Cathedrals as agents of psychological health and wellbeing within secular societies: Assessing the impact of the Holly Bough service in Liverpool Cathedral

This study was designed to test the hypothesis that events like the Holly Bough service held in Liverpool Cathedral on the Fourth Sunday of Advent that attracts a wide range of participants, including regular churchgoers and occasional (sometimes annual) visitors, contributes significantly to the psychological health and wellbeing of these participants. At the Holly Bough service held in 2019 a total of 383 participants (139 males, 229 females, and 15 individuals who preferred not to say) completed a recognised measure of psychological health and wellbeing (the Oxford Happiness Questionnaire) while they were waiting for the service to begin and again during a five- minute organ improvisation just before the close of the service. The data demonstrated a significantly higher score on the Oxford Happiness Questionnaire at time two than at time one, suggesting that the experience of the service functioned as an agent of psychological health and wellbeing.


Introduction
The developing science of cathedral studies draws attention the multiple functions fulfilled by cathedrals within the local areas in which they are located. Coleman (2019) succinctly voices, 'the double identities of English cathedrals as places of religious engagement but also as locations of heritage'. Recent studies have suggested that cathedrals may be making a positive contribution to psychological health and wellbeing through both of these identities. The aim of the present paper is to explore the current status of the evidence to support such claims (among congregations, among visitors, and among those attending special events), and then to propose an innovative research method to test the claim within one distinctive context that may unusually draw together into one event individuals whose attendance is motivated more by religious engagement and individuals whose attendance is motivated more by secular interests, namely the Holly Bough service held in Liverpool Cathedral on the Fourth Sunday of Advent, just a few days before Christmas.

Psychological health and wellbeing among congregations
As with any church, at the heart of cathedrals there is a congregation of people who meet for worship on a Sunday. The science of cathedral studies draws attention to ways in which these congregations may be reaching some people whom parish churches may be failing to reach, people who somehow just do not fit into the psychological profile of church congregations  and who may be more motivated than many churchgoers by asking religious questions (Francis & Lankshear, in press). Some commentators have speculated that people choose to worship in cathedrals to avoid commitment to richer engagement in church life. The science of cathedral studies challenges this speculation by listening to participants' own accounts of their motivations and by revealing the depth of bonding and bridging social capital within cathedral congregations.
Here are signs that cathedrals enhance both personal psychological health and wellbeing among participants in the Sunday congregation and social wellbeing within the local community (Francis & Williams, 2015a. Not only do cathedrals attract Sunday congregations, they also attract what ap Siôn refers to as 'hidden congregations'. For ap Siôn the hidden congregations are those people who regularly come back into the cathedral to light candles, to sit quietly, or to post prayers on the prayer board (ap Siôn, 2015a(ap Siôn, , 2015b(ap Siôn, , 2017. Insight into the personal benefits derived from such activities can be revealed by careful reading of the messages left in the visitors' books (Burton, 2015) or by careful analysis of the content of the prayer cards (ap Siön, 2015a). Members of these hidden congregations use the prayer board not only to voice (sometimes heart-rendering) petitions, but also (on occasion) to come back to express deeply felt gratitude and relief in light of the perceived outcomes (ap Siôn, 2015a). Here are signs that cathedrals enhance psychological health and wellbeing among the hidden congregations.
In her study of the significance of the Friends' associations for cathedrals today, Muskett (2015) draws attention to what may be considered as another form of hidden congregation. Drawing on the theory of passive participation in voluntary associations, Muskett concludes from her data that members of the Friends' associations benefit from an enhanced sense of social capitals generated through loyal networking within a distributed community of members near and far brought together by the values of the cathedral. Here are signs that cathedrals enhance psychological health and wellbeing among those who relate more remotely with the ongoing life of the cathedral community.

