Can one-to-one coaching improve selection success and who benefits most? The role of internship candidate generalised self-efficacy

ABSTRACT

This study presents a field quasi-experiment to examine whether one-to-one coaching can significantly impact an objective outcome: selection success. Furthermore, we examine whether all participants benefit equally from coaching. We examine whether coaching significantly impacts on selection success (i.e., whether candidates were successfully offered a job role) (N = 357) and whether interactions between coaching and generalised self-efficacy exist (n = 152). Participants were second year undergraduate students in an UK university who were applying for a one year work internship. Our analysis indicates that coaching positively impacted on selection success and specifically, individuals lower in generalised self-efficacy benefitted from coaching. We contribute to the literature on coaching by examining the effectiveness of coaching in relation to an objective outcome. We also advance our understanding of the role of individual differences in coaching by testing the interaction effects for a well-recognised individual difference variable: generalised self-efficacy. Finally, we contribute to the literature regarding the design of career development support in terms of how best to support people based on their individual differences.

KEYWORDS:

• Coaching
• coaching effectiveness
• learning and performance
• generalised self-efficacy

Practice points

This article provides evidence of the efficacy of coaching, enabling coaches to justify the investment organisations and individuals make in coaching. It also provides further insight into the individual difference factors that influence who benefits from coaching, helping coaches to understand how best to support their clients. Finally, it highlights how coaching can be used in a particular context.

• Coaching can impact on objective criterion, strengthening its value as an intervention.

• Coaches may assess coachees’ generalised self-efficacy to tailor their coaching, enhancing effectiveness.

• Coaching can enhance selection performance, supporting its use as a career development tool.

Can coaching positively impact on objective criteria and who benefits the most from coaching? If we knew the answers to these questions then leaders, HR managers and coaches could make informed decisions regarding when and for whom coaching is an effective developmental intervention. The impact of coaching has been explored on a wide variety of outcomes (see Graßmann et al. (2020) for a review) however, there is still a shortage of empirical evidence to validate the impact of coaching on objective criteria, a gap that has been highlighted by a number of scholars as requiring urgent attention (Athanasopoulou & Dopson, 2018; Jones et al., 2016). This paper addresses this gap by examining the impact of coaching on an objective outcome: selection success.

If establishing that coaching works to influence objective outcomes is step one, step two is to provide a more nuanced understanding of the factors influencing the success of a coaching engagement, and this includes aspects of the coachee. There are gaps in our knowledge of who benefits most from coaching (Bozer & Jones, 2018; Jones et al., 2021). While the impact of individual differences has been explored extensively in a range of learning and development contexts (Bell et al., 2017; Colquitt et al., 2000), attention has only recently turned to exploring the role of individual differences in coaching effectiveness (De Haan et al., 2016, 2019; Jones et al., 2014, 2021; Stewart et al., 2008; Terblanche & Heyns, 2020). In this paper, we contribute to this growing area of research and investigate the impact of generalised self-efficacy (GSE) on the effectiveness of coaching.

In doing so, we make three important contributions to the literature. Firstly, within the specific coaching effectiveness literature, we provide one of the few studies to examine the impact of coaching on an objective performance criterion within a field setting. Secondly, we advance our understanding of the role of individual differences in the effectiveness of coaching, an important step in developing the relatively nascent literature on coaching. Finally, we contribute to the literature on career development beyond traditional interventions, such as CV writing and interview skills workshops, by exploring the impact of one-to-one coaching on selection success in a job search context.

Selection success

An important objective criterion within the work context is an individual’s ability to be successfully hired during a selection process, otherwise known as selection success. The majority of research that has explored the impact of interventions on selection success has generally focused on traditional career advice functions, such as interview and CV writing workshops (Taylor & Hooley, 2014) and practical activities such as networking and gaining experience (e.g., Jackson & Bridgstock, 2021; Jackson & Tomlinson, 2021). While this research has identified how best to develop the necessary job application skills, selection success also relies heavily on a number of interpersonal characteristics, for example, having the self-confidence and self-belief to be able to sell oneself effectively throughout the selection process (Manroop & Richardson, 2016).

