Koopmans et al.

BMC Public Health 2014, 14:513

http://www.biomedcentral.com/1471-2458/14/513

RESEARCH ARTICLE Open Access

Responsiveness of the individual work performance

questionnaire
Linda Koopmans1,2,4, Jennifer K Coffeng1,4, Claire M Bernaards1,2*, Cécile RL Boot1,4, Vincent H Hildebrandt1,2,
Henrica CW de Vet1,3 and Allard J van der Beek1,4

Abstract
Background: Individual work performance is an important outcome measure in studies in the workplace. Nevertheless,
its conceptualization and measurement has proven challenging. To overcome limitations of existing scales, the Individual
Work Performance Questionnaire (IWPQ) was recently developed. The aim of the current study was to gain insight into
the responsiveness of the IWPQ.
Methods: Data were used from the Be Active & Relax randomized controlled trial. The aim of the trial was to investigate
the effectiveness of an intervention to stimulate physical activity and relaxation of office workers, on need for recovery.
Individual work performance was a secondary outcome measure of the trial. In total, 39 hypotheses were formulated
concerning correlations between changes on the IWPQ scales and changes on similar constructs (e.g., presenteeism)
and distinct constructs (e.g., need for recovery) used in the trial.
Results: 260 Participants completed the IWPQ at both baseline and 12 months of follow-up. For the IWPQ scales, 23%,
15%, and 38%, respectively, of the hypotheses could be confirmed. In general, the correlations between change scores
were weaker than expected. Nevertheless, at least 85% of the correlations were in the expected direction.
Conclusions: Based on results of the current study, no firm conclusions can be drawn about the responsiveness of the
IWPQ. Several reasons may account for the weaker than expected correlations. Future research on the IWPQ’s
responsiveness should be conducted, preferably in other populations and intervention studies, where greater changes
over time can be expected.

Background technical tasks central to his or her job” [1]. It is now gen-
Individual work performance, defined as “employee be- erally agreed upon that, in addition to task performance,
haviours or actions that are relevant to the goals of the the IWP domain consists of contextual performance and
organization” [1], is an important outcome measure in counterproductive work behaviour [2-4]. Contextual per-
studies in the workplace. The conceptualization of IWP formance can be defined as “behaviours that support the
has a long history, and many frameworks have been organizational, social and psychological environment in
proposed to describe the construct domain of IWP [e.g., which the technical core must function” [5]. Counterpro-
1–3]. In the field of occupational health, for example, the ductive work behaviour can be defined as “behaviour that
main focus is on sickness absenteeism or presenteeism, i.e., harms the well-being of the organization” [3].
work absence or losses in IWP due to health impairments. Considering the diversity in conceptual frameworks of
In the field of work and organizational psychology, trad- IWP, it is not surprising that numerous instruments have
itionally, the main focus of the IWP construct has been on been developed to measure (aspects of) IWP. Numerous and
task performance, which can be defined as “the proficiency diverse behaviours, actions, or results are being applied as in-
with which individuals perform the core substantive or dicators of IWP [6]. In occupational health, numerous instru-
ments have been developed to measure sickness absenteeism
* Correspondence: [email protected] or presenteeism, such as the Work Productivity And Impair-
1
Body@Work, Research Center for Physical Activity, Work and Health, TNO-VU ment Questionnaire [7], Work Limitations Questionnaire [8],
University Medical Center, Amsterdam, The Netherlands and the WHO Health and Performance Questionnaire [9].
2
Expertise Center Life Style, TNO, PO Box 2215, Leiden, The Netherlands
Full list of author information is available at the end of the article Also, work and organizational psychologists have developed

© 2014 Koopmans et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the
Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public
Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this
article, unless otherwise stated.
Koopmans et al. BMC Public Health 2014, 14:513 Page 2 of 11
http://www.biomedcentral.com/1471-2458/14/513

numerous scales to measure task performance e.g., [10], study, we focus on the validity of a change score, which is
contextual performance e.g., [11], or counterproductive work estimated on the basis of two or more measurement
behaviour e.g., [12]. points [17]. The aim of the current study was to gain
However, all these scales show several limitations. insight into the responsiveness of the IWPQ.
Most strikingly, none of them measure all of the relevant
dimensions of IWP together. Thus, they do not measure Methods
the full range of IWP. Also, scales measuring different Participants
dimensions can include items overlapping in content Data were used from the Be Active & Relax “Vitality in
(antithetical items), creating unjust overlap between Practice” (VIP) randomized controlled trial [19]. The aim
these scales [13]. As a result, the content validity of of the Be Active & Relax trial was to investigate the effect-
these scales can be questioned. Furthermore, none of the iveness of an intervention to stimulate physical activity
scales appear suitable for generic use. The scales were and relaxation of office workers, on need for recovery. In
developed for specific populations, such as employees September 2011, an invitation was sent to 1,182 office em-
with health problems e.g., [7-9], or they were developed ployees of a financial service provider in The Netherlands,
and refined based on employees with a specific occupa- to participate in the project. A total of 412 employees
tion e.g., [10-12]. (response: 35%) from 19 departments completed the base-
The lack of consensus on how to conceptualize and line questionnaire and signed the informed consent form,
measure IWP is undesirable, because valid measurement and were included in the trial.
is a prerequisite for accurately establishing, for example, The trial included a 2x2 factorial design with four re-
predictors of IWP, or effectiveness of interventions to search arms. The four arms consisted of a combined social
improve IWP. To overcome the aforementioned limita- and physical environmental intervention, a social environ-
tions, the Individual Work Performance Questionnaire mental intervention only, a physical environmental inter-
(IWPQ) was recently developed [14,15]. The IWPQ is vention only and a control group. The social environmental
based on a three-dimensional conceptual framework of intervention consisted of Group Motivational Interviewing
IWP, which was developed after a systematic review of the (GMI). GMI is a counseling style that focuses on behav-
occupational health, psychology, and management litera- ioural change in groups and is derived from Motivational
ture [4]. This framework includes the aforementioned Interviewing at the individual level. GMI was delivered
dimensions of task performance, contextual performance, by the teamleaders of the departments. The teamlea-
and counterproductive work behaviour. The IWPQ is a ders received a two-day training by a GMI-professional.
generic instrument, thus, it is suitable for workers in all The trained teamleaders then gave three GMI-sessions
types of occupations (i.e., blue, pink, and white collar of 90 minutes each to their own team, within a period
workers) and workers with and without health complaints. of six weeks (i.e. three weeks between each session).
An important purpose of the IWPQ is to assess Two months after the final session, a booster session was
changes in IWP. For example, we may want to examine given by the teamleader. All sessions took place during
fluctuations in IWP over time (e.g., due to age), follow work hours. The main aim of these sessions was to stimu-
the effects of negative factors on IWP over time (e.g., late physical activity and relaxation. For the physical envir-
health problems), or identify successful methods to im- onmental intervention, at six departments, several VIP
prove IWP (e.g., intervention studies). In order to do (“Vitality in Practice”) zones were created: (1) the VIP Cof-
this, the IWPQ must be responsive to changes over time. fee Corner Zone (4 elements) – the coffee corner was
Responsiveness can be defined as “the ability of an modified by adding a bar table, bar chairs, a large plant
instrument to detect change over time in the construct and a giant wall poster (a poster visualizing a relaxing en-
to be measured” [16]. There is a lot of confusion about vironment, e.g. wood, water and mountains); (2) the VIP
the concept of responsiveness, and many different defi- Open Office Zone (2 elements) – the office was modified
nitions and measures have been proposed over the past by introducing exercise balls and curtains to divide desks
decades [17]. For example, the definition of responsive- in order to reduce background noise; (3) the VIP Meeting
ness has been clouded by a lack of distinction between Zone (2 elements) – conference rooms were modified by
cross-sectional and longitudinal validity. Secondly, it has placing a standing table and a giant wall poster (a poster
been clouded by a lack of distinction between the effect visualizing a relaxing environment, e.g. wood, water and
of an intervention, and the correlation of changes in the mountains); and (4) the VIP Hall Zone (3 elements) - table
instrument with changes in other instruments [18]. Also, tennis tables were placed and lounge chairs were intro-
responsiveness is often examined using inappropriate out- duced in the hall for informal meetings. In addition, foot-
come measures, such as effect sizes or standardized steps were placed on the floor in the entrance hall to
response mean [17]. Perhaps as a result of this unclarity, promote stair walking. By means of stimulating physical
responsiveness is a seldom examined issue. In the current activity and relaxation, work-related outcomes (e.g.,
Koopmans et al. BMC Public Health 2014, 14:513 Page 3 of 11
http://www.biomedcentral.com/1471-2458/14/513