Psychological health and wellbeing among visitors
More research has been given to cathedrals as visitor attractions than to cathedrals as places of worship. Early studies were particularly concerned with visitor motivation (Winter & Gasson, 1996;Jackson & Hardman, 1995;Voase, 2007;Williams, Francis, Robbins, & Annis, 2007;Gutic, Caie, & Clegg, 2010;Olsen, 2013;Francis, Annis, & Robbins, 2015) and concern with visitor motivation continues (see Riegel & Lindner, 2020). Some recent studies have focused on exploring visitor experiences (Francis, Williams, Annis, & Robbins, 2008;Hughes, Bond, & Ballantyne, 2013;Bond, Packer, & Ballantyne, 2015). Research on cathedral visitors has generally distinguished between visitors characterised as pilgrims and visitors characterised as secular tourists. For example, Williams, Francis, Robbins, and Annis (2007) distinguished among three groups operationalised in terms of frequency of church attendance: those who never attend a place of worship, those who attend a place of worship less often than weekly (styled occasional attenders), and those who attend a place of worship on a weekly basis.
Of particular relevance for the present paper is research concerned with the effect of visiting cathedrals on the outlook and perceived psychological health and wellbeing of the visitors themselves. In an early study reported by Winter and Gasson (1996) among 814 visitors to four English cathedrals (Coventry, Ely, Lichfield, and Wells), visitors named the perceived benefit of peace and quiet. Finding a sense of peace was also identified in the study of 514 visitors to St Davids Cathedral reported by Williams, Francis, Robbins, and Annis (2007). This study also reported a significant association between personal church attendance and identifying such a sense of peace. While 50% of those who never attended church agreed that they felt a sense of peace for their visit, the proportion rose to 81% among occasional churchgoers, and to 88% among weekly churchgoers. Interpreting visitor-perceived benefit from visiting Canterbury Cathedral, Bond, Packer, and Ballantyne (2015) employed factor analysis to distinguish among five benefits that they characterised as: connecting spiritually and emotionally, discovering new things, engaging mentally, interacting and belonging, and relaxing and finding peace.
Working within the broader field of environmental psychology, Ysseldyk, Haslam, and Morton (2016) tested the thesis that immersion within three different built environments (cathedral, castle, and shopping centre) or immersion within three different virtual environments (cathedral, mosque, and museum) would have significantly different effects on reported levels of self-esteem. Their data indicated that higher levels of self-esteem were reported among Christians immersed within the cathedral environment, but not for atheists.
Here are signs that cathedrals enhance psychological health and wellbeing among visitors, and especially among visitors rooted within the Christian tradition.

Psychological health and wellbeing among participants at special events
Alongside the Sunday congregation and the hidden congregation, cathedrals attract different styles of congregations for special events and for special occasions. Among these special events and special occasions, the specific event that has received most attention is the Christmas carol service, from both qualitative perspectives (Phillips, 2010;Murphy, 2016;Muskett, 2017;Coleman, Bowman, & Sepp, 2019) and quantitative perspectives (Walker, 2012a(Walker, , 2012b(Walker, , 2013(Walker, , 2015a(Walker, , 2015bFrancis, Edwards, & ap Siôn, in press;Francis, Jones, McKenna, Pike, & Williams, under review). A particular feature of these studies is that they draw attention to the distinctive profile of the individuals who come to these services. Here is an event in which committed churchgoers and those who come to the cathedral as secular tourists mingle.
One stream of the quantitative studies conducted among those who attend carol services has explored the participants' self-perceived impact of attendance at such services.
Using the four component model of spiritual wellbeing proposed by Fisher (1998Fisher ( , 2004Fisher ( , 2010Fisher ( , 2011Fisher ( , 2016) that speaks of the personal domain, the communal domain, the environmental domain, and the transcendental domain, Francis, Jones, McKenna, Pike, and Williams (under review) found that over half of the participants surveyed at the pre-Christmas service in Liverpool Cathedral reported that the service had helped them over all four domains: the service had helped 66% to feel better about themselves (personal domain), 64% to feel better about their relationship with other people (communal domain), 54% to feel better about their relationships with the world (environmental domain), and 56% to feel better about their relationship with God (transcendental domain). Here are signs that cathedrals enhance psychological health and wellbeing among those who attend special events.

