Anna Sutton ✉
(School of Psychology, University of Waikato, Hamilton, New Zealand)
(The Green Monkey, Amsterdam, Netherlands)
Group coaching has potential to enhance individual and organisational outcomes but demonstrating its effectiveness remains difficult. Practitioners are often caught between trying to demonstrate ROI and knowing that many of the benefits to clients are not captured by this bottom-line. In this paper, we evaluate a group coaching programme by assessing its impact on self-awareness development and work engagement. Using a pre- and post-test design, we find that group coaching improves self-awareness outcomes (reflective self-development, acceptance of self and others, proactivity at work) and employee engagement. Further, coaching sessions provided a supportive environment in which to explore the sometimes-challenging emotions associated with increased self-awareness.
self-awareness, group coaching, engagement, coaching
Accepted for publication: 11 January 2022
01 February 2022
© the Author(s)
Published by Oxford Brookes University
Workplace coaching is now well established as a means of improving organisational and individual-level outcomes, including financial results and goal achievement, well-being and problem-solving, and specific skills such as leadership or technical skills (Jones, Woods, & Guillaume, 2016). While coaching has traditionally involved individual one-on-one sessions between a coach and a client, there is increasing interest in the development and evaluation of group coaching. Group coaching is described as having significant potential but also some complex challenges in evaluation and research (O’Connor, Studholme, & Grant, 2017). These include the complexity of dealing with group dynamics and providing clarity to multiple stakeholders, as well as coaches needing considerable group facilitation skills in addition to their individual coaching skills.
Empirical studies on group coaching are still rather scarce (Kets de Vries, 2014), but have started to build a picture of the more complex dynamics involved in group as opposed to individual coaching. The benefits of group coaching often centre around the development of social capital and networks, making use of group dynamics and group trust to enhance outcomes. For example, a recent study of executive group coaching for women demonstrated the value that coachees placed on the sense of belonging to the group and fostering connections (Bonneywell & Gannon, 2021). This building of connections results in informal networks beyond the formal coaching sessions that may provide continuing development (Gray, De Haan, & Bonneywell, 2019) and the group itself may act as a catalyst for change (Nacif, 2021). Indeed, group coaching has recently been conceptualised in terms of critical pedagogy as a uniquely beneficial approach to coaching that can lead to collective empowerment and social justice change (Meldrum, 2021).
Utilising the group’s knowledge and experience helps individuals develop their goals and solutions (Mühlberger & Traut-Mattausch, 2015). Group coaching allows coachees to ‘benchmark’ their experiences with others, as well as deepen their learning and reflection by discussing with others in a supportive environment (Gray et al., 2019). Perhaps particularly valuable from an organisational perspective are findings that show group coaching helps coachees to think about strategic level issues (Gyllensten, Henschel, & Jones, 2020) and to adopt a more systems-level understanding (Gray et al., 2019).
Group coaching clearly has significant benefits but does not come without drawbacks. Some have criticised the move towards group coaching as nothing more than a cost-saving device. Certainly, group coaching can be a lower cost and also, by its very nature, means that less time is given to individual coachees (Lawrence, 2019). There may also be reluctance to discuss confidential issues in a group (Mühlberger & Traut-Mattausch, 2015). A further challenge in evaluating group coaching is in clarifying what is meant by the term.
Here, we report on an evaluation of group rather than team coaching where participants, although drawn from a single larger team in the organisation, were not interdependent nor engaged in coaching with a single shared goal. Similarly to Nacif (2021) we define group coaching as a collaborative small-group process taking place over a limited time period, using coaching principles to work with a group of individuals on their personal goals. As such, it is more than simply a group facilitation of activities or a series of exercises in a workshop. Alexander et al. (2020) suggest that the purpose of coaching can be summarised as the creation of a self-directed learning environment and therefore, to make best use of the empowering potential of group coaching, coachees should be offered a space for active co-learning (Meldrum, 2021).
There are as yet relatively few evaluations of group coaching interventions (O’Connor & Cavanagh, 2016) and the majority of those published focus on the positive outcomes of these interventions, with little consideration of the possible downsides or negative outcomes from group coaching. In this paper, therefore, we measure the potential benefits as well as drawbacks of group coaching in an effort to contribute to the literature on effective group coaching. We focus on two main variables: self-awareness outcomes and work engagements.