sickness absenteeism, work engagement and individual Job satisfaction was assessed using one overall question
work performance) were expected to improve for the (“Overall, how satisfied are you with your job?”) on a rating
intervention groups compared to the control group. For scale from 1 (”highly dissatisfied”) to 5 (“very satisfied”). A
the purpose of the current study, data of all four groups single-item measure of job satisfaction has been found to
were taken together. This study was approved by the correlate highly with job satisfaction scales, and was there-
Medical Ethics Committee of the VU University Medical fore considered valid [23,24].
Center, Amsterdam, The Netherlands. Full details of the Work engagement was measured using the Utrecht
design of the Be Active & Relax trial have been reported Work Engagement Scale (UWES) [25]. The UWES con-
elsewhere [19]. sists of three scales (vigour, dedication, and absorption),
and a total of 17 items assessed on a rating scale from 1
(“never”) to 7 (“always”). The total score was calculated by
Measures adding the means of each scale, and dividing the sum by
As examining the responsiveness of the IWPQ was not a three. The psychometric properties of this questionnaire
main aim of the Be Active & Relax trial, measurement in- have been tested and results indicated an acceptable reli-
struments were included that represented important out- ability of vigour (α = 0.68-0.80), dedication (α = 0.91)
comes in the trial. Measurements took place at baseline absorption (α = 0.73-0.75), and the total score (α = 0.93),
(T0), and at 6 months (T1) and 12 months (T2) follow- as well as acceptable convergent validity [25].
up. Only the measurements at baseline and at 12 months Work ability was assessed using one question (“How do
(T2) were used to assess responsiveness of the IWPQ. you rate your current work ability compared to lifetime
best?”) from the Work Ability Index (WAI) [26], on a
Individual work performance was measured using rating scale from 1 (“completely unable to work”) to 10
the Individual Work Performance Questionnaire (IWPQ) (“at its best”). The single-item question is very strongly
[14,15]. The IWPQ consists of 18 questions in three scales: associated with the total WAI, and has shown good pre-
task performance (5 items), contextual performance (8 dictive validity [27].
items), and counterproductive work behaviour (5 items). Performance rating by the manager was assessed by
The IWPQ has a recall period of 3 months and a rating asking one self-report question (“How would your manager
scale from 0 (“seldom”) to 4 (“always”) for task and context- rate your overall job performance, compared to colleagues
ual performance, and 0 (“never”) to 4 (“often”) for counter- in a similar job?”) on a rating scale from 1 (“much worse”)
productive work behaviour. For the IWPQ subscales, a to 5 (“much better”). This question was adapted from the
mean score is calculated by adding the item scores, and div- WHO-HPQ [9] presenteeism question, and previously
iding their sum by the number of items in the subscale. used in The Netherlands Working Conditions Survey [28].
Hence, the IWPQ yields three subscale scores that range The reliability and validity of this question is unknown.
between 0 and 4, with higher scores reflecting higher task Self-rated work quality and quantity were assessed
and contextual performance, and higher counterproductive using one question each (“How do you rate the quality of
work behaviour. The psychometric properties of the IWPQ your own work?” and “How do you rate the quantity of
have been tested and results indicated good to excellent in- your own work?”) on a rating scale from 1 (”insufficient”)
ternal consistency for task performance (α = 0.78), context- to 5 (“excellent”). The reliability and validity of these
ual performance (α = 0.85) and counterproductive work questions is unknown.
behaviour (α = 0.79). The IWPQ has shown good face and Need for recovery (NFR) was assessed using the Need
structural validity [6,14,15], as well as sufficient convergent for Recovery after Work scale [29]. This Dutch version
validity and good discriminative validity [20]. of the Questionnaire on the Experience and Evaluation of
Presenteeism, which can be defined as “decreased on- Work (Dutch abbreviation: VBBA) consists of eleven di-
the-job performance due to the presence of health prob- chotomous items (yes/no), representing short-term effects
lems” [21], was assessed through self-report with the of a day at work. The NFR score is a percentage score (0
World Health Organization Health and Work Perform- to 100) of positive answers of those providing data for at
ance Questionnaire (WHO-HPQ) [9]. Presenteeism was least 8 of the 11 items. The Need for Recovery after Work
assessed by asking participants to rate their actual per- scale has shown good reliability (α = 0.86-0.88), construct
formance in relation to possible performance. The score validity, and responsiveness in The Netherlands [29-31].
represents percentage of performance, and has a lower Physical activity was assessed using the Short Question-
bound of 0 (total lack of performance) and an upper naire to Assess Health Enhancing Physical Activity
bound of 100 (top performance). The reliability and val- (SQUASH) [32]. Duration and intensity of active commut-
idity of the HPQ was examined for several occupations, ing, leisure time activities, sport activities, household activ-
and showed good convergent validity. However, poor ities, and physical activities at work (standing and walking),
validity was found for white collar workers [9,22]. were assessed. For each domain, employees were asked to
Koopmans et al. BMC Public Health 2014, 14:513 Page 4 of 11
http://www.biomedcentral.com/1471-2458/14/513

