Personality and Individual Differences 163 (2020) 110029

Contents lists available at ScienceDirect

Personality and Individual Differences

journal homepage: www.elsevier.com/locate/paid

Self-regulation is negatively associated with habit tendencies: A validation T

of the German Creature of Habit Scale
Rebecca Overmeyera, , Sophia Fürtjesb, Karen D. Erschec, Stefan Ehrlichb, Tanja Endrassa
⁎

a
Faculty of Psychology, Institute of Clinical Psychology and Psychotherapy, Chair for Addiction Research, Technische Universtität Dresden, Dresden, Germany
b
Division of Psychological and Social Medicine and Developmental Neuroscience, Faculty of Medicine, Technische Universität Dresden, Dresden, Germany
c
Department of Psychiatry, University of Cambridge, Cambridge, UK

ARTICLE INFO ABSTRACT

Keywords: Adaptive behavior depends on the process of self-regulation. If self-regulation fails, maladaptive behavior may
Questionnaire start to predominate, leading to impulsive actions, maladaptive habits, and compulsive behavior patterns. The
Habit tendency to develop habitual behavior differs between individuals and appears to be associated with different
Routine aspects of regulatory control. The present study examined associations of habitual propensity with self-regula-
Automaticity
tion, impulsivity, and compulsivity, elucidating possible underlying functional connections and evaluated the
Self-regulation
Bayesian lasso procedure
psychometric properties of a German version of the Creature of Habit Scale (COHS). Data of two independent
community samples was collected online. Dynamic connections with measures of regulatory control, namely
self-regulation, impulsivity, and compulsivity, were analyzed using a network approach. Psychometric analyses
were conducted using exploratory and confirmatory factor analyses utilizing a Bayesian lasso procedure.
Network analysis revealed that self-regulation was the most central node within the network of dynamic in-
teractions between habitual propensity, impulsivity, and compulsivity. Self-regulation exclusively had negative
associations with other measures. We replicated and validated the two-factorial structure of the COHS. Our
findings provide evidence that habitual propensity is interconnected with measures of regulatory control.
Information gathered using the COHS could be useful in identifying vulnerabilities and developing tailored
therapeutic strategies for psychopathologies associated with deficient regulatory control.

1. Introduction factors suspected to influence habit propensity, like intrinsic goal-di-

rectedness, or self-regulation, while also taking into account personality
Habits not only make actions more efficient, but maladaptive habits traits characterized by loss of control over behavior, namely impulsivity
can make actions dysfunctional. A typical example of such a dysfunc- and compulsivity (Hofmann, Schmeichel, & Baddeley, 2012; Robbins,
tional process is addiction, where individuals lose control over initia- Gillan, Smith, de Wit, & Ersche, 2012).
tion, duration and amount of their habitual drug use (American Habits are response patterns individuals exhibit repeatedly in a
Psychiatric Association, 2013). These processes presumably depend on specific situation (Wood & Rünger, 2016). They allow for highly effi-
the ability to self-regulate, as self-regulation refers to the deliberate or cient execution of routine acts, while attention can be focused else-
automatic use of mechanisms and metaskills for the modulation of where (Wood, Quinn, & Kashy, 2002). Habits are therefore established
behavior, attention, affect or thought (Heatherton, 2011; Karoly, 1993). to reduce the cognitive load of costly behavioral control over goal-di-
Understanding these processes and the factors influencing them is of rected behavior (Ouellette & Wood, 1998). Habitual response patterns
vital importance for developing treatments for habits that have become are learned by repetition (Lally, Van Jaarsveld, Potts, & Wardle, 2010).
dysfunctional and harmful and identifying individuals at risk for de- After repeated coupling of a stimulus with a specific response pattern
veloping maladaptive habits. The aim of the current study was to va- and subsequent reinforcement, the stimulus elicits the response pattern
lidate a German version of the Creature of Habit Scale (COHS), which automatically (Aarts, Verplanken, & Van Knippenberg, 1998). There-
assesses individual propensity for habit formation (Ersche, Lim, Ward, fore, repeated goal-pursuit in stable contexts leads to habit formation
Robbins, & Stochl, 2017). We assessed its association with internal (Ouellette & Wood, 1998). Hence, it is important to distinguish between

⁎
Corresponding author at: Rebecca Overmeyer, Technische Universität Dresden, Institute of Clinical Psychology and Psychotherapy, Chair for Addiction Research,
Chemnitzer Straße 46a, 01187 Dresden, Germany.
E-mail address: [email protected] (R. Overmeyer).

https://doi.org/10.1016/j.paid.2020.110029
Received 22 January 2020; Received in revised form 24 March 2020; Accepted 27 March 2020
0191-8869/ © 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
R. Overmeyer, et al. Personality and Individual Differences 163 (2020) 110029

