Content uploaded by Anthony M Grant
Author content
All content in this area was uploaded by Anthony M Grant on Feb 19, 2015
Content may be subject to copyright.
The Journal of Positive Psychology
Vol. 7, No. 4, July 2012, 334–348
Development and validation of the solution-focused inventory
Anthony M. Grant*, Michael J. Cavanagh, Sabina Kleitman, Gordon Spence, Michaela Lakota and
Nickolas Yu
Coaching Psychology Unit, School of Psychology, University of Sydney,
Sydney, NSW 2006, Australia
(Received 12 September 2011; final version received 22 May 2012)
Solution-focused coaching and solution-focused therapy are strengths-based approaches which emphasize
people’s resources and resilience and how these can be used in the pursuit of purposeful, positive change. The
Solution-focused Inventory (SFI) is a 12-item scale with three subscales: Problem Disengagement, Goal
Orientation and Resource Activation. Three studies in this article provide support for the validity of the SFI as a
measure of solution-focused thinking. The SFI negatively correlated with psychopathology and positively
correlated with measures of well-being, resilience and perspective taking. Test–retest reliability over 16 weeks was
0.84. Cronbach’s for the 12-item scale was 0.84. It also demonstrates sensitivity to purposeful change in that
participation in a leadership development coaching intervention was associated with significantly increased scores
on the SFI, whilst scores for the control group did not change.
Keywords: solution-focused; coaching; positive psychology; brief solution-focused therapy
Introduction
The solution-focused approach is a strengths-based
approach which emphasizes people’s resources and
resilience and how these can be used in the pursuit of
purposeful, positive change. It is now widely incorpo-
rated into a range of coaching and therapeutic
methodologies (e.g. Iveson, George, & Ratner, 2012;
Roeden, Maaskant, Bannink, & Curfs, 2011). Indeed,
a recent large-scale US-based study of family therapists
found that more than half of all family therapists listed
solution-focused therapy as one of the three most
common interventions used along with cognitive
behavioural therapy and Bowen’s family systems
theory (Bradley, Bergen, Ginter, Williams, & Scalise,
2010).
Solution-focused approaches are attracting increas-
ing interest in the scientific literature. A search of
PsycINFO in January 2012 revealed over 450 publica-
tions on solution-focused approaches. Of these, 126
involved quantitative data drawn from both therapeutic
and general populations, covering all age groups and
many industry sectors. The majority of these studies
suggest that the solution-focused approach is an effec-
tive means of producing change (for a formal review, see
Stams, Dekovic, Buist, & de Vries, 2006).
Despite this growing interest and practical appli-
cation, there has been little research into the mecha-
nisms by which solution-focused approaches operate.
One of the likely reasons for this is that there has not
yet been developed a validated theoretically grounded,
multifactorial instrument capable of reliably measuring
the key psychological mechanisms thought to underpin
solution-focused change. This article reports the devel-
opment and validation of an instrument (the Solution-
focused Inventory; SFI) designed to fill this gap.
Theoretical foundations to the solution-focused
approach
At its core, the solution-focused approach is decep-
tively simple. It assumes that people possess the
resources necessary to resolve their difficulties or
problems, and that time in the coaching or therapy
session is better spent identifying the desired solution
state and focusing on pathways to achieve that state,
rather than exploring the origins of the presenting
problem or the patterns of thought that create and
maintain it (Jackson & McKergow, 2002).
Theoretically, the solution-focused approach holds
that change is constrained or enabled by the way in
which the client (and therapist or coach) think and talk
about events. In other words, problems and solutions
are not things necessarily given in reality, but are
constructed in the discourse between the client and
others in the client’s world (Cavanagh & Grant, 2010;
de Shazer, 1988; O’Connell, 1998).
*Corresponding author. Email: anthony.grant@sydney.edu.au
ISSN 1743–9760 print/ISSN 1743–9779 online
ß2012 Taylor & Francis
http://dx.doi.org/10.1080/17439760.2012.697184
http://www.tandfonline.com
Downloaded by [University of Sydney] at 14:58 26 June 2012
Solution-focused thinking can be contrasted with
problem-focused approaches. A problem-focused
approach assumes that by understanding the causal
structure of a person’s difficulty, effective pathways to
action will emerge. In contrast, the solution-focused
approach eschews exploration of causal aetiology,
focusing instead directly on how to create the desired
change. Indeed, it holds that the complexity of life
means that a search for causal aetiology may well be
futile, and could even lead to a narrowing of possible
actions, eventual undermining of self-efficacy and
reductions in motivation and resilience (Cavanagh,
2006; Cavanagh & Grant, 2010; McKergow &
Jackson, 2005; McKergow & Stellamans, 2011).
At its extreme, problem-focused thinking may even
lead to a debilitating cycle of rumination – a persistent
cognitive focus on one’s problems, along with attempts
to identify the causes, meanings and consequences of
those experiences or problems (Nolen-Hoeksema &
Morrow, 1991; Trapnell & Campbell, 1999).
Rumination has been found to pose a risk factor for
both the onset of depression and depressive relapse
(Nolen-Hoeksema, 2000), and to negatively predict
subjective well-being and life satisfaction (Brown &
Ryan, 2003; Takano & Tanno, 2009), as well as leading
to reduced concentration, biased negative memory
retrieval and impaired problem-solving skills
(Lyubomirsky & Nolen-Hoeksema, 1995).
In contrast, solution-focused cognitive processing is
characterized by a style of thinking that eschews
excessive focus on problems and their causes. It focuses
on identifying approach goals and unnoticed resources
and finding multiple pathways to achieving those goals
(Cavanagh & Grant, 2010; Grant, 2001). By doing so it
enlarges the range of potential actions open to the
client.
Thus a solution-focused thinking style can be
expected to be associated with well-being and positive
affect. This is because reflecting on one’s goals and
thinking about ways to attain those goals tends to
stimulate pathways thinking (Snyder, Rand, &
Sigmon, 2002) and increases self-efficacy (Bandura,
1982), both of which are frequently associated with
well-being (Peterson, 2000; Sheldon, Elliot, Kim, &
Kasser, 2001; Sheldon, Kasser, Smith, & Share, 2002).
Furthermore, actively working on developing solutions
in the pursuit of personally valued goals is likely
to build self-efficacy, resilience and psychological
flexibility (Beasley, Thompson, & Davidson, 2003;
Kashdan & Rottenberg, 2010; Peterson, 2006).
By focusing on goals, resources and pathways, the
solution-focused approach attempts to facilitate disen-
gagement from problem-focused thinking and break
the debilitating cycles of rumination that often keep
clients focused on weakness and deficits (for further
discussion on these points see Robinson & Tamir,
2011; Cavanagh & Grant, 2010).
As yet, there is no reliable measure that is
specifically designed to assess the psychological factors
associated with solution-focused approaches. Given
the widespread and growing importance of solution-
focused interventions, this represents a serious gap in
evidence-based research. To address this issue, this
article reports on the validation of a measure
specifically designed to assess solution-focused pro-
cessing (Grant, 2003).
Application for cognitive-behavioural and positive
psychology approaches
As solution construction is a central task in many
approaches to fostering change, the development of an
instrument able to assess changes in orientation
towards solutions, resource identification and pro-
blem-focused thinking has application beyond research
into solution-focused coaching or solution-focused
therapy (Grant, Franklin, & Langford, 2002). Such
an instrument would be useful to assess the develop-
ment of solution-focused thinking, for example,
in cognitive-behavioural therapy where the therapist
typically turns the clients’ attention to the development
of strategies aimed at building positive coping skills
and personal resilience after dysfunctional or self-
limiting thoughts, feelings and behaviours have been
identified and addressed (Beck, 1995; Dryden, 1987;
Neenan & Dryden, 2002).
Such an instrument would also be of use in
assessing the psychological mechanisms underlining
positive psychology interventions. Because these focus
on building personal strengths and orienting the
client’s attention towards the construction of solutions
and related constructs such as hope, subjective and
psychological well-being (PWB) and psychological
flexibility (Grant & Spence, 2010; Kashdan &
Rottenberg, 2010; Peterson, 2006), and avoid pathol-
ogy-focused diagnosis and excessive problem analysis
(Biswas-Diener, 2010), they have much in common
with the solution-focused approach, and may well rely
on similar psychological mechanisms. Thus the devel-
opment of a short, reliable instrument capable of
identifying the processes involved in solution-focused
thinking may prove useful in understanding some of
the active components of cognitive-behavioural
and positive psychology approaches to therapy and
change.