report the frequency (i.e., times per week), duration of ac- performance, negative correlations were expected for
tivities (i.e., in minutes), and self-reported intensity (i.e., counterproductive work behaviour, and vice versa.
light, moderate or vigorous). Total scores for minutes per
week spent on light, moderate, and vigorous physical activ- Hypotheses with similar constructs
ities were calculated. The SQUASH scores have shown rea- The first 21 hypotheses (3 IWPQ scales × 7 constructs)
sonable reproducibility (r = 0.57-0.58) and validity against concern relationships of the IWPQ scales with similar
accelerometry (r = 0.45-0.67), which is comparable to other constructs. These constructs were classified as similar
physical activity questionnaires [32,33]. constructs, because these constructs were theoretically
General health and vitality were measured using the expected to correlate moderately with work performance,
Dutch version of the Rand-36 [34]. General health was or were found to correlate moderately with work perform-
measured by asking workers to indicate how they per- ance in previous research. For example, in a review by
ceived their general health, on a rating scale from 1 Judge et al. [38], the correlation between overall job satis-
(“poor”) to 5 (“excellent”). Vitality was measured with a faction and work performance was estimated to be 0.30.
scale of 5 items, asking workers to indicate how often they Therefore, the change in the IWPQ task and contextual
felt full of life, worn out, tired and full of energy, on a rat- performance scale was expected to correlate moderately
ing scale from 1 (“never”) to 6 (“always”). This scale was positive (0.30–0.50) with the change in presenteeism [20],
transformed to a 0–100 score, with higher scores indicat- job satisfaction e.g., [38], work engagement e.g., [39], work
ing higher vitality. The Dutch version of the Rand-36 has ability e.g., [40], performance rating by the manager [41],
shown good reliability for the vitality scale (α = 0.82) and work quality, and work quantity. The change in the IWPQ
had reasonable construct validity [34]. counterproductive work behaviour scale was expected to
Exhaustion was measured using the OLdenburg Burnout correlate moderately negative (−0.50–-0.30) with the
Inventory (OLBI) [35]. The OLBI consists of eight items change in presenteeism [20], job satisfaction e.g., [38],
on a 4-point scale ranging from 1 (“totally disagree”) to 4 work engagement e.g., [39], and work ability e.g., [40].
(“totally agree”). A mean score was calculated. The OLBI Based on literature, the change in the IWPQ counterpro-
has shown good reliability (α = 0.80-0.85) and reasonable ductive work behaviour scale was expected to correlate
convergent and discriminant validity in different occupa- weakly or not at all (−0.20–0.20) with the change in per-
tional groups [35,36]. formance rating by the manager, work quality, and work
Sickness absenteeism data were retrieved from company quantity [13].
records, for the year prior to the intervention (i.e. base-
line), and for the year of the intervention (i.e., 12 month Hypotheses with distinct constructs
follow-up). The score represents the number of workdays The last 18 hypotheses (3 IWPQ scales × 6 constructs)
absent per year. concern relationships of the IWPQ scales with distinct
constructs. These constructs were classified as distinct
constructs, because these constructs were theoretically
Correlations between change scores expected to correlate weakly or not at all with work per-
A construct approach of responsiveness testing [17] was formance, or were found to correlate weakly or not at all
applied in the current study, which means that hypothe- with work performance in previous research. For example,
ses were formulated concerning relationships between it was found that absenteeism is not strongly related to
changes on the IWPQ and changes on other instruments work performance [42,43]. Therefore, the change in the
used in the Be Active & Relax trial. Based on the literature, IWPQ task and contextual performance scale was
hypotheses concerning the relationships between changes expected to correlate weakly positive (0.20–0.30), and the
on the IWPQ scales and changes on other instruments change in the IWPQ counterproductive work behaviour
were formulated. In line with Cohen [37], we interpreted a scale weakly negative (−0.30–-0.20), with the change in
correlation coefficient over 0.50 as strong, 0.30 to 0.50 as need for recovery e.g., [36,44], physical activity e.g., [45],
moderate, 0.10 to 0.30 as weak, and below 0.10 as no rela- general health e.g., [21,46], vitality e.g., [47], and exhaus-
tion between constructs at all. When moderate correla- tion e.g., [48]. Finally, the change in each IWPQ scale was
tions were expected, based on literature, we classified expected to correlate weakly or not at all (−0.20– 0.20)
these constructs as similar constructs (e.g., presenteeism). with the change in sickness absenteeism [42,43].
When weak correlations or no correlations were expected,
we classified these constructs as distinct constructs (e.g., Data analysis
need for recovery). Based on the literature, expectations Pearson correlations between the change scores of each
were formulated per IWPQ scale, resulting in a total of 39 IWPQ scale and the change scores on the other constructs
hypotheses (3 IWPQ scales × 13 constructs). If positive were calculated for the change scores from baseline (T0)
correlations were expected for task and contextual to 12 months (T2). To examine the magnitude of the
Koopmans et al. BMC Public Health 2014, 14:513 Page 5 of 11
http://www.biomedcentral.com/1471-2458/14/513

changes, respondents were divided in three groups: general health as good (M = 3.35, SD = 0.85, on a 5-point
those who decreased at least one point on a construct, scale), and had an average BMI of 25.11 (SD = 4.07).
those who increased at least one point on a construct,
and those who changed less than one point on a con-
Descriptive statistics of the IWPQ scales and the
struct. For the decrease and increase groups, the mean
other constructs
change and SD of change were calculated for the IWPQ
Table 1 presents the mean scores and standard deviations
scales. Only participants who completed the IWPQ at
(SD) on the IWPQ scales and the other constructs at base-
both T0 and T2 were included in the data analysis. Ana-
line (T0) and 12 months (T2). It also reports the mean
lyses were conducted in SPSS 20.0 [49].
and standard deviation (SDchange) of the change scores on
the IWPQ scales and the other constructs from T0 to T2.
Results
Descriptive statistics of the participants Correlations between change scores
Of the 412 participants in the Be Active & Relax trial, 260 Table 2 presents the expected and observed correlations
participants (63%) completed the IWPQ at both baseline between the change scores of the IWPQ scales and the
and 12 months. The main reasons for loss-to-follow-up change scores of the other constructs. For task performance,
were changing job to a different employer and lack of mo- 85% of the correlations were in the expected direction, and
tivation. At baseline (n = 260), participants had a mean for contextual performance and counterproductive work
age of 43.2 years (SD = 9.9), worked 36 hours per week behaviour, 92% of the correlations were in the expected
(SD = 5.1), most were male (63%), and most were highly direction. However, in many cases, the correlations were
educated (79%). On average, participants rated their weaker than expected.