Self-awareness can be defined as “the extent to which people are consciously aware of their internal states and their interactions or relationships with others” (Sutton, 2016, p. 646) and is widely held to be an essential component of effective coaching. Coaches’ self-awareness, for example, has been identified by both coaches and clients as crucial to the formation of an effective coaching relationship and the development of self-awareness is key to a coach’s self-development process (O’Broin & Palmer, 2010).
Self-awareness is not just a valued outcome in itself, it has long been recognised as a necessary tool for any kind of self-development (Fenigstein, Scheier, & Buss, 1975) and a key element of specific interventions or practices designed to improve client outcomes. Improved awareness of the self is, for example, the root of better self-evaluations, which are key to building self-esteem (Dinos & Palmer, 2015). The benefits of executive coaching are understood to be based on improved self-awareness in the client as well as an increased ability to develop and achieve their goals and there is evidence that the sense of belonging and connection that is developed in group coaching can be particularly useful in enhancing both confidence and self-awareness (Bonneywell & Gannon, 2021).
There is substantial evidence from qualitative studies in a range of contexts that coaching results in increases in self-awareness. For example, a coaching programme delivered to educators was reported to lead to improved self-awareness for both the trainee coaches and their students (Patti, Holzer, Brackett, & Stern, 2015). Clients who were contemplating a career change described improvements in self-awareness, particularly around their strengths, and this was associated with an increase in self-worth (Archer & Yates, 2017). Similarly, executive coaching which aimed to prepare deans for their leadership roles within a university setting was also found to enhance self-awareness, an outcome that was particularly valued by the clients (Bertrand, 2019). And although the research on group coaching is still developing, there is evidence that it too is associated with increased self-awareness in coachees (Nacif, 2021). Improved self-awareness from group coaching sessions is reported to help coachees make changes in their behaviour, for example in becoming braver leaders or adopting clearer and more direct communication styles (Gyllensten et al., 2020).
The development of self-awareness is not always a smooth path, however. Referred to sometimes as the self-absorption paradox, increased self-knowledge may also be associated with negative affect or psychological distress (Trapnell & Campbell, 1999). Becoming more aware of oneself and interactions with others often involves accepting negative information as well as positive. For example, a coachee may come to the realisation that a fear of failure has resulted in the loss of promotion opportunities or feel guilty that their exclusively task-focused approach to work has upset some of their team members. These negative emotional responses may be particularly prevalent in group coaching, which has been described by coachees as sometimes feeling like ‘an emotional roller coaster’ that can lead to feeling vulnerable (Gray et al., 2019). Improved self-awareness, therefore, may result in both positive and negative outcomes: a sense of being “sadder but wiser’.
But measuring these outcomes of self-awareness in a way that can be generalised across different settings and contexts can be a challenge. The self-awareness outcomes questionnaire (SAOQ) was developed in response to this need for a quantitative measure of the emotions, behaviours and cognitions associated with improved self-awareness (Sutton, 2016). It distinguishes between four groups of outcomes (reflective self-development, acceptance of self and others, proactivity at work, emotional costs) and allows practitioners or researchers to evaluate the effectiveness of coaching interventions. The reflective self-development subscale measures the extent to which someone has developed a sense of attention to the self, using reflection to aid conscious, balanced learning about the self. The second subscale, acceptance of self and others, captures outcomes related to improved confidence and positive view of the self as well as increased understanding and acceptance of others. The SAOQ also includes a scale specific to work, representing a proactive approach to work and the ability to take a more objective view of workplace issues. These first three scales represent the beneficial outcomes of self-awareness while the final scale represents the common, though less desirable outcome of emotional costs associated with increased self-awareness, including a sense of guilt, fear or vulnerability. By including scales to measure both desirable and undesirable outcomes, the SAOQ can be used to evaluate the extent to which an intervention is able to support clients to greater self-awareness without causing undue psychological distress.
Ideally, a well-designed group coaching programme would aim to increase the positive outcomes of coaching in a supportive environment that protects against emotional costs. We therefore hypothesise that:
H1a: The group coaching intervention will increase reflective self-development outcomes.
H1b: The group coaching intervention will increase acceptance of self and others outcomes.
H1c: The group coaching intervention will increase proactivity at work outcomes.
H1d: The group coaching intervention will decrease emotional costs.