Paucity of existing solution-focused measures
To the best of our knowledge, there has been only one
peer-reviewed published measure specifically related to
solution-focused approaches; The Solution Building
Inventory (SBI; Smock, McCollum, & Stevenson,
2010). According to the authors, this scale attempted
The Journal of Positive Psychology 335
Downloaded by [University of Sydney] at 14:58 26 June 2012
to comprehensively measure all aspects of ‘solution-
building’ by clients in brief solution-focused therapy
(BSFT). Following a review of the literature, Smock
et al. (2010) developed a 22-item measure based on
three factors that, they argued, theoretically underpin
the solution-focused approach: (a) clearly identifying
the solution; (b) awareness of exceptions to the
problem and (c) developing a hope in the future.
Items included ‘I can recognize things that I can do,
even though it seems that the problem is someone else’;
‘I can recognize in others when things may be going
better for me’ and ‘There are times when I am really
proud of how I am able to handle difficult situations’.
However, exploratory factor analysis (EFA) of the
SBI yielded eight factors rather than the expected three
(Smock et al., 2010). When this analysis was repeated
with a second data set, the researchers were again
unable to confirm the hypothesized three-factor model.
Their data indicated the only viable model using their
items was a six-item, single-factor scale that did not
operationalize the proposed underlying theoretical
framework. Furthermore, its psychometric properties
are uncertain, and important properties such as test–
retest reliability and predictive validity were not
reported.
In order to further develop our understanding of
solution-focused approaches and more deeply explore
their utility in coaching, therapy and positive psychol-
ogy, a new theoretically grounded and empirically
validated measure that can assess processes central to
solution-focused thinking is needed. A brief, reliable
and valid measure of solution-focused thinking would
have potential to contribute to the field, both in terms
of assessing change in the psychological mechanisms
targeted by solution-oriented interventions, and more
broadly in terms of generating insight into the
psychological processes central to purposeful, positive
change. For such a measure to be useful to both
practitioners and researchers, it needs to be theoreti-
cally grounded, simple to administer and easy to score.
The aim of this study was to develop such a measure.
The SFI: Theoretical factor structure
In developing the SFI, a review of the solution-focused
literature identified two broad themes. These were,
firstly, an orientation towards solution construction
via the use of approach goals and active self-regulation
and secondly, an orientation towards noticing excep-
tions to the problem and utilizing resources and
strengths (see Grant, 2011 for a broader theoretical
and philosophical discussion of the underpinnings of
solution-focused interventions and the SFI).
In addition, we hypothesized a third factor con-
cerned with the person’s capacity to disengage from
problem-focused thinking. We reasoned that
solution-focused thinking is more than just goal setting
and resource awareness – it is also based on a mindset
that orients the person towards solutions and explicitly
away from problem-focused processing (see also
Biswas-Diener, 2010; Dweck, 2006). Thus, in addition
to assessing goal orientation (GO) and resource
utilization, a valid measure of solution-focused think-
ing should also include items that assess the relative
absence of problem-focused thinking.
The following three studies were conducted to
assess the factor structure, reliability and validity of the
SFI. Study one reports the exploratory and confirma-
tory factor analyses of the instrument. This study
sought to understand whether the factor structure was
stable and reflected the proposed underlying theoret-
ical framework, and which of three models best fit the
data. The models we examined were: (Model 1) a
unitary factor model (which suggests a single broad
latent factor underlying solution-focused behaviours);
(Model 2) a three independent, yet inter-correlated
factor model (e.g. solution-focused behaviours can be
viewed as being underlined by three factors obtained
from the correlations between original items) and
(Model 3) a hierarchical model with three factors
subordinate to a single higher order factor (e.g. a
second-order factor is derived on a basis of correla-
tions between first-order factors, supporting the
assumption of broad tendencies underlying three
conceptually related clusters of problem-focused
behaviours). Study two examined the validity of the
measure comparing it against a range of outcome
variables. Study three assessed the test–retest reliability
of the instrument as a whole and its responsiveness to
change following coaching.
Study one: Internal validation of the SFI
Method
Item construction
As discussed above, we suggest there are three key
factors theoretically deemed to be central to solution-
focused thinking: (1) a focus towards desired goal
states; (2) a focus on recognizing and utilizing strengths
and resources and (3) a focus on disengaging from
problems and problem-focused thinking. We named
these themes GO, resource activation (RA), and
problem disengagement (PD), respectively, and devel-
oped questionnaire items that reflected each theme.
Drawing on key themes reported in the solution-
focused practitioner literature, questionnaire items
were developed by the first and sixth authors (Grant,
2011): the GO items were designed to encapsulate the
key features of goal-focused self-regulation which
underpins an orientation towards solution construc-
tion (e.g. Locke & Latham, 2002). The RA items were
designed to reflect the core aspects of RA widely
336 A.M. Grant et al.
Downloaded by [University of Sydney] at 14:58 26 June 2012
reported in the solution-focused literature
(e.g. de Shazer, 1988; Furman & Ahola, 1992). The
PD items were designed to capture the key features of
problem-saturated thinking which impedes goal pur-
suit (e.g. van Randenborgh, Huffmeier, LeMoult, &
Joormann, 2010), and by reverse scoring these items, it
is possible to create an analogue measure of PD. There
were 14 items in the initial item pool which were scored
using a six-point scale (1 ¼strongly disagree;
6¼strongly agree).
Participants and procedure
Data were gathered in two stages. Stage one involved
the initial collection of data using a professional
sample (N¼242; 157 females; 85 males) of members
of the legal and health professions who were taking
part in an intervention study on leadership coaching
(mean age ¼41.92 years; standard deviation
(SD) ¼8.92) (Cavanagh, Spence, & Atkins, 2012). To
determine the factorial structure of the scale, the data
from this sample were subjected to EFA using the
maximum likelihood (ML) method with Promax rota-
tion. The analysis was conducted using SPSS v17.
In stage two, a second data set was collected from a
sample of undergraduate psychology students (Lakota,
2010). These students (N¼322; 226 females; 96 males)
participated in the study for course credit (mean
age ¼19.8 years; SD ¼4.34 years). These data were
used to confirm the factorial structure found in stage
one. The confirmatory factor analysis (CFA) using the
ML method was conducted using AMOS v7.0
(Arbuckle, 2006).
In both samples the questionnaire was completed in
small group settings, alongside other measures related
to different studies. Only data relevant to this scale are
reported here.
Results for stage one: EFA
Descriptive statistics and correlations for the 14-item
version of the SFI
Descriptive statistics for all the 14 items of the SFI for
the professional and university student samples are
presented in Table 1. Whilst the student sample has
slightly lower means on the items measuring PD, these
differences are not statistically significant.
Table 1. Descriptive statistics (minimum and maximum scores, mean and SD) for each item of the initial 14-item version of the
SFI scale in professional (N¼242) and university student (N¼322) samples.
University students Professional sample
Original item no. Minimum Maximum MSD Minimum Maximum MSD
SFI01 I tend to spend more time analysing
my problems than working on
possible solutionsÕ
1 6 3.98 1.20 1 6 3.21 1.16
SFI02 I tend to get stuck in thinking about
problemsÕ
1 6 3.83 1.23 1 6 2.96 1.13
SFI03 There is always a solution to every
problem
1 6 4.64 1.07 1 6 4.10 1.33
SFI04 I tend to focus on the negativeÕ1 6 4.40 1.27 1 6 3.69 1.26
SFI05 I’m not very good at noticing when
things are going wellÕ
1 6 4.31 1.27 1 6 3.90 1.28
SFI06 There are always enough resources to
solve a problem if you know where
to look
1 6 4.30 1.19 1 6 4.27 1.07
SFI07 I know how to use my personal
strengths
2 6 4.62 0.83 1 6 4.23 0.99
SFI08 Most people are more resilient than
they realise
2 6 4.93 0.79 2 6 4.54 0.87
SFI09 When things do go wrong I try to
learn from the experience
2 6 5.14 0.70 1 6 4.81 0.81
SFI10 Setbacks are a real opportunity to
turn failure into success
2 6 4.67 0.84 1 6 4.31 0.99
SFI11 I imagine my goals and then work
towards them
2 6 4.48 0.96 1 6 4.23 1.06
SFI12 I keep track of my progress towards
my goals
1 6 4.04 1.09 1 6 3.74 1.14
SFI13 I’m very good at developing effective
action plans
1 6 3.51 1.08 1 6 3.66 1.21
SFI14 I always achieve my goals 1 6 3.92 0.92 1 6 3.52 1.14
Note: Õindicates reverse-scored items.
The Journal of Positive Psychology 337
Downloaded by [University of Sydney] at 14:58 26 June 2012
Table 2 presents correlations between all the 14
items of the original SFI scale for each sample.
Correlations coefficients for the professional and
student samples are summarized in lower and upper
triangles, respectively.