Table 1 Mean scores (and SD) and mean change scores (and SDchange) on the IWPQ scales and the similar/distinct
constructs at baseline (T0) and 12 months (T2)
T0 T2 Change score
(baseline) (12 months) T2-T0
n = 412 n = 260 n = 260
Mean (SD) Mean (SD) Mean (SDchange)
IWPQ (0–4)
Task performance 2.46 (0.68) 2.63 (0.66) 0.17 (0.70)
Contextual performance 2.34 (0.71) 2.39 (0.79) 0.04 (0.69)
Counterproductive work behaviour 1.23 (0.65) 1.16 (0.66) -0.07 (0.64)
Similar constructs
Presenteeism (0–100) 76.58 (8.76) 75.87 (10.62) -0.79 (11.51)
Job satisfaction (1–5) 3.96 (0.73) 3.85 (0.75) -0.11 (0.80)
Work engagement (1–7) 4.91 (0.85) 4.84 (0.93) -0.07 (0.71)
Work ability (1–10) 7.79 (1.42) 7.70 (1.57) -0.08 (1.56)
Performance rating by the manager (1–5) 3.41 (0.81) 3.46 (0.81) 0.06 (0.81)
Self-rated work quality (1–5) 3.83 (0.79) 3.63 (0.87) -0.20 (0.95)
Self-rated work quantity (1–5) 3.87 (0.83) 3.74 (0.92) -0.12 (0.95)
Distinct constructs
Need for recovery (0–100) 32.20 (29.26) 27.78 (28.71) -2.40 (23.70)
Physical activity (min/week)
- Light 1810.10 (1363.68) 1603.23 (1618.94) -199.40 (1785.64)
- Moderate 281.81 (254.19) 350.94 (633.98) 72.66 (629.00)
- Vigorous 83.53 (160.15) 99.79 (272.90) 9.40 (266.15)
General health (1–5) 3.35 (0.85) 3.37 (0.84) 0.79 (1.53)
Vitality (0–100) 64.08 (18.84) 65.72 (17.97) 1.87 (15.17)
Exhaustion (1–4) 2.15 (0.48) 2.15 (0.46) 0.04 (0.40)
Sickness absenteeism (workdays absent per year) 7.55 (21.81) 7.37 (20.91) 0.55 (25.03)
Koopmans et al. BMC Public Health 2014, 14:513 Page 6 of 11
http://www.biomedcentral.com/1471-2458/14/513

Table 2 Pearson correlations (E = expected, O = observed) between change scores of the IWPQ scales and similar/
distinct constructs (n = 260)
IWPQ scale
Task performance Contextual performance Counterproductive work behaviour
Similar constructs
Presenteeism E: 0.30 – 0.50 E: 0.30 – 0.50 E: -0.50 – -0.30
O: 0.18 O: 0.22 O: -0.11
Job satisfaction E: 0.30 – 0.50 E: 0.30 – 0.50 E: -0.50 – -0.30
O: 0.12 O: 0.17 O: -0.24
Work engagement E: 0.30 – 0.50 E: 0.30 – 0.50 E: -0.50 – -0.30
O: 0.19 O: 0.29 O: -0.23
Work ability E: 0.30 – 0.50 E: 0.30 – 0.50 E: -0.50 – -0.30
O: 0.16 O: 0.26 O: -0.23
Performance rating by the manager E: 0.30 – 0.50 E: 0.30 – 0.50 E: -0.20 – 0.20
O: 0.16 O: 0.22 O: -0.02*
Work quality E: 0.30 – 0.50 E: 0.30 – 0.50 E: -0.20 – 0.20
O: 0.20 O: 0.18 O: -0.06*
Work quantity E: 0.30 – 0.50 E: 0.30 – 0.50 E: -0.20 – 0.20
O: 0.11 O: 0.19 O: 0.02*
Distinct constructs
Need for recovery E: -0.30 – -0.20 E: -0.30 – -0.20 E: 0.20 – 0.30
O: -0.15 O: -0.11 O: 0.16
Physical activity E: 0.20 – 0.30 E: 0.20 – 0.30 E: -0.30 – -0.20
- Light O: -0.09 O: -0.04 O: -0.07
- Moderate O: 0.03 O: 0.03 O: -0.07
- Vigorous O: -0.05 O: 0.00 O: -0.04
General health E: 0.20 – 0.30 E: 0.20 – 0.30 E: -0.30 – -0.20
O: -0.07 O: 0.08 O: 0.02
Vitality E: 0.20 – 0.30 E: 0.20 – 0.30 E: -0.30 – -0.20
O: 0.23* O: 0.29* O: -0.03
Exhaustion : -0.30 – -0.20 E: -0.30 – -0.20 E: 0.20 – 0.30
O: -0.23* O: -0.13 O: 0.23*
Sickness absenteeism E: -0.20 – 0.20 E: -0.20 – 0.20 E: -0.20 – 0.20
O: -0.14* O: -0.08* O: -0.09*
Hypotheses:
Confirmed 23% 15% 38%
In the right direction 85% 92% 92%
Note: E = expected correlation, O = observed correlation. * = Confirmed hypothesis.

