Content uploaded by Bruce Alexis Fernie
Author content
All content in this area was uploaded by Bruce Alexis Fernie on May 11, 2019
Content may be subject to copyright.
RUNNING HEAD: Metacognitive Model of Procrastination
!
1
A Metacognitive Model of Procrastination
Revision 1
Bruce A. Fernieab, Zinnia Bharuchaa, Ana V. Nikčevićc, Claudia Marinod, and
Marcantonio M. Spadae
aKing’s College London, Institute of Psychiatry, Psychology and Neuroscience,
Department of Psychology, London, UK
bHIV Assessment and Liaison Team, South London & Maudsley NHS Foundation
Trust, London, UK
cKingston University, Kingston upon Thames, UK
dDipartimento di Psicologia dello Sviluppo e della Socializzazione, Universita’ di
Padova, Padova, Italy
eLondon South Bank University, London, UK
October 2016
Compliance with Ethical Standards
All authors declare that they have no conflicts of interest. This study involved human
participants. All procedures performed in this study were conducted in accordance
with the ethical standards of the institutional research committee and with the 1964
Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent was obtained from all individual participants included in the study.
Author Notes
Correspondence should be addressed to: Bruce Alexis Fernie, Department of
Psychology, Institute of Psychiatry, Psychology and Neuroscience, King’s College
London, Henry Wellcome Building, De Crespigny Park, London, SE5 8AF, United
Kingdom. Tel. +44 (0)7779 300 427, fax +44 (0)20 7848 5310, e-mail
bruce.fernie@kcl.ac.uk
!
RUNNING HEAD: Metacognitive Model of Procrastination
!
2
Abstract
Background: procrastination refers to the delay or postponement of task or decision-
making initiation or completion and is often conceptualised as a failure of self-
regulation. Recent research has suggested that metacognitions play a role in
procrastination and that unintentional procrastination (UP), as opposed to intentional
procrastination (IP), may be the most problematic form of this behaviour. We aimed to
test a metacognitive model of procrastination that was grounded in the Self-Regulatory
Executive Function model. Methods: a convenience sample of 400 participants were
recruited and completed (at least partially) a battery of online questionnaires that
measured IP and UP, metacognitions about procrastination, depression, and Cognitive
Attentional Syndrome (CAS) configurations. Initially, we tested series of hypotheses to
establish the relationships between the experimental variables and to test whether CAS
configurations would independently predict UP when controlling for age, depression,
IP, metacognitions about procrastination, and whether an individual reported that they
had been diagnosed with a psychiatric disorder. Results: CAS configurations,
depression, and metacognitions independently predicted UP. Additionally, path
analysis revealed that the study data was an excellent fit to the proposed metacognitive
model of procrastination. Limitations: the study is cross-sectional. Conclusions: the
metacognitive model of procrastination presented in this paper can be used to generate
novel interventions to treat this problematic behaviour.
Keywords: Procrastination; metacognition; metacognitive model of procrastination;
Metacognitive Therapy; Self-Regulatory Executive Function model.
RUNNING HEAD: Metacognitive Model of Procrastination
!
3
Introduction
Procrastination
Most of us can recall a time in our lives when we have procrastinated, perhaps because
it is a nuanced concept that appears to be understood differently by different
individuals. Broadly speaking, the term ‘procrastination’ seems to be commonly used
to refer to an episode when an individual is ‘putting off’ or failing to complete an
activity (such as doing homework or filing a tax return) in any given moment.
Procrastination is a common behaviour, with the prevalence rates reported as high as
70% in students (Ellis & Knaus, 1977) and 20% in an adult sample (Harriott & Ferrari,
1996). Perhaps unsurprisingly, it has been found to be associated with diminished
academic and work performance, as well as poor mental health (Stöber & Joormann,
2001).
Some psychologists have conceptualised procrastination as a failure of self-
regulation (Baumeister & Heatherton, 1996; Baumeister, Heatherton, & Tice, 1994), in
other words a maladaptive attempt to manage behaviour or emotion. Some individuals
may believe that by postponing a task they will perform better (and successfully) at a
later date, however it is unlikely that this strategy consistently results in a successful
outcome (e.g., students submit assignments late, people fail to return their tax returns
on time, etc.). For this study, we define procrastination as the postponement or
avoidance of starting, engaging in, and/or completing a task or a decision-making
process, whether intentional or unintentional (Fernie, McKenzie, Nikčević, Caselli, &
Spada, 2015).
Conceptualizations and Models of Procrastination
Several different conceptualizations and models of procrastination have been proposed
in the extant psychological literature. For example, behaviourists have utilised operant
RUNNING HEAD: Metacognitive Model of Procrastination
!
4
conditioning to understand procrastination. This approach recruits avoidance behaviour
in the role of a maintaining factor for procrastination (Ferrari & Emmons, 1995).
Procrastination is reinforced because exposure to aversive stimuli (e.g., writing
challenging essays, cleaning filthy toilets, etc.) is avoided. This perspective has been
criticized for failing to account for individual differences amongst procrastinators
(Ferrari, Johnson, & McCown, 1995). From a more cognitive perspective, much like
Baumeister et al. (1994), Tuckman and Sexton (1989) also conceptualized
procrastination as a failure to self-regulate. In a similar manner, Ellis and Knaus (1977)
also postulated that procrastination was an illogical and non-goal directed behaviour but
emphasized the key role of irrational cognitions. Central to this Rationale-Emotive
Therapy perspective are the presence of two irrational beliefs: firstly, procrastinators
doubt their ability to complete a task and, secondly, they fear the possible negative
social consequences of failing to complete a task well. Further studies suggested that
particular cognitive constructs are implicated in procrastination. For example, the
perceived difficulty of a particular task (i.e., its level of ‘task aversiveness’) has been
shown to be associated with procrastination (Solomon & Rothblum, 1984), as well as
self-efficacy beliefs (Haycock, McCarthy, & Skay, 1998), self-esteem (Ferrari, 1994),
and perfectionism (Stöber & Joormann, 2001).
Later still, Steel and König (2006) presented a model of procrastination they
called Temporal Motivation Theory (TMT) in an attempt to synthesize several strands
of research. TMT can be represented by an equation that aims to calculate the perceived
utility ascribed to the initiation of, engagement with, and/or the completion of, a task,
arguing that this is a function of the likelihood that an individual will procrastinate in a
given situation. In calculating perceived utility, TMT employs several variables,
specifically: valence (i.e., the ‘amount’ of attraction or aversion an individual feels
RUNNING HEAD: Metacognitive Model of Procrastination
!