Inspection of correlations between the 14 items of
the original SFI scale suggests several robust patterns.
First, the items correlate positively with each other
across both samples. Second, there are several clusters
present within the correlations, with some items
correlating consistently higher with some group of
items, but not with the others. Overall, the correlations
range from 0.03 (p>0.05) to 0.65 (p<0.01) in the
professional sample and from 0.09 (p>0.05) to 0.72
(p<0.01) in the student sample. Third, correlations for
both samples are similar, but coefficients are slightly
more homogeneous for the student sample than for the
professional sample. Overall, however, these patterns
are consistent with our expectations, and suggest this
instrument is capturing a multidimensional construct,
with a possible broad second-order factor.
EFA: Determining the structure of the SFI
An EFA analysis was performed on the professional
sample. Although there were four eigenvalues higher
than one, the inspection of a scree plot clearly
indicated the presence of three distinct factors only.
Moreover, when four factors were extracted, one
factor was defined by two strong loadings only.
Thus, the three-factor solution model was adopted as
the most parsimonious.
Two items (‘I know how to use my personal
strengths’ and ‘When things do go wrong I try to learn
from the experience’) from the initial 14 items had very
low factor loadings and communality estimates. As the
problems with these two items were consistent across
both three- and four-factor solutions, we decided to
exclude them from further analysis. The EFA was then
repeated on the 12 retained items and the relevant
pattern matrix is presented in Table 3. Three resulting
factors accounted for 30%, 9.8% and 7.4% of the
variance, respectively, and this solution explained
47.2% of the total variance.
We tested the data set for univariate and multivar-
iate normalities. According to conventional criteria
(DeCarlo, 1997), all of the items were within the
accepted normal distribution parameters. We also
examined the suitability of the data for factor analysis
by using the Kaiser–Meyer–Olkin (KMO) measure of
sampling adequacy (Kaiser, 1970), as well as Bartlett’s
test of sphericity (Bartlett, 1950). The KMO measure
of sampling adequacy was 0.83, above the commonly
recommended value of 0.60, and Bartlett’s test of
sphericity was significant (
2
(91) ¼1167.06, p<0.01).
Interpretation of this model is as follows.
Factor 1: PD
This factor is defined by items that reflect a tendency
towards becoming enmeshed in, and over-emphasizing,
negative thinking about problems. These items were
reversed-coded creating an analogue measure of PD
and preserving the unidirectional nature of the overall
scale.
Factor 2: GO
Salient loadings on this factor are from the items that
reflect the tendency to envisage one’s goals, to set up
action plans to achieve them and to monitor progress –
aspects central to the solution-focused approach.
It should be noted that item 12 (‘I always achieve
my goals’) is the only item that assesses the actual
goal-related outcome, and it loaded least strongly than
the other GO factor items.
Table 2. Pearson product moment correlations between the initial 14 scale items in the professional (low triangle, N¼242) and
the university student (upper triangle, N¼322) samples.
1234567891011121314
1 1 0.64** 0.15** 0.46** 0.36** 0.10 0.26** 0.06 0.19** 0.20** 0.26** 0.21** 0.20** 0.20**
2 0.64** 1 0.17** 0.52** 0.38** 0.15** 0.32** 0.05 0.21** 0.20** 0.26** 0.12* 0.24** 0.31**
3 0.18** 0.20** 1 0.25** 0.22** 0.65** 0.26** 0.19** 0.26** 0.37** 0.30** 0.30** 0.32** 0.29**
4 0.50** 0.55** 0.38** 1 0.46** 0.22** 0.40** 0.16** 0.33** 0.33** 0.36** 0.27** 0.27** 0.30**
5 0.31** 0.41** 0.18** 0.45** 1 0.17** 0.33** 0.17** 0.26** 0.22** 0.19** 0.17** 0.19** 0.20**
6 0.10 0.29** 0.50** 0.33** 0.19** 1 0.26** 0.19** 0.29** 0.34** 0.28** 0.30** 0.33** 0.31**
7 0.35** 0.42** 0.17** 0.44** 0.41** 0.29** 1 0.23** 0.33** 0.28** 0.42** 0.35** 0.40** 0.40**
8 0.13 0.11 0.21** 0.26** 0.16* 0.27** 0.23** 1 0.30** 0.32** 0.19** 0.04 0.03 0.04
9 0.28** 0.22** 0.23** 0.30** 0.32** 0.24** 0.31** 0.30** 1 0.34** 0.42** 0.27** 0.29** 0.22**
10 0.29** 0.29** 0.39** 0.37** 0.30** 0.37** 0.26** 0.29** 0.57** 1 0.36** 0.27** 0.25** 0.24**
11 0.25** 0.24** 0.22** 0.26** 0.23** 0.24** 0.28** 0.20** 0.26** 0.36** 1 0.63** 0.49** 0.51**
12 0.20** 0.19** 0.16* 0.28** 0.22** 0.17** 0.26** 0.20** 0.19** 0.29** 0.72** 1 0.57** 0.50**
13 0.19** 0.21** 0.30** 0.36** 0.25** 0.27** 0.47** 0.20** 0.18** 0.22** 0.48** 0.49** 1 0.52**
14 0.16* 0.25** 0.24** 0.27** 0.19** 0.24** 0.40** 0.22** 0.20** 0.19** 0.32** 0.30** 0.50** 1
Note: *p<0.05; **p<0.01.
338 A.M. Grant et al.
Downloaded by [University of Sydney] at 14:58 26 June 2012
Factor 3: RA
This factor is defined by items that reflect positive
thinking about solutions and the perceived availability
of solutions.
Table 3 also displays correlations between the
three factors. Not surprisingly, the factors are positively
correlated (r’s range between 0.42 and 0.53), which
suggests the presence of common, higher order factor.
Descriptive statistics for the total 12-item SFI scale
Descriptive statistics and Cronbach’s reliability
estimates for the final 12-item SFI and the three
subscales are presented in Table 4.
Cronbach’s estimates for the overall 12-item scale
were good across both samples (¼0.83), and within an
acceptable range for each facet. That is, ’s of 0.78 and
0.68 were observed for PD and RA, respectively (for
both samples), whilst the GO ’s were 0.78 and 0.82 for
the professional and student samples, respectively.
A range of item–total correlations (from three reliability
analyses for each subscale) was found to be relatively
homogeneous (ranging between 0.33 and 0.64).
There was a good overall distribution for the total
score (i.e. 32–64 for the professional sample; 21–69 for
the student sample) and each of the subscales
(range ¼6–24), with a slight tendency towards negative
skewness and kurtosis. Overall, however, the data were
Table 3. Pattern matrix for the EFA (ML with Promax rotation) and item–total correlations (for each subscale) of the SFI
(N¼242).
Items PD GO RA h
2
Item–total correlation
1. I tend to spend more time analysing my problems
than working on possible solutionsÕ
0.830 0.004 0.151 0.587 0.47
2. I tend to get stuck in thinking about problemsÕ0.884 0.062 0.066 0.688 0.53
3. There is always a solution to every problem 0.039 0.071 0.749 0.490 0.45
4. I tend to focus on the negativeÕ0.556 0.007 0.281 0.551 0.64
5. I’m not very good at noticing when things are going
wellÕ
0.421 0.069 0.112 0.274 0.45
6. There are always enough resources to solve a problem
if you know where to look
0.085 0.054 0.764 0.493 0.43
7. Most people are more resilient than they realize 0.011 0.105 0.347 0.161 0.33
8. Setbacks are a real opportunity to turn failure into
success
0.152 0.137 0.404 0.338 0.52
9. I imagine my goals and then work towards them 0.005 0.824 0.009 0.675 0.54
10. I keep track of my progress towards my goals 0.037 0.942 0.114 0.775 0.48
11. I’m very good at developing effective action plans 0.009 0.485 0.231 0.401 0.53
12. I always achieve my goals 0.068 0.268 0.237 0.225 0.43
Factor correlations
1 0.421 0.507
1 0.476
Notes: Õindicates reverse-scored items. The authors permit free use of this scale for research, training and educational purposes.
Table 4. Descriptive statistics (minimum and maximum scores, mean and SD) and Cronbach’s reliability estimates for the
three subscales and the global SFI scale in the professional (N¼242) and the university student (N¼322) samples.