goal-directed and habitual instrumental behavior: Goal-directed beha- The COHS was developed to measure the tendency towards habitual
vior is guided by the known association between an action and its behavior in daily life and has been shown to efficiently assess the
consequences, the outcome is motivationally relevant at the time of propensity for habitual behavior in English speaking populations, with
decision-making (Dolan & Dayan, 2013). An example would be a per- satisfactory reliability (Ersche et al., 2017). The original version ex-
son's decision to abstain from the usage of plastic bags for the sake of hibited a two-factorial structure, distinguishing between routine beha-
the environment and take reusable ones to the supermarket for grocery vior and habitual automatic responses. The scales have been shown to
shopping. The first few times, conscious control likely needs to be al- be differentially associated with various measures of personality traits,
located to remember bringing the bags, maybe even forgetting them a like anxiety and compulsive as well as impulsive personality traits,
few times, and having to return to get them. Habitual behavior, how- goal-pursuit and cognitive flexibility. COHS Routine has been shown to
ever, is independent from the current motivational value of an outcome be inversely associated with sensation-seeking and impulsivity, COHS
and has been shaped by past reinforcement (Thorndike, 1911). Again, automaticity to be inversely correlated with goal-pursuit and positively
our example: After some time, in which a person had already taken the with impulsivity (Ersche et al., 2017, 2019). Both scales were asso-
reusable bags with them, they may not even have to consciously re- ciated with higher compulsivity as assessed with the obsessive-com-
member packing the reusable bags anymore. They just do, just as they pulsive inventory (OCI; Foa et al., 2002), a finding which has been
would with their wallet and keys. Therefore, goal-directed instrumental replicated (Ersche et al., 2019). They also appear to be negatively as-
control is characterized by active deliberation, high computational cost sociated with measures of cognitive flexibility (Lange & Dewitte, 2019),
and adaptive flexibility, while habitual instrumental control is marked which supports a possible connection to self-regulation. There is also
by automaticity, computational efficiency and inflexibility (Dayan, evidence suggesting that habitual tendencies, as measured by the
2009). Goals and consequences of repeated goal-directed actions COHS, explain more variance than the frequency of the behaviors in
therefore subsequently become less important and the behavior less question (Ersche et al., 2019).
susceptible to conscious control (Danner, Aarts, & de Vries, 2008; The present study had two aims: First, examine associations of self-
Robbins & Costa, 2017; Verplanken, 2006). This shift from goal-di- regulation, impulsivity, and compulsivity with habitual propensity. As
rected to habitual instrumental behavior is influenced by various in- outlined above, we expected that self-regulation should be negatively
ternal as well as external factors, e.g. high intrinsic goal-directedness, associated with habitual propensity, but also impulsivity and compul-
stress, and contact with stimulants (Corbit, Chieng, & Balleine, 2014; sivity. We also expected that it should influence the connection between
Linnebank, Kindt, & de Wit, 2018; Schwabe & Wolf, 2011). habitual propensity and impulsivity as well as compulsivity. Second, to
There is evidence that habitual behavior partitions into different assess habitual propensity, we furthermore aimed to develop and va-
sub-dimensions: Routines and automatisms are conceptualized as forms lidate a German version of the COHS. To this end, the COHS was
of habits that share critical features, but also distinctly differ from each translated into German and assessed online in two independent sam-
other concerning their function as well as their control over behavior ples. We expected to replicate the two-factorial structure of the ques-
(Ersche et al., 2017; Orbell & Verplanken, 2010). Routines can be tionnaire.
characterized as sequential action patterns that are deliberately exe-
cuted to make daily life more efficient, and can consciously be amended 2. Methods
if no longer appropriate (Clark, 2000). Habitual automatisms, however,
can be defined as cue-response associations and are a form of auto- 2.1. Study design and sample
maticity, but not synonymous (Orbell & Verplanken, 2010). They are
insensitive to short-term changes in goals (Wood, Labrecque, Lin, & 2.1.1. Sample size estimation
Rünger, 2014). Minimum sample size for factor analysis was estimated based on
The tendency towards developing such habitual behaviors differs simulation studies by Gagne and Hancock (2006), who proposed a
substantially between individuals (Lally et al., 2010). Furthermore, method that bases sample size estimation on measurement model
excessive habit formation appears to be prominent in mental disorders quality, or reliability, which can both be derived from the number of
characterized by deficient goal-directed control over action and re- indicators per factor and the factor loadings of each indicator. There-
sponse inhibition manifesting in high levels of compulsivity and/or fore, taking into account the number of indicators per factor (n = 11
impulsivity (Gillan, Robbins, Sahakian, van den Heuvel, & van Wingen, and n = 16, respectively) and the factor loadings of the original
2016). Compulsivity describes the failure to stop an ongoing in- questionnaire, we estimated a minimum sample size of N = 200.
appropriate behavior, while impulsivity describes the failure to inhibit
the initiation of behavior (Robbins et al., 2012). Substance use dis- 2.1.2. Sample
orders (SUD), eating disorders, trichotillomania and obsessive-com- The assessment of both samples included the COHS as well as so-
pulsive disorder (OCD) are just a few examples of disorders character- ciodemographic information, including age, education level and native
ized by deficient goal-directed control, impulsivity and compulsivity language. All participants were above 18 years of age and were native
(Gillan et al., 2011; Gillan et al., 2016; Hogarth, Chase, & Baess, 2012; speakers of German. Data were collected online and participants'
Robbins et al., 2012). However, even in healthy samples these con- identity remained anonymous to the research team.
nections persist: Compulsivity has been shown to be associated with a Sample 1 additionally included measures of obsessive-compulsive
preference for strict routines and a higher tendency to form habits and impulsive traits, self-regulatory capacity as well as two control
(Fineberg et al., 2010). Self-regulatory capacity, as well as impulsivity, items to check for attention (Meade and Craig, 2012). The order of the
have also been associated with habitual behavior (Carden & Wood, questionnaires was randomized across participants. Complete data from
2018; Torregrossa, Quinn, & Taylor, 2008). One could suspect, based on 439 individuals were collected online using the internet platform
these already established associations, deficient self-regulation to be the LimeSurvey (LimeSurvey Project Team, 2015). 199 participants were
cause of habitual behavior becoming problematic. In the case of weak excluded due to either false responding to the control items (n = 15),
self-regulation, desire may predominate behavior, leading to impulsive no fluency in German (n = 5), the presence of current or past self-
actions (Hofmann et al., 2012), and it may also increase the risk of reported mental disorders or intake of psychotropic substances within
habitual behavior spiraling out of control, leading to compulsive be- the past 3 months (n = 133), or neurological illness with possible in-
havior patterns (Gillan & Robbins, 2014). Although habits are regulated fluence on mood, cognition and behavior (n = 46). The final sample
by external cues, they require intact self-regulation for an individual to included 240 participants (mean age 26.5 years ± 8.16 standard de-
be able to adapt to changing environmental conditions and keep the viation (SD), 77.9% female; 83.8% had completed advanced education
balance between habitual efficiency and adaptability (Karoly, 1993). degrees). Participants could take part in a lottery to win 10 Euro.