Coaching as a developmental intervention is ideally placed to develop such interpersonal skills, as one-to-one coaching can be defined as a learning and development intervention that uses a collaborative, reflective, goal-focused conversation to achieve inter and intra-personal professional outcomes that are valued by the coachee (Bono et al., 2009; Smither, 2011). Research suggests that outcomes from coaching include increased self-awareness (Carter et al., 2017); self-confidence (Swart & Harcup, 2013) and resilience (Grant et al., 2010; Sardar & Galdames, 2018). However, the coaching literature has yet to investigate the impact of coaching on the important objective outcome of selection success. Therefore, to address this gap, we explore whether coaching enhances the chances of selection success for second year, full-time, undergraduate students seeking a 12-month work internship. We predict:

Hypothesis 1. Participation in one-to-one coaching will be associated with increased selection success.

The influence of generalised self-efficacy on the impact of coaching

In addition to exploring the impact of coaching on selection success, we also seek to explore who benefits most from coaching. Researchers have begun to explore the role of individual differences in the impact of coaching, such as the Big Five personality characteristics (Jones et al., 2014, 2021; Stewart et al., 2008; Terblanche & Heyns, 2020). These studies indicate that the topic of coachee individual differences is an avenue worthy of further exploration.

A useful framework provided to explain the role of individual differences in learning that can be applied to the coaching context is attribute-treatment interactions (ATIs). Attribute-treatment interactions suggest that individuals possessing certain characteristics may excel in one learning system yet struggle in another (Eysenck, 1996). Evidence suggests that attribute-treatment interactions are present in error training (Cullen et al., 2013; Gully et al., 2002); discovery learning and microteaching (Eysenck, 1996); computer-delivered training (Brown, 2001); psychomotor skills training (Herold et al., 2002) and e-learning (Orvis et al., 2010). Gully and Chen (2010) explain the theoretical reasoning behind the observed relationships between individual differences and learning outcomes. They propose that trainees actively regulate their motivation, emotion and learning processes. As such, trainees decide: what to attend to and determine how much effort to devote to the learning task; they actively engage or disengage from training and they are responsible for applying and transferring skills from training to the work environment. Individual differences influence these regulatory and motivational processes that determine whether trained content is learned, retained, applied and transferred. We aim to contribute to this literature by exploring whether GSE influences the impact of coaching on selection success.

Social cognitive theory highlights self-efficacy as a central mechanism with a wide explanatory power on diverse phenomena (Bandura, 1986). Researchers distinguish between perceived self-efficacy, which is conceptualised as a relatively malleable, task-specific belief and GSE, which is a relatively stable, trait-like generalised competence belief (Chen et al., 2000). In the present study, we are interested in the concept of GSE as Chen et al. (2000) propose that the concept of GSE is most useful when the performance under scrutiny is generalised. Our criterion is selection success and as such captures the participants’ performance in a range of activities including completion of application forms, interviews, and assessment centres. For many participants in the present study (undergraduate students), this will be the first time that they have participated in formal selection activities such as these. Therefore, the participants will have no previous experience from which to draw task-specific self-efficacy beliefs (Bandura, 1997). As such, we propose that their GSE beliefs are likely to be a more important influence on their performance in this context.

Research on GSE indicates that individuals higher in GSE have strong beliefs in their capabilities and set more challenging goals than those with lower GSE (Eden, 1988). Higher GSE is associated with investment of cognitive efforts and superior learning (Kyndt et al., 2016) and a lack of GSE has been linked to failure to fully realise individual career potential (Petruzziello et al., 2020). We therefore propose that GSE has important implications for selection success and warrants further exploration.

A handful of studies have explored the role of pre-training self-efficacy and training outcomes (Eden & Aviram, 1993; Gist et al., 1989, 1991; Stevens & Gist, 1997). Based on these studies, Fan and Lai (2014) argue that when teaching and learning supports the development of psychological resources needed to benefit from learning activities, individuals lower in self-efficacy benefit to a greater extent from this provision than those with higher self-efficacy, who may already possess these psychological resources. We seek to explore whether a similar effect is present for GSE and coaching as we propose that coaching is likely to similarly build coachees psychological resources.