No. of items M(SD) Minimum Maximum Skewness Kurtosis Cronbach’s
Professional sample
Total SFI scale 12 51.01 (7.64) 32 67 0.140 0.490 0.83
PD 4 16.52 (3.87) 6 24 0.070 0.367 0.78
GO 4 15.95 (3.14) 8 24 0.384 0.336 0.78
RA 4 18.55 (2.80) 9 24 0.141 0.486 0.68
University student sample
Total SFI scale 12 46.14 (8.03) 21 69 0.004 0.157 0.83
PD 4 13.77 (3.75) 4 24 0.195 0.217 0.78
GO 4 15.16 (3.68) 6 24 0.35 0.028 0.82
RA 4 17.21 (3.08) 4 24 0.198 0.181 0.68
The Journal of Positive Psychology 339
Downloaded by [University of Sydney] at 14:58 26 June 2012
distributed normally and no significant differences
were observed on the basis of gender. When both data
sets were combined, a small but a statistically signif-
icant positive correlation was observed between age
and scores on the SFI (r¼0.18; p<0.05).
The correlation between age and scores on the SFI
for the university student sample was (r¼0.21;
p<0.05), but there was no significant correlation
between age and scores on the SFI for the professional
sample (r¼0.03; ns).
In summary, the initial factor analysis using a
professional sample suggests that the SFI has good
psychometric properties and the emergent three factors
represent the underlying theoretical framework derived
from key themes from the literature on purposeful
positive change and BSFT (Miller, Hubble, & Duncan,
1996). The next step was to examine its factorial
stability.
Stage two: CFA using the university student sample
CFA was performed to determine whether the factorial
structure revealed in the EFA on the professional
sample would emerge using the student sample. To
examine the latent dimensionality of the SFI, three
models were compared. The models reflect three
alternative views of the solution-focused tendencies
(Figure 1). First, we postulated a one-factor broad
model (this was Model 1), reflecting a possibility that
the SFI captures a unidimensional construct. Second,
we postulated a three-correlated-factor model with
GO, RA and PD correlated factors (this was Model 2).
This model is based on the results received earlier using
EFA on the professional sample and it implies that the
SFI captures a multidimensional construct with three
first-order factors which capture relationships between
solution-focused behaviours. The results of the EFA
performed on the professional sample (a pattern of
robust correlations between three factors) suggested,
however, a possibility of a hierarchical model with
three first-order factors subordinate to a single higher
order factor. Thus, a two-level hierarchical model with
one second-order factor indexing broad solution-
focused thinking strategies and the three first-order
factors as in Model 2 were postulated (this was
Model 3).
Relevant fit indices are summarized in Table 5.
Chi-square (
2
) is one of the most commonly used
indexes of fit. Here small values relative to the degrees
of freedom are indicative of statistically non-significant
differences between the actual and the implied
matrixes, indicating that there is no discrepancy
between the hypothesized model and the data.
As this statistic is sensitive to sample size, in line with
the current best practice we used the root-mean-square
error of approximation (RMSEA) and its 90% confi-
dence interval (CI) to gauge approximate goodness of
SFI01 e1
SFI02 e2
SFI04 e3
SFI05 e4
SFT factor
SFI9 e5
SFI10 e6
SFI11 e7
SFI12 e8
SFI03 e9
SFI06 e10
SFI07 e11
SFI08 e12
nPFK
SFI01 e1
SFI02 e2
SFI04 e3
SFI05 e4
GFSR
SFI09 e5
SFI10 e6
SFI11 e7
SFI12 e8
SFT
SFI03 e9
SFI06 e10
SFI7 e11
SFI08 e12
nPFK
SFI01 e1
SFI02 e2
SFI04 e3
SFI05 e4
GFSR
SFI09 e5
SFI10 e6
SFI11 e7
SFI12 e8
SFT
SFI03 e9
SFI06 e10
SFI07 e11
SFI08 e12
SFTI
d1
d2
d3
Model 3Model 2Model 1
SFI factor
PD
PD
GO
GO
RA
RA
SFI
Figure 1.Graphical representation of Models 1–3.
Notes: In these diagrams, large circles represent latent factors, small circles represent error terms, single-headed straight arrows
index hypothesized factor loading and double-headed curved arrow index covariances/correlations between the factors.
SFI ¼Solution-focused Inventory; PD ¼problem disengagement; GO ¼goal orientation; RA ¼resource activation.
340 A.M. Grant et al.
Downloaded by [University of Sydney] at 14:58 26 June 2012
model fit in the population. Values lower than or
closeto 0.06 are taken to indicate good fit (Hu &
Bentler, 1999) and a narrow CI indicates good preci-
sion (MacCallum, Browne, & Sugawara, 1996).
We have also reported the relative likelihood ratio of
2
to degrees of freedom (
2
/df) statistic. It should be
noted that values of less than three are taken to
indicate good fit (Kline, 1998) when the sample size is
large. In addition, we used goodness-of-fit index (GFI)
to reflect a relative amount of covariance accounted
for by the model. Here, values greater than 0.90
suggest a reasonable-to-good fit (Kline, 1998). Finally,
we use the Tucker–Lewis index (TLI), an incremental
fit index, because it has been shown to be relatively
independent of sample size (Fan, Thompson, & Wang,
1999; Marsh, Balla, & McDonald, 1988).
Values greater than 0.90 are required, however, and
Hu and Bentler (1999) suggest a cut-off value of close
to 0.95.
Models 1 and 2 are nested, thus
2
will be used to
statistically compare the fit of the solutions. A differ-
ence in the chi-squared test (D
2
) was used to evaluate
alternative nested models. A statistically significant
value suggests that a less parsimonious model is a
significantly better fit than the initial model. Models 2
and 3, however, would produce identical-fit statistics as
these are mathematically identical solutions, and the
preference of one model over the other will be based on
a conceptual basis.
Model 1, the one-factor model, had the poorest fit
to data: 2
54 ¼511.579;
2
/df ¼9.474, the RMSEA
equals 0.162 (its 90% CI is 0.15–0.18). The TLI is
0.551 and the GFI 0.765 (see Table 5). Model 2, the
three-interrelated-factors model, had a substantially
better fit (diff 2
3¼364.27, p<0.001), and, overall this
model fits data reasonably well: 2
51 ¼147.31;
2
/df ¼2.89, the RMSEA equals 0.077 (its 90% CI is
0.062–0.091). The TLI is 0.90 and the GFI 0.925.
Replicating the results of the previous study, three
factors share consistent positive correlations: 0.33 (PD
and RA), 0.44 (PD and GO) and 0.53 (GO and RA).
Based on a pattern of these correlations and theory
articulated earlier, the two-level hierarchical model
(Model 3) with one second-order factor indexing broad
tendencies towards solution-focused thinking underly-
ing the three first-order factors (PD, GO and RA) was
postulated. Despite having the same fit indices as for
Model 2, Model 3 was accepted as the model which
best reflected current data and theory for solution-
focused tendencies.
1
The results are summarized in Table 6 and graph-
ically represented in Figure 1.
Factor 1: PD
As with the previous EFA analysis, this factor
captures four negative thinking items which were
reversed-coded. Similar to the results obtained in the
professional sample analysis, these four items have
solid loading on this factor (ranging from 0.533
to 0.798).
Factor 2: GO
As with the previous EFA analysis, only items that
reflect positive tendencies towards achieving one’s own
goals load on this factor. All four items have solid
loading on this factor (ranging from 0.685 to 0.766).
In the previous analysis using the professional sample,
item 12 (‘I always achieve my goals’) had a somewhat
lower loading on this factor (0.268). Within the student
sample, this item has a loading of 0.685, suggesting
that the low loading observed in the professional
sample may have been due to the characteristics of the
initial sample, rather than this item’s low content
validity. Further research is needed to confirm this
proposition.
Factor 3: RA
As with the professional sample, this factor is defined
by the four items that reflect an orientation towards
seeking and utilizing resources. However, as with the
results of the EFA analysis conducted using the
professional sample, item seven (‘Most people are
more resilient than they realize’) had a lower loading
on this factor (0.274) and a much lower communality
estimate than the other items (0.075). This suggests
Table 5. Fit statistics for the different models proposed to underlie the university student sample (N¼322).
Fit statistics
Model
2
df
2
/Ddf GFI TLI CFI RMSEA (90% CI)
1 511.58 54 9.474 0.765 0.551 0.633 0.162 (0.150–0.175)
2 147.31 51 2.888 0.925 0.900 0.923 0.077 (0.062–0.091)
Diff
2
364.27 3 121.42
3 147.31 51 2.888 0.925 0.900 0.923 0.077 (0.062–0.091)
Note: GFI ¼goodness-of-fit index; CFI¼comparative fit index; TLI¼Tucker–Lewis index; RMSEA¼root mean square error of
approximation.
The Journal of Positive Psychology 341
Downloaded by [University of Sydney] at 14:58 26 June 2012
that this item reflects a somewhat different construct
than the one captured by this factor’s other three items.