2
R. Overmeyer, et al. Personality and Individual Differences 163 (2020) 110029

Sample 2 comprised data from 602 female individuals was collected Reliability was assessed using McDonald's omega and Cronbach's alpha
online via the crowdsourcing platform clickworker (clickworker® (Cronbach, 1951; McDonald, 2013; Revelle & Zinbarg, 2009). Con-
GmbH, Essen, Germany), the study was set up via the online platform vergent and discriminant validity were examined using Kendall's tau
Labvanced (Finger, Goeke, Diekamp, Standvoß, & König, 2017). 134 (Kendall, 1938) correlations with measures of personality traits that
participants were excluded due to current or past self-reported mental have been linked to the individual propensity to habitual behavior.
health issues. This left a total sample of 468 participants (mean age Kendall's tau has been shown to be a better estimate of the correlation
29.3 years ± 7.05 SD, 88.7% had completed advanced education de- in the population, if the data is distributed non-normally (Howell,
grees). Participants received a financial compensation of 3.75 Euro for 2012). To analyze the dynamic connections between the assessed per-
their participation. sonality traits, we used a network approach and estimated a standar-
The ethics committee at the Technische Universität Dresden ap- dized Gaussian Graphical Model (GGM) using the graphical lasso as a
proved both study procedures (EK 310082018 and EK 135042018). regularization method, the tuning parameter was selected according to
the Extended Bayesian information criterion (Chen & Chen, 2008;
2.2. Measures Foygel & Drton, 2010; Friedman, Hastie, & Tibshirani, 2008; Lauritzen,
1996). Analysis was performed based on polychoric correlations
The COHS (Ersche et al., 2017) assesses the tendency for habitual (Epskamp & Fried, 2018). Edge weight, or correlation accuracy and
behavior in various situations in daily life on 27 items. The original stability of node centrality indices as measures of node importance were
version of the COHS was shown to exhibit a two-factorial structure, assessed using bootstrapping (see Epskamp, Borsboom, & Fried, 2018).
with factors termed routine (16 items) and automaticity (11 items). Re- All analyses were carried out with R (R Core Team, 2018), using the
liability has been shown to be satisfactory for both scales (COHS routine: packages psych, corpcor, GPArotation, nFactors, RGenData, Facto-
Cronbach's α = 0.89, McDonald's ω = 0.92; COHS automaticity: MineR, MASS, MCMCpack, msm, psychometric, qgraph and bootnet
Cronbach's α = 0.91, McDonald's ω = 0.86). The COHS was translated (Bernaards & Jennrich, 2005; Epskamp et al., 2018; Epskamp, Cramer,
into German and back into English by psychologists who were bilingual Waldorp, Schmittmann, & Borsboom, 2012; Fletcher, 2010; Jackson,
speakers of German and English. The retranslated version were com- 2011; Lê, Josse, & Husson, 2008; Martin, Quinn, & Park, 2011; Raiche,
pared to the original questionnaire. Differing items were discussed and 2010; Revelle, 2018; Ripley et al., 2013; Ruscio, 2018; Schafer et al.,
adapted (see Supplement I). 2017). The Bayesian lasso CFA procedure was carried out using scripts
The German version of the OCI, consisting of 18 items, was used to provided by Pan et al. (2017).
assess obsessive-compulsive symptoms using a sum score (Foa et al., 2002;
Gonner, Leonhart, & Ecker, 2007). Internal consistency has been shown 3. Results
to be satisfactory with Cronbach's α = 0.85 (Gonner et al., 2007). The
German version of the Barratt Impulsiveness Scale-11 (BIS) consists of 3.1. Exploratory factor analysis
30 items and is an instrument assessing impulsivity as a personality trait
(Patton, Stanford, & Barratt, 1995; Preuss et al., 2003). High im- Assumptions for EFA were met (see Supplement II). An initial ana-
pulsivity is characterized by rash, imprudent actions without con- lysis was conducted to extract the number of factors to retain. For
sideration of possible negative consequences and is operationalized sample 1, PA extracted two components, OC extracted four factors,
using a sum score. Internal consistency has been shown to be satisfac- MAP and CD extracted two, AF extracted one. We analyzed the data
tory with Cronbach's α = 0.83 (Stanford et al., 2009). The German using four and two factors. Compared to the two-factor solution, the
version of the Self-Regulation Scale (SRS) consists of 10 items and as- four-factor solution yielded more cross loadings and did not seem to
sesses the self-regulatory capacity for sustaining an action even if there adhere to meaningful constructs (see Table 1, Supplement II). For
are influences interfering with motivation and attention (Schwarzer, sample 2, results were ambiguous as well: PA extracted two compo-
Diehl, & Schmitz, 1999). High self-regulatory capacity is associated nents, OC and CD extracted three factors, MAP and AF extracted two.
with goal-directed behavior and perceived self-efficacy. Internal con- We calculated results for the three- and the two-factor solution. The
sistency has been shown to be satisfactory with Cronbach's α between three-factor solution reproduced the routine scale, but split up the au-
0.73 and 0.82 (Diehl, Semegon, & Schwarzer, 2006; Luszczynska, Diehl, tomaticity scale into two factors. As most of the items from the two
Gutiérrez-Dona, Kuusinen, & Schwarzer, 2004). automaticity factors had substantial cross-loadings, the split into two
separate constructs did not appear meaningful (see Table 2, Supplement
2.3. Data analysis II). Due to the more convincing results from the two-factor solutions,
two factors were retained in the analysis. Table 1 shows the factor
To validate the German version of the COHS in each sample, we first loadings after rotation. The clustering of the items replicated the factors
performed exploratory factor analysis (EFA) with oblique rotation from the original COHS routine and automaticity factors. Factors were
(oblimin). Analysis was conducted on a polychoric correlation matrix correlated (sample 1: ϕ = 0.18 [CI 0.06–0.29]; sample 2: ϕ = 0.074 [CI
owing to non-normality of the data (Holgado–Tello, Chacón–Moscoso, 0.027–0.122]). All squared differences between factor loadings of the
Barbero–García, & Vila–Abad, 2010), as assessed by Mardia's test two samples were below 0.03, suggesting strong replication (Osborne &
(Mardia, 1970). To extract the number of factors or components, the Fitzpatrick, 2012).
following techniques were used due to their comparably high accuracy
rate (Ruscio & Roche, 2012): parallel analysis for component extraction 3.2. Bayesian lasso confirmatory factor analysis
(PA), minimum average partial procedure (MAP), optimal coordinates
(OC), acceleration factor (AF) and comparison data (CD). To validate A CFA using a Bayesian lasso to reduce the inverse covariance
the factorial structure, we performed a confirmatory factor analysis matrix (Pan et al., 2017) was conducted in both samples. Goodness of
(CFA) using a Bayesian least absolute shrinkage and selection operator Fit for the proposed model was tested via posterior predictive (PP) p-
(lasso) to reduce the inverse covariance matrix, based on the procedure value (sample 1: PP p = .53; sample 2: PP p = .48) and the proposed
described by Pan, Ip, and Dube (2017). This procedure has the ad- two-factor model considered plausible, because the estimates were not
vantage of allowing the residuals to be correlated, after latent factors far from 0.50. The Bayesian estimates of the unknown parameters in the
have been included in the model, thus providing a better tradeoff be- factor-loading matrix (λ), their corresponding Highest Posterior Den-
tween fit and model complexity, an identifiable model, valid estimates sity Intervals (HPD) and standard errors (SE) are presented in Table 2.
of standard errors, and producing reliable results even in smaller See Fig. 1 for a path diagram of the COHS structure. All factor-loading
samples. We additionally performed CFA without Bayesian lasso. estimates were significant and appeared to be satisfactory in