5
towards the task – its level of task aversiveness), expectancy (i.e., a measure of how
likely an individual believes that a given task will yield utility), delay (i.e., the length of
time before the individual will experience the expected outcome), and gratification (i.e.,
an individual’s intolerance of the delay). Arguably all of these variables tap in to
several cognitive constructs that have been implicated in procrastination, namely task
aversion, self-efficacy, self-esteem, and perfectionism. However, TMT (alongside other
traditional CBT conceptualizations of procrastination that tend to emphasize the role of
the content of cognitions) appears to neglect the possible key role of cognitive
processes and attentional strategies in this behaviour.
Metacognitions, the Self-Regulatory Executive Function model, and
Procrastination
The concept of metacognitions refers to a higher-order thinking that embodies beliefs
concerning cognitive processes, attentional strategies, behaviours, and physical
sensations. Metacognitions play a central role in the Self-Regulatory Executive
Function (S-REF: Wells & Matthews, 1994, 1996) model of psychiatric disorders.
The S-REF model describes a cognitive architecture, consisting of three interacting
levels that are delineated into an automatic, low-level (or bottom-up) component, an
online stage that reflects conscious cognitive processes and attentional strategies, and
a higher-level that represents long-term memory and is where, according to the
model, metacognitive beliefs (or metacognitions) are stored. In clinical practice, when
working from a Metacognitive Therapy (MCT: Wells, 2011) perspective that was
built from the S-REF model, psychiatric disorders and emotional distress are
formulated using the Cognitive Attentional Syndrome (CAS) as a framework. The
CAS consists of cognitive processes (such as distraction, rumination, and worry),
maladaptive behaviours (e.g., avoidance), and attentional strategies (for example, self-
RUNNING HEAD: Metacognitive Model of Procrastination
!
6
focussed attention) that are governed by metacognitive beliefs (e.g., “My rumination
is uncontrollable” and “My worry keeps me safe”). According to MCT, psychiatric
disorder and emotional distress are the consequences of particular CAS configurations
that result in ‘perseveration’: i.e., sustained engagement in unhelpful processes, which
themselves represent self-regulation strategies that fail to modify maladaptive self-
knowledge and behaviour. Problematic CAS configurations can be characterized by a
particular relationship that individuals have with their thoughts, such that they are
treated as facts that represent an objective reality (in the terms of the S-REF model,
this is labelled ‘object-mode’) rather than mere transient mental events that are
separate from the self and the world (termed ‘metacognitive-mode’); in this mode,
thoughts are a form of potentially inaccurate representations that provide but a
shadowy impression of reality (Wells, 2011).
Intentional and Unintentional Procrastination
More recently, procrastination has been delineated into intentional and unintentional
domains (Fernie, Bharucha, Nikcevic, & Spada, 2016). Intentional procrastination (IP)
refers to the deliberate and conscious engagement in this behaviour, while unintentional
procrastination (UP) pertains to situations where it is perceived as involuntary. UP, but
not IP, has been shown to be associated with low mood and anxiety (Fernie et al.,
2016). The construct of IP aligns itself with the development of a self-report measure
called the Active Procrastination Scale (APS; Choi & Moran, 2009) and, of particular
relevance to this study, a sub-factor within the APS named ‘Intentional Decision to
Procrastinate’. A brief self-report measure of UP has also recently been developed and
validated, namely the Unintentional Procrastination Scale (UPS: Fernie et al., 2016).
Chu and Choi (2005) proposed that there are two types of procrastinator:
passive and active. They characterized a passive procrastinator in a manner that alludes
RUNNING HEAD: Metacognitive Model of Procrastination
!
7
to more ‘traditional’ conceptualizations of this behaviour. Passive procrastinators
typically leave tasks to the last minute despite their good intentions, attenuating
performance. Active procrastinators choose to delay task initiation or completion,
believing that this strategy may actually optimize performance. This delineation
between active and passive procrastinators is similar to the distinction between IP and
UP. However, in this paper, we test a model of procrastination in which problematic
procrastinators engage in both IP and UP due to presence of metacognitions that
activate maladaptive cognitive processes, such as distraction, rumination, and worry.
We argue and test in this paper the hypothesis that cognitive or ‘ego’ depletion
(Baumeister, Muraven, & Tice, 2000; Muraven & Baumeister, 2000) is key to
understanding UP. We propose that when engaging in IP, a problematic procrastinator
initiates a particular CAS configuration, consisting of distraction, rumination, and
worry, in a futile attempt to regulate their behaviour. As a result, UP becomes
unavoidable and perseverative because these cognitive processes and attentional
strategies consume significant mental resources and are inefficient means to achieve
specific goals or complete required activities. Consequently, the problematic
procrastinator has insufficient mental resources to allocate to task initiation and/or
completion, in other words they begin UP. Furthermore, engagement in rumination and
worry have been shown to result in negative affect (Berenbaum, Bredemeier,
Thompson, & Boden, 2012; Nolen-Hoeksema, 1991; Papageorgiou & Wells, 2009), a
consequence of which may be a lethargy that may further contribute to attenuated
performance. The use of distraction is also likely to be unhelpful. This attentional
strategy would result in the misallocation of the remnants of mental resources away
from task initiation and/or completion, deleteriously affecting performance.
Metacognitions and Procrastination
RUNNING HEAD: Metacognitive Model of Procrastination
!
