Learning and Individual Differences 59 (2017) 22–33

Contents lists available at ScienceDirect

Learning and Individual Differences

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

Development and initial validation of a multidimensional student MARK

performance scale
Daniel J. Cummings⁎, Arthur E. Poropat, Natalie J. Loxton, Nicola Sheeran
School of Applied Psychology, Griffith University. 176 Messines Ridge Road, Mt Gravatt, Queensland 4122, Australia

A R T I C L E I N F O A B S T R A C T

Keywords: Job performance has been recognised as multidimensional, and can be separated into task performance, orga-
Academic performance nisational citizenship behaviours, and counterproductive work behaviours. These dimensions of job performance
Personality have been applied to student performance in previous research. However, the qualitatively different contexts
Student citizenship behaviour between job and student performance mean that behaviours measuring multidimensional performance in the
University citizenship behaviour
work context may not be applicable to the university context. In two studies we sought to investigate whether
Counterproductive student behaviour
the dimensional structure of job performance could be replicated with student performance, and to develop
preliminary scales to measure multidimensional student performance. Results revealed that the three-factor
structure of job performance could be replicated in student performance, and that previous scales used to
measure the dimensions of student performance suffered issues in relation to construct definition and concept
overlap. Furthermore, we provided initial support for the validity of the multidimensional student performance
scale developed in this study, and correlated the subscales with measures of personality and grade point average.

1. Introduction 2012; Rotundo & Sackett, 2002) (e.g., staying back to do extra work to
help colleagues; Spector, Bauer, & Fox, 2010). Finally, counter-
Research on job performance in organisations has benefitted from productive work behaviours are behaviours that are detrimental to the
expanding the performance domain beyond core task performance into well-being or goals of an organisation or its stakeholders (e.g., bullying
extra-role behaviour (Hough & Oswald, 2008). It is now generally re- another staff member; Ones & Dilchert, 2013). Counterproductive work
cognised that there are three main, broad dimensions of job perfor- behaviours do not need to be intentionally harmful, and they may arise
mance: task performance, organisational citizenship behaviour, and from carelessness or habit (e.g., arriving late to work; Ones & Dilchert,
counterproductive work behaviour (Rotundo & Sackett, 2002; 2013).
Viswesvaran & Ones, 2000). The differentiation of these types of job In a recognition of the importance of academic behaviours beyond
performance has allowed a greater understanding of employee perfor- grade attainment, these broad categories of job performance have
mance, and the investigation of their unique antecedents, including begun to be investigated in tertiary education settings, transforming
personality and individual differences (e.g., Berry, Carpenter, & Barratt, organisational citizenship behaviour and counterproductive work be-
2012; Chiaburu, Oh, Berry, Li, & Gardner, 2011) and organisational haviour to student citizenship behaviour (SCB) and counterproductive
characteristics (e.g., Spector & Fox, 2010). student behaviour (CSB) respectively (e.g., Credé & Niehorster, 2009;
Task performance comprises those behaviours which contribute Meriac, 2012; Schwager et al., 2014; Zettler, 2011). The definition of
towards the technical core of an organisation—the production of a good SCB used in this study reflects other research on SCB (e.g.,Schwager
or provision of a service (e.g., selling widgets; Rotundo & Sackett, et al., 2014; Zettler, 2011) and has been directly adapted from Rotundo
2002). Generally these behaviours are formally recognised as being part and Sackett's (2002) definition of organisational citizenship behaviour:
of the job or role, though due to the changing nature of job roles and the being behaviour which positively contributes to the university's social
difficulty in comparing job descriptions across roles, this is not a de- or psychological environment. Similarly, following common definitions
fining characteristic (Rotundo & Sackett, 2002). Organisational citi- of counterproductive work behaviour (Ones & Dilchert, 2013) and CSB
zenship behaviour is considered to be behaviour that positively con- (Schwager et al., 2014; Zettler, 2011), in this study CSB was defined as
tributes to an organisation's social or psychological environment and is behaviours which are detrimental to the goals or well-being of the
outside of the core job tasks (Fox, Spector, Goh, Bruursema, & Kessler, university or its stakeholders. These behaviours may reflect disrespect,

⁎
Corresponding author.
E-mail address: dan.cummings@griffithuni.edu.au (D.J. Cummings).

http://dx.doi.org/10.1016/j.lindif.2017.08.008
Received 2 March 2017; Received in revised form 21 August 2017; Accepted 25 August 2017
1041-6080/ © 2017 Elsevier Inc. All rights reserved.
D.J. Cummings et al. Learning and Individual Differences 59 (2017) 22–33

disruption, and a disregard for university rules, processes, and norms. and use of unidimensional scales to measure CSB or SCB separately
However, the definition of task performance in organisations is not means that the scales have not been validated against other dimensions
directly transferable to the university domain, as they are qualitatively of student performance, causing a potential lack of clarity in constructs,
different. For example, an intuitive approach would suggest that stu- and concept overlap. For instance, some items measuring SCB (e.g., “I
dent task performance (STP) is approximately equivalent to fulfilling often complete study-related tasks last minute [r]”; Gehring, 2006) and
the formal degree requirements and academic achievement—this is the CSB (e.g., “Turned in work that was of poor quality—lower than your
approach many studies have taken (e.g., Schwager et al., 2014). How- true potential or ability”; Hakstian et al., 2002) appear to both re-
ever, it is important to note the differences between STP and task present poor work standards. While it is possible to develop scales se-
performance in organisations. In a work setting, the employee is paid by parately, and then ensure lack of concept overlap after scale develop-
the employer to perform their assigned duties, which directly con- ment, this has not yet been done for SCB and CSB scales.
tribute to the organisation's technical core—the production of a good or This problem is not restricted to the measurement of student per-
provision of a service. In contrast, in a university setting the student formance, as conceptual and empirical overlap of items has also been
completes the required tasks (e.g., attending class, writing assignments, found in the job performance literature (Spector, 2010). For instance,
and sitting exams) without direct recompense (instead, in many coun- the organisational citizenship behaviour item “Takes undeserved work
tries, the student pays the university for the privilege of studying there), breaks (r)” (Williams & Anderson, 1991) is tapping a similar construct
as they are considered to be related to the student's own self-im- to the counterproductive work behaviour item “Taken a longer break
provement. Furthermore, individual academic performance does not than you were allowed to take” (Bennett & Robinson, 2000). This causes
have a direct impact upon the university's technical core, in contrast to concerns about conceptual definitions of the constructs, and raises
task performance in organisations. Indeed, a large part of a university's questions on the broader construct validity of the scales or concepts
technical core is providing a service (education) to the students. (Fox et al., 2012; Spector, 2010). For example, recent research has
Consequently, student task performance should not “look” the same suggested that OCB-conscientiousness, one of the dimensions of orga-
as organisational task performance, in the same way that a customer's nisational citizenship behaviour originally outlined by Organ (1988),
behaviour would not look like an employee's behaviour. Using the ex- has items/behaviours which overlap with counterproductive work be-
ample of a gym (another self-improvement institution), while a gym haviour (Fox et al., 2012) and has conceptual and empirical overlap
employee's task performance may include behaviours relating to com- with task performance (Poropat & Jones, 2009), which calls into ques-
pleting assigned duties, such as cleaning the equipment, a gym client's tion its' validity as an organisational citizenship behaviour dimension.
task performance behaviours are likely to be completing their self-im- These overlapping items or constructs also have the effect of biasing
posed (or trainer imposed) exercises, not skipping gym days, adhering relationships. For instance, in a meta-analysis, Dalal (2005) found that
to a diet, and putting in effort to beat their personal best. when organisational citizenship behaviour and counterproductive work
As such, in order to incorporate the difference in context between an behaviour scales that contained overlapping items were removed from
organisational and academic setting, and consistent with the focus on the analysis, the correlation between organisational citizenship beha-
performance behaviours, rather than outcomes, in the present study STP viour and counterproductive work behaviour decreased from − 0.32 to
is defined as behaviours which are recognised as being part of the role − 0.19. Even more strikingly, Spector et al. (2010) meta-analytically
or requirements of a student, or that directly contribute to individual examined the effect of overlapping items, rating format (agreement
academic performance. This definition has parallels to items used to versus behavioural frequency), and rating source (self versus super-
measure task performance in organisations (e.g., “performs tasks that visor) on organisational citizenship behaviour and counterproductive
are expected of him/her”, “engages in activities that will directly affect work behaviour relationships. When controlling for these method fac-
his/her performance evaluation”; Williams & Anderson, 1991). tors, correlations between organisational citizenship behaviour and
This difference in context between an organisational and academic counterproductive work behaviour changed from negative, to zero or
setting causes different motivations and behavioural indicators across even positive. Similarly, when using scales specifically designed to avoid
all of the dimensions of student performance. For instance, some or- overlapping items and using a behavioural frequency rating format, Fox
ganisational citizenship behaviour scales contain items reflecting effort et al. (2012) found a positive correlation between organisational citi-
beyond that which is required (e.g., “Brought work home to prepare for zenship behaviour and counterproductive work behaviour.
next day”, Fox et al., 2012). Extra effort in job-related tasks may be As such, the investigation and measurement of multidimensional
considered an organisational citizenship behaviour as it positively student performance currently faces two main concerns. Firstly, there
contributes to a supportive work environment when it is beyond the have been no studies which have investigated whether the job perfor-
core job tasks (e.g., “Consistently takes the initiative to pitch in and do mance dimensions of task performance, organisational citizenship be-
anything that might be necessary to help accomplish team or organi- haviour, and counterproductive work behaviour can be replicated in
zational objectives, even if such actions are not normally part of own the academic domain. Related to this, it is not clear whether behaviours
duties”, Poropat & Jones, 2009), and as such may be considered to classified on a job performance dimension would also fit into the cor-
positively contribute to the organisation's psychological and social responding student performance dimension. Secondly, current scales
functioning. In contrast, a student putting in extra effort into their as- used to measure SCB and CSB have been developed independently of
sessment (or a gym client putting in extra effort to beat their personal each other, and there are concerns about conceptual overlap of items.
best), impacts the student but has negligible direct impact on the wider Therefore, we conducted two studies to address our two aims: first, to
university functioning. As such, the type of behaviours which would fit investigate whether the multidimensional structure of job performance
into each of the performance dimensions is likely to be different across could be replicated in student performance; and second, if multi-
organisational and university settings. dimensional student performance was found, to develop preliminary
Consequently, there are two main issues with the scales currently scales to measure these concepts. Study 1 sought to develop a set of
used to measure CSB and SCB. Firstly, some scales used were originally three brief, unidimensional scales, collectively titled the
developed in the work context and adapted to the university context Multidimensional Student Performance Scale (MSPS), to measure
(e.g., Allison, Voss, & Dryer, 2001; LeBlanc, 2014; Poropat, 2011). As multidimensional student performance. Study 2 sought to confirm the
mentioned previously, behaviours measuring multidimensional per- factor structure of the MSPS, compare it with current scales used to
formance in the work context may not be transferable to the university measure SCB and CSB, and provide evidence for the validity for the new
context. Secondly, previous scales have typically been developed to MSPS with respect to personality and grade point average (GPA).
only measure single dimensions of student performance (e.g., Gehring,
2006; Hakstian, Farrell, & Tweed, 2002). However, the development