We propose that individuals with lower GSE may doubt their ability to successfully secure an internship (Chen et al., 2000). The course of action selected is impacted by efficacy, with individuals with lower GSE having lower aspirations and weaker commitment to chosen actions (Kauppila & Tempelaar, 2016). Efficacy may also influence effort allocated to achieving goals (Chen et al., 2000); strategies used (Jiménez Ivars et al., 2014); reaction to failure (DeRue & Morgeson, 2007) and persistence in the face of obstacles (Eden, 1988). Therefore, in the context of pursuing an internship, an individual with lower GSE may limit their aspiration reflected in the number of applications they make, the amount of work they invest in preparing for the selection process and their reaction to rejection from recruiters. Consequently, we predict that coaching will be particularly beneficial for these individuals, as the coach can provide a source of feedback in relation to the level of challenge and aspiration in the goals set, encouraging coachees to challenge themselves (Grant, 2018). The coach may also use questioning techniques to raise awareness and challenge underlying faulty assumptions, consequently the coachee is encouraged to ensure that their self-judgments are based on evidence rather than perceptions of competence (Palmer & Szymanska, 2018). A likely consequence of this is that those coachees who doubt their ability to secure an internship due to lower GSE will experience an increase in selection success beliefs, consequently setting themselves more challenging goals linked to their applications.

Therefore, using the framework of attribute-treatment interactions, we predict that while individuals who are higher in GSE will also benefit from these processes, it is those who are lower in GSE who will experience greater benefits, as these individuals have the greatest developmental needs in these areas. Consequently, we hypothesise that:

Hypothesis 2. Individuals lower in GSE will benefit the most from coaching (indicated by increased selection success).

Method

Research design and participants

This study utilised a non-randomly assigned, quasi-experimental field design, conducted within a UK Business School. Details of second year, undergraduate students who had registered an interest in taking a work internship were gathered at the start of the academic year and formed the participants for this study. To successfully secure an internship, students had to apply to organisations offering internships of at least 36 weeks in duration to undergraduate students. Students then needed to progress through the assessment and selection process put in place by the organisation. This varied across organisations, but typically involved interviews, assessment centres including group activities and work sample tasks and psychometric tests. No internships were supplied by the university. The study was approved by the University of Worcester ethic committee (application CBPS20210005).

A total of 357 participants registered an interest in taking a work internship. The average age of the participants was 20.15 years old (SD = 1.97) and 56% of the participants were male. All participants registering an interest in an internship were offered coaching.

A total of 172 participants took up the opportunity of coaching, whilst 185 did not and therefore formed a comparison group. In the coached group, the number of coaching sessions ranged from one to eight (M = 2.09, SD = 1.31; see Table 1 for descriptive statistics for the whole sample).

Table 1. Descriptive statistics for all variables.

Variable	n	Mean	Median	SD	1	2	3	4	5	6
1. Age	165	20.15	20.0	1.97	–
2. Gender^c	357	–	–	–	.02	–
3. Number of coaching sessions	357	1.01	0.00	1.39	−.06	.13*	–
4. Generalised self-efficacy^a	152	4.15	4.13	.52	.11	−.07	−.06	(.87)
5. Internship commitment^a	159	5.68	5.78	.84	−.20*	−.07	0.01	.35**	(.87)
6. Selection success^b	357	–	–	–	.10	−.09	.16**	.10	.09	–

Notes: * p = < 0.05; **p = < 0.01. α is reported for generalised self-efficacy and internship commitment in parentheses on the diagonal.

^a Higher values indicate a greater degree of the variable.

^b Coded 0 = no, 1 = yes.

^c Coded 1 = male, 2 = female.

To test hypothesis two, all students interested in an internship were sent an internet-mediated questionnaire to complete before coaching commenced including questions on their age, GSE and commitment to securing an internship. Participant information was provided at the start of the questionnaire and before commencing the questionnaire participants provided informed consent electronically. A total of 165 participants returned the questionnaire (46.2% response rate) and 152 questionnaires were completed fully.

A number of checks were conducted to assess the theoretical equivalence of participants who were coached and those who were not. A chi-square test for independence (with Yates Continuity Correction) indicated that males were more likely to participate in coaching than females (χ² (1, n = 357) = 7.53, p = <.05). Direct logistic regression revealed that neither age (χ² (1, n = 165) = 1.887 p = >.05) nor GSE (χ² (1, n = 152) = 1.547 p = >.05) were significant predictors of whether participants were coached or not coached.