Besides the impact of group coaching on individual self-awareness and its associated outcomes, it is often important for coaches to be able to demonstrate that their practice results in outcomes understood and valued by the organisation. Grant (2012) noted that ROI is an unreliable measure of coaching effectiveness because of the difficulties in accurately calculating both the costs and the financial benefits of coaching. Instead, he called for a focus on more holistic measures, including well-being and engagement outcomes, to help identify the broader benefits of coaching.
Engagement can be defined as having energy and enthusiasm for work while demonstrating focused effort (Reis, Trullen, & Story, 2016) and is highly prized by many work organisations because of its association with both employee well-being (Sutton, 2020) and performance (Saks & Gruman, 2014). At the individual coaching level, strengths-based micro-coaching (Peláez, Coo, & Salanova, 2020) has been shown to improve employee engagement. Group coaching research, although scarce, has been shown to enhance well-being in terms of increased meaning, positive emotions and new perspectives (Nacif, 2021). And burnout, which is often conceptualised as a contrast to work engagement, has been demonstrated to reduce following a group coaching intervention for physicians (Alexander et al., 2020). We therefore hypothesise that a group based coaching intervention based on a strengths-based approach will have similar positive outcomes to the individual coaching interventions:
H2: The group coaching intervention will increase employee engagement.
In summary, in this paper we evaluate the effectiveness of a group coaching intervention. Entitled “Becoming your own CHO (Chief Happiness Officer)”, the programme had two main goals: to increase clients’ self-awareness and self-regulation, and to help them develop a personal development and happiness plan. Similarly to the process outlined by Schippers and Ziegler (2019), the intervention adopted a “life crafting” approach whereby participants identified their passions and values and developed a plan to utilise them more. In its structure and content, the programme adopts a generalised version of the model of collective career coaching proposed by Meldrum (2021), leading coachees through an iterative process of developing a vision, working through activities to identify barriers and develop critical awareness, co-operatively exploring options and steps, and transitioning to planned action. This model emphasises empowerment through positive group dynamics and support along with a shared understanding of the context in which coachees work.
The evaluation involved testing the effect of the group coaching on self-awareness outcomes and engagement using a pre- and post-test design. The chosen scales measure outcomes of benefit to the individual (improved self-image and confidence), to their colleagues and peers (increased understanding of others) and to the organisation as a whole (improved engagement).
A-priori power analysis using GPower (Faul, Erdfelder, Lang, & Buchner, 2007) indicated that a sample size of 23 would be sufficient to detect a large effect size in this pre- and post-test design (power = .95). Previous coaching interventions, ranging from videoconference interactions (Rosen, Berrios-Thomas, & Engel, 2016) through to individualised 12 week Rational Emotive Behavior Coaching with emergency workers (Wood, Wilkinson, Turner, Haslam, & Barker, 2021) have reported large effect sizes for changes in concepts related to the self-awareness outcomes measured here. For example, patients in the first study increased in confidence and self-knowledge and the emergency workers in the second study demonstrated a large effect size reduction in irrational beliefs about their work performance directly after the coaching intervention.
Although we initially aimed to recruit 30 participants to allow for drop out or missing data, recruitment was reduced due to COVID-19 restrictions towards the end of the study. Twenty-four employees (14 female, 10 male) from a large multinational financial company volunteered to take part, with 23 providing full data, and thus the study is adequately powered. Participants were highly educated, with university level education in a range of areas including Systems Engineering, Organisational studies and International Business Administration. They represented a range of career experience, with some having joined the organisation directly from university while others had worked elsewhere before, and with job tenure ranging from a few months to fifteen years.
Participation in both the coaching programme and the evaluation research was voluntary. The programme, designed and delivered by the second author, was advertised to members of a process improvement team and those who wished to take part signed up.
The training programme consisted of two, day-long interactive coaching workshops, separated by a minimum of a week with some pre-work before day 1 and homework exercises between day 1 and 2. The content was a combination of theory, practical examples, group and individual exercises, and individual coaching. Further details are provided in the Appendix.
Clients were informed of the nature of the research and assured of anonymity and confidentiality of results. If they wished to participate in the research, they completed questionnaires immediately at the beginning of the first training day (the pre-test measurement) and then again at the end of the second day (the post-test measurement). Of those who attended the coaching sessions, 24 completed the first survey and 23 completed the second.