23
D.J. Cummings et al. Learning and Individual Differences 59 (2017) 22–33

2. Study 1: Scale development included in a monthly email which is distributed to students at the
authors' University calling for research participation in various projects.
As using items developed for organisational citizenship behaviour Additionally, some students participated in exchange for partial course
and counterproductive work behaviour in a work context are unlikely credit for a psychology course in their second semester of university.
to be applicable in an academic context, Study 1a sought to generate Age ranged from 17 to 70 years (M = 23.07 years, SD = 8.40 years).
items that could be refined in the subsequent studies. Current students On average, participants had completed 1.46 years of tertiary study
were chosen as the most appropriate participants for item generation, (SD = 1.40 years, range = 0.0 to 8.5 years). Three-hundred and sixty-
as they are more likely to have experience and knowledge of a variety one (96.52%) participants were studying at an undergraduate level. The
of student behaviours, as opposed to staff members, who may only survey was anonymous, and all participants gave informed consent. The
observe a select representation of student behaviours. From this list of University Human Research Ethics review board granted research ap-
initial behaviours, Study 1b aimed to further reduce the number of proval, and the study abided by APA ethical guidelines.
items into three brief, unidimensional scales to measure multi-
dimensional student performance. 2.3.2. Materials and procedure
The 218 items from Study 1a were hosted on an online survey
2.1. Study 1a: Method platform. As agreement ratings are likely to reflect attitudes towards
behaviours as opposed to the frequency with which the behaviour oc-
2.1.1. Participants curs (Spector et al., 2010), this study used the same frequency anchors
Forty-one participants (58.5% female) from eight tertiary institu- as the counterproductive work behaviour (Spector et al., 2006) and
tions in Australia and 13 different fields (e.g., psychology, business, organisational citizenship behaviour checklists (Fox et al., 2012). An-
medicine, graphic design), were recruited via snowball convenience chors were 1 = Never, 2 = Once or twice, 3 = Once or twice per month,
sampling. Fifteen (26.6%) were postgraduate students, and 26 (63.4%) 4 = Once or twice per week, 5 = Every day. Participants were asked to
undergraduate students. Participants' ages ranged from 18 to 32 years indicate how often they had committed each behaviour in their current
(M = 23.90 years, SD = 4.08 years). The survey was anonymous, and degree.
all participants gave informed consent. The University Human Research
Ethics review board granted research approval, and the study abided by 2.4. Study 1b: Results and discussion
APA ethical guidelines.
Before conducting factor analysis, the 218 items from Study 1a were
2.1.2. Procedure further reduced by the researchers. Items were considered for elim-
Participants completed an online survey where they were given an ination based on three criteria: conceptual overlap, item variance and
explanation and examples of STP, SCB, and CSB, and told to list as many endorsement rates, and theoretical relationship with the proposed
examples of relevant behaviours as they could. Definitions were constructs. A number of items were identified as specific variations on a
adapted from those used by Spector et al. (2006) and Fox et al. (2012) broader concept (e.g., “Made fun of another student”, “Made fun of a
for their development of the organisational citizenship behaviour and student for struggling with the course content”), or with conceptual
counterproductive work behaviour checklists. The explanation of STP overlap (e.g., “Made fun of another student”, “Bullied another student
was “There are a number of behaviours which students may do which or staff member”, “Socially excluded another student”), and as such,
are important to their academic success (e.g., ‘Have and follow a study only the central item was retained. Following Fox et al. (2012), and as
plan’ or ‘When can't understand the material, ask the tutor or another recommended by Clark and Watson (1995), items identified with low
student for help’). Think about the people you study with or have stu- variance or endorsement rates were considered for removal (e.g., “Stole
died with in the past. Please list as many examples as you can remember from other students”, “Volunteered for community service organised
of these kinds of academic behaviours.” The definition of SCB was through the university”). We used a criteria of variance < 0.50, and
“Sometimes students at university may make extra efforts that go endorsement rate of < 20% as our cutoff for consideration for item
‘above and beyond the call of duty’ (e.g., ‘Volunteered to assist others removal. However, a strict enforcement of these criteria would remove
with study’ or ‘Volunteered to assist with a university event’). Think a number of behaviours which have been used in other student per-
about the people you study with or have studied with in the past. Please formance scales (e.g., Meriac, 2012) and which are likely to be of
list as many examples as you can remember of these kinds of ‘student central theoretical relevance to the proposed constructs (e.g., “Cheated
citizenship behaviours’”. The explanation of CSB was “Sometimes stu- in an exam or test”). As such, we also judged items based on theoretical
dents at university may engage in behaviours which go against the considerations, such as their conceptual relevance to the proposed
goals of the university (e.g., ‘Made fun of another student’ or ‘Turned up constructs, and their perceived generalisability internationally and to
to a class hungover’). Think about the people you study with or have other university cultures. After reducing items based on these criteria, a
studied with in the past. Please list as many examples as you can re- smaller subset of 56 items was chosen for exploratory factor analysis.
member of these kinds of ‘counterproductive student behaviours’”. Some of the behavioural items were highly skewed, and the beha-
vioural frequency format used in this study may be considered non-
2.2. Study 1a: Results and discussion interval. As such, to reduce any potential concerns with parameter es-
timation and factor extraction (Brown, 2006; Li, 2015), we treated in-
In total, 735 behaviours were given (362 for student task, 168 for dicators as ordinal variables, and used robust weighted least squares
SCB, 205 for CSB). Duplicate, non-behavioural (e.g., “Did not believe estimation in MPLUS with geomin rotation for our exploratory factor
they had the ability to learn or change”), and unworkable (e.g., “Ate a analysis, which is robust to violations of multivariate non-normality
steak with a knife and fork in the front row of a lecture theatre”) items and can be used with non-interval indicators (Brown, 2006). Scree plot
were removed. This left 218 items (78 for STP, 47 for SCB, 93 for CSB). inspection revealed potentially three or four factors. The fourth factor
contained no items with a sufficient primary loading, and as such, the
2.3. Study 1b: Method factor analysis was constrained to three factors. Based upon the rotated
factor scores for the three factor solution, items with high cross loadings
2.3.1. Participants (> 0.30) or no primary loadings (< 0.30) were eliminated, and factor
Three hundred and seventy four participants, of which 287 analysis run again without these items. The final 49 item factor struc-
(76.74%) were female and 87 (23.26%) were male, were recruited at a ture is presented in Table 1.
large Australian university via two means. First, the study details were As can be seen in Table 1, the three factors correspond to the three