3
R. Overmeyer, et al. Personality and Individual Differences 163 (2020) 110029

Table 1 magnitude. As expected, the factors were positively correlated to a

Oblimin rotated standardized loadings (pattern matrix) based upon polychoric small extent (sample 1: ϕ = 0.069 [HPD 0.016–0.137]; sample 2:
correlation matrix for samples 1 and 2. ϕ = 0.074 [HPD 0.027–0.122]). The current analysis revealed 2.14%
Sample 1 Sample 2 (n = 15) and 6.27% (n = 44) of significant residual covariances within
the two samples, respectively. In Bayesian lasso CFA, not every residual
COHS COHS COHS COHS automaticity covariance has to be equal to zero, but the amount of significant re-
routine automaticity routine
sidual covariance should be smaller than 10%. See Supplement III for
Item 01 0.54 0.03 0.57 −0.01 results from CFA without Bayesian lasso (Tables 3–6) as well as the
Item 02 0.41 0.00 0.57 0.02 standardized residual covariance matrix (Table 7) and the polychoric
Item 04 0.53 −0.06 0.49 −0.11 correlation matrix from sample 1 (Table 8).
Item 06 0.42 0.23 0.39 0.13
Item 07 0.69 −0.08 0.64 −0.12
Item 10 0.50 0.22 0.43 0.22 3.3. Reliability
Item 12 0.66 −0.12 0.66 −0.01
Item 13 0.62 0.02 0.59 0.12
Item 14 0.58 −0.09 0.51 0.01 McDonald's omega and Cronbach's alpha suggested satisfactory re-
Item 15 0.85 −0.06 0.83 −0.05 liability for COHS routine (sample 1: α = 0.85 [CI 0.82–0.88], ω = 0.86
Item 17 0.45 0.08 0.52 0.13 [CI 0.83–0.88]; sample 2: α = 0.84 [CI 0.82–0.86], ω = 0.84 [CI
Item 18 0.30 0.08 0.35 0.02 0.82–0.87]), as well as COHS automaticity (sample 1: α = 0.85 [CI
Item 20 0.51 0.04 0.50 −0.04
0.82–0.88], ω = 0.86 [CI 0.83–0.88]; sample 2: α = 0.83 [CI
Item 22 0.68 0.11 0.57 0.03
Item 24 0.26 0.16 0.32 0.16 0.80–0.85], ω = 0.83 [CI 0.81–0.86]).
Item 27 0.72 0.07 0.69 0.03
Item 03 −0.06 0.74 0.04 0.59
Item 05 −0.03 0.57 −0.09 0.59 3.4. Validity and network dynamics
Item 08 0.02 0.34 0.05 0.40
Item 09 0.04 0.49 0.03 0.57 The validity of the COHS subscales was examined in sample 1 and is
Item 11 0.00 0.75 0.02 0.73
displayed in Table 3. Correlations were small to moderate in magni-
Item 16 0.07 0.65 0.16 0.58
Item 19 −0.04 0.75 −0.10 0.72 tude, even when considering the different metric of the tau statistic
Item 21 0.20 0.33 0.13 0.34 (Gilpin, 1993). Fig. 2 depicts the network model we used to evaluate
Item 23 0.00 0.65 0.05 0.56 the dynamic connections between self-reported measures of routine,
Item 25 0.03 0.88 −0.02 0.84 habit automaticity, impulsivity, compulsivity, and self-regulation in
Item 26 0.00 0.71 0.03 0.53
sample 1. The edge weights (i.e. connections between nodes) can be
Note. COHS routine and COHS automaticity = Creature of Habit Scale (Ersche interpreted as partial correlations, controlled for the presence of all
et al., 2017) routine and automaticity subscales. other variables in the network (Borsboom & Cramer, 2013). The
strongest connection was the negative association between self-

Table 2
Parameter estimates of Bayesian Lasso confirmatory factor analysis for samples 1 and 2.
Item Parameter estimates for sample 1. Parameter estimates for sample 2.

λ HPD SE λ HPD SE

COHS routine 1 0.59* 0.48*

2 0.53 0.34 – 0.72 0.10 0.63 0.47 – 0.82 0.09
4 0.48 0.32 – 0.68 0.09 0.45 0.29 – 0.60 0.08
6 0.54 0.36 – 0.74 0.10 0.44 0.26 – 0.61 0.09
7 0.58 0.38 – 0.77 0.10 0.56 0.40 – 0.71 0.08
10 0.52 0.35 – 0.69 0.09 0.37 0.25 – 0.51 0.06
12 0.69 0.51 – 0.86 0.09 0.67 0.52 – 0.84 0.08
13 0.75 0.55 – 0.97 0.11 0.71 0.53 – 0.90 0.10
14 0.51 0.34 – 0.68 0.09 0.54 0.38 – 0.68 0.08
15 0.91 0.71 – 1.00 0.11 0.87 0.70 – 1.00 0.10
17 0.42 0.28 – 0.59 0.09 0.52 0.36 – 0.68 0.08
18 0.31 0.15 – 0.50 0.09 0.31 0.16 – 0.45 0.07
20 0.50 0.36 – 0.66 0.08 0.45 0.31 – 0.59 0.07
22 0.71 0.53 – 0.89 0.10 0.56 0.41 – 0.72 0.08
24 0.33 0.16 – 0.50 0.09 0.31 0.18 – 0.43 0.06
27 0.87 0.68 – 1.00 0.10 0.76 0.58 – 0.93 0.09
COHS automaticity 3 0.59* 0.62*
5 0.45 0.32 – 0.61 0.07 0.48 0.35 – 0.61 0.07
8 0.25 0.11 – 0.38 0.07 0.29 0.19 – 0.42 0.06
9 0.27 0.13 – 0.43 0.07 0.42 0.26 – 0.59 0.08
11 0.61 0.49 – 0.72 0.06 0.62 0.48 – 0.76 0.07
16 0.58 0.45 – 0.71 0.07 0.51 0.36 – 0.65 0.07
19 0.64 0.46 – 0.84 0.10 0.64 0.47 – 0.83 0.09
21 0.28 0.09 – 0.48 0.09 0.25 0.10 – 0.38 0.07
23 0.52 0.35 – 0.69 0.09 0.44 0.29 – 0.58 0.07
25 0.72 0.59 – 0.87 0.08 0.73 0.57 – 0.92 0.08
26 0.50 0.38 – 0.62 0.06 0.41 0.30 – 0.54 0.06

Note. λ = standardized factor loadings; HPD = Highest Posterior Density Intervals, characterize uncertainty associated with estimated parameter; SE = standard
error; the items with asterisks were fixed to identify the scale of the latent factor; COHS routine and COHS automaticity = Creature of Habit Scale routine and
automaticity subscales (Ersche et al., 2017).

4
R. Overmeyer, et al. Personality and Individual Differences 163 (2020) 110029

Item 27
Item 24

Item 22
Item 3

Item 20 0.87
0.33 Item 5

Item 18 0.71 0.59

Item 8
0.50 0.45
Item 17
0.31 0.25 Item 9
0.42
Item 15 0.27
0.91 0.07 Item 11
0.61
Item 14
0.51
Routine Automaticity 0.58 Item 16
0.75
Item 13
0.64
0.69 Item 19
0.28
Item 12 0.52
0.52 Item 21
0.58
Item 10
0.54 0.72
Item 23
Item 7 0.48 0.50
Item 25
0.53
Item 6
0.59
Item 26
Item 4

Item 2
Item 1

Fig. 1. Path diagram of the Creature of Habit Scale (Ersche et al., 2017) Bayesian lasso confirmatory factor analysis results in sample 1, including all items with their
respective standardized factor loadings on the subscales as well as the weak correlation between the two subscales.