To control for equivalence of groups, we also compared the two groups on level of commitment towards an internship, as motivation and commitment have been identified as important predictors of outcomes in the extant training literature (Kraiger & Ford, 2021; Salas & Cannon-Bowers, 2001) and motivation is associated with uptake of personal development activities (Major et al., 2006). A direct logistic regression revealed however that internship commitment did not predict whether participants opted into coaching or not (χ² (1, n = 159) = 2.181 p = >.05). The coached and non-coached group therefore do not differ in age, in terms of the independent variable of GSE or in terms of their commitment to the outcome variable of securing an internship. However, as males were more likely to opt to be coached, this variable was included as a control in all subsequent analyses.

One-to-one coaching intervention

Both authors delivered the one-to-one coaching and to ensure competence in coaching, both authors completed tertiary coaching certification courses. The coaching sessions were structured utilising Whitmore’s GROW (i.e., Goals, Reality, Options, Will) model (2017). GROW has been used in a number of empirical coaching studies (for example Grant et al., 2010). The GROW model is a behaviourist approach to coaching, and works by providing a structured approach to the coaching conversation which allows the coachee to gain an increased awareness of their aspirations, a greater understanding of their current situation, explore the possibilities open to them and the actions they need to take to progress towards achieving their aspirations. The benefit of utilising a structured approach to coaching, such as GROW, in an experimental setting, is that it enables a degree of control and consistency to be applied across the coaching sessions. During the first session the participants’ goals were explored and documented, including agreement on how the participant could assess when they have achieved each goal. Participants then selected which goal they would like to work on first. Using a combination of active listening, Socratic, open questioning and reflecting back, each goal was explored in detail including the participants’ current ‘reality’ in relation to the goal, barriers that may have hindered their goal achievement in the past and the ‘options’ available to them to aid goal achievement. The participant would then agree on next steps they would implement to help them work towards achieving their goal following the coaching session. Each new coaching session would start with an update in which the participant would report their progress on agreed action points. If action points had not been achieved then these would be explored in detail utilising active listening, Socratic, open questioning and reflecting back. As and when the participant felt they had sufficiently explored each goal, attention would turn to a new goal and the same process would be followed.

To further facilitate consistency, throughout the duration of the research, both researchers met to engage in peer supervision and discuss the progress of the intervention (Hawkins & Smith, 2013). This ensured that, as far as possible, the coaching was consistent and accurately represented coaching as defined in this study.

Measures

Selection success. The Internship Manager provided confirmation of students who secured an internship. A total of 114 students (31.9%) secured an internship and 243 students (68.1%) did not secure an internship.

Generalised self-efficacy. Generalised self-efficacy was measured using the Chen et al. (2001) eight item GSE scale. Answer responses were provided on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). An example item is ‘I will be able to achieve most of the goals that I have set for myself’. The reliability coefficient for the present study was α = .87. To ensure that no expectancy bias entered into the coaching relationship the coaches were blind to the students’ level of GSE until after all coaching had been completed.

Internship commitment. Internship commitment was measured using an amended version of the Mowday et al. (1979) organisational commitment questionnaire (OCQ), where the word ‘organisation’ was replaced with ‘internship’. The short version of the scale was used, consisting of nine items with responses measured on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). An example item from the amended scale is ‘I am willing to put in a great deal of effort beyond that normally expected in order to be successful in obtaining an internship’. The alpha reliability coefficient for this sample was α = .87.

Analytical approach

Logistic regression analysis was utilised to assess the impact of the independent variables on selection success. The number of coaching sessions a participant attended was utilised as the independent variable.

To test hypothesis two, the data were split at the median of GSE with separate logistic regressions conducted to assess whether coaching has a differential impact on selection success for those higher and lower in generalised self-efficacy. Whilst it is acknowledged that using two logistic regression analyses to compare groups assumes that heterogeneity is the same across groups (which may not be the case), this assumption also applies to conducting one logistic regression with an interaction term involving a categorical or ordinal variable (Mood, 2010). Conducting one logistic regression including the interaction term of GSE and number of coaching sessions was explored, but was not possible due to unacceptable levels of multicollinearity in the model (Pallant, 2020), even when the variables were rescaled via centring (Aitken & West, 1991). It was therefore deemed appropriate to continue with two separate logistic regression analyses.