24
D.J. Cummings et al. Learning and Individual Differences 59 (2017) 22–33

Table 1 consistent with the current studies' conceptualisation of STP, but in-
Factor loadings of items. consistent with the placement of similar items in CSB or SCB in previous
scales (e.g., Gehring, 2006; Hakstian et al., 2002). Correlations between
Item Factor 1 Factor 2 Factor 3
the factors were −0.25 between STP and CSB, 0.28 between STP and
Set aside time specifically for studying 0.82 0.13 −0.07 SCB, and 0.28 between CSB and SCB.
Started studying for exams early (more than 0.78 0.11 0.01
1 week before exam date)
3. Study 2: Confirmatory factor analysis and validity
Kept up to date with lectures and readings 0.77 0.02 −0.05
Revised material throughout the semester 0.75 − 0.02 0.07
Used the text book or readings to support the 0.73 − 0.02 0.02 The exploratory factor analysis in Study 1b found three broad fac-
lecture material tors of multidimensional student performance, which matched with the
Made and followed a study timetable or plan 0.73 0.25 0.00 definitions of STP, SCB, and CSB used in this study. For the remainder
Set goals related to your university work 0.67 0.12 0.13
of this article, in order to differentiate between SCB and CSB measured
Wrote up summaries of study notes or lectures 0.65 0.20 0.07
Left assignments until the last minute − 0.60 0.12 0.22 by other scales, the scales developed from this series of studies will be
Proof-read your assignment 0.59 − 0.07 0.23 referred to as MSPS-STP, MSPS-SCB, and MSPS-CSB, and MSPS when
Aimed to do enough work to just pass − 0.54 0.18 0.13 referring to them collectively.
Replied promptly to university communications 0.53 0.04 0.19
The first aim of Study 2 was to confirm the factorial structure of
Wrote assessment due dates in a calendar, diary, 0.49 − 0.01 0.29
or planner these dimensions in a different sample, and thus use these items as a
Completed practice questions for exam 0.49 − 0.07 0.26 preliminary MSPS. The second aim of Study 2 was to examine the re-
preparation lationship between the MSPS and currently used SCB and CSB scales. At
Did extra research or work because you were 0.48 0.17 0.27 the moment, there appears to be one primary candidate scale used for
interested in the topic
SCB and one for CSB. Gehring's (2006) SCB (Gehring-SCB) scale has
Caught up on any missed material 0.45 − 0.07 0.20
Read feedback on an assignment 0.44 − 0.02 0.20 four-factors (27 items total), similar to the five factors outlined for
Checked emails, notifications, and other methods 0.36 − 0.13 0.18 organisational citizenship behaviour by Organ (1988), with the ex-
of communication ception of courtesy. Gehring-SCB contains the factors of altruism (e.g.,
Cheated in an exam or test 0.02 0.98 0.01
“I help fellow students who have difficulties comprehending the subject
Wrote negative comments about university or − 0.05 0.74 0.00
staff on websites or social media
matter”), conscientiousness (“I often complete study-related tasks last
Attended a lecture or tutorial hungover 0.07 0.74 −0.14 minute [r]”), civic virtue (e.g., “I do not mind acquiring new work
Plagiarised whole or part of an assignment − 0.01 0.77 −0.02 methods [e.g., presentation styles] if that is important for my studies”),
Made fun of another student 0.08 0.75 −0.05 and sportsmanship (e.g., “I attend lectures as often as I can”). Using a
Ignored teaching staff's instructions during class − 0.05 0.75 0.02
14 item Dutch version of Gehring-SCB, Schwager et al. (2014) found
Littered on campus 0.03 0.74 −0.10
Talked over the top of other people in class − 0.01 0.74 0.01 poor model fit for a single factor solution, but good model fit for a four
Lied to get an extension on an assignment − 0.13 0.70 −0.08 factor solution, with a single higher order factor.
Convinced another student to neglect their − 0.14 0.70 0.08 The main scale that is used for the investigation of CSB is one which
university work to do something else
has taken items from the Californian Psychological Inventory which
Slept in a class − 0.17 0.64 0.02
Went to the university bar prior to class 0.01 0.60 −0.01
may be applicable to CSB (Hakstian et al., 2002). Due to items which
Lied to avoid class − 0.19 0.58 0.00 are not unequivocally academic related, and questionable psychometric
Talked negatively about the quality of the 0.20 0.58 −0.01 properties of the scales due to cross-loadings, the original 40-item scale
teaching staff, classes, or university developed by Hakstian et al. (2002) has been refined by both Marcus,
Distracted others in a class − 0.22 0.58 0.21
Lee, and Ashton (2007) (26-items, eight factors; 2007) and Meriac
Used inappropriate language in class − 0.06 0.57 0.12
Made noise in a quiet study area of the library − 0.12 0.56 0.24 (2012) (19 items, two factors; 2012). As Meriac's (2012) version is more
Had an overdue library book 0.22 0.55 −0.18 parsimonious, we used this version as our CSB comparison scale in this
Left a class early − 0.20 0.39 0.12 study (Meriac-CSB). Meriac-CSB's two factors are labelled disengage-
Listened and supported a student who was having 0.00 − 0.08 0.79 ment (e.g., “Turned in work that was of poor quality-lower than your
personal problems
Shared study material or subject notes − 0.05 0.04 0.72
true potential or ability”) and cheating (e.g., “During an exam, brought
Helped another student when they didn't 0.12 − 0.05 0.72 in crib-notes or other aids that were not officially permitted”). The two-
understand or struggled with the material factor solution had acceptable model fit, and significantly better model
Picked up class material or provided class notes 0.01 0.13 0.66 fit than a one-factor solution, though of note, the latent factors were
for a peer when they were absent
only correlated 0.40 (Meriac, 2012).
Approached students who looked like they 0.01 0.00 0.62
needed assistance These aforementioned scales are not without their limitations. For
Provided a classmate with material or stationary 0.04 0.20 0.62 instance, Gehring-SCB contains behaviours relating to individual aca-
because they had forgotten theirs demic performance, such as “I often complete study-related tasks last
Participated in study groups with other students 0.13 0.00 0.56 minute (r)” and “I do not mind acquiring new work methods (e.g.,
Defended a student who was being teased − 0.09 0.27 0.52
Shared your positive university experiences with 0.27 0.09 0.50
presentation styles) if that is important for my studies”. In an academic
others setting, these behaviours are unlikely to have a substantive impact upon
Encouraged quiet students to contribute to 0.17 0.27 0.45 the social or psychological environment of the university. As such, their
discussions categorisation as a SCB must be questioned. Furthermore, these beha-
Proof-read another students' assignment 0.02 0.19 0.42
viours are likely to be considered part of the role or requirements of
Picked up rubbish and placed it in the bin 0.06 0.08 0.33
being a student, and are directly related to individual academic
Factor loadings ≥ 0.30 are indicated in bold. achievement. Therefore, they are likely more appropriately categorised
as STP. For instance, Study 1b of the current article found that a con-
theoretically proposed dimensions of student performance—Factor 1 ceptually similar item (“Left assignments until the last minute”) loaded
being STP, Factor 2 being CSB, and Factor 3 being SCB. Of note, items negatively on STP. Similarly, Meriac-CSB contains items related to low
tapping low personal standards (“Aimed to do enough work to just personal standards or disengagement (e.g., “Turned in work that was of
pass”) and poor time management (“Left assignments until the last poor quality—lower than your true potential or ability”). It is not clear
minute”) had negative loadings on the STP factor, which would be how this behaviour negatively impacts upon the goals or well-being of
the university, and is likely more appropriately considered negative STP

25
D.J. Cummings et al. Learning and Individual Differences 59 (2017) 22–33

(as found in Study 1b of the current article with the item “Aimed to do between openness and MSPS-SCB. However, the relationship between
enough work to just pass”). conscientiousness and SCB is less clear. Though conscientiousness is the
As such, we sought to examine the correlations between items in strongest predictor of organisational citizenship behaviour, definitions
Gehring-SCB and Meriac-CSB with the MSPS to investigate potential and scales of organisational citizenship behaviour contain items either
conceptual crossover in previous scales. The finding that items have a directly related to conscientiousness (and consequently outside of the
higher correlation with a MSPS other than its intended construct would performance domain, e.g., “Is one of my most conscientious em-
indicate concept overlap, and therefore poor construct validity of pre- ployees”, Podsakoff, MacKenzie, Moorman, & Fetter, 1990) or putting
vious scales. In this particular research, an item-level analysis has two in extra work (e.g., “Brought work home to prepare for next day”; Fox
main benefits over conducting a subscale-level analysis. Firstly, an et al., 2012). While putting in extra work may be considered organi-
item-level analysis may reveal specific problematic items hidden by a sational citizenship behaviour in the work context, in the university
subscale-level analysis. Secondly, an item-level analysis is likely to context these behaviours are more likely to be considered STP. As such,
provide a clearer and easier to interpret demonstration of construct the relationship between conscientiousness and MSPS-SCB was unclear,
overlap, as opposed to a subscale-level analysis which by necessity is and no specific predictions were made.
broader, and often more ill-defined, with multiple behavioural types Finally, self-reported GPA was reported by a smaller subset of stu-
falling under a particular subscale. We expected that both Gehring-SCB dents. As GPA can be considered an objective outcome measure of STP,
and Meriac-CSB would contain items with higher correlations with a we expected positive correlations between MSPS-STP and GPA, though
different MSPS than their intended construct (e.g., “Turned in work that no specific predictions were made regarding GPA, MSPS-CSB, and
was of poor quality—lower than your true potential or ability” would MSPS-SCB.
have a higher correlation with MSPS-STP than MSPS-CSB), indicating
construct contamination. 3.1. Study 2: Method
Furthermore, we sought to examine the incremental validity of the
MSPS over existing measures of Gehring-SCB and Meriac-CSB. As the 3.1.1. Participants
MSPS is expected to be more valid due to a reduction in construct Participants were recruited at a large Australian university via the
overlap, and they include an additional scale specifically measuring same method as Study 1b (research participation email and course
STP, we predicted that the MSPS would explain additional variance to credit), one year later. However, in this study, participants were also
Gehring-SCB and Meriac-CSB. able to enter a prize draw to win a $100 gift voucher, and were able to
The last purpose of Study 2 was to validate the MSPS with external receive an automated personality report upon completion of the survey.
criteria, providing evidence for concurrent validity. Due to the large In total, there were 425 participants (76.94% female, 21.65% male,
literature on personality and job performance (including counter- 1.41% other). Age ranged from 15 to 64 years (M = 22.48 years,
productive work behaviour and organisational citizenship behaviour), SD = 7.05 years). The majority of students were studying at an un-
the Big Five appeared to be an appropriate choice. Based on relation- dergraduate level (92.00%), and ranged across 66 different degrees or
ships between conscientiousness and both academic (Poropat, 2009) degree combinations (i.e., dual degrees). On average, participants had
and job performance (Hurtz & Donovan, 2000), we expected con- completed 1.88 semesters of their current degree (SD = 1.71 semesters,
scientiousness to be correlated with MSPS-STP. Though openness is not range 0 to 9 semesters). The survey was anonymous, and all partici-
a significant predictor of job performance (Hurtz & Donovan, 2000), pants gave informed consent. The University Human Research Ethics
after conscientiousness, it is the second strongest personality predictor review board granted research approval, and the study abided by APA
of academic performance, likely due to the relationship between ethical guidelines.
openness and learning motivation (Poropat, 2009). As such, we ex-
pected positive correlations between openness and MSPS-STP. 3.1.2. Materials
Conscientiousness, agreeableness, and neuroticism have been found 3.1.2.1. Multidimensional student performance scale. The 49 final items
to be correlated with counterproductive work behaviour in the job in the factor analysis of Study 1b were used as a preliminary MSPS (18
performance domain (Berry et al., 2012). We had similar justifications items for MSPS-STP, 19 items for MSPS-CSB, 12 items for MSPS-SCB).
for their relationship with MSPS-CSB. As conscientiousness is related to Participants were asked to indicate how often they committed each
following rules and social decorum (Roberts, Lejuez, Krueger, behaviour in their current degree, ranging from 1 = Never to 5 = Every
Richards, & Hill, 2014), and CSB is related to rule violation and conflict, day. Consistent with previous research on SCB and CSB (e.g., Zettler,
we expected conscientiousness to be negatively correlated with MSPS- 2011), and due to the limitations with expecting accurate teacher
CSB. As people low in agreeableness may be manipulative, incon- ratings of frequencies of out-of-class student behaviours, MSPS was self-
siderate, and cheat to get ahead (Mount, Ilies, & Johnson, 2006), we rated.
expected a negative correlation between agreeableness and MSPS-CSB.
As people scoring high on neuroticism are more likely to experience 3.1.2.2. Gehring-student citizenship behaviour. We used Zettler's (2011)
negative emotions, which may cause them to react with counter- version of Gehring's (2006) SCB scale as the comparison scale for SCB.
productive work behaviour (Spector, 2010), we expected neuroticism The scale was translated to English from German by Ingo Zettler and his
to be positively correlated with MSPS-CSB. colleagues, with some input from the first author of this study. Zettler's
Correlations between personality and organisational citizenship (2011) version of Gehring's (2006) SCB scale contains four factors:
behaviour appear to be of a smaller magnitude than those of counter- altruism (8 items), conscientiousness (4 items), sportsmanship (8
productive work behaviour (Berry et al., 2012; Chiaburu et al., 2011). items), and civic virtue (7 items). Following other studies which have
Despite this, conscientiousness, agreeableness, and openness appear to used Gehring's (2006) scale (e.g., Schwager et al., 2014; Zettler, 2011)
be the best predictors of organisational citizenship behaviour (Chiaburu and scale guidelines, an agreement rating format was used, ranging
et al., 2011). People scoring high on agreeableness are more likely to be from 1 = Strongly disagree to 6 = Strongly agree. Higher scores indicate
helpful and considerate towards others, and as such, we expected higher SCB.
agreeableness to be positively correlated with MSPS-SCB. The correla-
tion between openness and organisational citizenship behaviour has 3.1.2.3. Meriac-counterproductive student behaviour. Meriac's (2012)
been explained as possibly due to adaptability, sharing knowledge, and version of Hakstian et al.’s (2002) inventory of CSB was used, as it is
being aware of their social environment (Chiaburu et al., 2011). These the most parsimonious version. Meriac's (2012) version contained an
links between openness and organisational citizenship behaviour are item (“Did slow or sloppy work”) which the current authors considered
equally as valid for SCB, and so we expected a positive correlation not unequivocally related to the academic domain, so it was not