Self-awareness outcomes were measured using the self-awareness outcomes questionnaire (SAOQ: Sutton, 2016), a 38 item self-report questionnaire with four subscales representing three beneficial outcomes (reflective self-development (RSD), acceptance (Acc) and proactivity at work (Pro)) and one negative outcome (emotional costs (Costs)) associated with self-awareness. Example items include “I learn about myself and how I see the world” (RSD), “I have compassion and acceptance for others” (Acc), “I see my work life as something I have power to affect” (Pro), “I have had to revisit difficult past experiences” (Costs). Items are scored on a 5 point scale of frequency from 1 = Never to 5 = Almost always and scale scores are calculated as the mean of the item responses.
Work engagement was measured using the nine-item Utrecht Work Engagement Scale (UWES: Schaufeli and Bakker, 2003). An example item is “At my work, I feel bursting with energy” and items are scored on a scale from 0 = Never to 6 = Always/Every day. Scale scores were calculated as the mean of item responses.
Participants were also given the opportunity to provide an open-ended response to the question “Have you noticed any changes in your life as a result of this training program? If so, please describe them here.’
Scale reliability was checked using Cronbach alpha. The RSD, costs and engagement scales had satisfactory reliability at both pre- and post-test (α> .7). Reliability was slightly lower for the pre-test acceptance scale (α =.65), but still acceptable. The proactive scales had lower reliabilities at both time points (α = .47 and .59 at pre- and post-test respectively). Although lower than commonly accepted cut-offs, we have retained the scales in this study for three reasons First, the sample size in this study, although adequately powered for the detection of within-group change, is not large enough to justify the amendment of a previously published measure with good psychometric properties. Second, as Taber (2018) notes, although .7 is often used as a cut-off for acceptability, alphas above .4 are also described as “acceptable” and “sufficient” in the literature. Finally, the scales were developed to assess the breadth of outcomes associated with self-awareness and while deleting some items might improve reliability, the scale would no longer properly represent these broad outcomes (see Schmitt, 1996). (However, given that low reliabilities can lead to under- or over-estimating effects, we did also check the robustness of the findings by repeating the analyses reported below using revised acceptance and proactive scales. With items removed from each scale to improve reliability, we found the same pattern of results.)
As responses to the open-ended question were relatively brief, we provide a simple thematic content analysis to give context and depth to the quantitative findings.
Table 1 shows the descriptive statistics for all variables (mean, SD, Cronbach alpha for each scale as well as the bivariate correlations). To evaluate whether the coaching programme had the hypothesised effects on participants, paired sample t-tests were conducted on the pre- and post-test scores. Assumption checks revealed that the difference scores for the acceptance and proactive at work scales were non-normal, therefore we used the Wilcoxon signed-rank test instead of a t-test for these two scales.
There was a significant improvement in the positive outcomes related to self-awareness, with increases in pre- to post-test scores for reflective self-development (t(22) = -5.98, p < .001, d = -1.25), acceptance (Z = -4.2, p < .001) and proactivity at work (Z = 3.94, p < .001). The mean of positive ranks for the acceptance scale was 12 and the mean of negative ranks was 0, while for proactivity at work these values were 10.5 and 0 respectively, both indicating large positive effects. Notably, participants reported a significant decrease in the emotional costs associated with self-awareness (t(22) = 3.93, p < .001, d = .82). Participants also reported an increase in engagement levels (t(22) = -5.74, p < .001, d = -1.20). Figures 1 and 2 illustrate these changes. All of these effect sizes can be considered large using Cohen’s (1988) criteria. All hypotheses were supported: the group coaching intervention improved positive outcomes while decreasing negative outcomes.
Note. RSD = reflective self-development, Acc = acceptance, Pro = proactive at work, Cost = emotional costs, Eng = engagement* p<.05, **p<.01, ***p<.001
Note. RSD = reflective self-development, Acc = acceptance of self and others, Pro = proactive at work, Cost = emotional costs.
Twenty-one out of the 23 participants who completed the final questionnaire provided a brief response to the open-ended question asking about changes they had noticed as a result of the coaching sessions. Thematic analysis resulted in the identification of three main themes (awareness, tools and concrete plans, positive attitudes and approach), summarised in Table 2 and discussed below with illustrative quotes from participants. Participant numbers are given in brackets after each quote.