Results

Table 1 shows the descriptive statistics for all variables.

Table 2 reports the analysis to test hypothesis one. A hierarchical logistic regression was conducted with number of coaching sessions as the independent variable and selection success as the dependent variable. Gender was included in Block 1 as a control variable.

Table 2. Main effect of number of coaching sessions on selection success.

					95% confidence interval for Exp (β)
	β	SE	Wald	p	Exp (β)	Lower	Upper
Constant	−.83	.17	24.35	.00	.44	–	–
Number of coaching sessions	.26	.08	9.83	.00	1.29	1.10	1.52
Gender	−.48	.24	4.10	.04	.62	.39	.99
Log likelihood	434.52
Degrees of freedom	2
χ²	12.71**
Pseudo R² (Cox & Snell R squared)	.04

The logistic regression revealed that the model was statistically significant χ² (2, n = 357) = 12.71, p = .002, distinguishing between participants who achieved selection success and those who did not. Adding number of coaching sessions to the model made a significant increase to the predictive power of the model χ² (1, n = 357) = 10.02, p = .002. The model explained between 3.5% (Cox & Snell R squared) and 4.9% (NagelKerke R squared) of the variance in selection success. The number of coaching sessions was a significant unique predictor of selection success (β = .26, SE = .08, Wald = 9.83, df = 1) p = .002. The odds ratio was 1.29, indicating that for every additional coaching session, participants are nearly 1.3 times more likely to secure an internship.

Therefore, hypothesis one was supported as the results indicate that receiving coaching significantly increases the likelihood of achieving selection success and that this likelihood increased as the number of coaching sessions increased.

Table 3 reports the results of tests of hypothesis two. The sample was split at the median of GSE (4.13). A total of 84 participants scored below the median value and 68 participants scored above the median value. Two hierarchical logistic regressions were run to assess whether the number of coaching sessions predicted selection success in both the lower and higher GSE samples. Gender was controlled for in each analysis.

Table 3. Interaction effect of number of coaching sessions and generalised self-efficacy on selection success.

	Lower generalised self-efficacy							Higher generalised self-efficacy
					95% C.I. for Exp (β)							95% C.I. for Exp (β)
	β	SE	Wald	p	Exp (β)	Lower	Upper	β	SE	Wald	p	Exp (β)	Lower	Upper
Constant	−1.22	.46	7.03	.01	.29	–	–	−.58	.41	2.00	.16	.56	–	–
Number of coaching sessions	.57	.16	8.43	.00	1.76	1.20	2.58	.26	.16	2.63	.11	1.30	.95	1.76
Gender	−.62	.52	1.44	.23	.54	.20	1.48	−.88	.54	2.67	.10	.42	.15	1.19
Log Likelihood	92.22							83.28
Degrees of freedom	2							2
χ²	11.73**							5.02
Pseudo R² (Cox & Snell R squared)	.13							.07

Utilising number of coaching sessions as the independent variable, when GSE was lower, the model was statistically significant χ² (2, n = 84) = 11.73 p = .003, explaining between 13.0% (Cox & Snell R squared) and 18.4% (NagelKerke R squared) of the variability in selection success. The addition of number of coaching sessions to the model significantly increased the predictive power of the model χ² (1, n = 84) = 10.70 p = .001. Number of coaching sessions was a significant unique predictor (β =  .57, SE = .20, Wald = 8.43, df = 1) p = .004. The odds ratio was 1.76, which indicates that for every additional coaching session, a participant with lower GSE was nearly 1.8 times more likely to secure an internship.

