Content uploaded by Tim Theeboom
Author content
All content in this area was uploaded by Tim Theeboom on May 01, 2014
Content may be subject to copyright.
This article was downloaded by: [Tim Theeboom]
On: 16 September 2013, At: 09:46
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
The Journal of Positive Psychology: Dedicated to
furthering research and promoting good practice
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/rpos20
Does coaching work? A meta-analysis on the effects
of coaching on individual level outcomes in an
organizational context
Tim Theebooma, Bianca Beersmaa & Annelies E.M. van Vianena
a Department of Work and Organizational Psychology, University of Amsterdam,
Weesperplein 4, 1018 XA, Amsterdam, The Netherlands
Published online: 13 Sep 2013.
To cite this article: Tim Theeboom, Bianca Beersma & Annelies E.M. van Vianen , The Journal of Positive Psychology
(2013): Does coaching work? A meta-analysis on the effects of coaching on individual level outcomes in an organizational
context, The Journal of Positive Psychology: Dedicated to furthering research and promoting good practice, DOI:
10.1080/17439760.2013.837499
To link to this article: http://dx.doi.org/10.1080/17439760.2013.837499
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
Does coaching work? A meta-analysis on the effects of coaching on individual level outcomes
in an organizational context
Tim Theeboom*, Bianca Beersma and Annelies E.M. van Vianen
Department of Work and Organizational Psychology, University of Amsterdam, Weesperplein 4, 1018 XA, Amsterdam,
The Netherlands
(Received 12 February 2013; accepted 15 August 2013)
Whereas coaching is very popular as a management tool, research on coaching effectiveness is lagging behind.
Moreover, the studies on coaching that are currently available have focused on a large variety of processes and outcome
measures and generally lack a firm theoretical foundation. With the meta-analysis presented in this article, we aim to
shed light on the effectiveness of coaching within an organizational context. We address the question whether coaching
has an effect on five both theoretically and practically relevant individual-level outcome categories: performance/skills,
well-being, coping, work attitudes, and goal-directed self-regulation. The results show that coaching has significant posi-
tive effects on all outcomes with effect sizes ranging from g= 0.43 (coping) to g= 0.74 (goal-directed self-regulation).
These findings indicate that coaching is, overall, an effective intervention in organizations.
Keywords: coaching; coaching effectiveness; coaching interventions; coaching outcomes; meta-analysis
Introduction
The use of coaching methodologies as a means of
enhancing performance and development in organizations
has increased substantially over the last two decades.
Since its foundation in 1995, the International Coach
Federation (ICF) has seen its member count grow to over
20,000 members in over 100 countries in 2012 (ICF,
2012) and the total annual revenue from coaching is esti-
mated at roughly $2 billion globally (ICF, 2012). Coach-
ing can be defined as a ‘result-oriented, systematic
process in which the coach facilitates the enhancement
of life experience and goal-attainment in the personal
and/or professional lives of normal, non-clinical clients’
(Grant, 2003, p. 254).
While coaching is often considered as a useful tool
for individual and organizational development (Grant,
Passmore, Cavanagh, & Parker, 2010), the lack of a sys-
tematic empirical review of research on the outcomes of
coaching makes it prone to skepticism regarding its
effectiveness (Bono, Purvanova, Towler, & Peterson,
2009; Bozer & Sarros, 2012). This skepticism seems
valid in the light of the high costs of coaching. In a sur-
vey completed by 428 coaches, Bono et al. (2009) found
that the average hourly fee for coaches was $237. Har-
vard Business Review even stated that ‘coaches aren’t
monks bound to a vow of poverty, and they can earn up
to $3500 an hour’(Couto & Kauffman, 2009, p. 1).
However, the current literature on coaching is inconclu-
sive on whether these high financial costs outweigh the
benefits that coaching potentially has for organizations
(Leonard-Cross, 2010). Thus, whereas coaching is very
popular as a management tool and organizations are
apparently willing to pay large amounts of money for it,
an empirical review of coaching effectiveness is lagging
(Bozer & Sarros, 2012).
In addition, most studies on coaching are conducted
by practitioners. While these studies can provide
valuable insights, most practitioners are not trained in
research methods. As a result, validated outcome mea-
sures are seldom used and the studies generally lack a
firm theoretical foundation (Grant, 2013). Consequently,
extant research on coaching has focused on a large vari-
ety of processes and outcome measures (Latham, 2007)
and this disjointed state of the literature on coaching
hinders the establishment of a theoretical framework for
future research (Spence & Oaedes, 2011; Sue-Chan &
Latham, 2004).
All in all, there is a strong need for a quantitative
summary and integration of existing coaching research.
To date, several good literature reviews, rather than
empirical reviews, have been published (Brock, 2008;
Feldman & Lankau, 2005; Grant, 2001; Grant & Cava-
nagh, 2004; Grant, Passmore, et al., 2010; Kampa-Kok-
esch & Anderson, 2001; Passmore & Fillery-Travis,
2011). To our knowledge only De Meuse, Dai, and Lee
(2009) conducted a quantitative review. Their review
assessed the success of coaching in terms of its return on
investment (ROI). Although ROI can be an indicator of
*Corresponding author. Email: t.theeboom@uva.nl
© 2013 Taylor & Francis
The Journal of Positive Psychology, 2013
http://dx.doi.org/10.1080/17439760.2013.837499
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
the effectiveness of coaching as a change methodology,
it also has some serious limitations which we will
discuss below.
With the meta-analysis we present in this article, we
aim to provide a comprehensive quantitative review and
to answer the question: Does coaching
1
work when
provided in an organizational context by professionally
trained coaches? Furthermore, by means of a systematic
review and integration of the types of coaching outcomes
that were included in prior studies, we aim to give an
initial impetus to further theoretical development of
coaching research. That is, by organizing coaching
outcomes that emerge from the literature into meaningful
categories, we aim to enable future studies to build on or
extend theory that explains the paths and processes that
lead to these different categories of outcomes.
Coaching
Coaching has its roots in a multitude of disciplines,
including philosophy, sociology, anthropology, sports,
communication science and even natural sciences (Brock,
2008). However, in terms of the number of articles
published in peer-reviewed journals, sub-disciplines of
psychology have shown to be the most fruitful areas of
research on coaching (Grant, Passmore, et al., 2010).
Initially, most of the research on coaching was conducted
within areas such as sports psychology (e.g. Gallwey,
1974; Whitmore, 1992) and clinical psychology (e.g.
Berg & Szabo, 2005; De Shazer, 1988). More recently,
research on coaching has particularly flourished in two
strongly related areas of psychology that emerged within
the last two decades: positive psychology (Seligman &
Csikszentmihalyi, 2000) and coaching psychology
(Passmore, 2010).
Positive psychology focuses on studying three aspects
that constitute the scientific notion of happiness: positive
emotion, meaning, and engagement (Seligman, 2007).
Coaching psychology focuses on studying behavior,
cognition, and emotion within coaching practice to deepen
understanding of coaching processes and to enhance
coaching techniques (Passmore, 2010). While there are
some debates about how these areas of research are related
to each other (some authors see coaching psychology as a
sub-discipline or an applied form of positive psychology;
e.g. Grant & Cavanagh, 2007), both areas share their
focus on performance enhancement, positive aspects of
human nature, and the strengths of individuals (Linley &
Harrington, 2005). Therefore, positive psychology seems
to offer a robust framework for researching coaching and
as such may constitute ‘one of the solutions to the lack of
a theoretical framework in the coaching field’(Freire,
2013, p. 428)
While different psychological sub-disciplines initially
developed their own specific conceptual framework,
more recent literatures have gradually moved towards a
generally accepted definition of coaching. Kilburg (1996)
originally defined coaching of executives in organiza-
tions as ‘a helping relationship between a managerial-
client and a consultant that follows a formally defined
coaching agreement’. Grant (2003) transformed this
definition into a more general one and defined coaching
as a ‘result-oriented, systematic process in which the
coach facilitates the enhancement of life experience and
goal-attainment in the personal and/or professional life of
normal, non-clinical clients’(p. 254).
This latter definition encompasses several important
features, namely: it can be applied to a multitude of
coaching domains (e.g. personal coaching and organiza-
tion coaching) and coaches (executive and non-execu-
tive), it emphasizes the self-directedness of the coaching
process, and it recognizes coaching as a systematic
process rather than just being empathic and ‘having good
conversations’(Leonard-Cross, 2010). Furthermore,
although the differences and similarities between coach-
ing and therapy are still a topic of debate (Bono et al.,
2009; Brunning, 2006; Hart, Blattner, & Leipsic, 2001),
Grant’s(2003)definition of coaching distinguishes
coaching from therapy by its focus on a non-clinical
population.
Coaching effectiveness: beyond return of investment
measures
The literature on coaching has grown exponentially in
the last 15 years. Whereas only 93 articles were
published in the years between 1937 and 1999, the total
number of articles and dissertations on coaching reached
634 in 2011 (Grant, 2013) and the number has been stea-
dily growing ever since. However, the bulk of articles
still consists of descriptive papers and/or case studies as
well as practitioner articles primarily aimed at emphasiz-
ing the benefits of certain coaching interventions (De
Meuse et al., 2009).
Additionally, this predominantly practitioner-
generated research has resulted in an overemphasis on
ROI measures. ROI as a measure of coaching effective-
ness is appealing because it provides some direct insight
into the tangible benefits of coaching interventions
(Fillery-Travis & Lane, 2006; Leonard-Cross, 2010).