8
General metacognitions, measured using the Metacognitions Questionnaire 30 (MCQ-
30: Cartwright-Hatton & Wells, 1997; Wells & Cartwright-Hatton, 2004), have been
found to be significant predictors of both decisional and behavioural procrastination
(Spada, Hiou, & Nikcevic, 2006). Later still, metacognitions specific to procrastination
(Fernie & Spada, 2008) have been shown to be significant predictors of, again, both
decisional and behavioural procrastination (Fernie, Spada, Nikčević, Georgiou, &
Moneta, 2009). Furthermore, three impaired factors of attentional control (focusing,
shifting, and flexible control of thoughts), as well as two general metacognitive factors
(negative beliefs concerning thoughts about uncontrollability and danger and lack of
cognitive confidence), were found to be significantly and positively associated with
decisional procrastination (Fernie et al., 2015). In line with the S-REF model and
supported by earlier research (Fernie et al., 2016), we propose that IP is, to some extent,
a voluntary strategy aimed at self-regulation that is fuelled by metacognitions. Research
has shown that positive metacognitions about procrastination (e.g., “When I
procrastinate, I am unconsciously mulling over difficult decisions”) were more strongly
associated with IP than negative metacognitions about procrastination (e.g., “My
procrastination is uncontrollable”) when controlling for negative affect, while the
relative strengths of these relationships were inverted when the bivariate associations
between metacognitions and UP were examined (Fernie et al., 2016).
Study Aims and Hypotheses
Our first aim was to test several hypotheses to provide further evidence for the
proposed metacognitive model of procrastination. These were that: (1) positive
metacognitions about procrastination will be positively associated with both IP and UP
(this would replicate the findings of an earlier study: Fernie et al., 2016); (2) levels of
depression would be significantly associated with UP, maladaptive CAS
RUNNING HEAD: Metacognitive Model of Procrastination
!
9
configurations, and negative metacognitions about procrastination; (3) problematic
CAS configurations, consisting of distraction, rumination, and worry result in cognitive
depletion and therefore should be significantly and positively correlated with UP; and
(4) CAS configurations of this type will predict significant additional variance in UP
when controlling for negative affect, as well as other variables that have been found to
be associated with the dependent variable.
By testing these hypotheses, and collating our findings with those from earlier
research that investigated the role of metacognitions in procrastination, we would then
be able to address the primary aim of this study: i.e., to use path analysis to test a
metacognitive model of procrastination. In this model, positive metacognitions about
procrastination activate IP and then, if problematic CAS configurations are brought
online (fuelled by negative metacognitions about procrastination), this will result in
negative affect, before leading to UP.
Methods
Participants
A convenience sample of 400 (309 females; mean age = 30.9 years; SD = 12.6; range
18 to 72 years) participants was recruited for this study, via a university research
volunteer email circular and social media, and completed a battery of online
questionnaires, but only 305 participants completed all measures. This means that
nearly a quarter of individuals who started to participate in the study did not go on to
complete it, however this dropout rate is not unusual in web-based research (see
Birnbaum, 2004 for a discussion of the issues associated with web-based research).
Eligibility criteria required that participants: (1) were at least 18 years of age, (2)
possessed adequate English language skills, and (3) consented to participate. Due to the
significant number of international students enrolled at the university, the participants
RUNNING HEAD: Metacognitive Model of Procrastination
!
10
were from all continents, except Antarctica. However, the majority of participants
(53.2%) reported being British or having dual citizenship (one nationality of which was
British). The ethnicity of the sample was reasonably diverse, although the majority of
the participants self-identified as White (73.5%). While 20.1% of the sample reported
their ethnicity as Asian, 2.3% described themselves as Black, 2.3% as having a mixed
ethnicity, 0.3% stated they were Arab, and the remaining participants preferred not to
say. We anticipated that the sample would be international and that for many of the
participants English would not be their first language. This revealed itself to be a
reasonable supposition, with 36.2% of the sample reporting a first language other than
English. However, 95.7% of the participants reported that they were either ‘confident’
or ‘very confident’ in their ability to read English, while 94.3% rated their ability to
write English at these same confidence levels. None of the participants reported
inadequate English reading or writing skills.
We also wanted to be able to describe our sample in terms of (1) their mental
health and (2) their exposure to psychological therapies, because this could mean that
some participants were familiar with some of the concepts expressed by the items that
comprise the experimental measures used in this study, potentially risking a response
bias. Firstly, 9.3% disclosed that they had a mental health diagnosis. Of these, 73.0%
reported that they had been diagnosed with depression and/or an anxiety disorder.
Secondly, 24.5% of the entire sample stated that they had previous, or were currently
engaged in, psychological therapy. Of these, 47.0% had been exposed to CBT and
20.4% reported that their therapy was or is, at least in part, to help with problematic
procrastination. However, perhaps in part because problematic procrastination is not a
clinical diagnosis in either the latest version of the Diagnostic and Statistical Manual
(Association, 2013) or the International Statistical Classification of Disease and Related
RUNNING HEAD: Metacognitive Model of Procrastination
!
11
Health Problems (Organization, 2004), we were unable to evaluate whether this
represented an approximately equivalent rate to individuals in the wider population who
seek help for procrastination.
Materials
Emotional measures
The Patient Health Questionnaire 9 (PHQ-9; Kroenke, Spitzer, & Williams, 2001) was
used to assess depressive symptoms. The PHQ-9 is a nine-item scale designed to
measure depressive symptoms and possesses good psychometric properties, with
respondents indicating the level they have experienced the symptoms described by the
items (e.g., “Feeling down or depressed”) over the preceding two-weeks on a four-
point, Likert-type scale that ranges from ‘not at all’ (scoring zero) to ‘nearly all the
time’ (scoring three). Responses to all items are summed together, meaning that higher
scores indicate the presence of greater levels of depressive symptoms (Kroenke et al.,
2001).
Procrastination measures
The ‘Intentional Decision to Procrastinate’ (IDP) factor of Active Procrastination
Scale (APS) was used as a measure of IP. The full APS consists of a total of 16-items
equally distributed over four factors (Choi & Moran, 2009). The IDP factor contains
items such as “I intentionally put off work to maximize my motivation” and “To use
my time more efficiently, I deliberately postpone some tasks”. Participants are
required to indicate the extent to which they agree with such statements on a four-
point, Likert-type scale ranging from ‘disagree’ (scoring one) to ‘agree’ (scoring
four). The responses are summed, so that higher scores reflect greater levels of IP.
The IDP factor of the APS has been reported to possess good validity and adequate
internal consistency (Choi & Moran, 2009). We used the Unintentional
RUNNING HEAD: Metacognitive Model of Procrastination
!
12
Procrastination Scale (UPS: Fernie et al., 2016) to measure UP. The UPS consists of
six-items such as “Often I mean to be doing something, but it seems that sometimes I
just don’t get round to it” and “I really want to get things finished in time, but I rarely
do”. Much like the IDP, participants indicate their strength of belief in the items on a
four-point, Likert-type scale, ranging from ’do not agree’ (scoring one) to ‘agree very
much’ (scoring four). Again responses are totalled and higher scores indicate greater
levels of UP. The UPS possesses discriminant, construct, and concurrent validity, as
well as good internal consistency (Fernie et al., 2016).