26
D.J. Cummings et al. Learning and Individual Differences 59 (2017) 22–33

included. The resulting scale contained 7 items in the disengagement Table 2

factor, and 11 items in the cheating factor. Rating format followed that Factor loadings in confirmatory factor analysis.
in the original article (Hakstian et al., 2002), and ranged from
Variable STP CSB SCB
1 = Never even considered it to 6 = Did it three or more times.
Participants were asked to indicate how often they committed each Student task performance scale
behaviour in their current degree. Higher scores indicate higher CSB. Revised material throughout the semester 0.76
Made and followed a study timetable or plan 0.69
Set aside time specifically for studying 0.68
3.1.2.4. Personality. Personality was measured using Johnson's (2014) Set goals related to your university work 0.66
120 item version of the IPIP NEO (Goldberg et al., 2006), which Started studying for exams early (more than 1 week before 0.63
measures extraversion, openness, agreeableness, conscientiousness, and exam date)
neuroticism by 24 items each. Participants rated how much each item Kept up to date with lectures and readings 0.62
Used the text book or readings to support the lecture 0.62
was accurate for them on a scale of 1 = Very inaccurate to 5 = Very
material
accurate. Higher scores, indicate greater levels of the trait. Wrote up summaries of study notes or lectures 0.61
Wrote assessment due dates in a calendar, diary, or planner 0.54
3.1.2.5. Grade point average. Participants were instructed to provide Did extra research or work because you were interested in 0.54
the topic
their current GPA for their current degree (GPA ranges from 0 to 7, with
Completed practice questions for exam preparation 0.52
4 being a passing grade). Two-hundred and eighty-nine (68%) Caught up on any missed material 0.52
participants provided their GPA. Read feedback on an assignment 0.51
Proof-read your assignment 0.51
3.1.3. Procedure Replied promptly to university communications 0.45
Counterproductive student performance scale
The survey was hosted on an online survey platform. University
Plagiarised whole or part of an assignment 0.81
ethics was granted for the study. Cheated in an exam or test 0.72
Lied to get an extension on an assignment 0.71
3.2. Study 2: Results Wrote negative comments about university or staff on 0.71
websites or social media
Distracted others in a class 0.69
3.2.1. Factor structure and scale characteristics of the multidimensional Talked over the top of other people in class 0.69
student performance scale Ignored teaching staff's instructions during class 0.69
Confirmatory factor analysis was conducted in MPLUS using robust Littered on campus 0.69
weighted least squares estimation with indicators treated as ordered Convinced another student to neglect their university work 0.69
to do something else
nominal variables. Other analyses were conducted in IBM SPSS 22. For
Used inappropriate language in class 0.64
the confirmatory factor models, the latent factors were allowed to Lied to avoid class 0.61
covary. As the χ2 statistic has been found to be overly stringent (Brown, Went to the university bar prior to class 0.59
2006), we used a combination of fit indices to judge model fit Had an overdue library book 0.56
Made noise in a quiet study area of the library 0.56
(Hu & Bentler, 1999): the root mean squared error of approximation
Slept in a class 0.51
(RMSEA), comparative fit index (CFI), and the Tucker Lewis index Made fun of another student 0.51
(TLI). Note, these are the only fit indices provided by MPLUS for robust Attended a lecture or tutorial hungover 0.51
weighted least squares estimation with covariates. As with other studies Talked negatively about the quality of the teaching staff, 0.50
and common procedures (e.g., Brown, 2006; Krupić, Corr, Ručević, classes, or university
Left a class early 0.45
Križanić, & Gračanin, 2016), we judged good model fit to be
Student citizenship performance
RMSEA < 0.05, moderate model fit to be 0.05–0.08, marginal fit to be Listened and supported a student who was having personal 0.74
0.08–0.10, and > 0.10 to be poor fit. For CFI and TFI we judged good problems
model fit as > 0.95, acceptable model fit as > 0.90, and poor model Picked up class material or provided class notes for a peer 0.74
fit < 0.90. when they were absent
Shared study material or subject notes 0.73
The initial model (using the same items as Study 2, with single la- Helped another student when they didn't understand or 0.69
tent factor loadings) revealed poor fit based on CFI and TLI (χ2(1124) struggled with the material
= 2373.73, p < 0.001, RMSEA = 0.51, CFI = 0.87, TLI = 0.87). Provided a classmate with material or stationary because 0.69
Though the items “Left assignments until the last minute” and “Aimed they had forgotten theirs
Approached students who looked like they needed 0.68
to do enough work to just pass” had significant loadings on STP,
assistance
modification indices suggested cross loadings on CSB and SCB. As these Defended a student who was being teased 0.67
items appeared to be problematic, they were dropped from subsequent Participated in study groups with other students 0.63
analysis. After this, the model had acceptable fit (χ2(1031) = 1871.56, Proof-read another students' assignment 0.62
p < 0.001, RMSEA = 0.04, CFI 0.91, TLI 0.91). However, the item Encouraged quiet students to contribute to discussions 0.58
Shared your positive university experiences with others 0.55
“Checked emails, notifications, and other methods of communication” Picked up rubbish and placed it in the bin 0.51
had a relatively low r2 (0.10), and a comparatively low estimate/
standard error (6.30) so was removed from analysis. Final model fit Note. STP = Student task performance. CSB = Counterproductive student behaviour.
indices indicated acceptable fit (χ2(986) = 1746.58, p ≤ 0.001, SCB = Student citizenship behaviour.
RMSEA = 0.04, CFI 0.92, TLI 0.92). Factor loadings are presented in
Table 2. items, we correlated the MSPS with the items in Meriac-CSB (Table 4)
The average inter-item correlation for MSPS-STP was 0.31, for and Gehring-SCB (Table 5). For Meriac-CSB, the disengagement sub-
MSPS-CSB it was 0.28, and for MSPS-SCB it was 0.36. As can be seen in scale revealed two items with stronger correlations with MSPS-STP than
Table 3 (which provides scale descriptives and a correlation matrix of with MSPS-CSB. As can be seen from Table 5, all subscales in Gehring-
the study variables), the MSPS had relatively low correlations with each SCB contained items with higher correlations with MSPS-CSB or MSPS-
other, suggesting that the three constructs are distinct. STP than MSPS-SCB. This is particularly the case with the facets of
conscientiousness and sportsmanship, with all items exhibiting poten-
3.2.2. Comparison with existing student performance scales tial construct contamination.
To examine the construct validity of Meriac-CSB and Gehring-SCB In order to examine incremental validity of the MSPS in predicting

27
 D.J. Cummings et al.

Table 3
Descriptors and correlations of study variables.

Variable α M SD 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

1. MSPS-SCB 0.88 30.00 8.47

2. MSPS-STP 0.87 50.50 10.01 0.43⁎⁎⁎
3. MSPS-CSB 0.87 28.66 7.97 0.35⁎⁎⁎ − 0.06
4. Gehring-SCB 0.83 109.93 14.38 0.50⁎⁎⁎ 0.54⁎⁎⁎ −0.20⁎⁎⁎
5. Gehring-SCBVir 0.76 22.29 5.82 0.50⁎⁎⁎ 0.29⁎⁎⁎ 0.08 0.69⁎⁎⁎
6. Gehring-SCBAlt 0.71 34.71 5.50 0.41⁎⁎⁎ 0.25⁎⁎⁎ −0.14⁎⁎ 0.73⁎⁎⁎ 0.28⁎⁎⁎
7. Gehring-SCBCon 0.52 16.80 3.17 0.28⁎⁎⁎ 0.66⁎⁎⁎ −0.22⁎⁎⁎ 0.67⁎⁎⁎ 0.34⁎⁎⁎ 0.28⁎⁎⁎
8. Gehring-SCBSpo 0.67 36.12 5.48 0.21⁎⁎⁎ 0.47⁎⁎⁎ −0.36⁎⁎⁎ 0.78⁎⁎⁎ 0.26⁎⁎⁎ 0.45⁎⁎⁎ 0.54⁎⁎⁎
9. Meriac-CSB 0.85 36.22 11.82 0.08 − 0.26⁎⁎⁎ 0.64⁎⁎⁎ − 0.31⁎⁎⁎ −0.08 − 0.14⁎⁎ − 0.37⁎⁎⁎ − 0.37⁎⁎⁎
10. Meriac-CSBDis 0.71 18.96 6.67 − 0.04 − 0.30⁎⁎⁎ 0.41⁎⁎⁎ − 0.27⁎⁎⁎ −0.16⁎⁎⁎ − 0.04 − 0.38⁎⁎⁎ − 0.28⁎⁎⁎ 0.84⁎⁎⁎