However, the ROI measure has some serious limitations.
For instance, the factors included in the most frequently
used calculation of the ROI metric (benefits –costs/
costs × 100) are highly idiosyncratic and tend to ignore
context variables such as team input. Therefore, it is
often impossible to determine the degree to which the
financial benefits can be directly attributed to the coach-
ing intervention (Grant, Passmore, et al., 2010, p. 26).
Additionally, performance measures (direct benefits) are
seldom available and a narrow focus on ROI and other
2T. Theeboom et al.
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
performance-related measures neglects other –more
indirect –ways in which organizations could potentially
benefit from coaching, such as employee well-being and
health. Therefore, the current meta-analysis investigates
coaching effectiveness by looking at well-validated, more
distal indicators of functioning in addition to perfor-
mance measures: well-being, coping, work and career-
related attitudes, and goal-directed self-regulation.
The overall goal of coaching in a work context is to
optimize a person’s work-related functioning. First,
individuals in organizations function better if they feel
well, that is, if their basic needs are fulfilled (Deci &
Ryan, 1985,2000) and if they do not struggle with
health-related problems as caused by their job (Burton
et al., 1999). With regard to the latter, work-related stress
affects over 20% of workers in the European Union
(Brunn & Milczarek, 2007) which costs organizations
about 20,000 million Euros per year. Indeed, research
has evidenced negative relationships between ‘soft’
outcomes, including individual health and well-being,
and ‘hard’performance measures (Bond & Bunce, 2003;
Wright & Cropanzano, 2000). Specifically, Duijts, Kant,
van den Brandt, and Swaen (2008) found that a coaching
intervention in medical health care and educational orga-
nizations resulted in a positive change in employees’
well-being as indicated by a decrease in sickness-related
absenteeism and burnout and an increase in life satisfac-
tion. Furthermore, coaching could also have a preventive
role via its effect on individual’abilities to cope with
stressors. Several studies have found coaching to have
positive effects on coping mechanisms such as resilience
(Grant, Curtayne, & Burton, 2009), mindfulness (Spence,
Cavanagh, & Grant, 2008), and the use of self-enhancing
attributions (Moen & Skaalvik, 2009).
Another way in which coaching could benefit
organizational effectiveness is via its potentially benefi-
cial effects on employees’work- and career-related
perceptions and attitudes (e.g. job satisfaction). Coaching
may facilitate the cognitive ‘reframing’of work experi-
ences and attitudes which is a central aspect of multiple
coaching programs (Grant, 2001). In order to function
better, employees may initially try to change their work
tasks and interactions, but if organizational boundaries
put constraints on making actual changes, a coach could
help them adjust their job perceptions (Wrzesniewski &
Dutton, 2001). Indeed, a preliminary study showed that
coaching positively influenced work- and career-related
attitudes (Bozer & Sarros, 2012) which, in turn, may
positively affect work performance (Judge, Thoresen,
Bono, & Patton, 2001; Organ & Ryan, 1995).
Finally, coaching enhances organizational effective-
ness through its potentially beneficial effect on employ-
ees’goal-directed self-regulation (Grant, 2003). Most
coaching programs target either one or more stages of
goal-directed self-regulatory processes. For example,
several studies have shown that coaching can positively
influence goal-attainment expectancy (Evers, Brouwers,
& Tomic, 2006; Moen & Skaalvik, 2009) and goal-
progression and commitment (Green, Oades, & Grant,
2006). Because the relationship between goal-setting on
the one hand and motivation and performance on the
other hand is well established (Locke & Latham, 2002),
improving employees’self-regulation by enhancing their
ability to set and strive for goals is yet another way in
which organizations could benefit from coaching.
All in all, coaching could benefit organizations by
enhancing employees’performance and skills, well-
being, coping, work attitudes, and goal-directed self-
regulation. Because extant empirical studies (see Method
section below) lack a clear conceptual framework for
classifying coaching outcomes, we categorized available
outcomes into these outcome dimensions, which are both
theoretically and practically relevant and also well
established within the psychological literature.
Method
Data collection
In order to decide which studies to include in our
analyses, we undertook an extensive literature search,
which consisted of five phases based on the PRISMA
statement for reporting meta-analysis (Moher, Liberati,
Tetzlaff & Altman, 2009). A PRISMA flowchart is
displayed in Figure 1.
First, we searched several (social) science databases
(Google Scholar, JSTOR, Mendely, PsycINFO, Science-
Direct, and Springerlink) using the following keywords:
workplace coaching, developmental coaching, executive
107 Potential abstracts identified through electronic database searching,
mailing-lists and email requests
107 Full articles
screened
69 Articles excluded because of
cross-sectional study design or lack
of quantitative data analysis
38 Articles
assessed for
inclusion criteria
20 Articles excluded because the
coaching intervention did not match
our definition of coaching, and there
were possible confounds and/or
insufficient data.
18 Studies
included in the
meta-analysis
Figure 1. Search strategy used for the inclusion of studies
considered in the present meta-analysis.
The Journal of Positive Psychology 3
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
coaching, effects of coaching, and outcomes of coaching.
Initially, we searched for articles (i) that included
quantitative data on the effects of coaching; (ii) in which
coaching was provided by professionally trained external
coaches or trained peers; and (iii) in which the coaches
belonged to a non-clinical population. Second, we
performed both backward and forward searches on the
studies we retrieved in phase 1. Third, we contacted sev-
eral scholars known to be active in the field of coaching
psychology research in order to retrieve (yet) unpub-
lished results. At the same time, we sent out a request
for published and unpublished studies via the mailing-list
services of Academy of Management, the Society for
Industrial and Organizational Psychology, and the
European Association of Social Psychology.
In the fourth phase, we screened the 107 full articles
retrieved in the first three phases and excluded all cross-
sectional studies since these do not allow controlling for
most threats to internal validity (Cook & Campbell,
1979). Furthermore, we excluded all studies in which the
authors did not perform quantitative analyses (e.g. case
studies). In the fifth and final phases, we used a priori
conceptual and methodological criteria to decide which
of these studies would be included in the final analysis.
For example, we excluded all studies in which the
described coaching process did not match our definition
of coaching or for which a description of the coaching
process could not be obtained.
In addition to Grant’s (2003) definition of coaching,
we took one more characteristic of coaching into account
when selecting studies, namely that the coaching was
provided by a (professionally trained) coach with no
formal authority over the coachee. The main reason for
this is that research on mentoring has shown that a
mentor’s formal authority over the protégée can affect
the way in which the protégée behaves during coaching
sessions (Mullen, 1994; Tepper, 1995; Waldron, 1991).
Furthermore, the goals in managerial coaching (in which
a manager coaches an employee) are often strongly
related to organizational performance (Cox, Bachkirova,
& Clutterbuck, 2009) which might influence the degree
to which the coaching process is self-directed. Another
reason for excluding studies on managerial coaching is
that our aim for this research was to set a first step for
answering the question whether the high costs of hiring
external professionally trained coaches are justifiable for
both individual clients and organizations. Thus, we
excluded all studies in which the coach had a formal
authority over the coachee.
We also excluded all studies in which the influence
of other interventions (e.g. leadership development pro-
grams) could not be ruled out as a confounding factor.
Because we aim to provide more insight into the use-
fulness of coaching in organizations, we only included
studies conducted in a work or educational context.
Studies conducted in an educational context (e.g.
undergraduate students) were included because the cli-
ents’characteristics and needs are similar to those of
clients in organizational settings, that is, they are simi-
lar with respect to their demographics and the chal-
lenges they face (working in teams, meeting deadlines
etc.). Finally, we excluded all studies in which not
enough statistical information was available or could
not be made available after contacting the correspond-
ing authors.
This selection process resulted in a total of 18 studies
included in the final analysis. All studies that were
included in the final analysis are indicated with an * in
our list of references. An overview of basic characteris-
tics of these studies is displayed in Table 1.
Outcome categorization
The categorization of outcomes was conducted in three
steps. In the first step, the first and the second authors
mutually assigned all study outcomes into one of the
following categories: performance/skills, well-being,
coping, work attitudes, or goal-directed self-regulation
while the third author categorized the outcomes inde-
pendently of the first two authors. In the second step,
the interrater agreement was calculated. The interrater
agreement (Cohens’κ) was 0.80, which is considered
to be large (see Landis & Koch, 1977). In the third
step, the authors discussed their discrepancies and
agreed on a final categorization. Several studies in the
meta-analysis included multiple measures within the
same outcome category (e.g. measures of stress and
burnout both fall within the well-being category). An
average of these effect sizes was included in order to
prevent violation of the independent sample assumption
(Hunter & Schmidt, 2004).
The performance/skills category includes both subjec-
tive and objective outcome measures that either directly
reflect performance (e.g. number of sales, supervisory
rated job performance) or reflect the demonstration of
behaviors needed for an organization to be effective (e.g.
transformational leadership behaviors). The well-being
category includes both subjective and objective outcome
measures that are a direct representation of peoples’
well-being, health, need fulfillment, and affective
responses. Examples of these are measures of psychopa-
thology (e.g. Depression Anxiety and Stress Scale;
Lovibond & Lovibond, 1995) and burnout (e.g. Maslach
Burnout Inventory; Maslach & Jackson, 1986).