Finally, we used the Metacognitions about Procrastination Scale (MaPS) as a
measure of meta-beliefs about procrastination (Fernie et al., 2009). The MaPS consists
of two, eight-item factors that assess positive metacognitions about procrastination
(PMP) and negative metacognitions about procrastination (NMP). An example item of
an item from the PMP factor is “When I procrastinate, I am unconsciously mulling over
difficult decisions” and for the NMP factor is “My procrastination is uncontrollable”. In
terms of concurrent validity, NMP has been shown to be significantly correlated with
both general and decisional procrastination, but PMP is only significantly associated
with the latter (Fernie et al., 2009). The MaPS uses the same response-format as the
UPS. Higher scores on either factor (which are summed separately) indicate a greater
endorsement of positive and/or negative metacognitions about procrastination. The
factors have been shown to possess good internal consistency (Fernie et al., 2009).
Measure of Cognitive Attentional Syndrome configurations
Participants were asked three questions about the extent that they worry, ruminate,
and/or try to distract themselves during an episode of procrastination. Participants were
required to respond on five-point, Likert-type scale ranging from “I do not worry about/
ruminate about/ distract myself from my procrastination or I do not procrastinate”
RUNNING HEAD: Metacognitive Model of Procrastination
!
13
(scoring zero) to “I always worry about/ ruminate about/ distract myself from my
procrastination” (scoring four). Thus higher scores suggest greater levels of worry,
rumination, and/or distraction activated in response to an episode of procrastination.
The responses to each question were converted to z-scores and used to calculate a mean
measure of problematic CAS configuration activation (termed CAS mean).
Procedure
Potential participants were directed to the study website containing the questionnaires.
The first two pages of this provided information regarding the purpose of the study,
describing that responses were anonymous, and that consent would be assumed once
participants click on the ‘submit’ button that followed the battery of questionnaires. In
the pages following this information, participants were presented with a series of
questions to ascertain their demographic details, their exposure to psychotherapy
(subsequently subjected to binary coding), and their current mental wellbeing (e.g., “Do
you have a psychiatric diagnosis?”). Participants were not required to record their
names. Once more participants were informed that, by clicking on the submit button
that followed the questionnaires, they were consenting to participate in the study. Prior
to this, they were informed that once they click the submit button, it would not be
possible to withdraw their data from the study because it was uploaded in an
anonymous form.
Statistical Analyses
Firstly, the distribution of the study data was assessed for non-normality. Secondly
statistical differences in the study variables were examined between males and females,
those with psychiatric diagnoses, and those exposed or naïve to psychological therapy.
Four of the experimental hypotheses were evaluated using correlational and
hierarchical regression analyses. Finally path analysis was used to test a metacognitive
RUNNING HEAD: Metacognitive Model of Procrastination
!
14
model of procrastination that was based on earlier work by Fernie et al. (2016) and used
UPS as the ‘end-point’ of the model because this variable has been shown to be
strongly associated with negative affect (Fernie et al., 2016).
Results
Distribution of Data and Non-Parametric Difference Tests Between Demographic
Variables
A series of Kolmogorov–Smirnov normality tests were conducted on the data from all
the experimental measures and these revealed that all of the variables were non-
normally distributed. As a consequence, a series of non-parametric Mann-Whitney U
tests were conducted on the data to establish whether the gender of participants, those
with psychiatric diagnoses, or individuals who had been exposed to psychological
therapy differed on all experimental measures. These tests suggested that males
significantly endorsed NMP more than females (Mann-Whitney U = 8638, n1 = 69, n2 =
252, p < .05). They also indicated that individuals that self-reported having a
psychiatric diagnosis had significantly higher levels of depression (Mann-Whitney U =
2333.5, n1 = 32, n2 = 273, p < .0001), engaged in more UP (Mann-Whitney U = 2466.5,
n1 = 32, n2 = 302, p < .0001), more strongly endorsed NMP (Mann-Whitney U = 2928,
n1 = 32, n2 = 289, p < .01), and showed greater activation of maladaptive CAS
configurations (Mann-Whitney U = 2560.5, n1 = 32, n2 = 274, p < .0001). This finding
suggested the need to control for psychiatric diagnosis in the regression analysis that
sought to test the fourth study hypothesis. Furthermore, these Mann-Whitney tests
revealed that participants who had been exposed to psychological therapy endorsed
NMP to a significantly less extent than those who had not (Mann-Whitney U = 8663, n1
= 92, n2 = 229, p < .05), hinting that talking therapies directly or indirectly targeted
these beliefs.
RUNNING HEAD: Metacognitive Model of Procrastination
!
15
Descriptive Statistics, Internal Consistencies of the Study’s Experimental
Measures, and Testing the First, Second, and Third Study Hypotheses
Table 1 shows the descriptive statistics and Cronbach’s alphas for all experimental
measures. We generated Spearman’s Rho correlation analyses to explore the
relationships between some of the participants’ demographics and the experimental
measures (see Table 1). The UPS significantly and positively correlated with all
variables (supporting the first, second, and third study hypotheses) except age, which
revealed a significant and negative relationship, suggesting that people engage in less
UP as they age. The variables that were significantly associated with UPS indicated that
we needed to control for them in the later regression analysis that tested the fourth
study hypothesis.
Hierarchical Regression Analysis with UPS as the Outcome Variable to Test the
Fourth Study Hypothesis
The data concerning participants’ ages, as well as their scores from the IDP, the PHQ-9,
both factors of the MaPS, their mean CAS z-scores, and whether a participant had
received a psychiatric diagnosis were assessed for their suitability for regression
modeling. No evidence of multicollinearity in the dataset: (1) no correlations greater
than r = .9 were identified between the predictor variables used in the regression
analysis, (2) the ranges of the Tolerance Index (TI) were 1.00 to 0.54 (i.e., no TIs were
calculated below 0.20), and (3) the Variance Inflation Factors (VIF) for all predictor
variables were less than 10 (the highest VIF was 1.84). Additionally, the Durbin-
Watson test suggested that the assumption of independent errors is tenable.