28
11. Meriac-CSBChe 0.87 17.26 7.18 0.17⁎⁎⁎ − 0.15⁎⁎ 0.67⁎⁎⁎ − 0.25⁎⁎⁎ 0.03 − 0.20⁎⁎⁎ − 0.25⁎⁎⁎ − 0.35⁎⁎⁎ 0.87⁎⁎⁎ 0.46⁎⁎⁎
12. Extraversion 0.89 79.72 13.60 0.11⁎ 0.18⁎⁎⁎ 0.01 0.24⁎⁎⁎ 0.18⁎⁎⁎ 0.15⁎⁎ 0.20⁎⁎⁎ 0.19⁎⁎⁎ − 0.05 −0.06 − 0.03
13. Agreeableness 0.84 85.96 11.74 0.08 0.16⁎⁎⁎ −0.33⁎⁎⁎ 0.34⁎⁎⁎ 0.05 0.37⁎⁎⁎ 0.24⁎⁎⁎ 0.32⁎⁎⁎ − 0.25⁎⁎⁎ −0.13⁎⁎ − 0.30⁎⁎⁎ 0.06
14. Openness 0.81 79.42 11.32 0.14⁎⁎ 0.24⁎⁎⁎ −0.14⁎⁎ 0.33⁎⁎⁎ 0.10⁎ 0.38⁎⁎⁎ 0.14⁎⁎ 0.29⁎⁎⁎ − 0.12⁎ −0.02 − 0.17⁎⁎⁎ 0.13⁎⁎ 0.28⁎⁎⁎
15. Conscientiousness 0.88 82.57 12.61 0.04 0.48⁎⁎⁎ −0.33⁎⁎⁎ 0.43⁎⁎⁎ 0.12⁎ 0.19⁎⁎⁎ 0.60⁎⁎⁎ 0.46⁎⁎⁎ − 0.41⁎⁎⁎ −0.38⁎⁎⁎ − 0.33⁎⁎⁎ 0.26⁎⁎⁎ 0.30⁎⁎⁎ 0.11⁎
16. Neuroticism 0.90 71.24 14.86 − 0.05 − 0.13⁎⁎ 0.11⁎ − 0.21⁎⁎⁎ −0.12⁎ − 0.07 − 0.23⁎⁎⁎ − 0.21⁎⁎⁎ 0.19⁎⁎⁎ 0.24⁎⁎⁎ 0.10⁎ − 0.42⁎⁎⁎ − 0.03 − 0.04 − 0.50⁎⁎⁎
17. GPAa 5.27 0.98 − 0.10 0.27⁎⁎⁎ −0.28⁎⁎⁎ 0.25⁎⁎⁎ −0.09 0.21⁎⁎⁎ 0.31⁎⁎⁎ 0.35⁎⁎⁎ − 0.27⁎⁎⁎ −0.21⁎⁎⁎ − 0.24⁎⁎⁎ − 0.04 0.18⁎⁎ 0.10 0.42⁎⁎⁎ − 0.20⁎⁎⁎

Note. MSPS = Multidimensional Student Performance Scale. SCB = Student citizenship behaviour. STP = Student task performance. CSB = Counterproductive student behaviour. Vir = Virtue. Alt = Altruism. Con = Conscientiousness.
Spo = Sportsmanship. Dis = Disengagement. Che = Cheating. GPA = Grade point average.
a
n = 289.
⁎
p < 0.05.
⁎⁎
p < 0.01.
⁎⁎⁎
p < 0.001.
Learning and Individual Differences 59 (2017) 22–33
D.J. Cummings et al. Learning and Individual Differences 59 (2017) 22–33

Table 4
Correlations between items in the Meriac-CSB scale and the Multidimensional Student Performance scale.

MSPS- MSPS- MSPS-

STP SCB CSB

Cheating
For a term paper or essay, copied a portion of text (e.g., sentence or paragraph) directly from a book or article without citing the reference. − 0.13⁎⁎ 0.09 0.48⁎⁎⁎
Handed in an assignment and/or project that contained passages (e.g., sentences or paragraphs) that had been copied from someone else. − 0.13⁎⁎ 0.12⁎ 0.50⁎⁎⁎
During an exam, brought in crib-notes or other aids that were not officially permitted. − 0.15⁎⁎ 0.01 0.53⁎⁎⁎
During an exam, quickly looked at, and got information from, a classmate's paper. − 0.08 0.13⁎⁎ 0.47⁎⁎⁎
During an exam, briefly glanced at another person's paper. − 0.13⁎⁎ 0.08 0.37⁎⁎⁎
Knowingly helped a classmate during an examination by allowing him/her to see your exam paper. − 0.06 0.16⁎⁎⁎ 0.50⁎⁎⁎
Submitted a class paper or project that was not your own work. − 0.11⁎ 0.13⁎⁎ 0.56⁎⁎⁎
Received help from others on homework assignments (e.g., essays or projects) that were supposed to have been done independently. − 0.09 0.18⁎⁎⁎ 0.38⁎⁎⁎
Before a ‘make-up’ exam, asked the other students in the class who had already written the exam for information on the content of the exam. − 0.06 0.12⁎ 0.39⁎⁎⁎
Examined a copy of an upcoming exam which was taken without permission from the professor. − 0.16⁎⁎ 0.12⁎ 0.52⁎⁎⁎
During a ‘closed-book’ exam, arranged your books (or notes) on the floor in such a way that you were able to discreetly glance at them. − 0.11⁎ 0.10⁎ 0.52⁎⁎⁎
Disengagement
Failed to return library books on time. − 0.05 0.03 0.27⁎⁎⁎
Slept through your early morning class. − 0.17⁎⁎⁎ − 0.02 0.20⁎⁎⁎
Showed up for class late. − 0.12⁎ − 0.04 0.24⁎⁎⁎
Turned in work that was of poor quality-lower than your true potential or ability. − 0.24⁎⁎⁎ − 0.05 0.15⁎⁎
Came to an exam poorly prepared because of lack of effort and study. − 0.35⁎⁎⁎ − 0.13⁎⁎ 0.20⁎⁎⁎
Missed afternoon classes because you took an extra long lunch. − 0.18⁎⁎⁎ 0.02 0.41⁎⁎⁎
Did not do your share of the work in a cooperative group project. − 0.17⁎⁎⁎ 0.05 0.35⁎⁎⁎

Note. MSPS = Multidimensional Student Performance Scale. SCB = Student citizenship behaviour. STP = Student task performance. CSB = Counterproductive student behaviour.
Bolded items have a higher correlation with a different multidimensional student performance scale than their expected construct.
⁎
p < 0.05.
⁎⁎
p < 0.01.
⁎⁎⁎
p < 0.001.

Table 5
Correlations between items in the Gehring-SCB scale and the Multidimensional Student Performance scale.

MSPS-STP MSPS-SCB MSPS-CSB

Conscientiousness
I organize upcoming learning sessions in advance 0.49⁎⁎⁎ 0.35⁎⁎⁎ 0.09
I try to stay on schedule when working on a group project as best as I can. 0.38⁎⁎⁎ 0.08 − 0.29⁎⁎⁎
I always show up on time to meetings when working with other students on a group project 0.36⁎⁎⁎ 0.14⁎⁎ − 0.32⁎⁎⁎
I often complete study-related tasks last minute (r). 0.45⁎⁎⁎ 0.11⁎ − 0.15⁎⁎
Sportsmanship
I attend lectures as often as I can. 0.42⁎⁎⁎ 0.19⁎⁎⁎ − 0.24⁎⁎⁎
I select lectures/courses primarily based on interest, not effort. 0.08 0.08 − 0.13⁎⁎
I attend additional lectures if they supplement my studies in a meaningful way. 0.37⁎⁎⁎ 0.24⁎⁎⁎ − 0.12⁎
I accept unfavourable conditions (e.g., time or place of events), if an event/course is important to me. 0.16⁎⁎⁎ 0.11⁎ − 0.23⁎⁎⁎
I attend lectures irregularly if attendance is not mandatory (r). 0.30⁎⁎⁎ 0.10⁎ − 0.17⁎⁎⁎
When I am unhappy with my studies, I often talk about transferring to another university (r). 0.16⁎⁎ − 0.05 − 0.26⁎⁎⁎
If I want to attend a specific course, I can accept a waiting period. 0.22⁎⁎⁎ 0.20⁎⁎⁎ − 0.11⁎
I am often absent from study-related meetings (e.g., study group) without any obvious reason (r). 0.30⁎⁎⁎ 0.05 − 0.37⁎⁎⁎
Altruism
I am seldom willing to help fellow students who struggle with their studies (r). 0.01 − 0.02 − 0.09
I help fellow students who have difficulties comprehending the subject matter. 0.27⁎⁎⁎ 0.42⁎⁎⁎ − 0.07
I am seldom willing to help out fellow students with my own knowledge (e.g., language or computer skills) (r). −0.02 0.01 − 0.10⁎
I am always willing to partner up with fellow students if they cannot overcome the workload on their own. 0.13⁎⁎ 0.40⁎⁎⁎ 0.01
In lecture rooms, I let disabled students go in first. 0.26⁎⁎⁎ 0.08 − 0.29⁎⁎⁎
I support fellow students when they experience personal problems. 0.21⁎⁎⁎ 0.40⁎⁎⁎ − 0.05
I am always willing to complete minor tasks for my fellow students. 0.17⁎⁎⁎ 0.34⁎⁎⁎ 0.04
I encourage fellow students when they are downhearted. 0.27⁎⁎⁎ 0.40⁎⁎⁎ − 0.12⁎
Civic virtue
I get actively involved in the general student welfare (e.g., in demonstrations against high tuition fees). 0.09 0.40⁎⁎⁎ 0.19⁎⁎⁎
I take part in panel discussions (e.g., feedback talks) so as to support the exchange between students, as well as between students and 0.16⁎⁎⁎ 0.32⁎⁎⁎ 0.09
faculty.
I do not mind acquiring new work methods (e.g., presentation styles) if that is important for my studies. 0.32⁎⁎⁎ 0.19⁎⁎⁎ − 0.25⁎⁎⁎
I actively participate in the design and planning of study facilities at university 0.10⁎ 0.32⁎⁎⁎ 0.20⁎⁎⁎
I keep myself informed about new developments in policy on higher education. 0.19⁎⁎⁎ 0.26⁎⁎⁎ 0.02
I take advantage of additional courses offered by the university (e.g., foreign language courses). 0.15⁎⁎ 0.30⁎⁎⁎ 0.08
I point out interesting courses to others. 0.32⁎⁎⁎ 0.44⁎⁎⁎ − 0.02

Note. MSPS = Multidimensional Student Performance Scale. SCB = Student citizenship behaviour. STP = Student task performance. CSB = Counterproductive student behaviour.
Bolded items have a higher correlation with a different multidimensional student performance scale than their expected construct. Bold and italicised items have weak correlations
(< 0.20) with all multidimensional student performance scales.
⁎
p < 0.05.
⁎⁎
p < 0.01.
⁎⁎⁎
p < 0.001.