The coping category includes outcome measures
related to the ability to deal with present and future job
demands and stressors. Examples of these are measures
of self-efficacy (e.g. General Self-Efficacy Scale;
Schwarzer & Jerusalem, 1995) and mindfulness (e.g.
Mindful Attention Awareness Scale; Brown & Ryan,
4T. Theeboom et al.
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
Table 1. Study characteristics and outcome overview of studies included in the meta-analysis.
Study nIntervention
Nr.
sessions Outcomes Design
Bozer and Sarros (2012) 96 Cognitive behavioral
coaching
11 Self-reports
Career satisfaction
Job commitment
Job Performance
Self-Awareness
Supervisory ratings
Job performance
Self-awareness
Task performance
RCT
Cerni, Curtis, and
Colmar (2010)
14 Epstein constructive
thinking intervention
10 Staff-ratings
Transformational leadership
QEF
Egan and Song (2005) 103 Coaching on goal-setting
and goal achievement
Unknown Self-reports
Job satisfaction
Organizational commitment
Performance goal orientation
Performance rating
Supervisory ratings
Performance rating
QEF
Finn (2007) 32 CB-SFC 6 Self-reports
Developmental planning
Developmental support
Openness to new behaviors
Positive affect
Self-efficacy
QEF
Grant (2003) 20 CB-SFC group coaching 10 Self-reports
Anxiety
Depression
GAS
Insight
Self-reflection
Stress
Quality of life
WSD
6
Grant (2008) 29 CB-SFC peer coaching 5 Self-reports
Anxiety
Cognitive hardiness
Depression
GAS
Learning
Personal insight
Well-being
WSD
(Continued)
The Journal of Positive Psychology 5
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
Table 1. (Continued).
Study nIntervention
Nr.
sessions Outcomes Design
Grant et al. (2009) 41 CB-SFC 10 Self-reports
Anxiety
Depression
Gas
Resilience
Stress
Well-being
Work
RCT
Grant, Green, and
Rynsaardt (2010)
65 CB-SFC 10 Self-reports
Agency
Anxiety
Cognitive hardiness
Depression
Hope
Interpersonal problems
General health questionnaire
GAS personal
GAS work
Resilience
Stress
Workplace well-being
RCT
Green et al. (2007) 44 CB-SFC 10 Self-reports
Anxiety
Cognitive hardiness
Hope
Depression
Stress
RCT
Green et al. (2006) 50 CB-SFC in groups 9 Self-reports
Agency
Autonomy
Environmental mastery
Goal striving
Negative affect
Personal growth
Positive affect
Positive relations with others
Purpose in life
Satisfaction with life
Self-acceptance
Total hope
RCT
Kochanowski, Seifert,
and Yukl (2010)
84 Coaches provided
Feedback on leadership
style
10 Self-reports
Leadership behavior
QEF
Luthans and
Peterson (2004)
67 Coaches discussed
discrepancy between
executives’own
performance ratings and
ratings of employees
1 Self-reports
Anxiety
Depression
GAS
3
Goal difficulty
WSD
(Continued)
6T. Theeboom et al.
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
Table 1. (Continued).
Study nIntervention
Nr.
sessions Outcomes Design
Organizational commitment
Satisfaction with coworkers
Satisfaction with supervision
Satisfaction with work
Self-insight scale
Self-reflection scale
Stress
Turnover intentions
Quality of Life Inventory
Moen and Skaalvik (2009) 19 Coaching aimed at self-
actualization and
promotion of resources
4 Self-reports
Attribution of failure to
ability
Attribution failure to
circumstances
Attribution failure to effort
Attribution failure to strategy
Attribution success to ability
Attribution success to
circumstances
Attribution success to effort
Attribution success to
strategy
Autonomy
Competence
Goal clarity
Goal commitment
Goal difficulty
Goal feedback
Goal Strategy
Need satisfaction at work
Relatedness
Self-efficacy
QEF
Peterson (1993) 100 Individual coaching for
effectiveness program
50 Self-reports
Effectiveness
Supervisory ratings
Effectiveness
Ratings by coach
Effectiveness
WSD
Poepsel (2011) 28 Online CB-SFC 4 Self-reports
Goal attainment
Hope
Subjective well-being
RCT post-
test only
Smither, London, Flautt,
Vargas, and Kucine (2003)
1243 Coaching based on 360-
degree feedback
3 Self-reports
Elaboration of self-reports
Elaboration of peer ratings
QEF
(Continued)
The Journal of Positive Psychology 7
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
2003). The work attitudes category includes outcome
measures related to cognitive, affective, and behavioral
responses toward work and career, such as job
satisfaction (e.g. Job Description Index; Smith, Kendall,
& Hulin, 1969), organizational commitment (e.g. Organi-
zational Commitment Scale; Porter, Steers, Mowday &
Boulian, 1974), and career satisfaction (e.g. Career
Satisfaction Scale; Greenhaus, Parasuraman & Wormley,
1990). Finally, the goal-directed self-regulation category
includes all outcome measures related to goal-setting,
goal-attainment, and goal-evaluation. This category also
includes goal-attainment scaling (GAS) measures, which
are increasingly popular in coaching settings (see Spence
[2008] and Peterson [1993] for overviews of the use of
GAS in coaching research).
Calculating the effect sizes
One of the most challenging steps in a meta-analysis is
combining the effect sizes of different studies in one
Table 1. (Continued).
Study nIntervention
Nr.
sessions Outcomes Design
Elaboration of supervisory
ratings
Performance
Rating by peers
Performance
Supervisory ratings
Performance
Rating by authors
Goal specificity
Spence et al. (2008)15
a
Coaching on goal-setting
and goal achievement
4 Self-reports
Anxiety
Depression
Environmental mastery
Mindfulness
Reflection
Rumination
Satisfaction with life
Self-Acceptance
Stress
RCT
Spence and Grant (2007) 40 CB-SFC 10 Self-reports
Autonomy
Environmental mastery
Goal commitment
Goal progression
Negative affect
Personal Growth
Positive affect
Positive relations with others
Purpose in life
Self-acceptance
Subjective well being
RCT
Notes: RCT = randomized control trial; QEF = quasi-experimental field study in which participants were non randomly allocated to experimental and
control groups; WSD = within-subjects design without control group, which includes both pre- and post-intervention measures; CB-SFC = cognitive
behavioral solution-focused coaching; and GAS = goal-attainment scaling.
a
Only the C-MT condition (see original) was included for the calculation of effect sizes.
8T. Theeboom et al.
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
analysis (McGaw & Glass, 1980). Since effect sizes
based on means are easily interpretable and the studies
in our analysis employed a large variety of outcome
measures to assess the impact of coaching interventions,
the use of an effect size index based on standardized
means was the logical choice (Borenstein, Hedges,
Higgins, & Rothstein 2009; Cohen, 1988). Since the
most popular index, Cohen’sd, tends to overestimate the
population effect size when small samples are included
in the analysis, we chose to use Hedges gwhich can still
be interpreted as the mean difference expressed in
standard deviation units and applies a simple correction
to overcome this bias (Hedges, 1981).
Effect sizes can be defined in relation to pre-interven-
tion scores, post-intervention scores, or difference scores.
In theory, it is possible to choose either definition
because effect sizes can be transformed into a common
effect size index by using the correlation between pre-
and post-intervention scores to estimate the sampling
bias (Morris & DeShon, 1997). Unfortunately, the corre-
lations between pre- and post-intervention scores were
seldom provided and the amount of studies that used a
mixed design outnumbered the amount of studies that
used a within-subject design. Therefore, effect sizes
based on the pooled (experimental and control groups)
standard deviations of the post-intervention scores were
chosen as the referent effect size index. By doing so, the
estimation of parameters was minimized since only
the pre-intervention/post-intervention correlations for the
(minority of) studies that used a within-subject design
needed to be estimated
2
. Finally, effect sizes based on
post-intervention standard deviations are likely to be
slightly biased downward, and are thus the most conser-
vative choice (Carlson & Schmidt, 1999).
Meta-analytic procedure and statistical analyses
The Hedges and Olkin (1985) approach to meta-analysis
was used to calculate the effect sizes. Comparisons with
other commonly applied methods such as by Hunter and
Schmidt (1990) and Rosenthal (1991) suggest that differ-
ences between the Hedges and Olkin methods and the
other methods are relatively small and only apply under
very specific circumstances. If anything, the Hedges and
Olkin method can be considered to be the most conser-
vative approach because it does not allow for the statisti-
cal corrections for artifactual sources of variance (e.g.
measurement error and restriction of range) that tend to
result in an inflation of effect size estimates (Borenstein
et al., 2009). Additionally, the Hedges and Olkin
approach seems to provide the most conservative esti-
mate of the (lower limit of) confidence intervals (John-
son, Mullen, & Salas, 1995), which can be used for
determining the statistical significance of effect sizes.
After selecting the general approach for the meta-
analysis, the statistical model for the meta-analysis has
to be designated. In terms of the model for the meta-
analysis, the (conservative) random-effect model was
adopted as recommended by the National Research
Council (1992). As opposed to the fixed effect model,
the random-effect model allows that the true effect size
varies from study to study based on both the variability
of the independent variable (e.g. intensity or duration of
intervention) and differences in the samples of the
research population such as age, educational background,
and type of job of the coaches (Borenstein et al., 2009;
Hedges & Cooper, 1994).