Furthermore, histograms and normality plots suggested that the residuals were normally
distributed and plots of the regression-standardized residuals against the regression-
RUNNING HEAD: Metacognitive Model of Procrastination
!
16
standardized predicted values provided evidence that the assumptions of linearity and
homoscedascity were met.
A six-step hierarchical regression analysis was conducted with UPS as the
outcome variable (see Table 2) to test the fourth study hypothesis. Participants’ age was
entered as a predictor on the first step, self-reported diagnosis of a psychiatric disorder
on the second, PHQ-9 on the third, IDP on the fourth, both PMP and NMP on the fifth,
and CAS mean on the final step. PHQ-9, PMP, NMP, and CAS mean remained
independent, significant predictors in the final model. At any step following the first,
participants’ age was a non-significant predictor of UPS, and at no point in the
hierarchical regression model was current or previous exposure to psychological
therapy an independent predictor of UP. The final model explained 37% of the variance
in UPS scores, supporting the fourth study hypothesis.
Path Analysis of the Metacognitive Model of Procrastination
The primary aim of this study sought to test our proposed metacognitive model of
procrastination (see Figure 1). We used the Lavaan package (Rosseel, 2012), running
on the R Studio package (R-Studio, 2012), to test the pattern of relationships of our
proposed model using path analysis (bootstrap = 1000). We were restricted to using
scores obtained from 305 of the 400 participants because cases were excluded if any
were missing data. Our model revealed that all estimated coefficients were significant
at the p < 0.001 level and explained 46% of the total variance in UPS, indicating that it
was an excellent fit of the observed data. To further evaluate the goodness of fit of the
model, we calculated the R-Square of each endogenous variable: IDP = 13%, CAS
mean = 40%, PHQ-9 = 16%, and UPS = 24%. Again, this suggested that our model was
at least a good fit of the observed data. Additionally all the indirect paths were
significant at a minimum of the 1% level (not shown in Figure 1). PMP had an indirect
RUNNING HEAD: Metacognitive Model of Procrastination
!
17
effect on CAS mean via IDP (β = .07, p < .01), IDP had an indirect effect on PHQ-9 via
CAS mean (β = .08, p < .01), and CAS mean had an indirect effect on UPS via PHQ-9
(β = .20, p < .001).
Discussion
Addressing the Aims of the Study
The primary aim of this study was to test a metacognitive model of procrastination that
was developed within the framework of the S-REF model and synthesized the findings
from several studies that had implicated a role for metacognitions in this often
problematic behaviour (Fernie et al., 2016; Fernie et al., 2015; Fernie & Spada, 2008;
Fernie et al., 2009; Spada et al., 2006). Prior to evaluating the model, we found
evidence to support all four of the study’s hypotheses. Firstly, PMP were significantly
and positively associated with both IP and UP. Secondly, PHQ-9 scores were
significantly associated with UP, maladaptive CAS configurations, and NMP,
providing further evidence for the key roles played by the CAS and metacognitions in
negative affect, and thus for the S-REF model. Thirdly, we found that problematic CAS
configurations were significantly and positively correlated with UP. Fourthly, CAS
configurations of this type predicted significant additional variance in UP when
controlling for the participants’ ages, their current or previous exposure to
psychological therapy, IDP, PMP, and NMP. These findings (particularly those that
resulted from the testing of the third and fourth hypotheses) lent support to our
contention that UP could be significantly explained by the depletion of mental
resources that had been misallocated to problematic CAS configurations consisting of
distraction, rumination, and worry. Indeed, earlier research had suggested that impaired
performance is associated with such cognitive and attentional processes (see Hofmann,
Schmeichel, & Baddeley, 2012 for a discussion).
RUNNING HEAD: Metacognitive Model of Procrastination
!
18
Arguably, we found good evidence that our proposed metacognitive model of
procrastination was, at the very least, a good fit of the data obtained by this study. Our
model proposes that PMP lead to IP that, alone, is not necessarily harmful. However, in
the case of problematic procrastinators, if in response to IP problematic CAS
configurations are brought online alongside NMP (e.g., “My procrastination is
uncontrollable”), this worsens mood and perservative UP results. This is because
engagement in cognitive processes such as worry and rumination drains mental
resources, reducing the ability to perform and further lowering mood. These now
depleted mental resources not only reinforce NMP, they are also often misallocated
through maladaptive attentional strategies (i.e., distraction) that make it even more
difficult to initiate or complete the task at hand, further contributing to UP and low
mood.
We suggest that our model has the capacity to offer a framework that can be
used to interpret a broad range of the findings from the extant literature on
procrastination. For example, our model clearly frames UP as a failure of self-
regulation (Baumeister & Heatherton, 1996; Baumeister et al., 1994; Tuckman &
Sexton, 1989). Additionally, intuitively it seems that the depletion of mental resources
resulting from the problematic CAS configurations, intrinsic to our model, is likely to
increase the perceived difficulty of a task (i.e., task aversiveness: Solomon &
Rothblum, 1984; Steel & König, 2006), worsen mood (Stöber & Joormann, 2001), and
lead to the perseveration of UP, consequently harming self-efficacy (Haycock et al.,
1998) and self-esteem (Ferrari, 1994) beliefs. It also seems that our model would
reinforce at least the first of the two beliefs that Ellis and Knaus (1977) argued were
strongly endorsed by problematic procrastinators: i.e. they doubt their ability to
complete a task. If maladaptive CAS configurations deplete mental resources,
RUNNING HEAD: Metacognitive Model of Procrastination
!
19
procrastinators will be less able to complete tasks. Furthermore, some studies have
reported a degree of success in reducing procrastination using paradoxical techniques
where participants were ‘allowed’ to procrastinate (e.g., Shoham-Salomon, Avner, &
Neeman, 1989). Such an intervention may indirectly challenge NMP and reduce the
activation of problematic CAS configurations.
Unlike Choi and Moran (2009)’s delineation of procrastinators into two distinct
sub-types (i.e., active and passive) that may seem to share a conceptual overlap with the
constructs of IP and UP, our proposed model does not see these types of behaviours as
representing a categorical difference between individuals. Instead, our findings suggest
that IP leads to UP via particular CAS configurations, NMP, and low mood. In fact, this
is consistent with the correlation analyses conducted by Choi and Moran (2009) that
found a significant and positive correlation between IDP and their measure of passive
procrastination.