29
D.J. Cummings et al. Learning and Individual Differences 59 (2017) 22–33

Table 6 significant.
Hierarchical regression for Gehring-SCB, Meriac-CSB, and multidimensional student
performance subscales in predicting GPA.
3.2.3. Relationships with personality and GPA
Variable B SE B β ΔR2 Bivariate correlations with personality and GPA were used to assess
the concurrent validity of the MSPS, which can be seen in Table 3.
Step 1 0.14⁎⁎⁎ MSPS-CSB had negative correlations with conscientiousness, agree-
Semesters completed −0.063 0.032 − 0.114⁎
ableness, and openness, and a positive correlation with neuroticism.
Age 0.027 0.009 0.176⁎⁎
Gender −0.176 0.135 − 0.074 MSPS-STP had positive correlations with conscientiousness, agree-
Meriac-CSB −0.013 0.005 − 0.155⁎ ableness, extraversion, and openness, and a negative correlation with
Gehring-SCB 0.011 0.004 0.162⁎⁎ neuroticism. MSPS-SCB was positively correlated with openness and
Step 2 0.06⁎⁎⁎ extraversion, but was not significantly correlated with agreeableness.
Semesters completed −0.058 0.031 − 0.106
Bivariate correlations suggested that MSPS-STP was positively, and
Age 0.018 0.009 0.121⁎
Gender −0.100 0.132 − 0.042 MSPS-CSB was negatively correlated with GPA.
Meriac-CSB 0.000 0.007 0.001
Gehring-SCB 0.013 0.005 0.183⁎ 3.3. Study 2: Discussion
MSPS-STP 0.022 0.007 0.225⁎⁎
MSPS-SCB −0.026 0.009 − 0.221⁎⁎
MSPS-CSB −0.014 0.010 − 0.118 The results of Study 2 confirmed the three factor structure of the
MSPS, while the low correlations between the scales indicated that they
Note. MSPS = Multidimensional Student Performance Scale. SCB = Student citizenship were measuring distinct constructs. This suggests that the MSPS does
behaviour. STP = Student task performance. CSB = Counterproductive student beha- measure multidimensional student performance.
viour. Previous scales measuring SCB and CSB may suffer from issues re-
⁎
p < 0.05.
⁎⁎
p < 0.01.
lating to construct validity, as an item level analysis revealed a number
⁎⁎⁎
p < 0.001. of items in both scales with higher correlations with a multidimensional
student performance dimension other than their purported factor.
GPA, the Gehring-SCB, Meriac-CSB, along with age, gender, and se- Gehring-SCB had a negative correlation with both MSPS-CSB and
mesters completed of the current degree were entered as the first step in Meriac-CSB. In contrast, MSPS-SCB (which was developed concurrently
a hierarchical regression, followed by the MSPS as a second step. As can with MSPS-CSB so as to eliminate overlapping concepts) has a positive
be seen in Table 6, the results indicated that overall, the MSPS showed correlation with MSPS-CSB and the (less problematic) cheating factor of
incremental validity over existing student performance scales. MSPS- Meriac-CSB. This result is similar to that found in the job performance
STP was a significant positive predictor of GPA while controlling for literature. Dalal (2005), Spector et al. (2010), and Fox et al. (2012)
other student performance dimensions and control demographic vari- found that after controlling for overlapping items, response format, and
ables. MSPS-SCB was a significant negative predictor, while Gehring- rating source, the commonly observed negative correlation between
SCB was a significant positive predictor. Neither MSPS-CSB nor Meriac- organisational citizenship behaviour and counterproductive work be-
CSB was a significant predictor in the regression model. haviour was reduced or even reversed, with Fox et al. (2012) suggesting
To further understand the non-significant results for MSPS-CSB and that there may be common organisational antecedents to the two
Meriac-CSB we conducted an additional stepwise regression which is constructs. This result appears to be replicated with the MSPS.
presented in Table 7. The results suggest that Meriac-CSB becomes non- As expected, MSPS-STP was positively correlated with con-
significant with the inclusion of MSPS-CSB, though MSPS-CSB remains scientiousness and openness, consistent with research on academic
performance (Poropat, 2009). However, MSPS-STP was also positively
correlated with extraversion and agreeableness, and negatively corre-
Table 7
Hierarchical regression for Meriac-CSB and multidimensional student performance sub-
lated with neuroticism. The reason behind this may be due to academic
scales in predicting GPA. motivation. Extraversion is positively correlated with academic en-
gagement (Komarraju & Karau, 2005), while agreeableness is nega-
Variable B SE B β ΔR2 tively correlated with amotivation (Komarraju, Karau, & Schmeck,
Step 1 0.112⁎⁎⁎
2009). This suggests that though people scoring high on these traits
Semesters completed − 0.065 0.032 − 0.117⁎ may not always achieve good results, they are likely to be engaged and
Age 0.029 0.009 0.192⁎⁎ motivated enough to exhibit task performance behaviours. In contrast,
Gender − 0.211 0.136 − 0.089 neuroticism is positively correlated with academic disengagement
Meriac-CSB − 0.017 0.005 − 0.204⁎⁎⁎
(Komarraju & Karau, 2005), suggesting that people high in neuroticism
Step 2 0.018⁎
Semesters completed − 0.068 0.032 − 0.123⁎ are less likely to complete academic related tasks.
Age 0.028 0.009 0.186⁎⁎ We found negative correlations between MSPS-CSB and agreeable-
Gender − 0.200 0.135 − 0.084 ness and conscientiousness, and positive correlations between neuroti-
Meriac-CSB − 0.008 0.006 − 0.090 cism and MSPS-CSB, as expected. The unexpected negative correlation
MSPS-CSB − 0.022 0.009 − 0.177⁎
between openness and MSPS-CSB may be due to learning orientations.
Step 3 0.051⁎⁎⁎
Semesters completed − 0.063 0.031 − 0.115⁎ As people high on openness are more likely to be intrinsically motivated
Age 0.021 0.009 0.140⁎ to know (Komarraju et al., 2009), they may be less likely to engage in
Gender − 0.112 0.133 − 0.047 behaviours such as cheating, as it would be counterproductive for their
Meriac-CSB − 0.001 0.007 − 0.010
learning.
MSPS-CSB − 0.021 0.010 − 0.175⁎
MSPS-STP 0.027 0.006 0.273⁎⁎⁎ There was a positive correlation between openness and MSPS-SCB,
MSPS-SCB − 0.016 0.008 − 0.134⁎ as expected. There was also a small but significant correlation between
extraversion and MSPS-SCB. One explanation for this may be that
Note. MSPS = Multidimensional Student Performance Scale. SCB = Student citizenship people high in extraversion are more likely to be engaged in university
behaviour. STP = Student task performance. CSB = Counterproductive student beha-
(Komarraju & Karau, 2005), and this may be even greater for the social
viour.
⁎
p < 0.05.
aspects of university. As such, a greater engagement with university
⁎⁎
p < 0.01. would lead to more behaviours at and relating to university, including a
⁎⁎⁎
p < 0.001. greater opportunity for SCB.

