala, 2016; Watts & Stenner, 2012). The chapters in this volume present several different ways of using data from our YARG study for analyses, ranging from analyses of our main findings in terms of five global prototypes and how prototypes from all countries resemble these to discussions of the relevance of how single items and statements are played out. In some chapters we also use the prototypes as categories that help us account for an individual level. In these cases, we may refer to a person of a prototype, i.e. an individual that correlates strongly with that specific prototype only and to a certain extent can be assumed to reflect the worldview pattern of that prototype.

These possibilities highlight the specific potential of Q-methodology as a tool for researching worldviews. Worldviews are expressed by people in their own unique ways, relative to their distinctive circumstances and one is not exactly similar to the another. Neither are they necessarily stable over time and place. Amidst all the uniqueness, the researcher can nevertheless find a shared pattern that allows a more abstract formulation. Q-methodology allows us to approach this empirically, where each Q-sort is the expression of a concrete individual, but each factor yields a more abstract prototype which helps to identify what is common to a worldview. In this project we find the methodological capacity to move between these levels important in order to generate a more valid representation of the phenomenon we call worldview. It is a quality different from both the virtues of representative surveys, and detailed hermeneutically informed qualitative methods. It corresponds with Wilhelm Dilthey’s hermeneutics and his principle of forming generalizations by moving from parts to wholes and back again. What may look like a weakness from conventional perspectives, is actually a strength.

1.4 FQS and Cross-Cultural Research

FQS is ideally suited for cross-cultural research. FQS ’s rich item pool and flexibility in their placement puts subjectivity at the center of investigation. This promotes methodological attentiveness to the many different ways of living, experiencing, and expressing religiosity and their idiosyncratic configurations. The 101 statements that Wulff (2019) compiled reflect major religious traditions including observations from subfields in the study of religions. The set of statements is also broad in the sense that in addition to ways of thinking and viewing, it covers experiential and emotional dimensions, and also practices and ways of doing things.

This version of FQS was developed in a North American context, but it has still been successfully used also in several studies with both religious and non-religious groups (Terho, 2013; Pennanen, 2013; Lassander & Nynäs, 2016; Kontala, 2016) . Still, any systematic multicultural study requires a thorough evaluation of instruments and methods before implementation. Therefore, teams of scholars from all involved countries contributed to evaluating Wulff’s initial FQS . With regard to the religious and spiritual worldviews in their cultural contexts, they proposed revisions of current statements and suggested new ones, as well as pointing out statements that they found to be problematic for some reason.

This evaluation provided us with a significant and extensive input for a revision of the FQS that took place during a seminar where a multicultural and multidisciplinary team together reviewed the response. Throughout this process, we strived to be attentive to more local forms of religiosity, non-religiosity and secular positions. This process resulted in the FQS-b (appendix 1), and only this version has been used in the YARG study. As a result of this process, we introduced for instance a new statement “Believes in some way, but does not view him- or herself as religious” (FQS28). This connotes a form of simultaneity and has been identified by the research team as central to Scandinavian religiosities (see also af Burén, 2015; van der Braak & Kalsky, 2017). Still, this statement proved to be salient for many participants in other areas as well, and came to contribute to defining religious subjectivities in China. Another new statement reflects to what extent “His or her sexuality is strongly guided by a religious or spiritual outlook” (FQS59), tapping into how many issues about morality are topical and form religiosities.

Despite all the efforts put into defining the specific items used in the FQS-set, this remains a challenging project. The particular expressions still need to be generalized for a variety of different contexts. The ambition to produce a version of the FQS that has multicultural validity requires modesty. A related problem is that the process of producing a valid Q-set tends to push the linguistic expressions to a level of sophistication that not necessarily echoes the ways in which young adults express worldviews. Using the word transcendence in the FQS-set is an example of this. It does not necessarily resonate with peoples’ everyday language. If the statements in the Q-set become too distanced from a real-life discourse , they may end up hard to comprehend, inviting participants to play a guessing game or make participants lose interest. To achieve true multi-cultural validity might involve item-by-item international, multilingual, and cross-cultural validation of all individuals (Wolf et al., 2020).

Validity and reliability of Q-methodology cannot be easily transferred from regular survey methodology. Like other research methods, Q-methodology has been the subject of critical discussions. One has e.g. addressed the need to further discuss how to perceive the role of language and discourses in contrast to both a more scientific approach and the focus on subjectivity (e.g. Billard, 1999; Druschke et al., 2019). It has also been claimed that the method has limited theoretic value due to the use of factor analysis, that the forced distribution distorts participants actual outlooks and preferences (Kampen & Tamás, 2014), and that the subjectivity and bias of the researcher is overlooked (Robbins & Krueger, 2000; Sneegas, 2020). In comparison with quantitative methods, Q-method has correctly been criticized for lacking the possibility for quantifying generalizations: it is not primarily concerned with which proportion of a larger population is associated with which prototype (see e.g. Thomas & Baas, 1992/1993).