For the task performance scale, 3 out of 13 (23%) with the change in vitality (r = 0.29), and weakly negative
hypotheses were fully confirmed. As expected, the change with the change in absenteeism (r = −0.08). Furthermore,
in task performance correlated moderately positive with the correlation between the change in contextual perform-
the changes in vitality (r = 0.23), moderately negatively ance and the changes in most of the similar constructs (e.g.,
with the change in exhaustion (r = −0.23), and weakly presenteeism, work engagement, work ability) approached
negative with the change in absenteeism (r = −0.14). the 0.30 correlation strength.
For the contextual performance scale, 2 out of 13 (15%) For the counterproductive work behaviour scale, 5 out of
hypotheses were fully confirmed. As expected, the change 13 (38%) hypotheses were fully confirmed. As expected, the
in contextual performance correlated moderately positive change in counterproductive work behaviour correlated
Koopmans et al. BMC Public Health 2014, 14:513 Page 7 of 11
http://www.biomedcentral.com/1471-2458/14/513

weakly with the changes in rating by the manager scores of task performance and light and intense physical
(r = −0.02), work quality (r = −0.06), work quantity (r = activity (r = −0.09 and −0.05, respectively), task perform-
0.02), and absenteeism (r = −0.09), and moderately positive ance and general health (r = −0.07), contextual performance
with the change in exhaustion (r = 0.23). and light physical activity (r = −0.04), and counterproduct-
In sum, 23%, 15%, and 38% of the hypotheses could be ive work behaviour and general health (r = 0.02).
confirmed for the IWPQ task performance, contextual
performance, and counterproductive work behaviour Magnitude of change
scales, respectively. As hypothesized, the correlations of Table 3 presents the number of participants that respect-
the IWPQ scales were slightly stronger with similar con- ively decreased or increased at least one point on a con-
structs than with distinct constructs, on average. However, struct from T0 to T2, and their corresponding mean
in general, the correlations between change scores were change and standard deviation of change on the IWPQ
weaker than expected. Nevertheless, most of the corre- scales. For example, of the 260 participants, 111 partici-
lations (at least 85%) were in the expected direction. pants reported a decreased need for recovery of at least
Exceptions were the correlations between the change one point. On average this group showed an increase in

Table 3 Number of participants (n) that respectively decreased or increased at least one point on a construct from T0
to T2, and their corresponding mean change and standard deviation of mean change (SDchange) on the IWPQ scales
IWPQ
Task performance Contextual performance Counterproductive work behaviour
Mean change Mean change Mean change
T0-T2 n (SDchange) (SDchange) (SDchange)
Similar constructs
Presenteeism Decreased 62 -0.07 (0.83) -0.16 (0.73) 0.09 (0.67)
Increased 61 0.18 (0.57) 0.23 (0.71) -0.08 (0.71)
Job satisfaction Decreased 56 0.07 (0.84) -0.17 (0.87) 0.21 (0.81)
Increased 36 0.24 (0.66) 0.16 (0.62) -0.42 (0.58)
Work engagement Decreased 19 -0.16 (1.09) -0.30 (0.79) 0.26 (0.77)
Increased 12 0.27 (0.94) 0.21 (0.73) -0.23 (0.62)
Work ability Decreased 77 0.04 (0.71) -0.17 (0.70) -0.07 (0.69)
Increased 78 0.32 (0.75) 0.24 (0.71) -0.15 (0.66)
Performance rating by the manager Decreased 39 -0.07 (0.62) -0.19 (0.63) -0.08 (0.62)
Increased 52 0.20 (0.81) 0.19 (0.70) -0.08 (0.69)
Self-rated work quality Decreased 78 -0.01 (0.80) -0.14 (0.74) -0.04 (0.74)
Increased 43 0.33 (0.67) 0.23 (0.73) -0.07 (0.61)
Self-rated work quantity Decreased 64 0.10 (0.79) -0.09 (0.80) -0.14 (0.72)
Increased 52 0.24 (0.57) 0.15 (0.65) -0.11 (0.52)
Distinct constructs
Need for recovery Decreased 111 0.27 (0.65) 0.12 (0.65) -0.11 (0.63)
Increased 91 0.10 (0.76) -0.07 (0.82) 0.02 (0.70)
Physical activity Decreased 79 0.23 (0.77) 0.09 (0.70) 0.03 (0.62)
Increased 181 0.15 (0.66) 0.02 (0.68) -0.11 (0.64)
General health Decreased 36 0.44 (0.74) -0.07 (0.77) -0.26 (0.70)
Increased 114 0.15 (0.63) 0.09 (0.68) -0.09 (0.58)
Vitality Decreased 95 0.03 (0.75) -0.16 (0.68) -0.02 (0.65)
Increased 108 0.30 (0.67) 0.18 (0.66) -0.10 (0.66)
Exhaustion Decreased 4 0.30 (0.99) 0.49 (0.43) -0.70 (1.09)
Increased 7 -0.34 (0.91) 0.29 (0.74) 0.43 (0.76)
Absenteeism Decreased 84 0.25 (0.71) 0.08 (0.68) -0.05 (0.67)
Increased 73 0.06 (0.66) -0.08 (0.70) -0.13 (0.62)
Koopmans et al. BMC Public Health 2014, 14:513 Page 8 of 11
http://www.biomedcentral.com/1471-2458/14/513