When GSE was higher, the model was not significant χ²(2, n = 68) = 5.02, p = .08, explaining between 7.1% (Cox & Snell R squared) and 9.8% (NagelKerke R squared) of the variability in selection success. The addition of number of coaching sessions did not result in a significant improvement to the model χ² (1, n = 68) = 2.68 p = .10. Number of coaching sessions was not a significant unique predictor for those higher in GSE (β = .26, SE = .16, Wald = 2.63, df = 1) p = .11.⁣

Therefore, hypothesis two was partially supported. We had predicted that all participants would benefit from coaching, although those lower in GSE would benefit the most. Our findings show however that only individuals lower in GSE benefited significantly from coaching.

Discussion

In this paper we sought to address the question: Can coaching positively impact on objective criterion and who benefits the most from coaching? Our findings have implications for the literature on coaching in several ways. Firstly, by contributing to our understanding of the effectiveness of coaching in terms of an objective outcome. Secondly, advancing our understanding of the role of individual differences in the effectiveness of coaching and finally, by elaborating the wider implications for the design of career development support to best suit candidates’ individual differences.

We proposed that coaching as a development technique is particularly suitable for enhancing selection success as it can foster self-confidence, raise awareness of strengths and enhance persistence in the face of rejection, all of which have been shown to influence selection success (Manroop & Richardson, 2016). Our findings supported hypothesis one and indicated that the number of coaching sessions a participant received was a significant predictor of selection success; the more coaching sessions the participants received, the higher the probability that they secured an internship. This finding is important as it provides evidence that coaching can have a positive impact on objective criterion in the form of selection success.

In relation to hypothesis two, we focused on GSE. Using the framework of attribute-treatment interactions, we had predicted that individuals who are lower in GSE would experience greater benefits from coaching and this prediction was based on our reasoning that individuals who are lower in GSE may set fewer challenging goals and may give up more easily following rejection (Nicholson et al., 2013). Working with a coach encourages the candidate to set suitably challenging goals and develop focused action plans to commit attention and efforts towards behaviours that will support the achievement of these goals. Our findings partially supported our prediction, indicating that the impact of coaching was greater for those lower in GSE and selection success significantly improved in line with the number of coaching sessions for this group. This finding sheds light on our understanding of the individual differences that influence coaching outcomes.

Interestingly, we had anticipated that, to a lesser extent, individuals higher in GSE would also benefit from receiving coaching. Our findings indicate that this was not the case, as coaching did not significantly contribute to selection success for participants higher in GSE. Our findings therefore support the conclusions drawn by Fan and Lai (2014) who note that in certain learning conditions, individuals higher in GSE did not appear to benefit, as these individuals already possess high levels of the psychological resources that certain interventions aim to develop. Research findings such as ours highlight that learners should not be considered as a homogenous group in relation to the outcomes derived from learning activities (Cullen et al., 2013).

Practical implications

Our research offers a number of important implications for practice. Our findings indicate that one-to-one coaching is a useful methodology in a career development context. We demonstrated that increasing the number of coaching sessions participants engaged in significantly enhanced their likelihood of securing a 12-month internship. We do however, emphasise that in our view, coaching is only likely to have a positive impact on selection success if candidates engage in coaching on a voluntary basis and are committed to the target outcome (Jones & Andrews, 2019).

Our findings also indicate that individuals who are lower in GSE are most likely to benefit from coaching. Given that learning and development budgets are often restricted and one-to-one coaching can be an expensive developmental intervention, our findings provide important data which can be used to inform evidence-based decisions on who will benefit the most from coaching. Organisations could consider screening employees based on GSE to assist in selecting whom coaching is offered to.

Finally, we proposed that the reason why individuals lower in GSE benefitted more from coaching is because the coach can support the coachee to set challenging goals and in reflecting on, raising awareness of and challenging negative thought patterns. It may therefore be beneficial for coaches to understand the GSE of their coachee prior to the start of coaching. For coachees lower in GSE, it may be particularly important that the coach is prepared to explore and potentially challenge the goals set by the coachee to ensure they provide an aspirational challenge that is likely to motivate behaviour change. It is also likely that when coaching individuals with lower GSE, techniques that facilitate reflection, raise awareness and question faulty assumptions and cognitions are likely to be particularly impactful in terms of assisting individuals with lower GSE to identify strengths and regulate negative thought processes that may have a detrimental impact on their ability to achieve their goals.