Clinical Implications
Studies testing the efficacy of CBT for procrastination seem sparse. Indeed, the authors
of this study were only able to identify a few relevant published studies. Firstly, a
proposal for a RCT that plans to compare group-delivered versus Internet-based CBT
for procrastination (Rozental, Forsström, Nilsson, Rizzo, & Carlbring, 2014) and,
secondly, a study that reported a case series that evaluated a brief CBT intervention in a
small sample (n=7) utilising an uncontrolled design that reported successful outcomes
(Karas & Spada, 2009). Thirdly, another study reported that a group CBT intervention
was beneficial in reducing academic procrastination (Toker & Avci, 2015) whilst,
fourthly, a study that compared Acceptance and Commitment Therapy (arguably a third
wave competitor to MCT) to CBT suggested that both were at least mildly affective in
reducing problematic aspects of procrastination (Wang et al., 2015)
RUNNING HEAD: Metacognitive Model of Procrastination
!
20
However, we propose that treatments based on our model have significant
potential. The model that we have presented in this paper is grounded in S-REF theory
from which MCT was developed. A recent meta-analysis provided preliminary
evidence that MCT is more effective than traditional CBT in treating anxiety and
depression (Normann, van Emmerik, & Morina, 2014), which we suggest bodes well
for the potential efficacy of MCT-derived interventions for problematic procrastination.
Our model suggests several targets that could be addressed by MCT interventions.
Firstly, PMP and NMP could be challenged using restructuring techniques or
behavioural experiments. Secondly, the metacognitions hypothesized to govern the
problematic CAS configuration activated in response to IP could also be targeted with
the aim of reducing worry and rumination. Thirdly, a reduction of the use of distraction
could be achieved by improving selective attention by using Attention Training
Technique (Wells, 2007; Wells, White, & Carter, 1997).
On a side note, our analyses also provided further support for the strength of the
psychometric properties of the MaPS (Fernie et al., 2009). The PMP factor of the MaPS
revealed ‘acceptable’, and the NMP factor ‘excellent’, internal consistency.
Furthermore, evidence for the concurrent and discriminant validity of the MaPS was
generated. The PMP factor was significantly associated with IDP and UPS, while the
NMP factor was only significantly related to UPS and not IDP.
Limitations
This study is subject to several limitations that will have to be addressed by future
research. First, social desirability, self-report biases, context effects, and poor recall
may have contributed to errors in the self-report measurements. Second, a cross-
sectional design was adopted and this does not allow causal inferences. Third, this
study utilized self-report measures to assess subjective experience and meta-awareness
RUNNING HEAD: Metacognitive Model of Procrastination
!
21
and as such, like much cognitive research, there is always doubt whether we are
measuring the constructs we intend. Fourth, there were issues with the sample
characteristics: the majority of participants were female, nearly three quarters identified
themselves ethnically as ‘White’, and approximately half of the sample, in terms of
nationality, identified as British. This impacts on our ability to generalize these findings
to other ethnicities and nationalities, though a significant proportion of participants self-
reported as non-white and non-British. Fifth, the lack of homogenous sample
nationality risked leading to increased error measurements due to the self-report
measures all being written in English; however, participants’ ratings of their language
abilities suggested that very few lacked confidence in reading or writing in English.
Sixth, this study recruited a convenience sample and potential confounders such
as socio-economic status and education were not controlled for. Seventh, the
psychometric properties of the MaPS have, to date, not been extensively tested in
empirical studies. This requires the acknowledgement that the use of the MaPS in this
study potentially weakens its findings. Finally, an eighth limitation could lie in in the
similarities between the IDP and the PMP factor of the MaPS: i.e., are they measuring
the same construct? We argue that they are not. IDP measures the frequency of IP
behaviour whilst the PMP factor of the MaPS assesses the strength of positive meta-
beliefs about procrastination. Indeed, whilst these two variables were significantly
associated, this was found only to be a moderate correlation (r = .35), suggesting that
these are related, but different, constructs.
Conclusions
This study’s results are attenuated by several limitations, but we suggest that this paper
presents significant evidence that describes a novel and potentially clinically relevant
metacognitive model of procrastination from which new treatments, based on MCT,
RUNNING HEAD: Metacognitive Model of Procrastination
!
22
could be developed. We argue that problematic procrastination is a harmful behaviour
that, not only deleteriously affects students’ academic performance, but also reduces the
productivity of those in employment. It prevents individuals from achieving their full
potential, negatively impacts on their mental health, and is sometimes harmful to their
interpersonal relationships. Research has suggested that it affects a significant number
of people and therefore an effective treatment could be widely beneficial, both on
individual and societal levels. However, to justify such optimistic speculations, further
research is needed to test the validity of the metacognitive model of procrastination that
we have presented. For example, it may be of interest to test the model in samples of
individuals with a primary clinical diagnosis that often incorporates procrastination as
part of its presentation, such as addictive disorders. Fundamentally, however, the
accuracy and robustness of our model will depend on research that tests whether
modifications of the specified metacognitions, as well as improvements in selective
attention, alters levels of UP. This might be achieved by conducting a pilot study that
uses interventions derived from the model presented in this paper to disrupt the here
within implicated maintaining factors of procrastination.
RUNNING HEAD: Metacognitive Model of Procrastination
!
23
Acknowledgements
Author BAF receives salary support from the National Institute for Health Research
(NIHR) Mental Health Biomedical Research Centre and Dementia Research Unit at
South London and Maudsley NHS Foundation Trust and King’s College London. The
views expressed are those of the author and not necessarily those of the NHS, the NIHR
or the Department of Health.
Author BAF would also like to acknowledge the support of Emma Milam and Sarah
Thacker in the writing of this paper.
RUNNING HEAD: Metacognitive Model of Procrastination
!
24
References
Association, A. P. (2013). Diagnostic and Statistical Manual of Mental Disorders
(Fifth ed.): American Psychiatric Association.
Baumeister, R. F., & Heatherton, T. F. (1996). Self-regulation failure: An overview.
Psychological inquiry, 7(1), 1-15.
Baumeister, R. F., Heatherton, T. F., & Tice, D. M. (1994). Losing control: How and
why people fail at self-regulation: Academic Press.