30
D.J. Cummings et al. Learning and Individual Differences 59 (2017) 22–33

Agreeableness was not significantly correlated with MSPS-SCB, controlling for STP, SCB detract from a student's own individual per-
contrary to predictions. In contrast, agreeableness was significantly formance. Alternatively, it is possible that those who engage in SCB are
correlated with Gehring-SCB. There are two possible methodological already struggling with the content, thus directing their focus outwards
explanations for this pattern of results. Firstly, Gehring-SCB contained instead of towards their studies. Future research would need to be
items with high negative correlations with MSPS-CSB. Due to the conducted to uncover the reason for this relationship. Overall, the
aforementioned relationship between MSPS-CSB and agreeableness, it MSPS increased the variance explained for GPA over existing measures
may be that these overlapping items are inflating the correlation be- of SCB and CSB, and control demographic variables, providing evidence
tween Gehring-SCB and agreeableness. Another explanation is due to for the incremental validity of the MSPS in predicting GPA.
the rating format. Gehring-SCB uses an agreement rating format, while The fact that GPA was self-reported is a limitation of this study, as
MSPS-SCB uses a behavioural frequency rating format. In studies on there is no assurance of honesty in reporting. However, collecting data
agreeableness and organisational citizenship behaviour, when using the on student achievement from official records would require the student
organisational citizenship behaviour checklist (developed to eliminate participants to provide identifying information (e.g., a student
overlapping items with counterproductive work behaviour, and using a number). We felt this may cause the participants to be less truthful in
behavioural frequency rating format), correlations between agreeable- their responses, particularly as some items were tapping academic
ness and organisational citizenship behaviour have been found to be misconduct (cheating, plagiarism). As the primary focus of this study
marginal (r = 0.04; Cohen, Panter, Turan, Morse, & Kim, 2014). As was on the student performance scales, rather than GPA, we felt that
such, it is likely that it is a combination of both item overlap and re- anonymous responding to encourage honesty was the most appropriate
sponse format which are contributing to the differences in statistical option. Furthermore, any unreliability in the GPA measure would pre-
relationships between agreeableness, and MSPS-SCB and Gehring-SCB. sumably affect all scales equally (personality, the multidimensional
While this explains the statistical pattern of results, from a theore- performance scales, Gehring-SCB, and Meriac-CSB), therefore should
tical standpoint, the lack of association between agreeableness and not favourably bias any particular scales, particularly in estimates of
MSPS-SCB (and organisational citizenship behaviour, in the case of incremental validity.
Cohen et al., 2014) still seems counterintuitive, as agreeableness is Another limitation caused by the anonymous collection of data was
associated with kindness and helping people, concepts which seem to that we are unable to assess whether there was any overlap in partici-
be behaviourally tapped by MSPS-SCB. One possibility is that the re- pants from Study 1b to Study 2, which may have increased the stability
lationship between agreeableness and SCB is context dependent, such of the factor structure from Study 1b to Study 2. However, as the data
that while people high in agreeableness may have a desire and will- was collected a full year later, we would not anticipate a great degree of
ingness to help people, this only manifests as SCB under particular si- participant overlap, due to the participants' who were recruited in re-
tuational circumstances. This would explain the significant correlations turn for course credit no longer being eligible for that recruitment
between agreeableness and organisational citizenship behaviour (par- method, the fact that the student cohort would be different, with some
ticularly when using an agreement rating format, which also captures students graduating and other students commencing, and the fact that
attitude to the behaviour) which become marginal when a behavioural the participants were sampled from a much larger potential student
rating format is used. Though the authors are unaware of any research population (over 40,000 students). Regardless, we would recommend
specifically looking at the moderating effect of situational context on this study be replicated in other student populations.
agreeableness and organisational citizenship behaviour relationships,
this importance of the situational context on trait and performance 4. General discussion
relationships is consistent with research which has found that the si-
tuational context moderates the relationship between honesty-humility This series of studies sought to investigate the dimensional structure
(another personality trait) and counterproductive work behaviour of student performance and develop preliminary scales to measure
(Zettler & Hilbig, 2010). Consistent with this interpretation, it is of note multidimensional student performance. The results indicated that the
that generally, the correlations between personality and MSPS-SCB three factor structure of task performance, organisational citizenship
were smaller than that for personality and MSPS-STP or MSPS-CSB, a behaviour, and counterproductive work behaviour seen in job perfor-
finding which is similar to that found in the counterproductive work mance can be replicated in student performance. However, previous
behaviour and organisational citizenship behaviour literature (Berry scales that have been used to measure CSB and SCB appear to have
et al., 2012; Chiaburu et al., 2011). It is possible that in general, con- construct overlap, causing concerns with construct validity. As the
textual factors play a bigger part in SCB than for STP and CSB. MSPS in this study was developed concurrently using factor analytic
Finally, as expected, MSPS-STP was positively correlated with GPA, techniques, these concerns are reduced. Furthermore, the MSPS showed
and was of a higher magnitude than meta-analytically derived corre- concurrent validity with measures of personality and GPA, and incre-
lations between intelligence and conscientiousness in predicting GPA mental validity for predicting GPA over previous CSB and SCB scales.
(Poropat, 2009). This relationship remained significant after controlling As such, though preliminary, the MSPS appears to be reliable and valid
for MSPS-SCB, MSPS-CSB, existing measures of SCB and CSB, and measures of multidimensional student performance.
control demographic variables, suggesting that MSPS-STP captures
unique variance over other measures of student performance. 4.1. Implications
Neither MSPS-CSB nor Meriac-CSB predicted GPA in the first re-
gression model, though they both had significant bivariate correlations The research presented in this article has a number of important
with GPA. A subsequent regression model indicated that Meriac-CSB theoretical and practical considerations. Firstly, previously researchers
became a non-significant predictor when controlling for demographics investigating SCB and CSB needed to choose from a diverse set of non-
and MSPS-CSB, though MSPS-CSB remained significant. This indicates optimal scales. This has led to a proliferation of different scales and
that while MSPS-CSB and Meriac-CSB are highly correlated, MSPS-CSB scale refinements used to measure these constructs across various stu-
explains additional variance to Meriac-CSB. dies. The lack of consistency in scale use makes comparing findings
Gehring-SCB was a significant positive predictor of GPA, while across studies difficult, particularly when considering potential con-
MSPS-SCB was a significant negative predictor in the regression models. struct contamination in the scales used.
Gehring-SCB showed a large number of items exhibiting construct Secondly, the finding that student performance is multidimensional
overlap, which may explain the divergence from MSPS-SCB. However, has implications for researchers investigating student achievement,
the results for MSPS-SCB and GPA, are difficult to explain due to the outcomes, and academic behaviours more broadly. For instance, while
lack of significant bivariate relationship. It is possible that after the relationship between STP and GPA was in the expected direction,

31
D.J. Cummings et al. Learning and Individual Differences 59 (2017) 22–33

unexpectedly, SCB was a negative predictor of GPA when controlling by other universities, and therefore it may be considered a CSB.
for the other scales. The reasons for this are currently not clear, how- However, most items were chosen for their anticipated generalisability,
ever it does indicate that researchers investigating academic achieve- and any poorly performing items in subsequent studies can easily be
ment and related constructs may benefit from expanding the research identified by researcher judgement, or post-hoc by considering cor-
domain past behaviours and outcomes traditionally used. There is a rected item-total correlations when assessing scale reliability. Similarly,
general dearth of research on the multiple dimensions of student per- some items may need to be reworded due to the specific institutional or
formance, particularly when compared to their related constructs in cultural context of the research. Any researchers using these scales will
organisations. be the most appropriate people to make this decision, though we cau-
Thirdly, as with job performance, the recognition that multiple tion that any changes are made with the overall construct definitions in
types of behaviour comprise the academic performance domain has mind. A less likely possibility is that the dimensional structure of
important practical considerations in the consideration of student per- multidimensional student performance will be different in other uni-
formance. For instance, traditionally, students are admitted to uni- versities or countries. However, as the dimensional structure of the
versity and considered successful based upon their cognitive ability and MSPS is both theoretically and empirically driven, we consider this
their academic achievement, which makes a degree of sense. While STP possibility minimal.
does not directly contribute to the university's technical core, it does Finally, this study used a behavioural rating format, consistent with
indirectly affect the technical core in the same way a customer affects similar scales in an organisational setting (Fox et al., 2012;
an exclusive business. For instance, high STP may cause increased re- Spector & Fox, 2010), as it is likely less influenced by attitudes towards
tention, reduced administration costs, and increased graduate outcomes the behaviour than an agreement rating format (Spector & Fox, 2010).
and therefore university reputation. However, a university's culture and However, the behavioural frequency rating format comes with its own
reputation is more than the academic output of its students, and often limitations. For instance, some of the items included in this study (e.g.,
includes components such as inclusion, social responsibility, respect, or “read feedback on an assignment”, “cheated on an exam or test”) are
integrity. As such, and similar to employees, a good student is more unlikely to be able to be demonstrated every day. As such, it is possible
than their academic output, and SCB and CSB are important when that the students may not use the full rating-scale, or alternatively, do
considering student quality. There is increasing consideration of using not interpret the rating format entirely literally, and instead (for in-
attributes beyond academic potential in the selection of students stance), interpret every day to mean every opportunity. Future research
(Schmitt, 2012), and consequently, the framework provided in the may benefit from examining the interpretation of rating anchors and
current series of studies may provide a basis of further research on different response formats, and how this may affect outcomes.
student characteristics. Consequently, there may be times when a behavioural rating format
Finally, the framework provided in this study may assist in further is inappropriate, three of which seem of particular relevance to the
research into university functioning more generally. For instance, in an academic context. Firstly, when investigating behaviours with a very
organisational setting, research has investigated not only the individual low prevalence, a behavioural rating format may not provide enough
level characteristics which lead for task performance, organisational variance when considering individual items. This may be of particular
citizenship behaviour, and counterproductive work behaviour, but also concern for those wishing to investigate CSBs. Secondly, in contexts
the organisational characteristics and work processes that lead to these where participants have not had sufficient time or opportunity to
behaviours (Fox et al., 2012). Similarly, by understanding the uni- complete the behaviours (e.g., before commencing a degree), an
versity characteristics which facilitate STP, SCB, and CSB, we may be agreement or likelihood format is likely to be more appropriate.
able to encourage desirable behaviours and discourage undesirable Thirdly, for researchers wishing to investigate the possible attitudinal
behaviours, which have broader implications for the student experi- antecedents of these behaviours, a combination of rating formats in-
ence. cluding behavioural as well as attitudinal or likelihood may be most
useful.
4.2. Limitations and future directions
4.3. Conclusion
We note two main limitations for this research. Firstly, this study
treated STP, SCB, and CSB as unidimensional constructs, and did not In conclusion, this study found that the three factor structure for job
investigate potential lower order facets. This was intentional, as we performance in organisations can be replicated for student performance
wanted to discover the broad dimensions of student performance and in universities. Additionally, we provided preliminary scales which can
develop preliminary scales for each. Investigating the lower order be used to measure these concepts, which showed both reliability and
structure of these dimensions before assessing the broad dimensional validity. Future researchers may benefit from considering multiple di-
structure would be premature, and could lead to problems regarding mensions of student performance in their research, as opposed to fo-
construct definitions. Additionally, there is debate in the job perfor- cusing on academic achievement.
mance literature about the lower order facets of organisational citi-
zenship behaviour and counterproductive behaviour, and the validity Acknowledgements
and utility of considering these lower order facets (Berry,
Ones, & Sackett, 2007; Hoffman, Blair, Meriac, & Woehr, 2007; We would like to acknowledge the assistance of Ingo Zettler, Inge
Ones & Dilchert, 2013; Poropat & Jones, 2009). As such, any investiga- Schwager, and John Meriac for the provision of scale items. We would
tion of lower order facets of CSB and SCB should be conducted from an also like to acknowledge the editor's and the anonymous reviewers'
emic (bottom-up) rather than etic (imposing structures from other time and helpful comments.
contexts–i.e., organisational) approach. If such an investigation is
conducted, we would recommend adding additional items to the core References
scales outlined in this study, ensuring that any new items are most
strongly correlated with their proposed construct, in order to ensure Allison, B. J., Voss, R. S., & Dryer, S. (2001). Student classroom and career success: The
construct stability. role of organizational citizenship behavior. Journal of Education for Business, 76(5),
282–288. http://dx.doi.org/10.1080/08832320109599650.
A second limitation of this study is that it was only validated in one Bennett, R. J., & Robinson, S. L. (2000). Development of a measure of workplace de-
university, in Australia. It seems possible that some items will be in- viance. Journal of Applied Psychology, 85(3), 349–360. http://dx.doi.org/10.1037/
appropriate in other universities or countries. For instance, the item 0021-9010.85.3.349.
Berry, C. M., Carpenter, N. C., & Barratt, C. L. (2012). Do other-reports of
“Proof-read another students' assignment” may be considered collusion