task performance (Mchange = 0.27, SDchange = 0.65), an IWPQ, and its good discriminant validity, suggest that it is
increase of contextual performance (Mchange = 0.12, likely that the IWPQ scales can also detect changes within
SDchange = 0.65), and a decrease in counterproductive groups over time.
work behaviour (Mchange = −0.11, SDchange = 0.63). A total Possibly, low responsiveness of the IWPQ could be
of 91 participants reported an increased need for recovery. caused by ceiling and floor effects in the scales. Although
On average this group showed a slight increase in task previous examination of the IWPQ using Rasch analysis
performance (Mchange = 0.10, SDchange = 0.76), a slight has shown that the items of the IWPQ are relatively well-
decrease in contextual performance (Mchange = −0.07, distributed over the scales, persons continue to score rela-
SDchange = 0.82), and a slight increase in counterproduct- tively high on task performance (ceiling effect), and low
ive work behaviour (Mchange = 0.02, SDchange = 0.70). on CWB (floor effect); [15]. This could be caused by the
Similarly, for example, a total of 81 participants tendency of persons to evaluate and present themselves in
reported a decrease in presenteeism of at least one point. a socially desirable, favorable way [51,52]. As a conse-
On average this group showed a slight decrease in quence of the ceiling and floor effects, it becomes hard to
task performance (Mchange = −0.07, SDchange = 0.83), a detect further improvements in task performance, and fur-
decrease in contextual performance (Mchange = −0.16, ther decreases in CWB. Thus, the ability to detect changes
SDchange = 0.73), and a slight increase in counterproduct- at the high part of the task performance scale, and low
ive work behaviour (Mchange = 0.09, SDchange =0.67). A part of the CWB scale, may be diminished.
total of 61 participants reported an increase in present- Another possible reason for the lower than expected
eeism. On average this group showed an increase in correlations may lie in the study population. As said
task performance (Mchange = 0.18, SD change = 0.57) an before, the population in the current study consisted of
increase in contextual performance (Mchange = 0.23, relatively healthy, well-functioning office workers who, in
SDchange = 0.71), and a slight decrease in counterpro- general, scored high on constructs such as general health,
ductive work behaviour (Mchange = −0.08, SDchange = presenteeism, and job satisfaction, and low on constructs
0.71). As can be observed in the above examples, most such as need for recovery, exhaustion, and sickness absen-
of the time, a decrease or increase in a construct was teeism. This makes it hard to obtain or detect any further
associated with a corresponding decrease or increase in improvements in this population. Despite the use of an
the IWPQ scales. intervention, small changes on the constructs over the 12-
month intervention period were obtained. When examin-
Discussion ing the scatterplots of the change scores, low spread on
The aim of the current study was to examine the many constructs can be observed (i.e., dots clustered in
responsiveness of the IWPQ, i.e., the ability of the IWPQ the middle), and this can cause deflated correlations [17].
to detect change over time. A total of 39 hypotheses Finally, a reason for the lower than expected correla-
were formulated concerning the relationships between tions may be that the intervention was not effective
changes on the IWPQ and changes on similar constructs enough to obtain changes in IWP. The primary aim of the
(e.g., presenteeism) and distinct constructs (e.g., need for Be Active & Relax study was to investigate the effective-
recovery) used in the Be Active & Relax trial. Although ness of an intervention to stimulate physical activity and
most of the correlations between change scores were in relaxation of office workers, on need for recovery [19].
the expected direction, most were weaker than expected. Indirectly, an increase in physical activity and relaxation
Several reasons may account for this. were expected to improve IWP. However, it may be that
First, the IWPQ questions may not be sensitive enough the intervention was not specific or intense enough to
to pick up changes in IWP over time. Also, it is hard to obtain improvements in IWP. Despite the fact that the
say how a change from answer categories “regularly” to intervention was not directly targeted at IWP, and despite
“often” can be achieved. What needs to be done to accom- high baseline levels on the constructs, a statistically sig-
plish a change from “regularly” to “often,” e.g., in keeping nificant increase in tasks performance (B = 0.2, 95% CI
your work results in mind? And what does this change 0.0; 0.4), and a statistically significant decrease in context-
mean? In sum, the questions of the IWPQ scales may lack ual performance (B = −0.3, 95% CI −0.4; 0.1), were
discriminative ability. However, in the developmental detected in the Be Active & Relax study [53]. The decrease
phase of the IWPQ scales, Rasch analysis [50] was in contextual performance could be explained by the fact
performed to make sure that those items with a high that participants in the intervention groups were stimu-
discrimination parameter (i.e., high slope) were retained in lated to engage in physical activity and relaxation during
the IWPQ 1.0 [14,15]. Also, in the validation phase of the the workday, and this possibly could have reduced taking
IWPQ scales, the IWPQ 1.0 was able to discriminate on extra work tasks, for example. Thus, this study showed
between known groups [20]. The fact that items with a that the IWPQ is able to detect statistically significant
high discrimination parameter were included in the changes in individual work performance over time.
Koopmans et al. BMC Public Health 2014, 14:513 Page 9 of 11
http://www.biomedcentral.com/1471-2458/14/513

Whether this change is a relevant change remains a ques- other constructs, because it was expected that many
tion for future research. participants would show no changes, and based on results
in earlier studies with similar questionnaires. For the
Assessment of responsiveness constructs used in the current study, previous research
As stated in the Introduction, there is a lot of confusion has shown that, for example, the cross-sectional correl-
about the concept over responsiveness, and many different ation between IWP and work engagement ranges between
definitions and measures have been proposed over the r = 0.30-0.50 e.g., [56]. It is therefore questionable whether
past decades [17]. In addition, or perhaps, as a result, correlations of r = 0.30-0.50 between their change scores
responsiveness is a seldom examined issue. For example, can reasonably be expected. Such high correlations
Abma et al. [54] reviewed the measurement properties of between change scores would likely be obtained for identi-
five self-report (health-related) work functioning instru- cal constructs, rather than similar (but not identical)
ments; the EWPS, WLQ, SPS, WPS, and LEAPS. For all constructs.
five instruments, the methodological quality of responsive-
ness testing was poor, or not studied. Of the instruments Recommendations for future research
used in the current study, only the responsiveness of the The current study provides merely a first step towards
Need for Recovery Scale was examined. Based on effect gaining insight into the responsiveness of the IWPQ. The
sizes, the responsiveness of this scale appeared to be good responsiveness of the IWPQ should be further examined
[30]. However, the responsiveness of the other question- in future research, to determine whether its responsive-
naires used in the current study remains unknown. This is ness is truly low, or whether the low responsiveness found
a limitation of the responsiveness testing process, because in the current study was caused by limitations of the
responsiveness of a new questionnaire is tested against current study. We therefore recommend examining the
change scores of existing questionnaires, whose respon- responsiveness of the IWPQ in different populations, pref-
siveness is also unknown, and may be poor. erably in populations with low(er) baseline levels on the
No golden standard or clear guidelines seem to exist for constructs, where large(r) changes on the constructs over
the assessment of responsiveness and the interpretation of time can be expected. Suggestions for such populations
results. De Vet and colleagues [17] stated that responsive- could be a sample of workers with work-related musculo-
ness is often examined based on inappropriate outcome skeletal health problems, mental health problems, and/or
measures, such as effect sizes or standardized response low job satisfaction. An intervention study, which is
mean. They advise that responsiveness should be seen as a directly aimed at improving IWP, could obtain greater
form of longitudinal validity, using either a criterion changes in these populations, making it easier to detect
approach (if a gold standard is available) or a construct changes in IWP and related constructs. Suggestions for
approach (testing hypotheses of change scores). such a study could be an intervention focusing on man-
In addition to the lack of clarity on how responsiveness agerial style, technological improvements at work, and/or
should be tested, there are no clear guidelines as to what job skills training. Also, the responsiveness of the IWPQ
the strength of correlations between change scores should should preferably be examined using other measurement
be. A final reason for the large percentage of unconfirmed instruments of which the responsiveness is known. In
hypotheses in the current study, may be that the hypothe- addition, more information on the smallest detectable
sized correlations (r = 0.30-0.50) were too high to begin change and the minimally important change of the IWPQ,
with. In line with Cohen [37], we interpreted a correlation would further aid the interpretation of the responsiveness
coefficient over 0.50 as strong, 0.30 to 0.50 as moderate, of the IWPQ. Finally, the responsiveness of questionnaires
0.10 to 0.30 as weak, and below 0.10 as no relation deserves greater attention, and clear guidelines for asses-
between constructs at all. Often, Cohen’s guidelines are sing and interpreting responsiveness should be adopted.
used for cross-sectional correlations, i.e., when a correl- The guidelines proposed by Terwee et al. [18], Mokkink
ation between two different measurement scores obtained et al. [57], and De Vet et al. [17] could provide a good
at the same point in time is examined (thus, there is only starting point for this.
one measurement). When it comes to correlations be-
tween change scores (multiple measurements), it is based Conclusion
on two measurements, and a double measurement error Based on results of the current study, no firm conclusions
is involved. Due to this double measurement error, it can be drawn about the responsiveness of the IWPQ. Over-
seems reasonable that lower correlations may be expected. all, most of the correlations between changes on the IWPQ
This issue has been addressed by other researchers. For scales and changes on other constructs were in the expected
example, Abma et al. [55] examined the responsiveness of direction, although not as high as expected. This might indi-
the Work Role Functioning questionnaire, and they cate low responsiveness of the IWPQ. However, the weaker
hypothesized correlation sizes around 0.20 to 0.30 with than expected correlations may also be accounted for by
Koopmans et al. BMC Public Health 2014, 14:513 Page 10 of 11
http://www.biomedcentral.com/1471-2458/14/513