Baumeister, R. F., Muraven, M., & Tice, D. M. (2000). Ego depletion: A resource
model of volition, self-regulation, and controlled processing. Social cognition,
18(2), 130.
Berenbaum, H., Bredemeier, K., Thompson, R. J., & Boden, M. T. (2012). Worry,
anhedonic depression, and emotional styles. Cognitive Therapy and Research,
36(1), 72-80.
Birnbaum, M. H. (2004). Human research and data collection via the Internet.
Psychology, 55(1), 803.
Cartwright-Hatton, S., & Wells, A. (1997). Beliefs about Worry and Intrusions: The
Meta-Cognitions Questionnaire and its Correlates. Journal of Anxiety
Disorders, 11(3), 279-296. doi: 10.1016/S0887-6185(97)00011-X
Choi, J. N., & Moran, S. V. (2009). Why not procrastinate? Development and
validation of a new active procrastination scale. J Soc Psychol, 149(2), 195-
211. doi: 10.3200/SOCP.149.2.195-212
Chu, A. H., & Choi, J. N. (2005). Rethinking procrastination: positive effects of
"active" procrastination behavior on attitudes and performance. J Soc Psychol,
145(3), 245-264. doi: 10.3200/SOCP.145.3.245-264
RUNNING HEAD: Metacognitive Model of Procrastination
!
25
Ellis, A., & Knaus, W. J. (1977). Overcoming procrastination: Or how to think and
act rationally in spite of life's inevitable hassles: Institute for Rational Living.
Fernie, B. A., Bharucha, Z., Nikcevic, A. V., & Spada, M. M. (2016). The
Unintentional Procrastination Scale. Journal of Rational-Emotive &
Cognitive-Behavior Therapy.
Fernie, B. A., McKenzie, A.-M., Nikčević, A. V., Caselli, G., & Spada, M. M. (2015).
The Contribution of Metacognitions and Attentional Control to Decisional
Procrastination. Journal of Rational-Emotive & Cognitive-Behavior Therapy,
1-13.
Fernie, B. A., & Spada, M. M. (2008). Metacognitions about procrastination: A
preliminary investigation. Behavioural and Cognitive Psychotherapy, 36(03),
359-364.
Fernie, B. A., Spada, M. M., Nikčević, A. V., Georgiou, G. A., & Moneta, G. B.
(2009). Metacognitive Beliefs About Procrastination: Development and
Concurrent Validity of a Self-Report Questionnaire. Journal of Cognitive
Psychotherapy, 23(4), 283-293. doi: 10.1891/0889-8391.23.4.283
Ferrari, J. R. (1994). Dysfunctional procrastination and its relationship with self-
esteem, interpersonal dependency, and self-defeating behaviors. Personality
and Individual Differences, 17(5), 673-679.
Ferrari, J. R., & Emmons, R. A. (1995). Methods of procrastination and their relation
to self-control and self-reinforcement: An exploratory study. Journal of Social
Behavior and Personality, 10(1), 135.
Ferrari, J. R., Johnson, J. L., & McCown, W. G. (1995). Procrastination Research:
Springer.
RUNNING HEAD: Metacognitive Model of Procrastination
!
26
Harriott, J., & Ferrari, J. R. (1996). Prevalence of procrastination among samples of
adults. Psychological Reports, 78(2), 611-616.
Haycock, L. A., McCarthy, P., & Skay, C. L. (1998). Procrastination in college
students: The role of self‐efficacy and anxiety. Journal of Counseling &
Development, 76(3), 317-324.
Hofmann, W., Schmeichel, B. J., & Baddeley, A. D. (2012). Executive functions and
self-regulation. Trends Cogn Sci, 16(3), 174-180. doi:
10.1016/j.tics.2012.01.006
Karas, D., & Spada, M. M. (2009). Brief cognitive-behavioural coaching for
procrastination: A case series. Coaching: An International Journal of Theory,
Research and Practice, 2(1), 44-53.
Kroenke, K., Spitzer, R. L., & Williams, J. B. (2001). The PHQ-9: validity of a brief
depression severity measure. J Gen Intern Med, 16(9), 606-613.
Muraven, M., & Baumeister, R. F. (2000). Self-regulation and depletion of limited
resources: Does self-control resemble a muscle? Psychol Bull, 126(2), 247.
Nolen-Hoeksema, S. (1991). Responses to depression and their effects on the duration
of depressive episodes. J Abnorm Psychol, 100(4), 569-582. doi:
10.1037/0021-843x.100.4.569
Normann, N., van Emmerik, A. A., & Morina, N. (2014). The efficacy of
metacognitive therapy for anxiety and depression: a meta-analytic review.
Depress Anxiety, 31(5), 402-411. doi: 10.1002/da.22273
Organization, W. H. (2004). International statistical classification of diseases and
health related problems (The) ICD-10. World Health Organization.
RUNNING HEAD: Metacognitive Model of Procrastination
!
27
Papageorgiou, C., & Wells, A. (2009). A prospective test of the clinical metacognitive
model of rumination and depression. International Journal of Cognitive
Therapy, 2(2), 123-131.
R-Studio. (2012). R Studio: integrated development environment for R: Version 0.97.
Rosseel, Y. (2012). lavaan: An R package for structural equation modeling. Journal
of Statistical Software, 48(2), 1-36.
Rozental, A., Forsström, D., Nilsson, S., Rizzo, A., & Carlbring, P. (2014). Group
versus Internet-based cognitive-behavioral therapy for procrastination: Study
protocol for a randomized controlled trial. Internet Interventions, 1(2), 84-89.
doi: 10.1016/j.invent.2014.05.005
Shoham-Salomon, V., Avner, R., & Neeman, R. (1989). You're changed if you do and
changed if you don't: Mechanisms underlying paradoxical interventions.
Journal of consulting and clinical psychology, 57(5), 590.
Solomon, L. J., & Rothblum, E. D. (1984). Academic Procrastination - Frequency and
Cognitive-Behavioral Correlates. Journal of Counseling Psychology, 31(4),
503-509. doi: Doi 10.1037//0022-0167.31.4.503
Spada, M. M., Hiou, K., & Nikcevic, A. V. (2006). Metacognitions, Emotions, and
Procrastination. Journal of Cognitive Psychotherapy, 20(3), 319-326. doi:
10.1891/jcop.20.3.319
Steel, P., & König, C. J. (2006). Integrating theories of motivation. Academy of
Management Review, 31(4), 889-913.