Heterogeneity between studies was quantified by an
assessment of both the classical Cochran Qstatistic
(1954) and the I
2
statistic as proposed by Higgins and
Thompson (2002), see also Higgins, Thompson, Deeks,
& Altman, 2003). While the Qstatistic serves as a test
of significance for between-study heterogeneity, the value
for I
2
represents the proportion of between-study vari-
ance in effect sizes that can be attributed to between-
study heterogeneity rather than within-study variability
(Borenstein et al., 2009). When the value for I
2
is large
(see Higgins & Thompson, 2002 for some guidelines on
interpretation), one of the possible explanations for this
is the existence of moderating variables. In what should
be considered exploratory analyses due to the relatively
small amount of studies, we explored the influence of
two of these potential (methodological) moderating vari-
ables in order to provide guidance for the future method-
ological approach of coaching research. First, following
the example of Fusar-Poli et al. (2012), we conducted
subgroup analyses for sets of studies that were different
in terms of study design (mixed designs vs. within-
subject designs). Second, we examined the influence of
the number of coaching sessions using meta-regression
analysis. Finally, following recent recommendations by
Sterne et al. (2011), we assessed the risk for publication
bias (or small study bias) by visually inspecting funnel
plots and by applying the regression intercept of Egger,
Smith, Schneider, and Minder (1997).
Software for the analysis and statistical corrections
The software that was used for the analysis was
comprehensive meta-analysis. CMA was developed by
Borenstein, Hedges, Higgins, and Rothstein (2005) and
is based on the Hedges and Olkin’s(1985) approach to
meta-analysis. CMA offers advantages over other soft-
ware in terms of its flexibility to handle multiple data
entry formats (e.g. data from within-subject designs and
mixed designs) and its intuitive approach to sensitivity
analysis and the detection of between-study heterogene-
ity (Borenstein et al., 2009).
The Journal of Positive Psychology 9
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
Results
Aggregated effect sizes and overall between-study
heterogeneity
Table 2contains the weighted effect sizes (aggregated
over outcomes) per study.
The point estimate of the overall weighted effect size
(aggregated over all studies and outcomes) was
significant (g=0.66, 95% CI, 0.39–0.93, p= 0.000),
suggesting that coaching, in general, has a significant
positive effect across the range of outcome measures we
examined. The relatively large point estimate of gin the
study by Peterson (1993) encouraged us to perform a sen-
sitivity analysis in which the analysis was repeated while
excluding the results of this study. Although the overall
weighted point estimate of gdropped, it remained signifi-
cant (g= 0.51, 95% CI, 0.34–0.69, p< 0.000). Thus,
excluding the study by Peterson (1993) did not alter our
conclusions regarding the point estimate of the overall
weighted effect size. According to the tentative criteria
set by Higgins and Thompson (2002), the heterogeneity
in effect sizes was statistically significant and large in
magnitude (Q= 130.05; p< 0.000; I
2
= 86.93). As men-
tioned, this substantial variance in effect sizes encourages
the consideration of moderating variables. Therefore, we
further explored the heterogeneity in effect sizes.
Effect sizes per outcome category
The main goal of this meta-analysis was to provide
insight into the effects of coaching on various individ-
ual-level psychological outcomes. Table 3contains the
results for all outcome categories: performance and
skills, well-being, coping, work attitudes, and goal-direc-
ted self-regulation.
The results indicate that coaching interventions have
significant positive effects on all outcome categories:
performance and skills (g= 0.60, 95% CI, 0.04–0.60,
p= 0.036), well-being (g= 0.46, 95% CI, 0.28–0.62,
p< 0.001), coping (g= 0.43, 95% CI, 0.25–0.61,
p< 0.001), work attitudes (g= 0.54, 95% CI, 0.34–0.73,
p< 0.001), and goal-directed self-regulation (g= 0.74,
95% CI, 0.42 –1.06, p< 0.001). Based on our earlier
sensitivity analysis (see above), we repeated the analysis
for the performance and skills outcome while excluding
the results of the study by Peterson (1993). Although the
weighted point estimate of the effect size for
performance and skills dropped considerably, it remained
significant (g= 0.19, 95% CI, 0.04–0.32, p= 0.013). We
will discuss the implications of this finding in the
discussion section. Finally, we note that the significance
of the Qstatistics and the moderate-to-high values of I
2
for both the performance and skills outcomes and the
goal-directed self-regulation outcomes indicate that the
influence of between-study heterogeneity is especially
apparent for these outcome categories.
The influence of study design
We explored the differences in effect size patterns
between studies that used a mixed design vs. studies that
used a within-subject design. The results of this analysis
are shown in Table 4.
Table 2. Weighted effect sizes per study aggregated over outcomes.
CI (95%)
Study nSessions Study design gLower Upper p-value
Bozer and Sarros (2012) 96 11 RCT 0.36 −0.12 0.83 0.145
Cerni et al. (2010) 14 10 QEF 0.08 −0.15 0.30 0.501
Egan and Song (2005) 103 Unknown QEF 0.69 0.29 1.08 0.001
Finn (2007) 32 6 QEF 1.17 0.34 1.99 0.006
Grant (2003) 20 10 WSD 0.84 0.31 1.38 0.002
Grant (2008) 29 5 WSD 0.71 0.23 1.19 0.004
Grant et al. (2009) 41 10 RCT 0.63 0.00 1.25 0.049
Grant et al. (2010) 65 10 RCT 0.25 −0.34 0.83 0.406
Green, Grant, and Rynsaardt (2007) 44 10 RCT 0.61 0.02 1.20 0.042
Green et al. (2006) 50 9 RCT 0.70 0.13 1.28 0.016
Kochanowski et al. (2010) 84 10 QEF 0.25 −0.16 0.66 0.239
Luthans and Peterson (2004) 67 1 WSD 0.74 0.47 1.01 0.000
Moen and Skaalvik (2009) 19 4 QEF 0.82 −0.10 1.75 0.080
Peterson (1993) 100 50 WSD 2.33 1.93 2.72 0.000
Poepsel (2011) 28 4 RCT 0.92 0.16 1.70 0.018
Smither et al. (2003) 1243 3 QEF 0.13 0.00 0.27 0.049
Spence et al. (2008)15
a
4 RCT 0.10 −0.39 0.58 0.695
Spence and Grant (2007) 40 10 RCT 0.29 −0.33 0.91 0.355
Total 2090 –– 0.66 0.39 0.93 0.000
Notes: N= analyzed sample size; g= Hedges’g; CI (95%) = 95% confidence interval for g; RCT = randomized control trial; QEF = quasi-experimen-
tal field-study in which participants were non randomly allocated to experimental and control groups; and WSD = within-subjects design without con-
trol group, which includes both pre- and post-intervention measures.
a
Only the C-MT condition (see original) was included for the calculation of effect sizes.
10 T. Theeboom et al.
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
The weighted point estimates of the effect sizes seem
to indicate that the effect sizes of studies that used a
within-subject design are larger than the effect sizes of
the studies that used an independent-group design. Sub-
group analysis indicated that the overall effect size
(aggregated over outcomes) of studies that used a
within-subjects design (g=1.15, 95% CI, 0.46–1.83)
was significantly larger than the effect size of studies that
used a mixed design (g= 0.39, 95% CI, 0.22–0.56,
p= 0.036). These results imply that study design has a
considerable influence on the relation between coaching
interventions and individual-level outcomes (Hunter &
Schmidt, 2004). We will further reflect on these results
in the discussion section.
The influence of the number of coaching sessions
Table 5displays the differences in effect sizes for
studies that differ in terms of the number of coaching
sessions. The choice for the comparison of studies
with less or equal to five sessions vs. more than five
sessions emerged from the available study data with
some studies reporting a maximum of five coaching
sessions whereas other studies reporting more than this
maximum.
The weighted point estimates of the effect sizes show
a mixed picture. Although a larger number of coaching
sessions seems to be beneficial for both coping and
goal-directed self-regulation outcomes, the reversed
pattern is observed for work/career attitudes and perfor-
mance/skills (higher effect sizes for a smaller number of
sessions). A meta-regression in which the number of
coaching sessions was entered as a predictor of the
weighted effect sizes (aggregated over outcomes)
revealed no significant effects. These results indicate that
the number of coaching sessions is not related to the
effectiveness of the interventions.
Table 3. Weighted effect sizes of coaching interventions on all outcome categories.
CI (95%)
kN gLower Upper pQI
2
Performance/skills 6 2007 0.60 0.04 1.16 0.036 112.24
*
95.55
Well-being 10 564 0.46 0.28 0.62 0.000 7.72 0.00
a
Coping 10 1703 0.43 0.25 0.61 0.000 6.36 43.01
Attitudes 7 507 0.54 0.34 0.73 0.000 8.64 30.51
Self-regulation 11 789 0.74 0.42 1.06 0.000 81.27
*
53.38
Notes: k= number of studies included in the analysis; N= total sample size in kstudies; g= Hedges’g; CI = 95% random effects confidence intervals;
and Q = Cochran Q statistic. I
2
= the proportion of total variation in the estimates of treatment effect that is due to heterogeneity between studies.
*
indicates that between-study heterogeneity significant at α= 0.000.
a
The I
2
was truncated to zero because the Q statistic used for the computation of I
2
was smaller than its degrees of freedom.