characteristics of the intervention study, such as the rela- health and work performance questionnaire (HPQ). J Occup Environ Med
tively healthy, well-functioning study population, and an 2003, 45:156–174.
10. Williams LJ, Anderson SE: Job satisfaction and organizational commitment
intervention study that was not primarily aimed at IWP. as predictors of organizational citizenship and in-role behaviors. J Manag
Nevertheless, the IWPQ was able to show statistically sig- 1991, 17(3):601–617.
nificant changes in IWP during baseline and 12 months 11. Podsakoff PM, MacKenzie SB: A second generation measure of organizational
citizenship behavior. Bloomington: Indiana University; 1989.
follow-up. Future research should provide more information 12. Bennett RJ, Robinson SL: Development of a measure of workplace
about the smallest detectable change, the minimally import- deviance. J Appl Psychol 2000, 85(3):349–360.
ant change, and the responsiveness of the IWPQ in other 13. Dalal RS: A meta-analysis of the relationship between organizational
citizenship behavior and counterproductive work behavior. J Appl Psychol
populations and intervention studies. 2005, 90:1241–1255.
14. Koopmans L, Bernaards CM, Hildebrandt VH, van Buuren S, van der Beek AJ,
Competing interests de Vet HCW: Development of an individual work performance
The authors declare that they have no competing interests. questionnaire. Int J Product Perform Manag 2013, 62(1):6–28.
15. Koopmans L, Bernaards CM, Hildebrandt VH, Van Buuren S, Van der Beek AJ,
Authors’ contributions De Vet HCW: Improving the individual work performance questionnaire
All authors have made substantial contributions to conception and design of using rasch analysis. J Appl Meas 2014, 15(2):160–175.
the article. JKC and CRLB are responsible for data acquisition. LK is 16. Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW, Knol DL, Bouter LM,
responsible for data analysis. All authors have contributed to interpretation De Vet HCW: The COSMIN study reached international consensus on
of the data, and have been involved in drafting the manuscript or revising it taxonomy, terminology, and definitions of measurement properties for
critically for important intellectual content. All authors have given approval health-related patient-reported outcomes. J Clin Epidemiol 2010, 63:737–745.
of the final version of the article, and agree to be accountable for all aspects 17. De Vet HCW, Terwee CB, Mokkink LB, Knol DL: Measurement in medicine.
of the work in ensuring that questions related to the accuracy or integrity of New York, United States of America: Cambridge University Press; 2011.
any part of the work are appropriately investigated and resolved. 18. Terwee CB, Dekker FW, Wiersinga WM, Prummel MF, Bossuyt PM: On
assessing responsiveness of health-related quality of life instruments:
Acknowledgements guidelines for instrument evaluation. Qual Life Res 2003, 12(4):349–362.
All authors were funded by Body@Work, Research Center for Physical 19. Coffeng JK, Hendriksen IJM, Duijts SF, Proper KI, Van Mechelen W, Boot CRL:
Activity, Work and Health, TNO-VU University Medical Center, Amsterdam, Development of the be active & relax vitality in practice (VIP) project
The Netherlands. and design of an RCT to reduce the need for recovery in office
employees. BMC Public Health 2012, 12:592.
Author details 20. Koopmans L, Bernaards CM, Hildebrandt VH, De Vet HCW, Van der Beek AJ:
1 Construct validity of the individual work performance questionnaire.
Body@Work, Research Center for Physical Activity, Work and Health, TNO-VU
University Medical Center, Amsterdam, The Netherlands. 2Expertise Center J Occup Environ Med 2014, 56(3):331.
Life Style, TNO, PO Box 2215, Leiden, The Netherlands. 3Department of 21. Schultz AB, Edington DW: Employee health and presenteeism: a
Epidemiology and Biostatistics, EMGO + Institute for Health and Care systematic review. J Occup Rehabil 2007, 17(3):547–579.
Research, VU University Medical Center, Amsterdam, The Netherlands. 22. Kessler RC, Ames M, Hymel PA, Loeppke R, McKenas DK, Richling DE, Stang
4
Department of Public and Occupational Health, EMGO + Institute for Health PE, Ustun TB: Using the world health organization health and work
and Care Research, VU University Medical Center, Amsterdam, The performance questionnaire (HPQ) to evaluate the indirect workplace
Netherlands. costs of illness. J Occup Environ Med 2004, 46(6):S23–S37.
23. Wanous JP, Reichers AE, Hudy MJ: Overall job satisfaction: How good are
Received: 5 March 2014 Accepted: 14 May 2014 single-item measures? J Appl Psychol 1997, 82(2):247–252.
Published: 27 May 2014 24. Nagy MS: Using a single-item approach to measure facet job satisfaction.
J Occup Organ Psychol 2002, 75:77–86.
References 25. Schaufeli WB, Bakker AB: Utrecht work engagement scale: Preliminary manual.
1. Campbell JP: Modeling the performance prediction problem in industrial Utrecht, The Netherlands: Occupational Health Psychology Unit, Utrecht
and organizational psychology. In Handbook of industrial and University; 2003. Report No.: Version 1.
organizational psychology. Volume 1. 2nd edition. Edited by Dunnette MD, 26. Ilmarinen J: The work ability index (WAI). Occup Med (Lond) 2007, 57:160.
Hough LM. Palo Alto, CA, US: Consulting Psychologists Press; 1990:687–755. 27. Ahlstrom L, Grimby-Ekman A, Hagberg M, Dellve L: The work ability index
2. Viswesvaran C, Ones DS: Perspectives on models of job performance. Int J and single-item question: associations with sick leave, symptoms, and
Sel Assess 2000, 8(4):216–226. health - a prospective study of women on long-term sick leave. Scand J
3. Rotundo M, Sackett PR: The relative importance of task, citizenship, and Work Environ Health 2010, 36(5):404–412.
counterproductive performance to global ratings of performance: a 28. Koppes LLJ, De Vroome EMM, Mol MEM, Janssen BJM, Van den Bossche
policy-capturing approach. J Appl Psychol 2002, 87(1):66–80. SNJ: The Netherlands working conditions survey. PlantijnCasparie, Almere: The
4. Koopmans L, Bernaards CB, Hildebrandt VH, Schaufeli WB, De Vet HCW, Van Netherlands Organisation for Applied Scientific Research; 2008.
der Beek AJ: Conceptual frameworks of individual work performance: a 29. Van Veldhoven M, Broersen S: Measurement quality and validity of the
systematic review. J Occup Environ Med 2011, 53(8):856–866. “need for recovery scale”. Occup Environ Med 2003, 60(Suppl 1):i3–i9.
5. Borman WC, Motowidlo SJ: Expanding the criterion domain to include 30. De Croon EM, Sluiter JK, Frings-Dresen MH: Psychometric properties of the
elements of contextual performance. In Personnel Selection in need for recovery after work scale: Test-retest reliability and sensitivity
Organizations. Edited by Schmitt N, Borman WC. San Francisco, CA: Jossey to detect change. Occup Environ Med 2006, 63(3):202–206.
Bass; 1993:71–98. 31. Van Veldhoven MJ, Sluiter JK: Work-related recovery opportunities: testing
6. Koopmans L, Bernaards CM, Hildebrandt VH, De Vet HCW, van der Beek AJ: scale properties and validity in relation to health. Int Arch Occup Environ
Measuring individual work performance: Identifying and selecting Health 2009, 82(9):1065–1075.
indicators. Work: J Prev, Assessment Rehabil. pre-press 26 June 2013. 32. Wendel-Vos GC, Schuit AJ, Saris WH, Kromhout D: Reproducibility and
7. Reilly MC, Zbrozek AS, Dukes EM: The validity and reproducibility of a relative validity of the short questionnaire to assess health-enhancing
work productivity and activity impairment instrument. physical activity. J Clin Epidemiol 2003, 56(12):1163–1169.
Pharmacoeconomics 1993, 4(5):353–365. 33. Wagenmakers R, van den Akker-Scheek I, Groothoff JW, Zijlstra W, Bulstra SK,
8. Lerner D, Amick BC, Rogers WH, Malspeis S, Bungay K, Cynn D: The work Kootstra JW, Wendel-Vos GCW, Van Raaij JJAM, Stevens M: Reliability and
limitations questionnaire. Med Care 2001, 39(1):72–85. validity of the short questionnaire to assess health-enhancing physical
9. Kessler RC, Barber C, Beck A, Berglund P, Cleary PD, McKenas D, Pronk N, activity (SQUASH) in patients after total hip arthroplasty. BMC Musculoskelet
Simon G, Stang P, Ustun TB, Wang P: The world health organization Disord 2008, 17(9):141.
Koopmans et al. BMC Public Health 2014, 14:513 Page 11 of 11
http://www.biomedcentral.com/1471-2458/14/513