Research question
While the cumulative evidence arising from the previous research reviewed above is largely consistent in supporting the view that people who come to cathedrals, either as worshippers or as tourists, tend to feel that their psychological health and wellbeing has been enhanced by such engagement, the research design from which such data have been derived is weak. For example, the recent study by Francis, Jones, McKenna, Pike, and Williams (under review) among individuals who attended the pre-Christmas service in Liverpool Cathedral was able to report that 66% of the participants had helped them to feel better about themselves. Reported self-perception of this nature is of considerable importance, but there are research designs that can generate a different, and stronger kind of evidence. Suppose that it were possible to take an objective measure of psychological health and wellbeing at the point when people entered the service and again at the point when they were getting ready to leave? Suppose that the service itself was regarded as an intervention deliberately positioned within the lives of those participants? Might it be possible to gauge the impact of this intervention on psychological health and wellbeing?

Empirical theology and positive psychology
The connection between cathedrals and psychological health and wellbeing is a matter of relevance for the field of empirical theology and for the field of the positive psychology of religion. Empirical theology is that branch of theology that takes seriously engagement with theories and methods shaped within the social sciences (see van der Ven, 1988van der Ven, , 1993van der Ven, , 1998Cartledge, 1999;Francis & Village, 2015). Within the Christian tradition, empirical theologians draw inspiration from one of the characteristic ways in which Jesus is portrayed in the Gospels as dealing with theological questions. As empirical theologians Jesus' followers were encouraged to learn about the Reign of God by employing techniques now so familiar to natural scientists. The challenge given to Jesus' followers was to go out into the fields and to observe the sower at work, to describe and to classify the kinds of soils (qualitative research techniques) and to count the yield (quantitative research techniques).
Then away from the fields, these empirical theologians were led into the kitchen to observe the baker at work. Later these empirical theologians were led into the wedding feast to observe the behaviour of the guests, and to identify the fundamental patterns that shape human interaction, human ambition, and human humility, now drawing on techniques so familiar to social scientists. There is good reason, then, for empirical theology to examine the evidence for such clear claims For empirical theology to be properly equipped to examine the evidence for such clear claims linking religious faith with human flourishing, psychological health and wellbeing, engagement is needed with the broad field of positive psychology. Such engagement begins with an assessment of the range of instruments proposed for the measurement of wellbeing (Gallagher & Lopez, 2019). Among the constructs considered by positive psychology, the notion of happiness has emerged as of special significance (Argyle, 2001;Ryan & Deci, 2001;Csikszentmihalyi, 2002;Lyubomirsky, Sheldon, & Schkade, 2005;Kristjansson, 2010). The notion of happiness has also played an important part within the developing positive psychology of religion.