Table 4. Weighted effect sizes of coaching interventions on all outcome categories for different study designs.
CI (95%)
kN gLower Upper pQI
2
Performance/skills
IDG 5 1372 0.18 0.04 0.32 0.013 5.51 27.43
Within-subjects 1 100 2.33 1.93 2.72 0.000 –
b
–
b
Well-being
IDG 7 394 0.39 0.18 0.60 0.000 3.91 0.00
a
Within-subjects 3 116 0.54 0.26 0.82 0.002 3.12 35.94
Coping
IDG 7 325 0.26 0.15 0.37 0.000 11.67 48.57
Within-subjects 3 116 0.47 0.20 0.74 0.001 2.16 7.45
Attitudes
IDG 6 169 0.48 0.29 0.67 0.000 6.23 19.77
Within-subjects 1 67 0.74 0.47 1.01 0.000 –
b
–
b
Self-regulation
IDG 9 1565 0.32 0.21 0.43 0.000 37.37*78.59
Within-subjects 2 116 1.33 0.85 1.81 0.035 –
b
0.00
a
Notes: k= number of studies included in the analysis; N= total sample size in kstudies; g= Hedges’g; CI = 95% random effects confidence intervals;
Q= Cochran Qstatistic; and I
2
= the proportion of total variation in the estimates of treatment effect that is due to heterogeneity between studies.
*
Indicates that between-study heterogeneity significant at α= 0.000.
a
The I
2
was truncated to zero because the Q statistic used for the computation of I
2
was smaller then it’s degrees of freedom.
b
The statistics could not be computed because not enough data was available.
The Journal of Positive Psychology 11
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
Publication bias
A visual evaluation of the funnel plot did not reveal
obvious evidence of publication bias. Additionally, the
Egger intercept was non-significant (p= 0.08). However,
a visual inspection of funnel plots as well as tests for
funnel asymmetry may produce unreliable results when a
small number of studies is included in the analysis,
especially when heterogeneity is substantial (Sterne
et al., 2011). Therefore, we will further address the issue
of potential publication bias in the discussion section.
Discussion
Summary of findings
This meta-analysis aimed to provide insight into the
possible beneficial effects of coaching within an organi-
zational context. We examined relationships between
coaching interventions and several individual-level out-
comes that are relevant for both individuals and organi-
zations. The results show that coaching has significant
positive effects on performance and skills, well-being,
coping, work attitudes, and goal-directed self-regulation.
In general, our meta-analytic findings indicate that
coaching is an effective tool for improving the function-
ing of individuals in organizations.
We should note, however, that an examination of the
between-study heterogeneity showed that the effects of
coaching interventions varied considerably between
studies (especially in the performance and skills and
goal-directed self-regulation outcome categories). This
heterogeneity could be –at least partially –attributed to
the relatively small number of studies in the analysis.
Alternatively, heterogeneity could also signal the
presence of moderating factors (Higgins & Thompson,
2002). The findings of our exploratory analyses indeed
suggest that research design could be one of these mod-
erating factors. Studies using a within-subject design
generally displayed stronger positive effects of coaching
interventions than studies using an independent-group
(only one study) or mixed design.
A possible explanation for this finding is that studies
using a mixed-design controlled for additional sources of
bias in comparison with studies using a within-subjects
design (Morris & DeShon, 2002). More specifically, the
addition of a control group allows the researcher to
control for the natural maturation of participants over
time and for selection effects (Cook & Campbell, 1979).
At the very least, these results show that the choice for a
specific study design has considerable implications for
the conclusions that can be drawn with regard to the
effectiveness of coaching interventions.
An examination of the results regarding the intensity
of the coaching intervention suggests that a greater
number of coaching sessions does not necessarily result
in stronger positive effects. This, somewhat
counter-intuitive, pattern of results might be caused by a
spurious correlation. That is, people with less serious or
complex problems may need fewer sessions and experi-
ence more positive effects of coaching than people with
serious and/or complex problems. Alternatively, it could
be explained by type of coaching interventions that were
applied in the majority of studies in which the number of
coaching sessions was small, namely solution-focused
coaching. Solution-focused coaching originates in brief
family therapy (de Shazer, 1988) and differs from other
Table 5. Weighted effect sizes of coaching interventions for studies that differ in terms of the number of coaching sessions.
CI (95%)
kN gLower Upper QI
2
Performance/skills
≤5 sessions 3 1329 0.26 0.02 0.53 0.00 58.57
>5 sessions 2 110 0.11
b
−0.09 0.32 0.00 0.00
a
Well-being
≤5 sessions 4 195 0.47 0.15 0.79 5.83 38.36
>5 sessions 7 296 0.46 0.22 0.69 0.00 0.00
a
Coping
≤5 sessions 5 1429 0.35 0.11 0.59 0.00 50.33
>5 sessions 5 182 0.54 0.30 0.79 0.00 0.00
a
Attitudes
≤5 sessions 3 126 0.67 0.41 0.94 19.14 44.39
>5 sessions 4 150 0.35 0.08 0.61 0.75 0.00
a
Self-regulation
≤5 sessions 6 1457 0.52 0.15 0.88 6.962 79.94
>5 sessions 5 215 1.02 0.67 1.36 0.00 32.18
Notes: k= number of studies included in the analysis; g= Hedges’g; CI = 95% random effects confidence intervals; Q= Cochran Qstatistic; and
I
2
= the proportion of total variation in the estimates of treatment effect that is due to heterogeneity between studies.
a
The I
2
was truncated to zero because the Qstatistic used for the computation of I
2
was smaller then it’s degrees of freedom.
b
One study reported an average of 50 coaching sessions and was therefore considered an outlier and excluded from the meta-regression analysis.
12 T. Theeboom et al.
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
forms of therapy and coaching by its premise that there is
no need for an extensive analysis and understanding of
problems in order to create solutions (Berg & Szabo,
2005; Grant & O’Connor, 2010). Therefore, it is possible
to jump directly to the ultimate aim of coaching, namely
the identification of solutions, potentially resulting in a
smaller number of sessions needed to make progression
(Kim, 2008). Future research could investigate whether
solution-focused coaching is indeed more effective than
other coaching approaches and whether specific coaching
effects also depend on significance and/or complexity of
coaches’problems.
Although the results should be interpreted with
caution because of the exploratory nature of the analy-
sis, the finding that coaching can be effective even
when the number of coaching sessions is relatively
small is encouraging for organizations and individuals
in need of coaching. However, while the difference in
the number of sessions does not seem to impact the
mean effect size, the examination of the heterogeneity
statistics does show that there is less variability in the
effect sizes for studies using a larger amount of ses-
sions. In other words, the robustness of the effects of
coaching seems to increase with the number of coach-
ing sessions. This finding corroborates research on adult
learning which suggests that deeper levels of learning
(e.g. transformative learning; Mezirow, 1991) only
occur when there are sufficient opportunities for critical
reflection and active experimentation.
Future research: a need for theoretical enrichment
It is our hope that future research will not only con-
tinue to examine whether coaching is effective, but also
respond to the need for more theoretical development
in coaching psychology. A strong theoretical framework
is expedient from both an empirical and a practical
perspectives (Grant, 2010; Spence & Oaedes, 2011).
For scholars working in the field of coaching psychol-
ogy, a strong theoretical foundation could purposefully
guide the construction of cumulative knowledge. For
practitioners, insight into how (rather than if) coaching
works can provide guidelines for the improvement of
extant coaching interventions and the development of
new interventions.
One way in which theoretical enrichment of the
coaching literature could be facilitated is by incorporat-
ing theoretical perspectives from several sub-disciplines
of psychology (Grant, 2010), particularly from research
into related fields of developmental interactions such as
therapy, mentoring, and training (D’Abate, Eddy, &
Tannenbaum, 2003). More specifically, the relative
theoretical richness of these fields may serve as a source
of inspiration for theoretical enrichment in four interre-
lated areas of coaching research: the design of coaching
interventions, individual characteristics of the coach and
the coachee, and the relationship between the coach and
the coachee. We provide some specific suggestions for
each of these areas below.
Research concerning the design of coaching interven-
tions may benefit from the literature on training and
mentoring which draws heavily on educational psychol-
ogy and theories on (adult) learning (e.g. theory on
transformative learning; Mezirow, 1991). Theories on
adult learning and its underlying mechanisms can
provide insights that are relevant for increasing the
‘transfer of coaching’(i.e. long-term effectiveness of
coaching interventions). Furthermore, Spence and Oaedes
(2011) have suggested that Deci and Ryans’(1985) self-
determination theory (SDT) is a valuable theoretical
framework for future research on the design of coaching
interventions. Central constructs of SDT such as goal-
setting, intrinsic motivation and the human needs of
competence, relatedness, and autonomy, are crucial for
facilitating durable change within coaches (Ryan, Lynch,
Vansteenkiste, & Deci, 2011).
Research concerning the characteristics of coaches
may find a valuable starting point in the therapy litera-
ture. For example, studies investigating the personal
characteristics of effective therapists have shown that
individual characteristics such as (perceived) empathy
are important predictors of therapy outcomes (Burns &
Nolen-Hoeksema, 1992; Elliott, Bohart, Watson, &
Greenberg, 2011). Recent work in the field of executive
coaching indeed suggests that non-specific factors such
as understanding, encouraging, and listening behaviors
of the coach may be better predictors of coaching effec-
tiveness than specific factors such as the coaching meth-
odology (de Haan, Culpin, & Curd, 2011). In this light,
the influence of constructs related to coaches’ability to
perceive and manage the emotional states of coaches,
such as emotional intelligence (Salovey & Mayer, 1989),
seems especially relevant to examine in future research.