regulation and impulsivity (pr = −0.41). Interestingly, the direct centrality indices of the GGM. The correlation stability coefficient (CS)
connections between impulsivity and automaticity as well as self-reg- quantifies the maximum proportion of cases that can be dropped while
ulation and routine did not meet the threshold for relevance, leaving maintaining a correlation of higher than 0.7 (Epskamp et al., 2018).
shortest indirect connections via self-regulation and impulsivity, re- The CS indicated that betweenness (CS(cor = 0.7) = 0) and closeness
spectively. Both routine and automaticity were positively connected to (CS(cor = 0.7) = 0.129) were not stable under bootstrap subsetting
compulsivity (pr = 0.23 and pr = 0.14, respectively), as well as to each cases and have to be interpreted with caution. Node strength and edges
other (pr = 0.14). The connection between automaticity and self-reg- performed better (CS(cor = 0.7) = 0.283; CS(cor = 0.7) = 0.363).
ulation was negative (pr = −0.29), as was the connection between
routine and impulsivity (pr = −0.31). The connection between self-
regulation and compulsivity was negative (pr = −0.20). See Fig. 2 in 4. Discussion
Supplement IV for bootstrap-estimated accuracy of edge weights. Self-
regulation appeared to be the most central node, as it had the highest The present study investigated the dynamic associations of self-
values for node strength, closeness and expected influence, which regulation, impulsivity, compulsivity and habitual propensity. As ex-
quantify how well a node is directly as well as indirectly connected to pected, self-regulation was found to be negatively associated with all
other nodes (Epskamp et al., 2018). See Fig. 1 and Table 9 in Supple- other measures of regulatory control. An exception was COHS routine,
ment IV for a visualization of the z-standardized and the raw values of which was connected to self-regulation via impulsivity. COHS routine
was negatively correlated with impulsivity. Both subscales of the COHS

Table 3
Kendall's tau correlations and their respective p-value between the two subscales of the COHS and measures of compulsive and impulsive personality traits as well as
self-regulatory capacity calculated for sample 1 (n = 240).
COHS routine COHS automaticity Compulsivity Impulsivity Self-regulation

tau p tau p tau p tau p tau p

COHS routine 1 –
COHS automaticity 0.15 0.001 1 –
Compulsivity 0.22 0.001 0.20 0.001 1 –
Impulsivity −0.16 0.001 0.10 0.020 0.03 0.489 1 –
Self-regulation −0.11 0.019 −0.28 0.001 −0.22 0.001 −0.29 0.001 1 –

Note. COHS routine and COHS automaticity = Creature of Habit Scale (Ersche et al., 2017) routine and automaticity subscale scores; Compulsivity = Obsessive-
Compulsive Inventory-Revised (Gönner, Leonhart, & Ecker, 2007) sum score; Impulsivity = Barratt Impulsiveness Scale-11 (Patton et al., 1995) sum score; Self-
regulation = sum score Self-Regulation Scale (Diehl et al., 2006).

5
R. Overmeyer, et al. Personality and Individual Differences 163 (2020) 110029

Fig. 2. Between-subject graphical lasso network with tuning

parameter selected using the extended Bayesian information
criterion. Nodes represent the traits impulsivity, compulsivity,
self-regulatory capacity as well as tendency for routine and
automatic behavior. Edges (connections) can be interpreted as
partial correlation coefficients. Red (dashed) lines represent
negative edges, green (solid) lines positive edges. Self-
regulation = Self-regulation scale (Schwarzer et al., 1999)
sum score, Routine = Creature of Habit Scale (Ersche et al.,
2017) routine subscale sum score, Compulsivity = Obsessive-
Compulsive Inventory (Gönner et al., 2007) sum score, Im-
pulsivity = Barratt Impulsiveness Scale (Preuss et al., 2003)
sum score, Automaticity = Creature of Habit Scale (Ersche
et al., 2017) automaticity subscale sum score. See the online
article for the color version of this figure. (For interpretation
of the references to color in this figure legend, the reader is
referred to the web version of this article.)

were positively associated with compulsivity. Self-regulation was direction. For example, high impulsivity may lead to less effective self-
shown to be the most central node within the network, therefore pro- regulation, leading to less control over automaticity.
viding evidence of influence on the connection between all other Self-regulation comprises three main components: monitoring if
measures. We also examined the psychometric properties of a German there is a need to self-regulate, motivation for self-regulation and self-
version of the COHS in two independent samples. We replicated and regulatory capacity (Baumeister & Heatherton, 1996; Carver & Scheier,
validated the two-factorial structure of the original version in both 1981). The SRS used in this study to examine self-regulation assesses
samples, using EFA and CFA analyses. All items consistently loaded on one specific component of self-regulation: attention control in goal
the expected factor. Both factors were weakly correlated, as expected. pursuit, which is also described as self-regulatory capacity (Diehl et al.,
The subscales were associated differentially with measures of reg- 2006). Habitual routines and automatic behaviors are often uncoupled
ulatory control, providing further evidence that both scales display from attentional processes and do not require conscious control to be
different aspects of habitual behavior and are relevant in behavioral executed (Evans & Stanovich, 2013). Conscious regulation becomes
regulation. Results therefore show that the German version of the necessary, if the behavior is no longer consistent with goals. If self-
COHS, with its distinction between routine behavior and automatic regulation is weak, the behavior may continue regardless (Wood et al.,
behavior, is a reliable, valid, and useful instrument. 2002). Individuals with low self-regulatory capacity may therefore run
the risk of their habits to get out of control. In this regard it comes as
4.1. Habit and behavioral regulation little surprise that it has been suggested that obsessive-compulsive
disorders may be associated with excessive habit formation, and com-
Dynamic interactions between psychological constructs are im- pulsivity described as a transdiagnostic trait, characterized by an im-
portant when considering individual differences and can be con- balance between goal-directed and habit-learning systems (Gillan et al.,
ceptualized within network analyses (Costantini et al., 2019). Our re- 2016; Gillan & Robbins, 2014). This also translates to subclinical
sults suggest that habitual propensity, regarding routine as well as samples (Chamberlain, Stochl, Redden, & Grant, 2018). The connection
automaticity, is intertwined with compulsivity, impulsivity, and self- to impulsivity, however, seems more complex. The COHS subscales
regulation in a network of behavioral regulation, with evidence that were not associated with impulsivity in the same way. COHS routine
self-regulation is the most influential node. Self-regulation controls how was shown to be negatively associated with impulsivity. It seems likely
strongly desire and impulsive action dominate behavior (Hofmann that individuals with a propensity to form routines also tend to have a
et al., 2012). The ability to self-regulate also influences habitual be- consistent lifestyle, plan aspects of daily life very carefully, and do not
havior, with the aim of preventing it to spiral out of control and fuel often act out of the spur of the moment tend to be less impulsive. These
compulsive behavior patterns (Gillan & Robbins, 2014). It appears that tendencies appear to describe the opposite of impulsivity, which is
although habits are regulated by external cues, they require intact self- often characterized as the tendency to act prematurely without fore-
regulation to enable adaptation to changing environments and maintain sight (Dalley, Everitt, & Robbins, 2011; Patton et al., 1995). On the
the balance between efficiency and adaptability. Interestingly, the link other hand, COHS automaticity is only indirectly linked to impulsivity,
between self-regulation and COHS routine was an indirect one, in- via self-regulation. Automaticity describes cue-response associations
dicating that routines are less dependent on self-regulation than auto- that occur without having to focus on the task at hand, allowing the
maticity. However, these connections may also go in the reverse individual to think about other things. It requires conscious self-