Within the field of the developing positive psychology of religion, Robbins and Francis (1996) and Francis, Jones, and Wilcox (2000) reviewed the current state of the literature and concluded that the problem they experienced in reconciling divergent findings resulted, at least in part, from the diversity of definitions and operationalisations of happiness used in that literature. In order to address this problem, their recommendation was to initiate a series of studies that could be linked by some common measures (or family of measures) used over diverse samples. In particular they proposed that such a programme of research could be integrated by focusing on the model of personal happiness proposed by Argyle and Crossland (1987) and Argyle, Martin and Crossland (1989) and operationalised in their instrument, the Oxford Happiness Inventory. The argument is that this model of happiness benefits from robust conceptualisation, sound operationalisation, and a growing body of empirical studies.
The operationalisation of happiness proposed by Argyle and Crossland (1987) and Argyle, Martin, and Crossland (1989) embraced three components of the construct: the frequency and degree of positive affect or joy; the average level of satisfaction over a period; and the absence of negative feelings, such as depression and anxiety. Working from this definition, they initially developed the Oxford Happiness Inventory for which they reported an internal reliability of .90 using alpha (Cronbach, 1951), and a 7-week test-retest reliability Persian (Bayani, 2008), and Portugese (Neto, 2001).
The Oxford Happiness Inventory comprised a set of 29 multiple-choice items for each of which participants were offered four options designed to reflect incremental levels of happiness, for example: I do not feel happy, I feel fairly happy, I am very happy, I am incredibly happy. In a subsequent study Hill and Argyle (2002) recognised the disadvantages presented by the bulky format of the Oxford Happiness Inventory for surveys requiring multiple instruments. Accordingly they proposed the Oxford Happiness Questionnaire that comprised a set of 29 Likert scale items, originally rated on a six-point scale and subsequently on the conventional five-point scale. In their foundation paper Hills and Argyle (2002) reported a high correlation between the two instruments (r = .80).
In their foundation study designed to establish a model for developing the positive psychology of religion, Robbins and Francis (1996) administered the Francis Scale of Attitude toward Christianity (Francis, Lewis, Philipchalk, Brown, & Lester, 1995) and the Oxford Happiness Inventory among a sample of 360 first-year undergraduate students in Wales. This study reported a significant positive correlation between religious affect and personal happiness. In their recent review of subsequent research building on this foundation study, Francis, Byrne, Lewis, and Sweetman (2020) identified studies employing the same two instruments among seven further samples that all confirmed the finding from the original study: 212 undergraduate students in the United States of America (Francis & Lester, 1997); 295 individuals ranging in age from late teens to late seventies in England (Francis & Robbins, 2000); 994 15-to 16-year-old secondary school students in England (Francis, Jones, & Wilcox, 2000); 496 members of the University of the Third Age in England (Francis, Jones, & Wilcox, 2000); 456 undergraduate students in Wales (Francis, Jones, & Wilcox, 2000); 89 undergraduate students in Wales (Francis, Robbins, & White, 2003), and 3,848 16to 19-year-old students in the Republic of Ireland (Francis & Lewis, 2016).
Subsequently, working within other faith traditions, the Oxford Happiness Inventory has been administered alongside the Katz-Francis Scale of Attitude toward Judaism (Francis & Katz, 2007) in three studies in Israel reported by Francis and Katz (2002)  Against this background, the aim of the present study is now to introduce the Oxford Happiness Questionnaire within the science of cathedral studies.