We need to carefully scrutinize our use of Q-methodology, recognizing its limitations and qualities. In order to do so we also have to address the different nature of Q-methodology. As Stenner et al. claim “the Q sort as a data-collection form is designed to maximize the expression of qualitative variation and to record it in numerical form” (Stenner et al., 2008, p. 218) revealing that Q-methodology is best conceived not as quantitative or qualitative, but as an inherent blend of both (Ramlo, 2021). The emerging results from a Q-study will be more like concluding that white tigers exist, than claiming that all crows are black. A central strength of the FQS is therefore its potential to expose new and emerging subjectivities and defining elements in these. This allows comparisons across samples and the possibility of further investigating these samples in large-scale surveys.

1.5 A Mixed-Method Approach

The development and international implementation of FQS was at the core of the YARG study. The novelty in our research was first and foremost defined by the implementation of the FQS , but essential to our project were also aspects such as how the FQS was designed in itself, and furthermore combined with other methods. We used a mixed method-approach that meant a combination of quantitative and qualitative research instruments. This allowed us to address our research questions more fully, but with other methods we could also assess the quality of the FQS .

The YARG study started with a survey including six item blocks assessing the participants’ current life situation, social life, sources for news and information, views and convictions, well-being and happiness and personal details (Appendix 3). The survey also included the Portrait Value Questionnaire (Schwartz, 1992, 2012; Schwartz et al., 2012). The survey had several functions. First, it provided essential demographic data that aided in our sampling for the FQS study. The FQS does not require a large number of respondents, but it is important to find enough respondents representing a variety of viewpoints. For this purpose, our initial survey (N ≈ 300/ country) enabled a broad selection of participants for FQS (n ≈ 45/country) with regard to gender, ethnicity, religious affiliation, language groups and class. These, in combination with other characteristics such as e.g., value priorities based on Schwartz’s PVQ , and field of study of the participants, guaranteed diversity among the study’s participants and their Faith Q-sorts. Since FQS is a new method, data from the survey has also helped us to explore the quality of FQS . How are individual prototypes reflected in measures of e.g. religiosity, demographics, moral attitudes, values and cultures?

Second, the survey also provided data for independent analyses. The part in the survey on media use has been at the core of analyses of religion and digitalization (Moberg & Sjö, 2020) . Several chapters in this volume are also rooted in independent analyses of parts of the survey. A good example of this is Chap. 10 in this volume, on prosociality, civic engagement and volunteering.

Schwartz’s PVQ on values has also been an important inclusion. From the perspective of sociology (e.g. Inglehart & Welzel, 2005; Maio, 2017), one can assume that individuals’ values are not necessarily so much influenced by their (religious) worldviews as they are by e.g. the socio-economic context they grow up in (see also Lassander, 2014) . The earlier idea that religious institutional affiliation translates into a particular value profile does not necessarily hold true, and there is ample ground for assuming that values play a more independent role in defining and directing how individuals self-identify in terms of religiosities.

Our mixed method approach also included interviews. The FQS sorting takes place face-to-face with the researcher giving the respondents the opportunity to ask questions and comment on the meaning of the statements. After the FQS sorting, a more extensive follow-up interview took place. During the interview, the respondents could explain in more detail why they had ranked certain statements as they had. The interviews were open and semi-structured. In contrast to the pre-formulated statements of the survey and FQS , the participants were encouraged during the interviews to share personal stories, express their own thoughts freely and with their own words. In this way, they could initially explain the lifeviews and priorities which they had previously expressed through the FQS-sort. Further, the interviews generated complementary knowledge about the interviewees’ thoughts on a limited set of themes and topics.

For our purposes in the YARG study and a wider comparability across all interviews, each individual interview followed a pre-defined general structure that was organized around three main themes of interest: (1) Interviewees’ experience of the FQS and thoughts around it, as well as his/her own personal engagement with religion/spirituality or similar positions of a secular character. (2) Interviewees’ personal history, self-understanding and current life situation. (3) Interviewees’ thoughts about the broader social and cultural contexts and communities that they are embedded and involved in. We developed detailed instructions for interviewing (see Appendix 4).

In a similar vein as the survey, the interviews had multiple functions. They were valuable to later interpretations of results from the FQS and for our evaluation of its quality. The recording started already when the FQS sorting was initiated, and also minor questions about the FQS procedure and FQS statements were made accessible for later analyses. Further, the interviews provided additional independent data that could be used as such or as part of other mixed method designs and strands. The interviews have been used to a great extent in the studies coming out of the YARG study already , (Klingenberg & Sjö, 2019; Moberg & Sjö, 2020) and are related to in several chapters in this volume as well.