6
R. Overmeyer, et al. Personality and Individual Differences 163 (2020) 110029

regulation to override an automatic response, which relies on detection role of self-regulation in these dynamic interactions suggests that in
of a need to regulate and then attentional deployment. Habit auto- pathological conditions with maladaptive habitual behaviors (e.g.
maticity has been associated with attention in the past using other smoking), treatments should target self-regulation. Information on ha-
measures, e.g. the Self-Report Habit Index (SRHI) correlated with at- bitual behavior gathered using the COHS could be useful in identifying
tentional bias to smoking cues in current smokers (Orbell & Verplanken, vulnerable individuals and developing tailored therapeutic strategies
2010). Interesting, with respect to the COHS automaticity subscale, is for certain psychopathologies.
the strong focus on eating-related items, especially in the context of an
indirect connection to impulsivity. Impulsivity has consistently been Financial support
related to various measures associated with overeating or binge eating,
in clinical as well as healthy samples (Claes, Nederkoorn, This work was funded by the Deutsche Forschungsgemeinschaft
Vandereycken, Guerrieri, & Vertommen, 2006; Meule, 2013; Rosval (DFG, German Research Foundation), grant numbers SFB 940 and TRR
et al., 2006). This potential association between COHS automaticity and 265.
overeating or binge eating could result from high levels of impulsivity
interfering with self-regulation. This could be of interest in future stu- Ethical standards
dies examining eating disorders.
The authors assert that all procedures contributing to this work
4.2. Outlook for further research comply with the ethical standards of the relevant national and in-
stitutional committees on human experimentation and with the
As the COHS is a new instrument, its usefulness needs to be further Helsinki Declaration of 1975, as revised in 2008.
evaluated in empirical studies, specifically regarding populations in
which habitual tendencies are deemed dysfunctional and self-regulation CRediT authorship contribution statement
is deficient, e.g. highly compulsive and impulsive populations. Also of
interest would be to examine the influence of extrinsic and not just Rebecca Overmeyer: Conceptualization, Methodology, Data cura-
intrinsic aspects of self-regulation as well as motivational aspects, as tion, Formal analysis, Visualization, Writing - original draft. Sophia
described by the self-determination theory (Deci & Ryan, 2008). Ad- Fürtjes: Conceptualization, Methodology, Data curation, Writing - re-
ditionally, the connection of routine and automaticity to the different view & editing. Karen D. Ersche: Conceptualization, Methodology,
components of self-regulation could be of relevance. For example, self- Writing - review & editing. Stefan Ehrlich: Conceptualization,
regulation critically depends on monitoring for the need to adapt a Supervision, Funding acquisition, Writing - review & editing. Tanja
behavior, also called performance monitoring (Ullsperger, Fischer, Endrass: Conceptualization, Methodology, Supervision, Funding ac-
Nigbur, & Endrass, 2014). Changes in brain activity associated with quisition, Writing - review & editing.
performance monitoring have been reported for both highly compulsive
but also impulsive individuals (e.g. Endrass & Ullsperger, 2014; Martin Declaration of competing interest
& Potts, 2009). To summarize, research connecting self-report to ex-
perimental designs, as well as studies aiming elucidate connections to None.
daily life using ecological momentary assessment is warranted.
Acknowledgements
4.3. Limitations
The authors would like to thank Tyler Bassett, Julia Hartl and Joe
Limitations of the current study include the relatively small sample King for the translation of the questionnaire; and Michael Höfler and
size of the first sample, and the restriction to female gender in the John Venz for helpful discussion on data analysis.
second sample. Also, not all items might be optimal for the subscales.
Some items consistently exhibited low factor loadings (λ ≈ 0.30; re- Appendix A. Supplementary data
levant for items 18 and 24 on the routine subscale, 8 and 21 on the
automaticity subscale), and it could be discussed if they measure as- Supplementary data to this article can be found online at https://
pects not directly related to the latent variables and if they are statis- doi.org/10.1016/j.paid.2020.110029.
tically meaningful (Tabachnick, Fidell, & Ullman, 2007). These items
should be further evaluated in larger and more heterogenic samples. References
Further, as most of the correlations with other measures of personality
were small to moderate in magnitude, results on network dynamics as Aarts, H., Verplanken, B., & Van Knippenberg, A. (1998). Predicting behavior from ac-
well as convergent and discriminant validity should be interpreted with tions in the past: Repeated decision making or a matter of habit? Journal of Applied
caution. Replication of these results in healthy as well as clinical po- Social Psychology, 28(15), 1355–1374.
American Psychiatric Association (2013). Diagnostic and statistical manual of mental
pulations might shed some light on the consistency of the structure of disorders. BMC Medicine, 17, 133–137.
these associations. However, consistent replication of the factor struc- Baumeister, R. F., & Heatherton, T. F. (1996). Self-regulation failure: An overview.
ture in two independent samples should be considered strong evidence Psychological Inquiry, 7(1), 1–15.
Bernaards, C. A., & Jennrich, R. I. (2005). Gradient projection algorithms and software for
for the model and psychometric quality of the COHS. Additionally, our arbitrary rotation criteria in factor analysis. Educational and Psychological
findings regarding the associations with impulsivity and compulsivity Measurement, 65(5), 676–696.
are in line with previous findings, suggesting reliable results (Ersche Borsboom, D., & Cramer, A. O. (2013). Network analysis: An integrative approach to the
structure of psychopathology. Annual Review of Clinical Psychology, 9, 91–121.
et al., 2017, 2019). Carden, L., & Wood, W. (2018). Habit formation and change. Current Opinion in Behavioral
Sciences, 20, 117–122.
5. Conclusions Carver, C. S., & Scheier, M. F. (1981). Attention and self-regulation: A control-theory ap-
proach to human behavior. Springer.
Chamberlain, S. R., Stochl, J., Redden, S. A., & Grant, J. E. (2018). Latent traits of im-
The German version of the COHS appears to be a reliable and valid pulsivity and compulsivity: Toward dimensional psychiatry. Psychological Medicine,
measure of daily behavior related to habitual routines and automatic 48(5), 810–821.
behavior. As the COHS subscales cover different aspects of habit pro- Chen, J., & Chen, Z. (2008). Extended Bayesian information criteria for model selection
with large model spaces. Biometrika, 95(3), 759–771.
pensity, it is not surprising that they appear to differ in their associa- Claes, L., Nederkoorn, C., Vandereycken, W., Guerrieri, R., & Vertommen, H. (2006).
tions with self-regulation, impulsivity and compulsivity. The central Impulsiveness and lack of inhibitory control in eating disorders. Eating Behaviors,