Stöber, J., & Joormann, J. (2001). Worry, procrastination, and perfectionism:
Differentiating amount of worry, pathological worry, anxiety, and depression.
Cognitive Therapy and Research, 25(1), 49-60.
RUNNING HEAD: Metacognitive Model of Procrastination
!
28
Toker, B., & Avci, R. (2015). Effect of Cognitive-Behavioral-Theory-based Skill
Training on Academic Procrastination Behaviors of University Students.
Educational Sciences: Theory & Practice, 5, 1157-1168.
Tuckman, B., & Sexton, T. (1989). Effects of relative feedback in overcoming
procrastination on academic tasks. Paper presented at the meeting of the
American Psychological Association, New Orleans, LA.
Wang, S., Zhou, Y., Yu, S., Ran, L.-W., Liu, X.-P., & Chen, Y.-F. (2015).
Acceptance and Commitment Therapy and Cognitive–Behavioral Therapy as
Treatments for Academic Procrastination A Randomized Controlled Group
Session. Research on Social Work Practice, 1049731515577890.
Wells, A. (2007). The attention training technique: Theory, effects, and a
metacognitive hypothesis on auditory hallucinations. Cognitive and
Behavioral Practice, 14(2), 134-138.
Wells, A. (2011). Metacognitive therapy for anxiety and depression: Guilford press.
Wells, A., & Cartwright-Hatton, S. (2004). A short form of the metacognitions
questionnaire: properties of the MCQ-30. Behav Res Ther, 42(4), 385-396.
doi: 10.1016/S0005-7967(03)00147-5
Wells, A., & Matthews, G. (1994). Attention and emotion: A clinical perspective:
Psychology Press.
Wells, A., & Matthews, G. (1996). Modelling cognition in emotional disorder: The S-
REF model. Behav Res Ther, 34(11), 881-888.
Wells, A., White, J., & Carter, K. (1997). Attention training: Effects on anxiety and
beliefs in panic and social phobia. Clin Psychol Psychother, 4(4), 226-232.
RUNNING HEAD: Metacognitive Model of Procrastination
!
29
Table 1: Means, SDs, ranges, and Cronbach’s alphas for all experimental variables and
a non-parametric correlation matrix
Mean
SD
Range
Cronbach's
alpha
2
3
4
5
6
7
1. Age
30.89
12.65
18-72
N/A
-.09
-.28**
-.18**
-.17**
-.13*
-.19**
2. IDP
15.59
5.2
4-28
0.66
.15**
.16**
.35**
-.06
.13*
3.PHQ-9
15.42
5.8
9-35
0.88
.48**
.22**
.44**
.43**
4. UPS
14.59
5.16
6-24
0.88
.21**
.46**
.50**
5.PMP
14.30
4.41
8-32
0.77
-.08
.16**
6. NMP
20.11
6.74
8-32
0.90
.60**
7. CAS
mean
.0013
0.83
-1.43-
1.78
N/A
Note. IDP = Intentional Decision to Procrastinate; PHQ-9 = Patient Health
Questionnaire 9; UPS = Unintentional Procrastination Scale; PMP = Positive
Metacognitions about Procrastination; NMP = Negative Metacognitions about
Procrastination; CAS mean = Cognitive Attentional Syndrome mean z-score; n=305 to
n=400; * = p < .05; ** = p < .01.
RUNNING HEAD: Metacognitive Model of Procrastination
!
30
Table 2: Six-step hierarchical regression analysis with UPS as the outcome variable
95% Confidence
Interval
Predictor
R2
Adjusted
R2
R2
Change
B
SE
β
LL
UL
Step 1
Age
-0.69
0.02
-.17**
-.11
-.02
.03**
.03**
.03**
Step 2
Age
-0.08
0.02
-.19**
-0.12
-0.03
EtPT
1.13
0.65
.10
-0.14
2.41
.04**
.03**
.01
Step 3
Age
-0.03
0.02
-.06
-0.07
0.02
EtPT
0.37
0.58
.03
-0.77
1.51
PHQ-9
0.42
0.05
.47**
0.33
0.51
.24**
.24**
.21**
Step 4
Age
-0.03
0.02
-.06
-0.07
0.02
EtPT
0.43
0.58
.04
-0.71
1.57
PHQ-9
0.41
0.05
.46**
0.32
0.50
IDP
0.10
0.05
.10
-0.00
0.20
.25**
.24**
.01
Step 5
Age
-0.02
0.02
-.04
-0.06
0.02
EtPT
0.03
0.54
.00
-1.03
1.09
PHQ-9
0.25
0.05
.28**
0.51
0.34
IDP
0.10
0.05
.10
-0.00
0.19
PMP
0.17
0.06
.14**
0.05
0.29
NMP
0.28
0.04
.37**
0.20
0.36
.36**
.35**
.11**
Step 6
Age
-0.01
0.02
-.03
0.05
0.03
EtPT
-0.08
0.53
-.01
-1.13
0.97
PHQ-9
0.23
0.05
.26**
0.14
0.33
IDP
0.07
0.05
.07
-0.03
0.17
PMP
0.14
0.06
.12*
0.02
0.26
NMP
0.20
0.05
.26**
0.11
0.29
CAS mean
1.17
0.37
.19**
0.44
1.91
.38**
.37**
.02**
Note. UPS = Unintentional Procrastination Scale; EtPT = Exposure to Psychological
Therapy; PHQ-9 = Patient Health Questionnaire 9; IDP = Intentional Decision to
Procrastinate; PMP = Positive Metacognitions about Procrastination; NMP = Negative
Metacognitions about Procrastination; CAS mean = Cognitive Attentional Syndrome
mean z-score; n = 305; * = p < .05; ** = p < .01.
RUNNING HEAD: Metacognitive Model of Procrastination
!
31
Figure 1: Standardized path coefficients for the metacognitive model of procrastination
Note. PMP = Positive Metacognitions about Procrastination; IDP = Intentional Decision
to Procrastinate; CAS = Cognitive Attentional Syndrome mean z-score; NMP =
Negative Metacognitions about Procrastination; PHQ-9 = Patient Health Questionnaire
9; UPS = Unintentional Procrastination Scale; n = 305; ** = p <.001.