Research concerning the characteristics of coaches
may explore the concept of ‘coachability’that originates
in the sports psychology literature. Coachability is a
multidimensional construct that reflects the combination
of personality traits (e.g. agreeableness, openness to
experience) and motivational components (e.g. achieve-
ment motivation) needed to improve functioning and
performance (Giacobbi, 2000). Furthermore, therapy
research has shown that outcome expectations and self-
efficacy of clients (coaches) play an important role in the
effectiveness of therapeutic interventions (e.g. Goldin
et al., 2012). These constructs will be of similar impor-
tance in the context of coaching.
Finally, studies on coaching and therapy have shown
that the relationship (working alliance) between a coach
and a coachee (therapist and client) has considerable
implications for the effectiveness of interventions (Baron
The Journal of Positive Psychology 13
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
& Morin, 2009; Del Re, Horvath, Flückiger, Symonds,
& Wampold, 2012). With this in mind, both the literature
on similarity attraction (Byrne, 1971) and interpersonal
trust (Mayer, Davis, & Schoorman, 1995) can be used as
theoretical frameworks to examine how functional
relationships between a coach and a coachee can be
established and sustained.
Limitations
Five limitations of the current study should be
mentioned. First, the majority of the studies included in
this meta-analysis relied on self-reports of outcome
measures. According to Peterson (1993), there is a con-
siderable inconsistency between self-reports and other-
reports (e.g. by the supervisor or coach) when evaluating
change in the coachee: self-reports tend to overestimate
the effects of coaching interventions. Hence, self-report
measures of performance seem troublesome (Podsakoff
& Organ, 1986). Therefore, future studies on coaching
should rely less on self-reports and should include other
sources for measuring coaching outcomes such as 360
feedback (see Smither, London, Flautt, & Fargas, 2003
for an example) as well as tangible results.
Another problem with self-reports is that it is difficult
to establish actual change on the outcome measure
(alpha change) rather than respondents’redefinition of
the rating scale (beta change) and/or the concept that is
measured (gamma change). Both beta and gamma
changes are due to a shifting conceptualization of the
outcome as a result of coaching (Peterson, 1993). It
should be noted, however, that also beta and gamma
changes can be conceived as relevant outcomes of
coaching. Transformative learning theory states that
existing belief systems and frames of reference need to
be challenged before deep-level changes will occur
(Mezirow, 1991). The ultimate aim of coaching is to
facilitate deep-level changes and learning (de Haan et al.,
2011). Therefore, more insight into alpha, beta, and
gamma changes and their underlying cognitive structures
(Thompson & Hunt, 1996) is needed because this may
help researchers and practitioners to better design a
coaching intervention and measure its impact.
A second limitation is that most studies in our
meta-analysis did not measure coaching effectiveness
over time (at multiple time-points), making it difficult to
assess the long-term impact of coaching interventions.
Third, the focus on individual-level benefits of coaching
in the studies included in our analysis neglected possible
‘spillover’effects that coaching could have on other
people within an organization. For example, if the
coachee is an executive and his or her coaching results
in improved leadership skills (the functioning of) subor-
dinates and co-workers will benefit as well. Future
research on the effectiveness of coaching could include
subordinate and co-worker perceptions so as to assess
the indirect effects of coaching.
A fourth limitation of this study is that the findings
are based on a relatively small number of studies.
Although we did not find any evidence for publication
bias, Sterne et al. (2011) noted that analyses for publica-
tion bias could produce unreliable results when the num-
ber of studies is small and heterogeneity across studies is
substantial. For this reason, our findings should be inter-
preted with caution. At the same time, however, our
findings consistently showed effectiveness of coaching
across a broad spectrum of outcome measures. Also, our
sensitivity analysis indicated that the removal of the
study by Peterson (1993) did not alter our conclusions.
Fifth, the 1243 participants in the study by Smither
et al. (2003) account for a large proportion of the total
number of participants in the studies we examined. How-
ever, since the effect sizes in this study were much smal-
ler than the average effect sizes over all studies, the
inclusion of the study of Smither et al. has resulted in a
conservative rather than optimistic estimation of the
effectiveness of coaching.
Finally, the general lack of empirical work on
coaching and its weak theoretical foundation has resulted
in a large variety of coaching interventions and
outcomes. As a consequence, the number of comparable
studies suitable for a meta-analytic synthesis was
relatively limited.
Conclusion
Despite its limitations, the current meta-analysis indicates
that coaching can be effectively used as an intervention
in organizations. Furthermore, this study has pointed out
several methodological issues that need to be addressed
in future studies on coaching effectiveness. The biggest
overall limitation of the coaching literature is the lack of
rigorous examinations showing the causal mechanisms
by which coaching interventions are effective. Thus, we
agree with Fillery-Travis and Lane (2006) that it is the
time to shift attention from the question ‘does it work?’
to ‘how does it work?’. This second question can only
be answered by building a firm theoretical framework
that can be used to identify the underlying mechanisms
and processes.
Notes
1. While some scholars explicitly distinguish personal
coaching from organizational coaching
2
(e.g. Grant, 2010),
we take the position that this distinction will be less clear in
practice. For example, when coaching concerns stress issues,
it often taps into the domain of work–life balance and the
intended changes and outcomes will thus affect both the pro-
fessional and the personal functioning of the coached. There-
fore, we do not make this distinction in our endeavor to
answer the question whether coaching is effective.
14 T. Theeboom et al.
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
2. For the transformation of effect sizes based on change
scores standard deviations (repeated measure designs) into
the referent effect size index (based on post intervention
score standard deviations), the comprehensive meta-
analysis (CMA) software’s’default option (r= 0.5) was
used. As recommended by Morris (2000) and Borenstein
et al. (2009), additional analyses based on different values
for r (i.e. r= 0.1 and r= 0.9) were conducted and these
demonstrated similar results. The interested reader can con-
tact the first author for more information.
References
Baron, L., & Morin, L. (2009). The coach–coachee relationship
in executive coaching: A field study. Human Resource
Development Quarterly, 20,85–106.
Berg, I. K., & Szabo, P. (2005). Brief coaching for lasting solu-
tions. New York, NY: W. W. Norton & Company, Inc.
Bond, F. W., & Bunce, D. (2003). The role of acceptance
and job control in mental health, job satisfaction, and
work performance. Journal of Applied Psychology, 88,
1057–1067.
Bono, J. E., Purvanova, R. K., Towler, A. J., & Peterson, D. B.
(2009). A survey of executive coaching practices. Person-
nel Psychology, 62, 361–404.
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein,
H. R. (2005). Comprehensive meta-analysis version 2
[Computer software]. Englewood, NJ: Biostat.
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein,
H. R. (2006). Comprehensive meta-analysis (Version
2.2.027) [Computer software]. Organizational Research
Methods, 11¸ 188–191.
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein,
H. R. (2009). Introduction to meta-analysis. Cornwall:
Wiley.
*Bozer, G., & Sarros, J. C. (2012). Examining the effectiveness
of executive coaching on coaches’performance in the
Israeli context. International Journal of Evidence Based
Coaching and Mentoring, 10,14–32.
Brock, V. G. (2008). Grounded theory of the roots and emer-
gence of coaching (Unpublished Doctoral dissertation).
International University of Professional Studies. Retrieved
from http://libraryofprofessionalcoaching.com/wp-content/
uploads/2011/10/dissertation.pdf
Brown, K. W., & Ryan, R. M. (2003). The benefits of being
present: Mindfulness and its role in psychological well-
being. Journal of Personality and Social Psychology, 84,
822–842.
Brunn, E., & Milczarek, M. (2007). Expert forecast on emerg-
ing psychosocial risks related to occupational safety and
health (European Risk Observatory Report). Retrieved from
http://osha.europa.eu/en/publications/reports/7807118/
Brunning, H. (2006). Executive coaching: Systems-psychody-
namic perspective. London: Karnac.
Burns, D. D., & Nolen-Hoeksema, S. (1992). Therapeutic
empathy and recovery from depression in cognitive-
behavioral therapy: A structural equation model. Journal
of Consulting and Clinical Psychology, 60, 441–449.
Burton, W. N., Conti, D. J., Chen, C. Y., Schultz, A. B., &
Edington, D. W. (1999). The role of health risk factors and
disease on worker productivity. Journal of Occupational
and Environmental Medicine, 41, 863–877.
Byrne, D. E. (1971). The attraction paradigm. New York, NY:
Academic Press.
Carlson, K. D., & Schmidt, F. L. (1999). Impact of experimental
design on effect size: Findings from the research literature
on training. Journal of Applied Psychology, 84, 851–862.
*Cerni, T., Curtis, G. J., & Colmar, S. H. (2010). Executive
coaching can enhance transformational leadership. Interna-
tional Coaching Psychology Review, 5,81–85.
Cochran, W. G. (1954). Some methods for strengthening the
common χ
2
tests. Biometrics, 10, 417–451.
Cohen, J. (1988). Statistical power analysis for the behavioral
sciences (2nd ed.). Hillsdale, NJ: Eribaum.