Procedure
When people came into the cathedral for the Holly Bough service the welcomers gave them a printed copy of the service and a white envelope containing the questionnaire and a pen. The welcomers invited participants to complete the questionnaire. This invitation was reinforced by the video screens organised around the cathedral to relay the service. The front page of the questionnaire carried the following message from the Dean

Preparing for Christmas in Liverpool Cathedral
As Dean of this wonderful cathedral I would like to listen to and to learn from the many people who come to our special services preparing for Christmas. By listening and learning I hope to be able to plan wisely for the future. This survey has been designed with two parts. I would like to invite you to complete part one while you are sitting and waiting for the service to begin. Then toward the end of the service there will be an opportunity for you to complete part two.
Everything you tell us is completely confidential and anonymous. Please feel free to answer as honestly as you can. We want to find out what people are really feeling and thinking about what we are doing as a Cathedral.
Please finish completing the questionnaire before you go, and leave the questionnaire on your seat.
Thank you for your help and cooperation.

Instrument
The questionnaire comprised two parts. Part one, completed while waiting for the service to start, explored demographic, background, and religious factors, and included both the Oxford Happiness Questionnaire (Hills & Argyle, 2002)

and the Francis Psychological
Type and Emotional Temperament Scales (Francis, 2005). Part two, completed toward the end of the service during an organ improvisation on the well-known Christmas carol 'In the bleak midwinter', explored perceptions of the service and the second presentation of the Oxford Happiness Questionnaire.
The Oxford Happiness Questionnaire as proposed by Hills and Argyle (2002) comprises 29 items, 17 with positive valency and 12 with negative valency. Examples of the positive items include: I feel that life is very rewarding; I am always committed and involved; I feel able to take anything on. Examples of negative items include: I am not particularly optimistic about the future; I don't feel particularly pleased with the way I am; I feel that I am not especially in control of my life. While originally developed for rating on a six-point scale, current usage has reverted to the conventional five-point Likert rating: agree strongly (5), agree (4), not certain (3), disagree (2), and disagree strongly (1). In the foundation paper Hills and Argyle reported a high level of internal consistency reliability in terms of the alpha coefficient (Cronbach, 1951): α = .91. In the first presentation of the Oxford Happiness Questionnaire (before the service began) participants were invited to rate each of the 29 items in respect of the prompt, 'How I have felt over the past week about me'. In the second presentation of the Oxford Happiness Questionnaire (during the organ improvisation on 'In the bleak midwinter') the participants were invited to rate each of the 29 items in respect of the prompt, 'How I am feeling about me now'.

Participants
A total of 383 participants at the Holly Bough service completed the Oxford Happiness Questionnaire both at time one and time two. These 383 participants comprised 139 males, 229 females, and 15 individuals who preferred not to say; 13 were under the age of 20, 35 in their twenties, 31 in their thirties, 45 in their forties, 93 in their fifties, 93 in their sixties, 60 in their seventies, 10 were aged seventy or over, and 3 preferred not to say. The majority described their present or most recent work as professional or semi-professional (78%), 9% as manual, 6% as student, and the remaining 7% preferred not to say. In terms of religious identity 84% described themselves as Christian, 14% as none, 1% as either Buddhist or Hindu, and 1% preferred not to say. The 84% who described themselves as Christian comprised 60% Anglican, 17% Catholic, 2% Methodist, and 1% Presbyterian. The remaining 4% included a few Orthodox, Pentecostal, Baptist, Welsh Presbyterian, and other unspecified.
In terms of attendance at acts of public worship (apart from occasions like weddings and funerals) 21% of the 383 participants reported weekly attendance, 8% at least once a month, 18% at least six times a year, 39% at least once a year, 13% never, and 1% preferred not to say.

Analysis
The data were analysed by SPSS employing the frequency, reliability, and t-test routines.

-insert table 1 about here -
The first step in data analysis concerned an examination of the psychometric properties of the Oxford Happiness Questionnaire in terms of the alpha coefficient (Cronbach, 1951) as an index of internal consistency reliability, the correlations between the individual items and the sum of the other 28 items as a measure of the contribution of each item toward the homogeneity of the scale, and the item endorsement (presented as the sum of the agree and agree strongly responses) as an index of the variability in item discrimination.
The data presented in table 1 demonstrate a high level of internal consistency reliability (α = .88). The majority of items correlated well with the total of the remaining items but with sufficient variability to reflect the width of the construct being accessed. The item discrimination varies between 23% and 95% for the positive items, and between 7% and 58% for the negative items, demonstrating a wide and helpful range. On the second administration, similar results were reported with an alpha coefficient of .91.
-insert table 2 about here -The second step in data analyses employed the t-test to examine the statistical significance between scores recorded on the two occasions. The data presented in table 2 demonstrate a significant increase at the .001 level of probability from a mean score of 102.4 at time one to a mean score of 103.7 at time two.