In short, the YARG mixed method approach has primarily meant a fixed sequential multiphase design where different parts have equal status (Leech & Onwuegbuzie, 2009). The main methods have been clearly defined in the YARG study already from the start of the research process, and the procedures with independent separate parts have been implemented accordingly. Yet, this fixed frame has allowed for emerging and dynamic elements and designs when it comes to single studies and analyses. As a whole, the YARG study has enabled a variety of research strands, namely the basic chain of research from posing a question and choosing data to analyzing and interpreting this (see Teddlie & Tashakkori, 2009).

1.6 The Research Process

The complex research process was made possible by an international research network. The YARG study as a whole was managed by a core team at Åbo Akademi University led by a principle investigator and senior researchers. Local implementation in respective country was administered by co-investigators, and commonly research assistants and/or PhD students were employed for the collection and handling of data. The YARG study was also supported by a scientific advisory board. Planning and training sessions, as well as seminars were held on a regular basis at Åbo Akademi University. The YARG data was collected during the years 2015–2016 in Canada, China, Finland, Ghana, India, Israel, Peru, Poland, Russia, Sweden, Turkey and the USA. These countries were chosen to represent a broad variety of national, cultural and linguistic contexts as well as a significant variety with regard to world religions, historical traditions, and contemporary religious developments and trends. The selection of countries covers all the main cultural value areas identified through the World Values Survey and “the Global Cultural Map” (Inglehart & Welzel, 2005).

As a contrast to this broad approach, our sampling in each country means a notable limitation to the project, since the YARG study focused on university students (ages appr. 18–26 representing a variety in terms of gender and fields of study). Turning to university students was a deliberate choice motivated by our interest in contemporary religious change . In comparison to previous generations, young people born after 1990 are not characterized by having been gradually accustomed to consumer culture and digital media during their lifetime. Instead, these social phenomena have constituted an inherent and unquestioned part of their childhood and youth (cf. Possamai, 2009). Palfrey and Gasser (2008) refer to this generation as ‘the born digital’ and the ‘digital natives.’ They have been raised during social and cultural conditions that are particularly relevant to a study of religious change. Beyer (2019, p. 278) similarly concludes about the identities of the so-called ‘millennials’ that have grown up with expanding global horizons and contexts that they are “better regarded as dynamic and contextual projects, as fluid nodes in networks of relations”.

Our assumption was that university students generally have relatively extensive capital in this respect, also in comparison with other young adults. However, already from Chap. 2 we can learn that university students still do not comprise a coherent group of people. In all countries, we have selected a small number of universities where the survey data has been collected. Universities with a specifically limited focus or character have not been included. Rather, we have initially tried to reach diversity among the respondents. The fact that the YARG study is based on convenience sampling means of course that there is no way to tell if the sample is representative of a larger population, and data from the survey does not allow us to draw conclusions about any specific population or to make valid statistical inferences. Rather, in conformity with the YARG study as a whole, the approach is more exploratory.

The YARG study was conducted in local languages in all countries. All material intended for the use of our respondents, including the presentation of the study, the consent form, the survey and the FQS, was translated from English into target languages: Arabic, Bengali, Mandarin Chinese, Finnish, French, Hebrew, Polish, Russian, Spanish, Swedish, and Turkish. A regular back-translation method guarantees comparability across cultures, and it is the most widely used technique to detect item bias in surveys, i.e., when some items in a test might function differently for different groups in a study (Brislin, 1970, 1980; Geisinger, 1994; Harkness, 2003; Lin et al., 2005; Plake & Hoover, 1979). Yet we used a double and back-translation process, since this has been promoted in order to gain even higher reliability in the translations (Hambleton, 1993, 1994; van de Vijver & Hambleton, 1996). A double and back-translation process enables the researcher to compare and select the “best” translation from two independent parallel translations (Harkness, 2003). Our experience was that this process also provides a higher degree of sensitivity to subtle ambiguities which proved to be essential when translating religious and spiritual vocabulary. This vocabulary can often be marked by biases emerging from the fact that religion in one culture or nation often is much more multifaceted than any translator may be aware of. The fact that we often received different proposals from our translators revealed that no translation is final, and in cases like ours, a multicultural academic team with a broad range of academic expertise in the field is needed to finally decide on its most adequate version.