Cook, T. D., & Campbell, D. T. (1979). Quasi-experimentation:
Design and analysis issues for field studies. Boston, MA:
Houghton Mifflin Company.
Couto, D., & Kauffman, C. (2009). What can coaches do for
you? Harvard Business Review,1–8. Retrieved from http://
hbr.org/2009/01/what-can-coaches-do-for-you/ar/1
Cox, E., Bachkirova, T., & Clutterbuck, D. (2009). The com-
plete handbook of coaching. London: Sage.
D’Abate, C., Eddy, E., & Tannenbaum, S. T. (2003). What’sin
a name? A literature-based approach to understanding men-
toring, coaching, and other constructs that describe devel-
opmental interactions. Human Resource Development
Review, 2, 360–384.
Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and
self-determination in human behavior. New York, NY: Ple-
num.
Deci, E. L., & Ryan, R. M. (2000). The ‘what’and ‘why’of
goal pursuits: Human needs and the self-determination of
behavior. Psychological Inquiry, 11, 227–268.
De Haan, E., Culpin, V., & Curd, J. (2011). Executive coaching
in practice: What determines helpfulness for clients of
coaching? Personnel Review, 40,24–44.
De Meuse, K. P., Dai, G., & Lee, R. J. (2009). Evaluating the
effectiveness of executive coaching: Beyond ROI? Coach-
ing: An International Journal of Theory, Research and
Practice, 2,117–134.
Del Re, A. C., Horvath, A. O., Flückiger, C., Symonds, D., &
Wampold, B. E. (2012). Therapist effects in the therapeutic
alliance-outcome relationship: A restricted-maximum likeli-
hood meta-analysis. Clinical Psychology Review, 32, 642–
649.
De Shazer, S. (1988). Clues: Investigating solutions in brief
therapy. New York, NY: W.W. Norton & Company Inc.
Duijts, S. F. A. P., Kant, I. P., van den Brandt, P. A. P., & Swa-
en, G. M. H. P. (2008). Effectiveness of a preventive
coaching intervention for employees at risk for sickness
absence due to psychosocial health complaints: Results of a
randomized controlled trial. Journal of Occupational &
Environmental Medicine, 50, 765–776.
*Egan, T., & Song, Z. (2005). A longitudinal quasi-experiment
on the impact of executive coaching. Paper presented at the
20th Annual Conference of the Society for Industrial and
Organizational Psychology, Los Angeles.
Egger, M., Smith, G. D., Schneider, M., & Minder, C. (1997).
Bias in meta-analysis detected by a simple, graphical test.
British Medical Journal, 315, 629–634.
Elliott, R., Bohart, A. C., Watson, J. C., & Greenberg, L. S.
(2011). Empathy. Psychotherapy, 48,43–49.
*Evers, W. J., Brouwers, A., & Tomic, W. (2006). A quasi-
experimental study on management coaching effectiveness.
Consulting Psychology Journal: Practice and Research, 58,
174–182.
Feldman, D. C., & Lankau, M. J. (2005). Executive coaching:
A review and agenda for future research. Journal of Man-
agement, 31, 829–848.
The Journal of Positive Psychology 15
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
Fillery-Travis, A., & Lane, D. (2006). Does coaching work or
are we asking the wrong question? International Coaching
Psychology Review, 1,23–35.
*Finn, F. A. (2007). Leadership development through executive
coaching: The effects on leaders’psychological states and
transformational leadership behavior (Unpublished
Doctoral dissertation). Queensland University of Technol-
ogy. Retrieved from http://eprints.qut.edu.au/17001/
Freire, T. (2013). Positive psychology approaches. In J. Pass-
more, D. Peterson, & T. Freire (Eds.), The Wiley-Blackwell
handbook of the psychology of coaching and mentoring
(pp. 426–442). West Sussex: Wiley-Blackwell.
Fusar-Poli, P., Bonoldi, I., Yung, A. R., Borgwardt, S., Kemp-
ton, M. J., Valmaggia, L., & McGuire, P. (2012). Predicting
psychosis: Meta-analysis of transition outcomes in individ-
uals at high clinical risk. Archives of General Psychiatry,
69, 220.
Gallwey, T. W. (1974). The inner game of tennis (1st ed.).
New York, NY: Random House.
Giacobbi, P. R. (2000). The athletic coachability scale: Con-
struct conceptualization and psychometric analyses (Doc-
toral dissertation). University of Tennessee, Knoxville.
Retrieved from http://sunzi.lib.hku.hk/ER/detail/hkul/
2688806
Goldin, P. R., Ziv, M., Jazaieri, H., Werner, K., Kraemer, H.,
Heimberg, R. G., & Gross, J. J. (2012). Cognitive reap-
praisal self-efficacy mediates the effects of individual cog-
nitive-behavioral therapy for social anxiety disorder.
Journal of Consulting and Clinical Psychology, 80, 1034–
1043.
Grant, A. M. (2001). Towards a psychology of coaching: The
impact of coaching on metacognition, mental health and
goal attainment (Unpublished Doctoral dissertation). Mac-
quarie University, Sydney. Retrieved from http://www.eric.
ed.gov/PDFS/ED478147.pdf
*Grant, A. M. (2003). The impact of life coaching on goal-
attainment, metacognition and mental health. Social Behav-
ior and Personality, 31, 253–264.
Grant, A. M. (2008). Personal life coaching for coaches-in-
training enhances goal attainment, insight and learning.
Coaching: An International Journal of Theory, Research
and Practice, 1,54–70.
*Grant, A. M. (2010). The development of coaching psychol-
ogy. International Coaching Psychology Review, 1,12–
22.
Grant, A. M. (2013). The efficacy of coaching. In J. Passmore,
D. Peterson, & T. Freire (Eds.), Handbook of the psychol-
ogy of coaching and mentoring (pp. 15–39). West Sussex:
Wiley-Blackwell.
Grant, A. M., & Cavanagh, M. J. (2004). Toward a profession
of coaching: Sixty-five years of progress and challenges for
the future. International Journal of Evidence-based Coach-
ing and Mentoring, 2,1–16.
Grant, A. M., & Cavanagh, M. J. (2007). Evidence-based
coaching: Flourishing or languishing? Australian Psycholo-
gist, 42, 239–254.
*Grant, A. M., Curtayne, L., & Burton, G. (2009). Executive
coaching enhances goal attainment, resilience and work-
place well-being: A randomized controlled study. The Jour-
nal of Positive Psychology, 4, 396–407.
*Grant, A. M., Green, L. S., & Rynsaardt, J. (2010). Develop-
mental coaching for high school teachers: Executive coach-
ing goes to school. Consulting Psychology Journal:
Practice and Research, 3, 151–168.
Grant, A. M., & O’Connor, S. A. (2010). The differential
effects of solution-focused and problem-focused coaching
questions: A pilot study with implications for practice.
Industrial and Commercial Training, 42, 102–111.
Grant, A. M., Passmore, J., Cavanagh, M., & Parker, H.
(2010). The state of play in coaching. International Review
of Industrial & Organizational Psychology, 25, 125–168.
*Green, L. S., Grant, A. M., & Rynsaardt, J. (2007). Evidence-
based life coaching for senior high school students: Build-
ing hardiness and hope. International Coaching Psychology
Review, 2,24–32.
*Green, L. S., Oades, L. G., & Grant, A. M. (2006). Cogni-
tive-behavioural, solution-focused life coaching: Enhancing
goal striving, well-being and hope. Journal of Positive Psy-
chology, 1, 142–149.
Greenhaus, J. H., Parasuraman, S., & Wormley, W. M. (1990).
Effects of race on organizational experience, job perfor-
mance evaluations, and career outcomes. Academy of Man-
agement Journal, 33,64–86.
Hart, V., Blattner, J., & Leipsic, S. (2001). Coaching versus
therapy: A perspective. Consulting Psychology Journal:
Practice & Research, 53, 229–237.
Hedges, L. V. (1981). Distribution theory for Glass’s estimator
of effect size and related estimators. Journal of Educational
Statistics, 6, 107–128.
Hedges, L. V., & Cooper, H. (1994). The handbook of research
synthesis. New York, NY: Sage.
Hedges, L., & Olkin, I. (1985). Statistical methods for meta-
analysis. San Diego, CA: Academic Press.
Higgins, J. P. T., & Thompson, S. G. (2002). Quantifying het-
erogeneity in a meta-analysis. Statistics in Medicine, 21,
1539–1558.
Higgins, J. P. T., Thompson, S. G., Deeks, J. J., & Altman,
D. G. (2003). Measuring inconsistency in meta-analyses.
British Medical Journal, 327, 557–560.
Hunter, J. E., & Schmidt, F. L. (1990). Methods of meta-analy-
sis: Correcting error and bias in research findings. New-
bury Park, CA: Sage.
Hunter, J. E., & Schmidt, F. L. (2004). Methods of meta-analy-
sis: Correcting error and bias in research findings (2nd
ed.). Thousand Oaks, CA: Sage.
International Coach Federation. (2012). Retrieved May, 16, 2012
from, http://www.coachfederation.org/about-icf/overview/
Johnson, B. T., Mullen, B., & Salas, E. (1995). Comparison of
three major meta-analytic approaches. Journal of Applied
Psychology, 80,94–106.
Judge, T. A., Thoresen, C. J., Bono, J. E., & Patton, G. K.