7
R. Overmeyer, et al. Personality and Individual Differences 163 (2020) 110029

7(3), 196–203. Hofmann, W., Schmeichel, B. J., & Baddeley, A. D. (2012). Executive functions and self-
Clark, F. A. (2000). The concepts of habit and routine: A preliminary theoretical synth- regulation. Trends in Cognitive Sciences, 16(3), 174–180.
esis. Occupational Therapy Journal of Research, 20(1_suppl), 123S–137S. Hogarth, L., Chase, H. W., & Baess, K. (2012). Impaired goal-directed behavioural control
Corbit, L. H., Chieng, B. C., & Balleine, B. W. (2014). Effects of repeated cocaine exposure in human impulsivity. The Quarterly Journal of Experimental Psychology, 65(2),
on habit learning and reversal by N-acetylcysteine. Neuropsychopharmacology, 39(8), 305–316.
1893. Holgado–Tello, F. P., Chacón–Moscoso, S., Barbero–García, I., & Vila–Abad, E. (2010).
Costantini, G., Richetin, J., Preti, E., Casini, E., Epskamp, S., & Perugini, M. (2019). Polychoric versus Pearson correlations in exploratory and confirmatory factor ana-
Stability and variability of personality networks. A tutorial on recent developments in lysis of ordinal variables. Quality & Quantity, 44(1), 153.
network psychometrics. Personality and Individual Differences, 136, 68–78. Howell, D. C. (2012). Statistical methods for psychology. Cengage Learning.
Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, Jackson, C. H. (2011). Multi-state models for panel data: The msm package for R. Journal
16(3), 297–334. of Statistical Software, 38(8), 1–29.
Dalley, J. W., Everitt, B. J., & Robbins, T. W. (2011). Impulsivity, compulsivity, and top- Karoly, P. (1993). Mechanisms of self-regulation: A systems view. Annual Review of
down cognitive control. Neuron, 69(4), 680–694. https://doi.org/10.1016/j.neuron. Psychology, 44(1), 23–52.
2011.01.020. Kendall, M. G. (1938). A new measure of rank correlation. Biometrika, 30(1/2), 81–93.
Danner, U. N., Aarts, H., & de Vries, N. K. (2008). Habit vs. intention in the prediction of Lally, P., Van Jaarsveld, C. H., Potts, H. W., & Wardle, J. (2010). How are habits formed:
future behaviour: The role of frequency, context stability and mental accessibility of Modelling habit formation in the real world. European Journal of Social Psychology,
past behaviour. British Journal of Social Psychology, 47(2), 245–265. 40(6), 998–1009.
Dayan, P. (2009). Goal-directed control and its antipodes. Neural Networks, 22(3), Lange, F., & Dewitte, S. (2019). Cognitive flexibility and pro-environmental behaviour: A
213–219. multimethod approach. European Journal of Personality, 33(4), 488–505.
Deci, E. L., & Ryan, R. M. (2008). Self-determination theory: A macrotheory of human Lauritzen, S. L. (1996). Graphical models. Vol. 17. Clarendon Press.
motivation, development, and health. Canadian Psychology, 49(3), 182–185. Lê, S., Josse, J., & Husson, F. (2008). FactoMineR: An R package for multivariate analysis.
Diehl, M., Semegon, A. B., & Schwarzer, R. (2006). Assessing attention control in goal Journal of Statistical Software, 25(1), 1–18.
pursuit: A component of dispositional self-regulation. Journal of Personality LimeSurvey Project Team (2015). LimeSurvey: An open source survey tool. Germany:
Assessment, 86(3), 306–317. LimeSurvey Project Hamburg.
Dolan, R. J., & Dayan, P. (2013). Goals and habits in the brain. Neuron, 80(2), 312–325. Linnebank, F. E., Kindt, M., & de Wit, S. (2018). Investigating the balance between goal-
Endrass, T., & Ullsperger, M. (2014). Specificity of performance monitoring changes in directed and habitual control in experimental and real-life settings. Learning &
obsessive-compulsive disorder. Neuroscience & Biobehavioral Reviews, 46, 124–138. Behavior, 46(3), 306–319.
Epskamp, S., Borsboom, D., & Fried, E. I. (2018). Estimating psychological networks and Luszczynska, A., Diehl, M., Gutiérrez-Dona, B., Kuusinen, P., & Schwarzer, R. (2004).
their accuracy: A tutorial paper. Behavior Research Methods, 50(1), 195–212. Measuring one component of dispositional self-regulation: Attention control in goal
Epskamp, S., Cramer, A. O., Waldorp, L. J., Schmittmann, V. D., & Borsboom, D. (2012). pursuit. Personality and Individual Differences, 37(3), 555–566.
qgraph: Network visualizations of relationships in psychometric data. Journal of Mardia, K. V. (1970). Measures of multivariate skewness and kurtosis with applications.
Statistical Software, 48(4), 1–18. Biometrika, 57(3), 519–530. https://doi.org/10.1093/biomet/57.3.519.
Epskamp, S., & Fried, E. I. (2018). A tutorial on regularized partial correlation networks. Martin, L. E., & Potts, G. F. (2009). Impulsivity in decision-making: An event-related
Psychological Methods, 23(4), 617. potential investigation. Personality and Individual Differences, 46(3), 303–308.
Ersche, K. D., Lim, T. V., Ward, L. H. E., Robbins, T. W., & Stochl, J. (2017). Creature of Martin, A. D., Quinn, K. M., & Park, J. H. (2011). MCMCpack: Markov chain Monte Carlo
Habit: A self-report measure of habitual routines and automatic tendencies in ev- in R.
eryday life. Personality and Individual Differences, 116, 73–85. https://doi.org/10. McDonald, R. P. (2013). Test theory: A unified treatment. Psychology Press.
1016/j.paid.2017.04.024. Meade, A. W., & Craig, S. B. (2012). Identifying careless responses in survey data.
Ersche, K. D., Ward, L. H. E., Lim, T. V., Lumsden, R. J., Sawiak, S. J., Robbins, T. W., & Psychological Methods, 17(3), 437–455. https://doi.org/10.1037/a0028085.
Stochl, J. (2019). Impulsivity and compulsivity are differentially associated with Meule, A. (2013). Impulsivity and overeating: A closer look at the subscales of the Barratt
automaticity and routine on the Creature of Habit Scale. Personality and Individual Impulsiveness Scale. Frontiers in Psychology, 4, 177.
Differences, 150, 109493. Orbell, S., & Verplanken, B. (2010). The automatic component of habit in health behavior:
Evans, J. S. B., & Stanovich, K. E. (2013). Dual-process theories of higher cognition: Habit as cue-contingent automaticity. Health Psychology, 29(4), 374.
Advancing the debate. Perspectives on Psychological Science, 8(3), 223–241. Osborne, J. W., & Fitzpatrick, D. C. (2012). Replication Analysis in Exploratory Factor
Fineberg, N. A., Potenza, M. N., Chamberlain, S. R., Berlin, H. A., Menzies, L., Bechara, A., Analysis: What it is and why it makes your analysis better. Practical Assessment,
... Hollander, E. (2010). Probing compulsive and impulsive behaviors, from animal Research & Evaluation, 17.
models to endophenotypes: A narrative review. Neuropsychopharmacology, 35(3), Ouellette, J. A., & Wood, W. (1998). Habit and intention in everyday life: The multiple
591–604. https://doi.org/10.1038/npp.2009.185. processes by which past behavior predicts future behavior. Psychological Bulletin,
Finger, H., Goeke, C., Diekamp, D., Standvoß, K., & König, P. (2017). LabVanced: A 124(1), 54.
unified JavaScript framework for online studies. Paper presented at the International Pan, J., Ip, E. H., & Dube, L. (2017). An alternative to post hoc model modification in
Conference on Computational Social Science (Cologne). confirmatory factor analysis: The Bayesian lasso. Psychological Methods, 22(4),
Fletcher, T. D. (2010). Psychometric: Applied psychometric theory. 687–704. https://doi.org/10.1037/met0000112.
Foa, E. B., Huppert, J. D., Leiberg, S., Langner, R., Kichic, R., Hajcak, G., & Salkovskis, P. Patton, J. H., Stanford, M. S., & Barratt, E. S. (1995). Factor structure of the Barratt
M. (2002). The Obsessive-Compulsive Inventory: Development and validation of a impulsiveness scale. Journal of Clinical Psychology, 51(6), 768–774.
short version. Psychological Assessment, 14(4), 485. Preuss, U., Rujescu, D., Giegling, I., Koller, G., Bottlender, M., Engel, R., ... Soyka, M.
Foygel, R., & Drton, M. (2010). Extended Bayesian information criteria for Gaussian (2003). Factor structure and validity of a german version of the barratt impulsiveness
graphical models. Paper presented at the advances in neural information processing scale. Fortschritte der Neurologie-Psychiatrie, 71(10), 527–534.
systems. R Core Team (2018). R: A language and environment for statistical computing. 2012. Vienna,
Friedman, J., Hastie, T., & Tibshirani, R. (2008). Sparse inverse covariance estimation Austria: R Foundation for Statistical Computing.
with the graphical lasso. Biostatistics, 9(3), 432–441. Raiche, G. (2010). nFactors: An R package for parallel analysis and non graphical solutions to
Gagne, P., & Hancock, G. R. (2006). Measurement model quality, sample size, and so- the Cattell scree test.
lution propriety in confirmatory factor models. Multivariate Behavioral Research, Revelle, W. (2018). Psych: Procedures for personality and psychological research. Evanston:
41(1), 65–83. Northwestern University.
Gillan, C. M., Papmeyer, M., Morein-Zamir, S., Sahakian, B. J., Fineberg, N. A., Robbins, Revelle, W., & Zinbarg, R. E. (2009). Coefficients alpha, beta, omega, and the glb:
T. W., & de Wit, S. (2011). Disruption in the balance between goal-directed behavior Comments on Sijtsma. psychometrika, 74(1), 145.
and habit learning in obsessive-compulsive disorder. American Journal of Psychiatry, Ripley, B., Venables, B., Bates, D. M., Hornik, K., Gebhardt, A., Firth, D., & Ripley, M. B.
168(7), 718–726. (2013). Package “mass”.
Gillan, C. M., & Robbins, T. W. (2014). Goal-directed learning and obsessive–compulsive Robbins, T., & Costa, R. M. (2017). Habits. Current Biology, 27(22), R1200–R1206.
disorder. Philosophical Transactions of the Royal Society, B: Biological Sciences, Robbins, T. W., Gillan, C. M., Smith, D. G., de Wit, S., & Ersche, K. D. (2012).
369(1655), 20130475. Neurocognitive endophenotypes of impulsivity and compulsivity: Towards dimen-
Gillan, C. M., Robbins, T. W., Sahakian, B. J., van den Heuvel, O. A., & van Wingen, G. sional psychiatry. Trends in Cognitive Sciences, 16(1), 81–91. https://doi.org/10.
(2016). The role of habit in compulsivity. European Neuropsychopharmacology, 26(5), 1016/j.tics.2011.11.009.
828–840. Rosval, L., Steiger, H., Bruce, K., Israël, M., Richardson, J., & Aubut, M. (2006).
Gilpin, A. R. (1993). Table for conversion of Kendall’s Tau to Spearman’s Rho within the Impulsivity in women with eating disorders: Problem of response inhibition, plan-
context of measures of magnitude of effect for meta-analysis. Educational and ning, or attention? International Journal of Eating Disorders, 39(7), 590–593.
Psychological Measurement, 53(1), 87–92. Ruscio, J. (2018). RGenData: Generates multivariate nonnormal data and determines how
Gonner, S., Leonhart, R., & Ecker, W. (2007). The German version of the obsessive- many factors to retain. R package version 1.0. https://CRAN.R-project.org/
compulsive inventory-revised: A brief self-report measure for the multidimensional package=RGenData.
assessment of obsessive-compulsive symptoms. Psychotherapie, Psychosomatik, Ruscio, J., & Roche, B. (2012). Determining the number of factors to retain in an ex-
Medizinische Psychologie, 57(9–10), 395–404. https://doi.org/10.1055/s-2007- ploratory factor analysis using comparison data of known factorial structure.
970894. Psychological Assessment, 24(2), 282.
Gönner, S., Leonhart, R., & Ecker, W. (2007). Das Zwangsinventar OCI-R-die deutsche Schafer, J., Opgen-Rhein, R., Zuber, V., Ahdesmaki, M., Silva, A. P. D., Strimmer, K., &
version des obsessive-compulsive inventory-revised. PPmP- Strimmer, M. K. (2017). Package “corpcor”.
Psychotherapie·Psychosomatik·Medizinische Psychologie, 57(09/10), 395–404. Schwabe, L., & Wolf, O. T. (2011). Stress-induced modulation of instrumental behavior:
Heatherton, T. F. (2011). Neuroscience of self and self-regulation. Annual Review of From goal-directed to habitual control of action. Behavioural Brain Research, 219(2),
Psychology, 62, 363–390. 321–328.