32
D.J. Cummings et al. Learning and Individual Differences 59 (2017) 22–33

counterproductive work behavior provide an incremental contribution over self-re- Marcus, B., Lee, K., & Ashton, M. C. (2007). Personality dimensions explaining re-
ports? A meta-analytic comparison. Journal of Applied Psychology, 97(3), 613–636. lationships between integrity tests and counterproductive behavior: Big five, or one
http://dx.doi.org/10.1037/a0026739. in addition? Personnel Psychology, 60(1), 1–34.
Berry, C. M., Ones, D. S., & Sackett, P. R. (2007). Interpersonal deviance, organizational Meriac, J. P. (2012). Work ethic and academic performance: Predicting citizenship and
deviance, and their common correlates: A review and meta-analysis. Journal of counterproductive behavior. Learning and Individual Differences, 22(4), 549–553.
Applied Psychology, 92(2), 410–424. http://dx.doi.org/10.1037/0021-9010.92.2.410. http://dx.doi.org/10.1016/j.lindif.2012.03.015.
Brown, T. A. (2006). Confirmatory factor analysis for applied research. New York, NY: Mount, M., Ilies, R., & Johnson, E. (2006). Relationship of personality traits and coun-
Guilford Publications. terproductive work behaviors: The mediating effects of job satisfaction. Personnel
Chiaburu, D. S., Oh, I. S., Berry, C. M., Li, N., & Gardner, R. G. (2011). The five-factor Psychology, 59(3), 591–622. http://dx.doi.org/10.1111/j.1744-6570.2006.00048.x.
model of personality traits and organizational citizenship behaviors: A meta-analysis. Ones, D. S., & Dilchert, S. (2013). Counterproductive work behaviors: Concepts, mea-
Journal of Applied Psychology, 96(6), 1140–1166. http://dx.doi.org/10.1037/ surement, and nomological network. In K. F. Geisinger, B. A. Bracken, J. F. Carlson,
a0024004. J.-I. C. Hansen, N. R. Kuncel, S. P. Reise, & M. C. Rodriguez (Vol. Eds.), APA handbook
Clark, L. A., & Watson, D. (1995). Constructing validity: Basic issues in objective scale of testing and assessment in psychology. Vol. 1. APA handbook of testing and assessment in
development. Psychological Assessment, 7(3), 309–319. http://dx.doi.org/10.1037/ psychology (pp. 643–659). Washington, DC: American Psychological Association.
14805-012. Organ, D. W. (1988). Organizational citizenship behavior: The good soldier syndrome.
Cohen, T. R., Panter, A. T., Turan, N., Morse, L., & Kim, Y. (2014). Moral character in the Lexington, MA: Lexington Books/DC Heath and Com.
workplace. Journal of Personality and Social Psychology, 107(5), 943–963. http://dx. Podsakoff, P. M., MacKenzie, S. B., Moorman, R. H., & Fetter, R. (1990). Transformational
doi.org/10.1037/a0037245. leader behaviors and their effects on followers' trust in leader, satisfaction, and or-
Credé, M., & Niehorster, S. (2009). Individual difference influences on self-focused and ganizational citizenship behaviors. The Leadership Quarterly, 1(2), 107–142. http://
other-focused counterproductive student behaviors. Personality and Individual dx.doi.org/10.1016/1048-9843(90)90009-7.
Differences, 47(7), 769–776. http://dx.doi.org/10.1016/j.paid.2009.06.018. Poropat, A. E. (2009). A meta-analysis of the five-factor model of personality and aca-
Dalal, R. S. (2005). A meta-analysis of the relationship between organizational citizenship demic performance. Psychological Bulletin, 135(2), 322–338. http://dx.doi.org/10.
behavior and counterproductive work behavior. Journal of Applied Psychology, 90(6), 1037/a0014996.
1241–1255. http://dx.doi.org/10.1037/0021-9010.90.6.1241. Poropat, A. E. (2011). The role of citizenship performance in academic achievement and
Fox, S., Spector, P. E., Goh, A., Bruursema, K., & Kessler, S. R. (2012). The deviant citizen: graduate employability. Education + Training, 53(6), 499–514. http://dx.doi.org/10.
Measuring potential positive relations between counterproductive work behaviour 1108/00400911111159467.
and organizational citizenship behaviour. Journal of Occupational and Organizational Poropat, A. E., & Jones, L. (2009). Development and validation of a unifactorial measure
Psychology, 85(1), 199–220. http://dx.doi.org/10.1111/j.2044-8325.2011.02032.x. of citizenship performance. Journal of Occupational and Organizational Psychology,
Gehring, F. (2006). University citizenship behavior. Development and validation of an in- 82(4), 851–869. http://dx.doi.org/10.1348/096317908x374139.
ventory on contextual performance at universities. (unpublished diploma thesis). Aachen, Roberts, B. W., Lejuez, C., Krueger, R. F., Richards, J. M., & Hill, P. L. (2014). What is
Germany: RWTH Aachen University. conscientiousness and how can it be assessed? Developmental Psychology, 50(5),
Goldberg, L. R., Johnson, J. A., Eber, H. W., Hogan, R., Ashton, M. C., Cloninger, C. R., & 1315–1330. http://dx.doi.org/10.1037/a0031109.
Gough, H. G. (2006). The international personality item pool and the future of public- Rotundo, M., & Sackett, P. R. (2002). The relative importance of task, citizenship, and
domain personality measures. Journal of Research in Personality, 40(1), 84–96. http:// counterproductive performance to global ratings of job performance: A policy-cap-
dx.doi.org/10.1016/j.jrp.2005.08.007. turing approach. Journal of Applied Psychology, 87(1), 66–80. http://dx.doi.org/10.
Hakstian, A. R., Farrell, S., & Tweed, R. G. (2002). The assessment of counterproductive 1037/0021-9010.87.1.66.
tendencies by means of the California psychological inventory. International Journal Schmitt, N. (2012). Development of rationale and measures of noncognitive college
of Selection and Assessment, 10(1–2), 58–86. http://dx.doi.org/10.1111/1468-2389. student potential. Educational Psychologist, 47(1), 18–29. http://dx.doi.org/10.1080/
00194. 00461520.2011.610680.
Hoffman, B. J., Blair, C. A., Meriac, J. P., & Woehr, D. J. (2007). Expanding the criterion Schwager, I. T. L., Hülsheger, U. R., Lang, J. W. B., Klieger, D. M., Bridgeman, B., &
domain? A quantitative review of the OCB literature. Journal of Applied Psychology, Wendler, C. (2014). Supervisor ratings of students' academic potential as predictors
92(2), 555–566. http://dx.doi.org/10.1037/0021-9010.92.2.555. of citizenship and counterproductive behavior. Learning and Individual Differences, 35,
Hough, L. M., & Oswald, F. L. (2008). Personality testing and industrial–organizational 62–69. http://dx.doi.org/10.1016/j.lindif.2014.07.005.
psychology: Reflections, progress, and prospects. Industrial and Organizational Spector, P. E. (2010). The relationship of personality to counterproductive work behavior
Psychology, 1(3), 272–290. http://dx.doi.org/10.1111/j.1754-9434.2008.00048.x. (CWB): An integration of perspectives. Human Resource Management Review, 21(4),
Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure 342–352. http://dx.doi.org/10.1016/j.hrmr.2010.10.002.
analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: Spector, P. E., Bauer, J. A., & Fox, S. (2010). Measurement artifacts in the assessment of
A Multidisciplinary Journal, 6(1), 1–55. http://dx.doi.org/10.1080/ counterproductive work behavior and organizational citizenship behavior: Do we
10705519909540118. know what we think we know? Journal of Applied Psychology, 95(4), 781–790. http://
Hurtz, G. M., & Donovan, J. J. (2000). Personality and job performance: The Big Five dx.doi.org/10.1037/a0019477.
revisited. Journal of Applied Psychology, 85(6), 869–879. http://dx.doi.org/10.1037// Spector, P. E., & Fox, S. (2010). Theorizing about the deviant citizen: An attributional
0021-9010.85.6.869. explanation of the interplay of organizational citizenship and counterproductive
Johnson, J. A. (2014). Measuring thirty facets of the Five Factor Model with a 120-item work behavior. Human Resource Management Review, 20(2), 132–143. http://dx.doi.
public domain inventory: Development of the IPIP-NEO-120. Journal of Research in org/10.1016/j.hrmr.2009.06.002.
Personality, 51, 78–89. http://dx.doi.org/10.1016/j.jrp.2014.05.003. Spector, P. E., Fox, S., Penney, L. M., Bruursema, K., Goh, A., & Kessler, S. R. (2006). The
Komarraju, M., & Karau, S. J. (2005). The relationship between the Big Five personality dimensionality of counterproductivity: Are all counterproductive behaviors created
traits and academic motivation. Personality and Individual Differences, 39(3), 557–567. equal? Journal of Vocational Behavior, 68(3), 446–460. http://dx.doi.org/10.1016/j.
http://dx.doi.org/10.1016/j.paid.2005.02.013. jvb.2005.10.005.
Komarraju, M., Karau, S. J., & Schmeck, R. R. (2009). Role of the Big Five personality Viswesvaran, C., & Ones, D. S. (2000). Perspectives on models of job performance.
traits in predicting college students' academic motivation and achievement. Learning International Journal of Selection and Assessment, 8(4), 216–226. http://dx.doi.org/10.
and Individual Differences, 19(1), 47–52. http://dx.doi.org/10.1016/j.lindif.2008.07. 1111/1468-2389.00151.
001. Williams, L. J., & Anderson, S. E. (1991). Job-satisfaction and organizational commitment
Krupić, D., Corr, P. J., Ručević, S., Križanić, V., & Gračanin, A. (2016). Five reinforcement as predictors of organizational citizenship and in-role behaviors. Journal of
sensitivity theory (RST) of personality questionnaires: Comparison, validity and Management, 17(3), 601–617. http://dx.doi.org/10.1177/014920639101700305.
generalization. Personality and Individual Differences, 97, 19–24. http://dx.doi.org/10. Zettler, I. (2011). Self-control and academic performance: Two field studies on university
1016/j.paid.2016.03.012. citizenship behavior and counterproductive academic behavior. Learning and
LeBlanc, C. J. (2014). Characteristics shaping college student organizational citizenship Individual Differences, 21(1), 119–123. http://dx.doi.org/10.1016/j.lindif.2010.11.
behavior. American Journal of Business Education, 7(2), 99–107. 002.
Li, C. H. (2015). Confirmatory factor analysis with ordinal data: Comparing robust Zettler, I., & Hilbig, B. E. (2010). Honesty-humility and a person-situation interaction at
maximum likelihood and diagonally weighted least squares. Behavior Research work. European Journal of Personality, 24(7), 569–582. http://dx.doi.org/10.1002/
Methods, 48(3), 936–949. http://dx.doi.org/10.3758/s13428-015-0619-7. per.757.

33