When this item was omitted from the model, the GFIs
improved, but only very slightly: 2
41 ¼112.551;
2
/df ¼2.745, the RMSEA equals 0.074 (its 90% CI
is 0.058–0.090). The TLI is 0.920 and the GFI 0.938.
Given that the content of this item holds some
theoretical importance and the fact that its omission
does not improve the model substantially, the decision
was made to retain it. One reason for this relatively low
loading may be that it is the only item that seeks an
opinion about ‘most people’ – a global statement –
rather than asking respondents to express an opinion
about one’s self.
Descriptive statistics for the total 12-item SFI scale
and its three subscales
Descriptive statistics and Cronbach’s reliability
estimates for the total 12-item SFI and the three
subscales are presented in Table 4. Cronbach’s
estimates are across both samples (¼0.83 for the
overall 12-item scale), and are within an acceptable
range for all three factors. The ’s for GO are slightly
higher for the student sample compared with the
professional sample (0.82 vs. 0.78) and are a result of
the improved factor loading pattern for this subscale.
As with the professional sample, there was a good
overall distribution for the total score (ranging from 21
to 69) and each of the subscales (ranging between 4 and
24) with a slight tendency towards negative skewness
and kurtosis. Overall, however, the data were distrib-
uted normally.
Study one: Discussion
Taken together our results suggest that the overall SFI
scale and its three subscales (PD, GO and RA) possess
adequate-to-good psychometric properties. In the
original professional sample, the three factors that
emerged were consistent with previous literature and
make good theoretical sense. For example, goal pro-
gression – moving towards a solution – requires
purposeful movement away from the undesirable
outcome or problem and the identification of, and
progression towards, a preferred outcome or solution
(for discussion on the implications of avoidance and
approach goals in purposeful positive change, see
Elliot, Sheldon, & Church, 1997; Locke, 1996).
Such purposeful change requires that the individual
sets goals, develops effective action plans and then
monitors and evaluates progress towards those goals,
adapting and changing action in line with feedback.
The monitor–evaluate–modify steps of this process
constitute a ‘cycle’ of self-regulated behaviours theo-
rized as important for successful behaviour change
(Carver & Scheier, 1998). The results of both factor
analyses and correlations between the factors (see
Table 3) support such propositions and suggest that
the processes of problem identification, resource iden-
tification and goal pursuit are discrete, but related at
the higher level processes.
The PD subscale consisted of the items such as
‘I tend to spend more time analysing my problems than
working on possible solutions’ and ‘I tend to get stuck
in thinking about problems’ (reverse scored). Such
items capture the essence of ‘problem saturation’ often
Table 6. Results of the CFA (ML) for the final 12 items of the SFI (N¼322).
First stratum PD GO RA h
2
1. I tend to spend more time analysing my problems than working on
possible solutionsÕ
0.743 – – 0.552
2. I tend to get stuck in thinking about problemsÕ0.798 – – 0.636
3. There is always a solution to every problem – – 0.807 0.652
4. I tend to focus on the negativeÕ0.682 – – 0.465
5. I’m not very good at noticing when things are going wellÕ0.533 – – 0.284
6. There are always enough resources to solve a problem if you know
where to look
– – 0.776 0.602
7. Most people are more resilient than they realise – – 0.274 0.075
8. Setbacks are a real opportunity to turn failure into success – – 0.496 0.246
9. I imagine my goals and then work towards them – 0.766 – 0.587
10. I keep track of my progress towards my goals – 0.777 – 0.603
11. I’m very good at developing effective action plans – 0.711 – 0.505
12. I always achieve my goals – 0.685 – 0.469
Second stratum
Solution-focused thinking 0.524 0.837 0.636
Notes: PD ¼problem disengagement; GO ¼goal orientation; RA ¼resource activation.
Õindicates reverse-scored items. The authors permit free use of this scale for research, training and educational purposes.
342 A.M. Grant et al.
Downloaded by [University of Sydney] at 14:58 26 June 2012
referred to in the solution-focused literature
(O’Connell, 1998), and aspects of rumination and
negative attentional bias often emphasized in the
cognitive-behavioural literature (Lyubomirsky,
Tucker, Caldwell, & Berg, 1999; Mathews, Ridgeway,
& Williamson, 1996). In contrast, RA items such as
‘There is always a solution to every problem’ and
‘There are always enough resources to solve a problem
if you know where to look’ reflect optimism, resource
identification, resilience and positive reframing that
typify the solution-focused approach (O’Hanlon,
1998). Finally, the GO subscale items (e.g. ‘I imagine
my goals and then work towards them’ and ‘I
keep track of my progress towards my goals’) appear
to reflect the core components of the goal-striving
process – being able to develop a clear image of a goal,
developing action plans, tracking progress and self-
efficacy – that are necessary for creating purposeful
positive change (Schmuck & Sheldon, 2001).
In summary, the initial factor analyses suggests that
the SFI has good psychometric properties and the
emergent three factors indeed represent the theoretical
framework deemed to underpin solution-focused
approaches to creating purposeful positive change
(Miller et al., 1996), and that the use of a single
aggregated score is in line with the model structure
indicated by Model 3. The next step was to explore
convergent validity.
Study two: Convergent validity
Method
In order to explore convergent validity, responses to
the SFI were correlated with responses to established,
related measures. In line with the model structure
suggested by Model 3, a total SFI score was calculated
by summing the scores of the three SFI subscales.
Satisfaction with life (SWL) was measured using
the SWL scale (SWLS; Diener, Emmons, Larsen, &
Griffin, 1985); PWB was assessed using the PWB scale
(PWBS; Ryff & Keyes, 1995); resilience was assessed
using the Cognitive Hardiness scale (CHS; Nowack,
1990); perspective taking (PT) using the Perspective-
taking scale (PTS; Davis, 1980) and the Depression,
Anxiety and Stress scale (DASS; Lovibond &
Lovibond, 1995) was used as a measure of
psychopathology.
We chose these measures because a solution-
focused mindset is characterized by a focus on positive
possibilities and a belief that goals, talents and abilities
can be developed through persistence. Because the
attainment of such outcomes is associated with well-
being (Sheldon & Elliot, 1999; Sheldon, Ryan, Deci, &
Kasser, 2004), we expected that people holding a
solution-focused mindset would experience greater life
satisfaction and PWB.
In addition, as people work to attain their goals,
they frequently have to overcome setbacks and develop
new pathways to their preferred outcome. Such pro-
cesses tend to develop resilience (Maddi, 2005), and
empirical research in coaching seems to support such
propositions (Grant, Curtayne, & Burton, 2009).
The development of a number of different solutions
also requires that the individual have the ability to
distance themselves from problems and view their
circumstances from a different perspective. As such, a
solution-focused mindset should also be associated
with PT capacity – the ability to generate and hold
different perspectives of self, others and the world.
Finally, we reasoned that people with low solution-
focused mindsets would be more prone to depression,
anxiety and stress. In short, we hypothesized that the
SFI would be positively correlated with SWL, PWB
and PT, and would be negatively correlated with
the DASS.
Participants and procedure
Data were drawn from the professional sample reported
in Study one. These participants (N¼242) were volun-
teers from the legal and health professions who were
taking part in an intervention study on leadership in
high-stress workplaces (Mean age ¼41.92 years
(SD ¼8.92); 157 females; 85 males). Questionnaires
were completed in small group settings.
Measures
SWL was assessed with SWLS (Diener et al., 1985),
which is a five-item scale that includes items such as:
‘The conditions of my life are excellent’ and ‘So far I
have gotten the important things I want in life’.
The SWLS is one of the most widely used measures of
well-being, with good psychometric properties and a
reported of 0.80 (Diener et al., 1985). Cronbach’s
for this study was 0.86.
PWB was assessed with a 36-item version of the
PWBS (Ryff & Keyes, 1995). The PWBS measures
well-being on six subscales: autonomy, environmental
mastery, purpose in life, personal growth, self-accep-
tance and positive relations with others. Internal
consistency () coefficients for the original six three-
item scales range are reported to range from 0.82 to
0.90 (Schmutte & Ryff, 1997) and; when the three-item
scales are summed to form a global PWB scale,
internal consistencies have been found to exceed 0.8
(Keyes & Ryff, 1998). Cronbach’s for this study
was 0.85.
Resilience was assessed using a 15-item of the CHS
(Nowack, 1990). This scale based on Kobasa’s (1979)
work measures the individual’s sense of personal
control, their propensity to rise to meet challenges
and their commitment to action. Nowack (1990)
The Journal of Positive Psychology 343
Downloaded by [University of Sydney] at 14:58 26 June 2012
reports an internal consistency of 0.83. Cronbach’s
for this study was 0.78.