(2001). The job satisfaction-job performance relationship:
A qualitative and quantitative review. Psychological Bulle-
tin, 127, 376–407.
Kampa-Kokesch, S., & Anderson, M. Z. (2001). Executive
coaching: A comprehensive review of the literature. Con-
sulting Psychology Journal: Practice and Research, 53,
205–228.
Kilburg, R. R. (1996). Toward a conceptual understanding and
definition of executive Coaching. Consulting Psychology
Journal: Practice and Research, 48, 134–144.
Kim, J. S. (2008). Examining the effectiveness of solution-
focused brief therapy: A meta-analysis. Research on Social
Work Practice, 18, 107–116.
*Kochanowski, S., Seifert, C. F., & Yukl, G. (2010). Using
coaching to enhance the effects of behavioral feedback to
managers. Journal of Leadership & Organizational Studies,
17, 363–369.
16 T. Theeboom et al.
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
Landis, J. R., & Koch, G. G. (1977). The measurement of
observer agreement for categorical data. Biometrics, 33,
159–174.
Latham, G. P. (2007). Theory and research on coaching prac-
tices. Australian Psychologist, 42, 268–270.
Leonard-Cross, E. (2010). Developmental coaching: Business
benefit–Fact or fad? An evaluative study to explore the
impact of coaching in the workplace. International Coach-
ing Psychology Review, 5,36–47.
Linley, P. A., & Harrington, S. (2005). Positive psychology
and coaching psychology: Perspectives on integration. The
Coaching Psychologist, 1,13–14.
Locke, E. A., & Latham, G. P. (2002). Building a practically
useful theory of goal setting and task motivation: A 35-year
odyssey. American Psychologist, 57, 705–717.
Lovibond, P. F., & Lovibond, S. H. (1995). The structure of
negative emotional states: Comparison of the depression
Anxiety stress scales (DASS) with the beck depression and
anxiety inventories. Behaviour Research and Therapy, 33,
335–343.
*Luthans, F., & Peterson, S. J. (2004). 360-degree feedback
with systematic coaching: Empirical analysis suggests a
winning combination. Human Resource Management, 42,
243–256.
Maslach, C., & Jackson, S. E. (1986). MBI: Maslach Burnout
Inventory manual (research edition) (1st ed. 1981). Palo
Alto, CA: Consulting Psychologists Press.
Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An
integrative model of organizational trust. Academy of Man-
agement Review, 20, 709–734.
McGaw, B., & Glass, G. (1980). Choice of the metric for effect
size in meta-analysis. American Educational Research
Journal, 17, 325–337.
Mezirow, J. (1991). Transformative dimensions of adult learn-
ing. San Francisco, CA: Jossey-Bass.
*Moen, F., & Skaalvik, E. (2009). The effect from coaching on
performance psychology. International Journal of Evidence
Based Coaching and Mentoring, 7,31–49.
Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009).
Preferred reporting items for systematic reviews and meta-
analyses: The PRISMA statement. Annals of Internal Medi-
cine, 151, 264–269.
Morris, S. B. (2000). Distribution of the standardized mean
change effect size for meta-analysis on repeated measures.
British Journal of Mathematical and Statistical Psychology,
53(1), 17–29.
Morris, S. B., & DeShon, R. P. (1997). Correcting effect sizes
computed from factorial ANOVA for use in meta-analysis.
Psychological Methods, 2, 192–199.
Morris, S. B., & DeShon, R. P. (2002). Combining effect size
estimates in meta-analysis with repeated measures and
independent-groups designs. Psychological Methods, 7,
105–125.
Mullen, E. J. (1994). Framing the mentoring relationship as an
information exchange. Human Resource Management
Review, 4, 257–281.
National Research Council (1992). Combining information:
Statistical issues and opportunities for research. Washing-
ton, WA: National Academy Press.
Organ, D. W., & Ryan, K. (1995). A meta-analytic review of
attitudinal and dispositional predictors of organizational
citizenship behavior. Personnel Psychology, 48, 775–802.
Passmore, J. (Ed.). (2010). Excellence in coaching: The indus-
try guide. London: Kogan Page.
Passmore, J., & Fillery-Travis, A. (2011). A critical review of
executive coaching research: A decade of progress and
what’s to come. Coaching: An International Journal of
Theory, Practice & Research, 4,70–88.
*Peterson, D. B. (1993). Measuring change: A psychometric
approach to evaluating individual training outcomes. Sym-
posium conducted at the Eighth Annual Conference of the
Society for Industrial and Organizational Psychologists,
San Francisco, CA.
Podsakoff, P. M., & Organ, D. W. (1986). Self-reports in orga-
nizational research: Problems and prospects. Journal of
Management, 12, 531–544.
*Poepsel, M. A. (2011). The impact of an online evidence-based
coaching program on goal striving, subjective well-being,
and level of hope (Unpublished Doctoral dissertation),
Minneapolis, MN. Harold Abel School of Social and
Behavioral Sciences. Retrieved from http://gradworks.
umi.com/3456769.pdf
Porter, L. W., Steers, R. M., Mowday, R. T., & Boulian, P. V.
(1974). Organizational commitment, job satisfaction, and
turnover among psychiatric technicians. Journal of Applied
Psychology, 59, 603–609.
Rosenthal, R. (1991). Meta-analytic procedures for social
research (Rev. ed.). Newbury Park, CA: Sage.
Ryan, R. M., Lynch, M. F., Vansteenkiste, M., & Deci, E. L.
(2011). Motivation and autonomy in counseling, psycho-
therapy, and behavior change: A look at theory and prac-
tice. The Counseling Psychologist, 39, 193–260.
Salovey, P., & Mayer, J. D. (1989). Emotional intelligence.
Imagination, Cognition and Personality, 9, 185–211.
Schwarzer, R., & Jerusalem, M. (1995). Generalized self-effi-
cacy scale. In J. Weinman, S. Wright, & M. Johnston
(Eds.), Measures in health psychology: A user’s portfolio.
Causal and control beliefs (pp. 35–37). Windsor: NFER-
NELSON.
Seligman, M. E. (2007). Coaching and positive psychology.
Australian Psychologist, 42, 266–267.
Seligman, M. E., & Csikszentmihalyi, M. (2000). Positive
psychology: An introduction. American Psychologist, 55(1),
5–14.
Smith, P. C., Kendall, L. M., & Hulin, C. L. (1969). The mea-
surement of satisfaction in work and retirement. Chicago:
Rand McNally.
*Smither, J. W., London, M., Flautt, R., Vargas, Y., & Kucine,
I. (2003). Can working with an executive coach improve
multisource feedback ratings over time? A quasi-experi-
mental field study. Personnel Psychology, 56,23–44.
Spence, G. (2008). New directions in evidence-based coaching:
Investigations into the impact of mindfulness training on
goal attainment and well-being (Unpublished Doctoral dis-
sertation). University of Sydney, Sydney. Retrieved from
http://ses.library.usyd.edu.au/bitstream/2123/2469/1/New%20
Directions%20in%20the%20Psychology%20of%20Coac
hing%2 0(Spence,%202006).pdf
Spence, G. B., Cavanagh, M. J., & Grant, A. M. (2008). The
integration of mindfulness training and health coaching: An
exploratory study. Coaching: An International Journal of
Theory, Research and Practice, 1, 145–163.
*Spence, G. B., & Grant, A. M. (2007). Professional and peer
life coaching and the enhancement of goal striving and
well-being: An exploratory study. Journal of Positive Psy-
chology, 2, 185–194.
Spence, G. B., & Oaedes, L. G. (2011). Coaching with self-
determination in mind: Using theory to advance evidence-
The Journal of Positive Psychology 17
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
based coaching practice. International Journal of Evidence
Based Coaching and Mentoring, 9,37–55.
Sterne, J. A., Sutton, A. J., Ioannidis, J., Terrin, N., Jones, D.
R., Lau, J., & Higgins, J. (2011). Recommendations for
examining and interpreting funnel plot asymmetry in meta-
analyses of randomised controlled trials. British Medical
Journal, 343(7818), 302–307.
Sue-Chan, C., & Latham, G. P. (2004). The relative effective-
ness of external, peer, and self-coaches. Applied Psychol-
ogy, 53, 260–278.
Tepper, B. J. (1995). Upward maintenance tactics in supervi-
sory mentoring and nonmentoring relationships. Academy
of Management Journal, 38, 1191–1205.
Thompson, R. C., & Hunt, J. G. (1996). Inside the black box
of alpha, beta, and gamma change: Using a cognitive-pro-
cessing model to assess attitude structure. Academy of
Management Review, 4, 655–690.
Waldron, V. R. (1991). Achieving communication goals in
superior-subordinate relationships: The multi-functionality
of upward maintenance tactics. Communications Mono-
graphs, 58, 289–306.
Whitmore, J. (1992). Coaching for performance. San Diego,
CA: Pfeiffer.
Wright, T. A., & Cropanzano, R. (2000). Psychological
well-being and job satisfaction as predictors of job per-
formance. Journal of Occupational Health Psychology,
5,84–94.
Wrzesniewski, A., & Dutton, J. (2001). Crafting a job: Employ-
ees as active crafters of their work. Academy of Manage-
ment Review, 26, 179–201.
18 T. Theeboom et al.
Downloaded by [Tim Theeboom] at 09:46 16 September 2013