Discussion and conclusion
The present study set out to assess the extent to which extant research findings could sustain the claim that cathedrals act as agents of psychological health and wellbeing within secular societies. Drawing on available data from the evolving science of cathedral studies, the body of extant research distinguished between two different constituencies with which cathedrals engage, the constituency rooted in the deeply religious and Christian character of cathedrals, and the constituency rooted in the broader cultural and heritage character of cathedrals. At one level these two constituencies may be characterised as concerned with worshippers and with tourists, but neither of these categories stands without further clarification. According to Williams, Francis, Robbins, and Annis (2007) tourists may include both religious pilgrims and secular tourists. Extant empirical research suggests that those visiting cathedrals may perceive some beneficial impact on aspects of psychological health and wellbeing (Winter & Gasson, 1996) although such impact may be more beneficial for religiously motivated pilgrims than for secular tourists (Williams, Francis, Robbins, & Annis, 2007), or for Christians rather than for atheists (Ysseldyk, Haslam, & Morton, 2016).
Clearly further more detailed and more finely calculated research is needed among visitors to cathedrals in order to test the psychological health and wellbeing hypothesis.
In terms of worshippers, the science of cathedral studies has identified three distinct constituencies. The first constituency concerns those who attend services on a regular basis.
Such people are continuous with, but in some ways different from those who attend parish churches (Lankshear, Francis, & Ipgrave, 2015;. Extant empirical research suggests that the depth of bonding and bridging social capital within cathedral congregations may enhance both psychological health and wellbeing for the participants and social wellbeing for the local community (Francis & Williams, 2015a).
Clearly further more detailed research is needed among regular cathedral congregations to test the psychological health and wellbeing hypothesis.
The second constituency concerns those to whom ap Siôn (2015a, 2015b, 2017) refers as 'hidden congregations'. Ap Siôn's own research explored the community of people who return time and time again to post prayers on the prayerboard or to light candles. Ap Siôn's analysis of the content of prayers posted for public reading suggests that this mechanism enhances wellbeing among the user communities. The term 'hidden congregations' may also be applied to the community of cathedral friends researched by Muskett (2015). Muskett concludes from her research that Friends' associations generate an enhanced sense of social capital that may in turn lead to enhanced wellbeing. Clearly further more detailed research is needed among the hidden congregations to test the psychological health and wellbeing hypothesis.
The third constituency concerns the kind of service that Walker (2017) styles as 'event', contrasting with the 'activities' that take place every week. Walker argues that some people belong to the Anglican Church through activities while others may belong through events. The 'events' best researched within the science of cathedral studies are the pre-Christmas carol services as evidenced by Walker (2012aWalker ( , 2012bWalker ( , 2013Walker ( , 2015aWalker ( , 2015b. Within the group of studies concerned with carol services, Francis, Jones, McKenna, Pike, and Williams, under review) found that over half of the participants at a pre-Christmas service in Liverpool Cathedral reported that the service had enhanced their sense of wellbeing over the four domains embraced by Fisher's (1998Fisher's ( , 2016 model of spiritual health. From our analysis of this study we concluded that the research design was weak, in the sense that it relied on the participants' feeling that the service had acted as an agent of psychological health and wellbeing for them. In order to address this weakness the present study was designed as a repeated-measure study. Participants completed the Oxford Happiness Questionnaire (Hills & Argyle, 2002) while they were waiting for the service to being and again toward the end of the service during a five-minute period of organ improvisation. This repeated-measure study found a significant increase in scores recorded on the Oxford Happiness Questionnaire between time one and time two.
This new repeated-measure study may offer the strongest evidence currently available within the science of cathedral studies for the thesis that cathedrals act as agents of psychological health and wellbeing within secular societies. This study, however, is not itself beyond criticism. Valid criticisms may question the adequacy of the Oxford Happiness Questionnaire as a relevant measure of psychological health and wellbeing; may question the value of what may be a short-lived change in psychological health and wellbeing as measured toward the end of the service; may question the simplicity of the statistical analysis; and may caution against generalisation on the basis of one location and one highly specific expression of a pre-Christmas cathedral event. These findings, nonetheless, are sufficiently intriguing for the study to deserve replication and extension within other locations and other contexts.
While the present study has framed the Oxford Happiness Questionnaire to focus on assessment of psychological health and wellbeing over two well-specified periods of time (comparing how participants have felt over the past week, and how they are feeling now), a subsequent replication study may wish to take a differently nuanced approach by applying the same prompt on both occasions referring to a longer period of time. For example, the prompt, 'How I have felt over the past six weeks about me' (posed both before the service and toward the end) would detect the extent to which participation in the service may have re-tuned evaluation of recent recalled experience of psychological health and wellbeing. Note: r = correlation between individual item and the sum of the other items; % = percentage of agree and agree strongly responses; * these items are reverse coded to compute r but not %