Psychopathology was assessed using a composite
score from the DASS (Lovibond & Lovibond, 1995).
The DASS-21 comprises three subscales measuring
depression, anxiety and stress and is designed to be
used with both clinical and non-clinical populations.
Internal consistency and test–retest reliability have
been found to be acceptable (r¼0.71–0.81; Brown,
Chorpita, Korotitsch, & Barlow, 1997). Cronbach’s
for this study was 0.89.
PT was assessed using the PTS which is a seven-
item subscale of The Empathy Questionnaire (Davis,
1980). Example items from this scale include: ‘I
believe that there are two sides to every question and
try look at them both’ and ‘I sometimes find it
difficult to see things from the ‘other guy’s’ point of
view’ (reverse scored). Davis (1980) reports coef-
ficients between 0.75 and 0.78, and a test–retest
reliability of 0.62. Cronbach’s for this study
was 0.77.
Study two: Results and discussion
The results of the correlational analysis are presented
in Table 7. As hypothesized, the SFI was negatively
correlated with psychopathology as measured by the
DASS (r¼0.44; p<0.001) and was positively corre-
lated with measures of well-being as measured by
SWLS (r¼0.38; p<0.001) and PWBS (r¼0.66;
p<0.001) and also positively correlated with resilience
as measured by CHS (r¼0.65; p<0.001). The SFI also
positively correlated with PT as measured by the PTS
(r¼0.33; p<0.001). In short, the convergent validity
of the SFI appears to be good.
Whilst the correlations between the SFI and SWLS,
PWBS, PTS and the DASS were as expected, the
relative magnitude of these correlations is of some
interest. The SFI was highly positively correlated with
both PWB (r¼0.66; p<0.001) and resilience (r¼0.65;
p<0.001), and both of those correlations were greater
than the correlation between the SFI and the PTS
(r¼0.33; p<0.001). Indeed, using a Fisher r-to-z
transformation to assess the significance of the differ-
ence between these correlation coefficients (Hinkle,
Wiersma, & Jurs, 1994), it was found that that the
difference between the correlation between the SFI and
PWB (r¼0.66) was significantly greater than the
correlation between the SFI and the PTS (r¼0.33)
(z¼4.96; p<0.001). In addition, the difference
between the correlation for SFI and resilience
(r¼0.65) was greater than the difference between the
SFI and the PTS (z¼4.77; p<0.001).
From these data, it would seem that solution-
focused thinking is indeed significantly related to PT,
but that the constructs of PWB and resilience are more
closely aligned with solution-focused thinking than PT.
One explanation for this finding may be PT is
important at certain points during the goal-striving
process (e.g. during solution generation and action
planning), but may not be as important as PWB and
resilience over a longer period of time when one is
actively striving for goal attainment. For example,
resilience – the ability to recover from setbacks and
preserve in the face of adversity – is more strongly
related to solution-focused thinking than PT. This may
reflect the greater importance that resilience has for
dealing with and overcoming the various hassles and
on-going problems of daily life. Similarly, key facets of
PWB (e.g. autonomy, purpose in life, personal growth,
environmental mastery; Ryff & Keyes, 1995) also
appear to be more important than PT and may be
critical for the successful enactment of positive pur-
poseful change.
Whilst these interpretations are somewhat specula-
tive, this discussion illustrates that the SFI scale
reported here may prove to be a useful tool for our
further understanding of the nature and processes
underlying solution-focused thinking. One avenue for
future research could be to explore the convergent
validity of the SFI with a range of other related
measures such as hope (Snyder et al., 1996), optimism
(Scheier, Carver, & Bridges, 1994) and cognitive
flexibility (Martin & Rubin, 1995).
Study three: Test–retest reliability and responsiveness
to coaching
Method
In order to explore the test–retest reliability of the SFI
and its sensitivity to change we utilized a subset of data
(N¼129) extracted from the same large leadership
development and coaching study used in Study one
(Cavanagh et al., 2012). This intervention study used a
between-subjects design with participants from three
Table 7. Correlations between SFI and other convergent
variables (N¼242).
SFI SWL PWB RES DASS
SWL 0.38** – – – –
PWB 0.66** 0.55** – – –
RES 0.65** 0.46** 0.69** – –
DASS 0.44** 0.38** 0.47** 0.61** –
PT 0.33** 0.04 0.43** 0.16* 0.07
Notes: SFI ¼Solution-focused Inventory;
SWL ¼Satisfaction with life; PWB ¼psychological well-
being; RES ¼resilience; DASS – depression, anxiety and
stress; PT ¼perspective-taking.
** and * indicate correlation significant at the 0.01 (two-
tailed) and 0.05 (two-tailed) levels.
344 A.M. Grant et al.
Downloaded by [University of Sydney] at 14:58 26 June 2012
cohorts randomly assigned to either an intervention
group (Group 1; n¼72) or a waitlist control group
(Group 2; n¼57). The data used in this study was
taken from the first cohort of participants.
The intervention consisted of a four-day leadership
development training workshop, followed by eight
sessions of leadership coaching over a time period of 16
weeks. The leadership development intervention used a
constructive developmental approach based on a four-
factor model of leadership (Cavanagh, 2008), which
sought to develop leaders working within complex,
stressful environments (i.e. public health and legal
professions). Facets of leadership emphasized in this
model included the need for leaders to take complex
integrative perspectives, to be mindful, to act purpose-
fully towards valued goals and to create positive
organizational contexts (Cavanagh, 2008). Both the
model and the training were informed by positive
psychology and solution-focused coaching approaches,
and the training consisted of a variety of didactic and
interactive learning activities (e.g. video-taped dialogue
sessions). The SFI measure was completed at Time 1
and Time 2 (i.e. 20 weeks later), along with a number
of other measures that are not reported here.
As the intervention was targeting constructs asso-
ciated with solution-focused thinking, we hypothesized
that participation in the leadership training and
coaching intervention would be associated with
increased scores on the SFI and that the scores for
the control group would not change.
Study three: Results and discussion
A repeated-measures ANOVA for the SFI showed a
significant time (Time 1, Time 2) by group (Group 1,
Group 2) interaction effect, F(1, 127); 6.17, p<0.05,
indicating that Group 1 (coaching) indeed had higher
scores on the SFI at the completion of the coaching
intervention at Time 2 compared to Group 2. Planned
contrasts indicated that scores for Group 2 did not
differ significantly from Time 1 to Time 2,
t(56) ¼1.11; ns. In addition, a significant correlation
was found between SFI scores at Time 1 and Time 2
for Group 2 (r¼0.84;p<0.001), indicating good test–
retest reliability, see Table 8 for means and SDs for
coaching (Group 1) and control groups (Group 2).
In sum, participation in the leadership coaching
intervention was associated with significantly increased
scores on the SFI, whilst the scores on the SFI for the
control group did not change. Whilst these findings
indicate the SFI has good test–retest reliability and is
responsive to interventions designed to increase solu-
tion-focused thinking, this will need to be confirmed by
future research.
Final discussion
Our aim in these studies was to develop a brief, reliable
measure of solution-focused thinking that would allow
practitioners and researchers to assess the extent to
which an individual’s thinking was oriented towards
solution construction. Such an instrument would
represent a useful contribution to the field, both in
terms of assessing outcomes of solution-focused inter-
ventions and in providing a means of better under-
standing the psychological processes central to
purposeful positive change across a variety of behav-
iour change settings. These studies provide initial
support for the reliability and validity of the SFI as a
measure of solution-focused thinking.
Two factorial analyses resulted in a final 12-item
scale comprising three subscales that we labelled PD,
GO and RA. These subscales demonstrated reason-
able-to-good internal reliability, and the scale as a
whole demonstrated good internal reliability. In addi-
tion, convergent validity and test–retest reliability were
also acceptable. Notably, the SFI appears to be
sensitive to changes following a leadership develop-
ment intervention.
Future research
The fact that the SFI comprises three subscales may
prove to be useful to researchers and practitioners
seeking to understand the differential impacts of
problem-focused or solution-focused thinking in the
change process. For example, the SFI could be used to
track changes in thinking styles as individuals engage
in the goal-striving process through participation in a
coaching or therapeutic programme. The general
expectation within the solution-focused literature is
that individuals engaged in solution-focused goal-
striving activities would naturally increase their levels
of solution-focused thinking (O’Connell, 1998), an
expectation that was supported by the findings
reported in Study three. Future research should
extend this line of enquiry and explore the links
between solution-focused thinking and degree to which
participants in such coaching programmes attain their
goals. The solution-focused literature (e.g. O’Hanlon,
1998), and to some extent the positive psychology
literature (e.g. Fredrickson, 2005), generally assumes
Table 8. Descriptive statistics for the SFI for coaching and
control groups (N¼129).