Future research

One of the biggest gaps in the coaching literature is related to the shortage of studies utilising objective outcome criteria (Athanasopoulou & Dopson, 2018). Our study has gone some way to address this gap by utilising selection success as our outcome measure, however further research is needed to extend our findings to other suitable objective outcomes.

An important step in coaching scholarship is understanding for whom coaching is most effective. We suggest that future research should continue to address this important area of research by exploring other coachee individual difference variables. Future research in this respect could also start to compare other coaching approaches. In the present study, we structured our coaching intervention using the GROW framework. Future research could explore whether differential effects are observed when alternative coaching approaches are adopted. As the coaching literature develops, more nuanced examinations of the factors influencing the effectiveness of coaching such as these are now needed.

A key characteristics of our sample was that participation in the coaching intervention was voluntary, indeed, this aspect of the study design provides a high level of ecological validity in that participation in coaching is generally voluntary as opposed to mandatory. While we were able to compare those participants in our study who participated in coaching compared to those who did not (in terms of age, gender, GSE and commitment towards gaining an internship), further research could expand this by exploring the variables that predict whether an individual decides to opt for coaching versus those who do not. Potential variables of interest may include personality (such as extraversion and conscientiousness), socioeconomic status and proactive personality.

Finally, future research could explore the mechanisms through which coaching impacts on objective outcomes. Following from this research, it would be interesting for future research to explore whether increases in GSE mediate the relationship between coaching and objective outcomes.

Limitations

In our study, while we were able to collect data from participants who did not receive coaching, this group self-selected out of coaching. Although we were able to demonstrate that there were no significant differences between the coached and not coached samples in the variables tested (i.e., age, gender, GSE, commitment towards gaining an internship), ideally, experiments should utilise randomised allocation to groups. Therefore, we propose future studies compare a coached group to a waitlist control group. This was not possible in the present study, as organisations hiring students for internships only recruit in a limited window of time. Allocating students who desire coaching to a waitlist may therefore have disadvantaged them in their internship search and was not considered ethical.

Conclusions

While research has indicated that coaching has a positive impact on learning and performance, little is known regarding the impact of coaching on objective outcomes and who benefits most from coaching. To address these gaps in the literature, we conducted a field-based, quasi-experiment to assess whether coaching has a significant impact on selection success for students seeking a 12-month work internship. Our findings indicate that coaching positively impacted on selection success and only individuals lower in GSE benefitted significantly from coaching. In addition to contributing to the literature regarding the impact of coaching on objective outcomes, our findings also advance our understanding of the role of individual differences in coaching and contribute to the design of careers support to best suit individual differences. Consequently providing evidence that coaching is impactful, however, not equally impactful for all.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Holly Andrews

Holly Andrews is an Associate Professor in Coaching and Behavioural Change at Henley Business School. Holly has worked in academia for more than 15 years, specialising in occupational psychology and coaching. She joined Henley Business School in February 2022, where she is Programme Director for the MSc in Coaching for Behavioural Change and leads research and impact within the coaching group. Prior to joining Henley Business School, Holly managed the Department of Management and Finance at Worcester Business School. In 2015 Holly achieved her PhD, which explored the impact of psychopathic traits on career pathways. This continues to be a key research interest of hers and she has engaged in extensive knowledge exchange work in this area, including public talks, media appearances and consulting work. Her current research stream focuses on evaluating the effectiveness of coaching in different contexts, where she has published in Academy of Management Learning and Education.

Rebecca J. Jones

Rebecca J. Jones is a Professor in Coaching and Director for the Henley Centre for Coaching at Henley Business School. Rebecca completed her PhD on coaching effectiveness at Aston University in 2016. She is particularly interested in using quantitative research methodology to explore the factors that influence the efficacy of coaching. She has published her research in journals such as Academy of Management Learning and Education, the Journal of Occupational and Organisational Psychology, the European Journal of Work and Organisational Psychology and the Journal of Managerial Psychology. Rebecca is passionate about evidence-based practice and teaching. In addition to teaching and research, Rebecca has extensive experience of coaching individuals at all levels, has worked with a range of corporate clients on consultancy projects. Rebecca is a frequent contributor to practitioner and professional bodies in addition to communicating her research to academic audiences, and is the host of the Coaching@Henley podcast.