Reflecting the quantitative findings that participants reported an increase in reflection and awareness, 12 responses mentioned an increase in their understanding or awareness of themselves: “I am more aware of what is important to me and what gives me energy and motivates me.” (P6) and “I would recommend it [the coaching] to anyone who wants to understand more about why they do what they do, and crystalize it in a meaningful way.” (P16)
This increased awareness was closely tied to a recognition of the specific tools they had learned and how the coaching had helped them to develop concrete plans for the future. Thirteen of the responses mentioned these tools and plans, for example, “the training has given me tools to assess my own emotions and actions” (P7) “I have a clearer image of my future, what gives me energy and I can start applying this in my life” (P1) and “[I have a] more clear view on what I really want to achieve, and how I can make incremental steps each day to make sure I reach my goals” (P12). These responses also included examples of specific steps that participants had already taken to achieve their goals, e.g. “After the training I immediately took a first action: I discussed with my lead a part of my job that I really didn't enjoy and we are now looking at if we can change this” (P24). These comments may provide some insight into why the proactive at work subscale was less reliable in this sample than others: the proactive steps participants were unique and perhaps not adequately captured by the items on the scale.
Several participants noted that they had developed a greater appreciation for the here and now alongside this new, more focused view of the future: “I am more aware of the good things I already have” (P9). A few participants noted specifically that they had a greater sense of meaning or purpose “As a result of my training I found my purpose that I verified with multiple people in my environment” (P7) or confidence “I have a more positive and confident mindset about my opportunities and ambitions” (P2).
The statistical analysis demonstrated that, despite the potential for increased feelings of vulnerability or emotional costs as a result of exploring personal issues in coaching, emotional costs actually decreased in this sample. A possible explanation for this may be found in the following quote “It was really moving on a personal level – and having a space to develop a purpose and motivation (and be vulnerable) was amazing.” (P16). This response implies a recognition that vulnerability is an important step in self-development but also that this coaching experience provided a safe space in which to do that exploration.
This study evaluated a group coaching intervention that aimed to improve clients’ self-awareness and self-regulation, and help them to enact a personal happiness and development plan. Results demonstrate that group coaching resulted in substantial improvements in the positive outcomes of self-awareness (reflective self-development, acceptance of self and others, proactivity at work) as well as reductions in emotional costs. The coaching also significantly enhanced employee engagement levels, which is likely to be of particular interest to organisations as engagement is known to be associated with work performance (Bakker, 2011).
Clients noted the importance of both their improved self-awareness and their learning of specific tools and techniques in developing their concrete action plans for improving happiness. Further supporting the quantitative measurements, clients reported a greater appreciation of their current situation and improved confidence, as well as highlighting the importance of having the supportive space to explore their self-development. Previous work has noted that increased self-awareness can lead to improved self-esteem and confidence (Dinos & Palmer, 2015) and this was confirmed in both the quantitative and qualitative results here.
Using a positive psychology approach to coaching involves developing practical interventions to improve well-being, such as coaching for happiness. But positive psychology is sometimes critiqued for taking “positivity” too far and avoiding any consideration of downsides or negatives (Held, 2004). In evaluating this group coaching intervention, therefore, we utilised a measure that specifically assesses the potential costs of improved self-awareness in order to provide a balanced evaluation of the effects. Increased self-awareness is often associated with emotional costs (Sutton, 2016) but this study demonstrates that effective coaching may mitigate against this. This group coaching intervention was able to improve clients’ positive outcomes while protecting against and even reducing the emotional costs.
This study also provides further evidence that the self-absorption paradox associated with increased self-awareness may be disentangled. Although previous studies have focused on personality variables (Trapnell & Campbell, 1999) or need for absolute truth (Simsek, 2013) to explain the paradox, we have shown that effective coaching can also encourage objective reflection while avoiding negative emotional costs. This demonstrates that the development of self-awareness must not necessarily be accompanied by negative outcomes and that effective coaching is key to this process.
As one of the first evaluations of group coaching, this study provides coaches and researchers with evidence that group coaching has measurable, positive outcomes. Group coaching can be complex and challenging but holds significant potential (O’Connor et al., 2017) and this study certainly provides evidence of some of this potential. The group coaching improved outcomes valuable to the individual (such as improved self-image and confidence), to their wider work colleagues (increased understanding of others) and to the organisation (improved engagement). These outcomes go beyond a simple ROI calculation and answer Grant’s (2012) call for a wider understanding of coaching effectiveness.