Time 1 Time 2
MSD MSD
Group 1 (coaching) 50.98 7.12 54.05 6.84
Group 2 (control) 52.17 8.58 52.86 6.90
The Journal of Positive Psychology 345
Downloaded by [University of Sydney] at 14:58 26 June 2012
that solution-focused thinking is a good thing and is
almost invariably associated with positive outcomes in
terms of well-being and goal attainment. We suspect
that this may be an over-simplistic view, and that
problem-focused thinking may have an important role
to play in purposeful positive change. However to date
little research has explored these issues.
In the only study we are aware of to explore the
differential impact of problem-focused versus solution-
focused coaching questions, Grant and O’Connor
(2010) found that the issues may be more complex
than often assumed. The Grant and O’Connor (2010)
study found that both the problem-focused and the
solution-focused conditions were effective at the
enhancing goal approach. However, the solution-
focused group experienced significantly greater
increases in goal approach compared with the pro-
blem-focused group. Problem-focused questions were
found to reduce negative affect and increase self-
efficacy but did not enhance positive affect. In
contrast, the solution-focused approach increased
positive affect and self-efficacy, decreased negative
affect, whilst also increasing participants’ insight and
understanding of the nature of the problem. We believe
that this area is ripe for future research.
Limitations
When interpreting the findings reported in this article,
it should be remembered that the data were collected
from a cohort of public health and legal professionals,
who were participating in a leadership development
study, and psychology undergraduates, who were
taking part in this research as part of their course
requirements. As such, these findings may be some-
what idiosyncratic and not representative of other
populations of interest (e.g. clinical populations or the
broader community). Furthermore, we used the data
collected from the same populations for the various
factor analyses, convergent validity tests and reliability
tests presented in this article. Future research that
focuses on other populations is recommended for the
further validation of the SFI, both to establish its
relevance for a broader cross-section of the population
and to permit further scale validation using indepen-
dent samples.
Summary
The data reported in this article provide preliminary
evidence that the SFI is a reliable measure of
solution-focused thinking. This has the potential to
help practitioners and researchers assess the extent to
which an individual’s thinking is oriented towards
solution construction, and may also be a useful
outcome measure for coaching and positive
psychology interventions. Future research using the
SFI has the potential to contribute to our under-
standing of both solution-focused approaches and
the broader positive psychological enterprise.
We hope that the SFI will prove to be a useful
tool in the development of such knowledge and look
forward to future developments in this area.
Acknowledgments
This research was funded by an Australian Research Council
Linkage Grant (project number LP0776814).
Note
1. To improve fit indices, two post hoc modifications
were considered. Two items were allowed to cross-load
on different factors; item four was allowed to also
load on PD (loading ¼0.195, p<0.01), while item 10
was allowed to also load on RA (loading ¼0.163,
p<0.05). However, these post hoc loadings were
low (2
49 ¼126.98; GFI ¼0.935, TLI ¼0.916,
RMSEA ¼0.07 (0.005–0.086)), and this model resulted
in only marginal improvement. Thus, Model 3 was
accepted as the final model.
References
Arbuckle, J.L. (2006). Amos 7.0 user’s guide. Chicago, IL:
SPSS.
Bandura, A. (1982). Self-efficacy mechanism in human
agency. American Psychologist, 37, 122–147.
Bartlett, M.S. (1950). Tests of significance in factor analysis.
British Journal of Statistical Psychology, 3, 77–85.
Beasley, M., Thompson, T., & Davidson, J. (2003).
Resilience in response to life stress: The effects of coping
style and cognitive hardiness. Personality and Individual
Differences, 34, 77–95.
Beck, J. (1995). Cognitive therapy: Basics and beyond.
London: The Guilford Press.
Biswas-Diener, R. (2010). Practicing positive psychology
coaching: Assessment, activities and strategies for success.
London: Wiley.
Bradley, P.D., Bergen, L.P., Ginter, E.J., Williams, L.M., &
Scalise, J.J. (2010). A survey of north american marriage
and family therapy practitioners: A role delineation study.
The American Journal of Family Therapy, 38, 281–291.
Brown, T.A., Chorpita, B.F., Korotitsch, W., & Barlow,
D.H. (1997). Psychometric properties of the Depression
Anxiety Stress Scales (DASS) in clinical samples.
Behaviour Research and Therapy, 35, 79–89.
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,
822–848.
Carver, C.S., & Scheier, M.F. (1998). On the self-regulation of
behavior. Cambridge: Cambridge University Press.
Cavanagh, M. (2006). Coaching from a systemic perspective:
A complex adaptive conversation. In D. Stober &
346 A.M. Grant et al.
Downloaded by [University of Sydney] at 14:58 26 June 2012
A.M. Grant (Eds.), Evidence-based coaching handbook
(pp. 313–354). Hoboken, NJ: Wiley.
Cavanagh, M. (2008). Creating flourishing leadership through
developmental coaching. First Australian Positive
Psychology and Well-being Conference, Sydney, 5–6
April, 2008.
Cavanagh, M., & Grant, A.M. (2010). The solution-focused
coaching approach to coaching. In E. Cox, T. Bachkirova,
& D. Clutterbuck (Eds.), Sage handbook of coaching
(pp. 34–47). London: Sage.
Cavanagh, M., Spence, G., & Atkins, P. (2012). Leadership in
high stress workplaces; an intervention study (Unpublished
manuscript). University of Sydney, Sydney, Australia.
Davis, M.H. (1980). A multidimensional approach to
individual differences in empathy. JSAS Catalogue of
Selected Documents in Psychology, 10, 85.
DeCarlo, L.T. (1997). On the meaning and use of kurtosis.
Psychological Methods, 2, 292–307.
de Shazer, S. (1988). Clues: Investigating solutions in brief
therapy. New York, NY: W.W. Norton & Co.
Diener, E., Emmons, R.A., Larsen, R.J., & Griffin, S. (1985).
The satisfaction with life scale. Journal of Personality
Assessment, 49, 71–75.
Dryden, W. (1987). Counseling individuals: The rational-
emotive approach. London: Taylor & Francis.
Dweck, C.S. (2006). Mindset: The new psychology of success.
New York, NY: Random House.
Elliot, A.J., Sheldon, K.M., & Church, M.A. (1997).
Avoidance personal goals and subjective well-being.
Personality and Social Psychology Bulletin, 23, 915–927.
Fan, X., Thompson, B., & Wang, L. (1999). The effects of
sample size, estimation methods, and model specification
on SEM fit indices. Structural Equation Modeling, 6,
56–83.
Fredrickson, B.L. (Ed.). (2005). The broaden-and-build
theory of positive emotions. New York, NY: Oxford
University Press.
Furman, B., & Ahola, T. (1992). Solution talk:
Hosting therapeutic conversations. New York, NY:
W.W. Norton & Co.
Grant, A.M. (2001). Rethinking psychological mindedness:
Meta-cognition, self-reflection and insight. Behaviour
Change, 18, 8–17.
Grant, A.M. (2003). The impact of life coaching on goal
attainment, metacognition and mental health. Social
Behavior and Personality, 31, 253–264.
Grant, A.M. (2011). The solution-focused inventory: A
tripartite taxonomy for teaching, measuring and concep-
tualising solution-focused approaches to coaching. The
Coaching Psychologist, 7, 98–106.
Grant, A.M., Curtayne, L., & Burton, G. (2009). Executive
coaching enhances goal attainment, resilience and work-
place well-being: A randomized controlled study. Journal
of Positive Psychology, 4, 396–407.
Grant, A.M., Franklin, J., & Langford, P. (2002). The self-
reflection and insight scale: A new measure of private self-
consciousness. Social Behavior and Personality, 30,
821–836.
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 & Commercial Training, 42, 102–111.
Grant, A.M., & Spence, G.B. (2010). Using coaching and
positive psychology to promote a flourishing workforce: A
model of goal-striving and mental health. In P.A. Linley,
S. Harrington, & N. Page (Eds.), Oxford handbook of
positive psychology and work (pp. 175–188). Oxford:
Oxford University Press.
Hinkle, D.E., Wiersma, W., & Jurs, S.G. (1994). Applied
statistics for the behavioral sciences. Boston, MA:
Houghton Mifflin Company.
Hu, L., & Bentler, P.M. (1999). Cutoff criteria for fit indexes
in covariance structure analysis: Conventional criteria
versus new alternatives. Structural Equation Modeling, 6,
1–55.
Iveson, C., George, E., & Ratner, H. (2012). Brief coaching:
A solution focused approach. New York, NY: Routledge.