8
R. Overmeyer, et al. Personality and Individual Differences 163 (2020) 110029

Schwarzer, R., Diehl, M., & Schmitz, G. (1999). Self-regulation scale. (Retrieved October, temporal dynamics of performance monitoring. Trends in Cognitive Sciences, 18(5),
28, 2006). 259–267.
Stanford, M. S., Mathias, C. W., Dougherty, D. M., Lake, S. L., Anderson, N. E., & Patton, J. Verplanken, B. (2006). Beyond frequency: Habit as mental construct. British Journal of
H. (2009). Fifty years of the Barratt Impulsiveness Scale: An update and review. Social Psychology, 45(3), 639–656.
Personality and Individual Differences, 47(5), 385–395. Wood, W., Labrecque, J. S., Lin, P.-Y., & Rünger, D. (2014). Habits in dual process models.
Tabachnick, B. G., Fidell, L. S., & Ullman, J. B. (2007). Using multivariate statistics. Vol. 5. Dual process theories of the social mind (pp. 371–385). .
Boston, MA: Pearson. Wood, W., Quinn, J. M., & Kashy, D. A. (2002). Habits in everyday life: Thought, emotion,
Thorndike, E. (1911). Animal intelligence: Experimental studies Macmillan. (New York). and action. Journal of Personality and Social Psychology, 83(6), 1281.
Torregrossa, M. M., Quinn, J. J., & Taylor, J. R. (2008). Impulsivity, compulsivity, and Wood, W., & Rünger, D. (2016). Psychology of habit. Annual Review of Psychology, 67,
habit: The role of orbitofrontal cortex revisited. Biological Psychiatry, 63(3), 253–255. 289–314.
Ullsperger, M., Fischer, A. G., Nigbur, R., & Endrass, T. (2014). Neural mechanisms and