To avoid creating a sense of being “sadder but wiser” as clients develop their self-awareness, coaches should work to create a supportive environment, a safe space in which clients can explore the sense of vulnerability that often results from introspection and self-exploration. In a group coaching scenario, this will involve not just developing trust between the coach and client, but also effective management of the group dynamics so that peers are involved in creating this support. By doing this, coaches can ensure that their clients are able to reap the rewards and even reduce the typical emotional costs associated with increased self-awareness.
It is important to recognise that self-development often involves courage to explore potentially distressing events or feelings (Sutton, Williams, & Allinson, 2015). We have found in this study that group coaching can provide a safe place for clients to do this, meaning that, rather than feeling greater costs as a result of their improved self-awareness, clients may actually report lower costs.
Of further note, the changes in clients’ self-awareness outcomes and engagement were not just statistically significant but had large effect sizes, implying substantial practical benefits and “real life” changes. Taken together, these results demonstrate the value of group coaching and can help to justify investment in these programmes.
As is often the case with intervention research which requires significant input from coaches, this study relied on a relatively small sample and awaits replication in future studies. It is our hope that the effect sizes found here will encourage other coaches and researchers to make this investment and continue to develop the evidence base for coaching. The small standard deviations we found indicate a fairly homogenous sample in terms of outcomes, and a larger sample would be able to explore individual differences in response to the coaching intervention in more detail. In addition, the use of a randomised control group design, although challenging to do in applied settings, is needed to truly identify the effects of the intervention.
The proactive at work subscale did not reach expected reliability levels in this sample and therefore the results related to this scale should be interpreted with caution. Qualitative comments indicated that clients were taking proactive steps at work, but that these were specific to their individual situation. It may be, therefore, that the scale items were not relevant enough to the individual’s context to represent proactivity adequately. Further research on this scale would be welcome.
The qualitative comments also highlighted a common theme of clients becoming more aware of and appreciative of the “here-and-now”. This is a distinctive element of mindfulness (Brown & Ryan, 2003) and we therefore recommend that future evaluation of this kind of happiness-focused group coaching may do well to include a measure of mindfulness.
Although we took steps to ensure the standardisation of approach and consistency across the different coaching groups, there remains the possibility that there were inter-group differences. Responding to the group dynamics is a key part of effective coaching, however, and complete standardisation is undesirable. We await further research using similar group-based coaching interventions to support or challenge these results. Similarly, although we used a pre-post design, the long-term effect of this group coaching intervention is not known. Examining whether these positive effects would be maintained, and what factors might influence this maintenance, weeks or months after the original coaching sessions would be a rewarding avenue for further research.
Group coaching is an effective means of improving outcomes of value to both the individual and the organisation. Evaluating coaching programmes using measures, which assess engagement and outcomes associated with self-awareness can provide coaches with simple, practical ways to demonstrate the value of their work to potential employers and clients.
The second author was employed by the organisation where she delivered the coaching programme reported in this paper but the research itself was not funded.
Anna Sutton is a Senior Lecturer in Organisational Psychology. Her research focuses on self-awareness, personality and authenticity at work, with an emphasis on improving employee well-being.
Cecile Crobach is founder of the organisational coaching and development company the Green Monkey. She specialises in coupling people development with data-driven process optimisation efforts.
• Participants will identify their personal values, their strengths, their drivers and energizers and their purpose. • Participants will gain insight into how their emotions work and what happens in their brain to better understand how emotions can influence their interaction with others. They will learn how they can influence their own emotions. • Participants will create a concrete individual development and happiness plan.
Groups of between 6-8 clients took part in a group coaching intervention. It consisted of 12 hours across 2 day-long sessions, with a minimum of one week between the sessions. In addition, participants were sent pre-work to complete as preparation for the first day and given homework between the first and second day.
The programme is a combination of theory, practical examples, group exercises and individual exercises. Through these elements, the coach provided the structure and support for clients to develop their goals and plans for attaining them. The coach provided further individual or group support and coaching as needed through the days.
The following figures provide some further information on the content of the pre-work, homework and face-to-face sessions. Further details may be requested from the second author.