Jackson, P.Z., & McKergow, M. (2002). The solutions focus:
The SIMPLE way to positive change. London: Nicholas
Brealey.
Kaiser, H.F. (1970). A second-generation little jiffy.
Psychometrika, 35, 401–415.
Kashdan, T.B., & Rottenberg, J. (2010). Psychological
flexibility as a fundamental aspect of health. Clinical
Psychology Review, 30, 885–878.
Keyes, C.L.M., & Ryff, C.D. (1998). Generativity in adult
lives: Social structural contours and quality of life
consequences. In D. McAdams & E. de St. Aubin (Eds.),
Generativity and adult development: Perspectives on caring
for and contributing to the next generation (pp. 227–263).
Washington, DC: American Psychological Association.
Kline, R.B. (1998). Principles and practice of structural
equation modelling. New York, NY: Guilford Press.
Kobasa, S.C. (1979). Stressful life events, personality, and
health: An inquiry into hardiness. Journal of Personality
and Social Psychology, 37, 1–11.
Lakota, M. (2010). Self-reflection, insight and subjective
wellbeing: A path through rumination, mindfulness and
constructive self-reflection (Unpublished honours thesis).
School of Psychology, University of Sydney.
Locke, E.A. (1996). Motivation through conscious goal
setting. Applied & Preventive Psychology, 5, 117–124.
Locke, E.A., & Latham, G.P. (2002). Building a practically
useful theory of goal setting and task motivation.
American Psychologist, 57, 705–717.
Lovibond, S.H., & Lovibond, P.F. (1995). Manual for the
depression anxiety stress scales. Sydney: Psychology
Foundation of Australia.
Lyubomirsky, S., & Nolen-Hoeksema, S. (1995). Effects of
self-focused rumination on negative thinking and inter-
personal problem solving. Journal of Personality & Social
Psychology, 69, 176–190.
Lyubomirsky, S., Tucker, K.L., Caldwell, N.D., & Berg,
K. (1999). Why ruminators are poor problem solvers:
Clues from the phenomenology of dysphoric rumina-
tion. Journal of Personality & Social Psychology, 77,
1041–1060.
MacCallum, R.C., Browne, M.W., & Sugawara, H.M.
(1996). Power analysis and determination of sample size
for covariance structure modelling. Psychological Methods,
1, 130–149.
Maddi, S.R. (2005). On hardiness and other pathways to
resilience. American Psychologist, 60, 261–262.
The Journal of Positive Psychology 347
Downloaded by [University of Sydney] at 14:58 26 June 2012
Marsh, H.W., Balla, J.R., & McDonald, R.P. (1988).
Goodness-of-fit in confirmatory factor analysis: The
effect of sample size. Psychological Bulletin, 103, 391–410.
Martin, M.W., & Rubin, R.B. (1995). A new measure of
cognitive flexibility. Psychological Reports, 76, 623–626.
Mathews, A., Ridgeway, V., & Williamson, D.A. (1996).
Evidence for attention to threatening stimuli in depression.
Behaviour Research & Therapy, 34, 695–705.
McKergow, M., & Jackson, P.Z. (2005). An introduction to
solutions focused coaching, consulting and facilitation
[Audio book – CD]. Cheltenham: Solutions Books.
McKergow, M., & Stellamans, A. (2011). Restoring the
client’s choice of action. InterAction – The Journal of
Solution Focus in Organizations, 3, 98–105.
Miller, S.C., Hubble, M.A., & Duncan, B.L. (1996).
Handbook of solution-focused brief therapy. San
Francisco, CA: Jossey-Bass.
Neenan, M., & Dryden, W. (2002). Life coaching: A
cognitive-behavioral approach. New York, NY: Brunner-
Routledge.
Nolen-Hoeksema, S. (2000). The role of rumination in
depressive disorders and mixed anxiety/depressive symp-
toms. Journal of Abnormal Psychology, 109, 504–511.
Nolen-Hoeksema, S., & Morrow, J. (1991). A prospective
study of depression and posttraumatic stress symptoms
after a natural disaster: The 1989 Loma Prieta earthquake.
Journal of Personality & Social Psychology, 61, 115–121.
Nowack, K. (1990). Initial development of an inventory to
asses stress and health. American Journal of Health
Promotion, 4, 173–180.
O’Connell, B. (1998). Solution-focused therapy. London:
Sage.
O’Hanlon, W. (1998). Possibility therapy: An inclusive,
collaborative, solution-based model of psychotherapy.
In M.F. Hoyt (Ed.), The handbook of constructive
therapies: Innovative approaches from leading practitioners
(pp. 137–158). San Francisco, CA: Jossey-Bass.
Peterson, C. (2000). The future of optimism. American
Psychologist, 55, 44–55.
Peterson, C. (2006). A primer in positive psychology.
New York, NY: Oxford University Press.
Robinson, M.D., & Tamir, M. (2011). A task-focused mind
is a happy and productive mind: A processing perspective.
In K. Sheldon, T. Kashdan, & M. Steger (Eds.), Designing
the future of positive psychology: Taking stock and moving
forward (pp. 160–174). Oxford: Oxford University Press.
Roeden, J.M., Maaskant, M.A., Bannink, F.P., & Curfs,
L.M. (2011). Solution-focused brief therapy with people
with mild intellectual disabilities: A case series. Journal of
Policy and Practice in Intellectual Disabilities, 8, 247–255.
Ryff, C.D., & Keyes, C.L.M. (1995). The structure of
psychological well-being revisited. Journal of Personality
and Social Psychology, 69, 719–727.
Scheier, M.F., Carver, C.S., & Bridges, M.W. (1994).
Distinguishing optimism from neuroticism (and trait
anxiety, self-mastery, and self-esteem): A re-evaluation of
the life orientation test. Journal of Personality and Social
Psychology, 67, 1063–1078.
Schmuck, P., & Sheldon, K.M. (Eds.). (2001). Life goals and
well-being: Towards a positive psychology of human striving.
Ashland, OH: Hogrefe & Huber.
Schmutte, P.S., & Ryff, C.D. (1997). Personality and well-
being: Reexamining methods and meanings. Journal of
Personality and Social Psychology, 73, 549–559.
Sheldon, K.M., & Elliot, A.J. (1999). Goal striving, need
satisfaction, and longitudinal well-being: The self-concor-
dance model. Journal of Personality & Social Psychology,
76, 482–497.
Sheldon, K.M., Elliot, A.J., Kim, Y., & Kasser, T. (2001).
What is satisfying about satisfying events? Testing 10
candidate psychological needs. Journal of Personality &
Social Psychology, 80, 325–339.
Sheldon, K.M., Kasser, T., Smith, K., & Share, T. (2002).
Personal goals and psychological growth: Testing an
intervention to enhance goal attainment and personality
integration. Journal of Personality, 70, 5–31.
Sheldon, K.M., Ryan, R.M., Deci, E.L., & Kasser, T. (2004).
The independent effects of goal contents and motives on
well-being: It’s both what you pursue and why you pursue
it. Personality & Social Psychology Bulletin, 30, 475–486.
Smock, S.A., McCollum, E.E., & Stevenson, M.L. (2010).
The development of the solution building inventory.
Journal of Marital and Family Therapy, 36, 499–510.
Snyder, C.R., Rand, K.L., & Sigmon, D.R. (2002). Hope
theory: A member of the positive psychology family.
In C.R. Snyder & S.J. Lopez (Eds.), Handbook of positive
psychology (pp. 257–276). London: Oxford University
Press.
Snyder, C.R., Sympson, S.C., Ybasco, F.C., Borders, T.F.,
Babyak, M.A., & Higgins, R.L. (1996). Development and
validation of the state Hope Scale. Journal of Personality
and Social Psychology, 70, 321–335.
Stams, G.J., Dekovic, M., Buist, K., & de Vries, L. (2006).
Efficacy of solution-focused brief therapy: A meta-
analysis. Gedragstherapie, 39, 81–94.
Takano, K., & Tanno, Y. (2009). Self-rumination, self-
reflection, and depression: Selfrumination counteracts the
adaptive effects of self-reflection. Behaviour Research and
Therapy, 47, 260–264.
Trapnell, P.D., & Campbell, J.D. (1999). Private self-
consciousness and the five-factor model of personality:
Distinguishing rumination from reflection. Journal of
Personality & Social Psychology, 76, 284–304.
van Randenborgh, A., Huffmeier, J., LeMoult, J., &
Joormann, J. (2010). Letting go of unmet goals: Does
self-focused rumination impair goal disengagement?
Motivation and Emotion, 34, 325–332.
348 A.M. Grant et al.
Downloaded by [University of Sydney] at 14:58 26 June 2012