Several ethical concerns were raised as part of the YARG study. For instance, religious and secular views are often considered to be sensitive information. The participants were well informed about the nature of the YARG study, that their participation was voluntary, and that they would remain anonymous. We especially emphasized that no teachers or parents would be able to request information about their participation in any form. The participants were informed that the data would be used for academic purposes only and by the YARG research team only, and that it would be securely archived after this. On the consent form that they signed and received a personal copy of, they were given contact information in the case they later would have questions or in the case that they would withdraw from the study at any point. These are standard regular procedures, but we also had to be extra careful. Even though it somewhat limited the factors we could include in our analyses, we could not for instance record in our final data set the fields of study or university belonging of our participants since this could put them at risk. This precaution was later in the project confirmed to be very important in some of the countries we included.

Overall, we have followed the national guidelines on the ethical principles of research in the humanities and social and behavioral sciences in Finland (2009) defined by the Finnish National Board on Research Integrity (TENK, 2009) as well as The European Code of Conduct for Research Integrity Revised Edition (ALLEA, 2017). In October 2015, we received approval for the YARG project as a whole from the Åbo Akademi University research Ethics committee and in addition, all co-investigators have followed corresponding national procedures. These vary a lot between the countries. In some national contexts, personal worldviews are also politically sensitive. We have therefore largely refrained from further descriptions of the universities that are involved in the YARG study and where the data has been collected.

The YARG network is extensive and it involves a variety of linguistic and academic contexts. In line with ethical guidelines for international projects, we have tried to promote the careers of involved researchers at all levels. Throughout the process, we have also put efforts into being inclusive in our dissemination process in order to recognize the important contribution of all network members. This is also reflected in our document on principles for YARG authors. It defines some guidelines and rules for dissemination of results, and emphasizes the role of co-authoring. Co-authoring can be a difficult process in international projects for many reasons. Yet, we find that it is worth the effort. At its best, it is a co-writing process where a fusion of horizons materializes in text and in new perspectives.

1.7 The Outline of the Volume

In general, the chapters in this volume that follow are divided into two sections. Especially chapter one, but also chapters two to four are essential to other chapters in this volume, since they together depict our approach, method and main results. Still, all chapters are expected to be read independently since they represent separate studies. We have therefore also needed to repeat certain parts of e.g. theories and methods, even though every chapter has its own focus.

The first section of chapters focuses primarily on what we can learn from the FQS data and explores the FQS-material from different viewpoints. First, however, and following this introductory chapter, we continue with a critical investigation of the young adults and university students that have been the participants in YARG. Like in every academic study, we need to be attentive to the limitations of the methods and approaches we use, and a reflection on our participants is therefore essential. What are the variations of worldviews in a cross-cultural perspective? How can these be further assessed, comprehended and conceptualized, and what are the implications of our results for different typologies of religions? We devote chapters to specific analyses of both respondents who initially seem to be positioned outside the main categories and types, and respondents who seem to hold secular and non-religious views. How are the ways of being (non-)religious distinct from each other? Both these groups are relevant and big. Aspects related to gender also provide lenses for general analyses based primarily on the FQS .

The second section of chapters includes more specific themes and cases, and parts where data from all parts of the mixed method approach also gets more attention. These chapters have a more thematic nature, and we bring up important aspects of being religious in relation to contextual aspects. This includes investigations of individualism and prosocial attitudes, and conservative and liberal values. We also look into how experiences of being discriminated against and being in a majority or minority position plays into worldviews and religiosities, as well as the relevance of the relationship between religion, public life and the state. In the concluding chapter, we bring the volume together, highlighting our main observations and looking forward towards where the YARG study and its varying findings may lead regarding comprehending religious, spiritual and secular subjectivities from a transnational perspective.

As pointed out in the beginning of this chapter, the YARG study is ambitious. This volume reflects an important but modest aspiration to de-center taken-for-granted categories and perspectives (see Bender et al., 2013). We move beyond European and American borders. We move beyond a focus on Christianity. We move beyond congregations and organizations. The cross-cultural mixed method approach we have developed in the YARG project has enabled this. Nevertheless, neither the FQS nor the mixed method design it is part of can in a satisfactory way meet all the ambitions and research questions we initially addressed, but we still think that this volume can help us to identify some new relevant aspects of what it means to be religious, spiritual, or secular today, and how these identifications are entangled in other processes.

In many of the chapters in this volume, the FQS and the mixed method design open up new vistas for research on religion in a global perspective. The FQS differs from regular scales and questionnaires, and it is designed to be sensitive to both commonalities in religious beliefs and practices, while at the same time exposing the limitations of thinking in terms of universal patterns and categories. It accommodates the possibility to identify a range of emerging and even ambiguous or conflicting subjective realities in a complex multicultural context. FQS thus has the potential to produce relevant systematically and empirically rooted observations that can feed into future research on religion and spirituality in nuanced ways.