34. Van der Zee KI, Sanderman R: Het meten van de algemene
gezondheidstoestand met de rand-36: Een handleiding. Groningen: Noordelijk
Centrum voor Gezondheidsvraagstukken, NCG; 2012. Report No.: 2nd ed.
35. Demerouti E, Bakker AB, Vardakou I, Kantas A: The convergent validity of
two burnout instruments: a multitrait-multimethod analysis. Eur J Psychol
Assess 2003, 19(1):12–23.
36. Demerouti E, Bakker AB: The oldenburg burnout inventory: A good alternative
to measure burnout and engagement. In Handbook of Stress and Burnout in
Health Care. Edited by Halbesleben JRB. Happauge, NY: Nova Science; 2008.
37. Cohen J: Statistical power analysis for the behavioral sciences. 2nd edition.
New Jersey: Lawrence Erlbaum Associates; 1988.
38. Judge TA, Bono JE, Thoresen CJ, Patton GK: The job satisfaction-job
performance relationship: a qualitative and quantitative review. Psychol
Bull 2001, 127(3):376–407.
39. Bakker AB, Demerouti E: Towards a model of work engagement. Career
Development International 2008, 13(3):209–223.
40. Van den Berg TIJ: The role of work ability and health on sustaining
employability. In PhD thesis. Rotterdam, The Netherlands: Erasmus University; 2010.
41. Harris MM, Schaubroeck J: A meta-analysis of self-supervisor, self-peer,
and peer-supervisor ratings. Pers Psychol 1988, 41:43–62.
42. Bycio P: Job performance and absenteeism: a review and meta-analysis.
Human Relations 1992, 45(2):193–220.
43. Morrow PC, McElroy JC, Laczniak KS, Fenton JB: Using absenteeism and
performance to predict employee turnover: early detection through
company records. J Vocat Behav 1999, 55:358–374.
44. Fritz C, Sonnentag S: Recovery, health, and job performance: effects of
weekend experiences. J Occup Health Psychol 2005, 10(3):187–199.
45. Pronk NP, Martinson B, Kessler RC, Beck AL, Simon GE, Wang P: The
association between work performance and physical activity,
cardiorespiratory fitness, and obesity. J Occup Environ Med 2004, 46(1):19–25.
46. Boles M, Pelletier B, Lynch W: The relationship between health risks and
work productivity. J Occup Environ Med 2004, 46(7):737–745.
47. Wright TA, Bonett DG, Sweeney DA: Mental health and work performance:
results of a longitudinal field study. J Occup Organ Psychol 1993, 66:277–284.
48. Wright TA, Cropanzano R: Emotional exhaustion as a predictor of job
performance and voluntary turnover. J Appl Psychol 1998, 83(3):486–493.
49. IBM Corp: IBM SPSS statistics for windows, version 20.0. ; 2011.
50. Rasch G: Probabalistic models for some intelligence and attainment tests.
Chicago: University of Chicago Press; 1960.
51. Podsakoff PM, MacKenzie SB, Lee J, Podsakoff NP: Common method biases
in behavioral research: a critical review of the literature and
recommended remedies. J Appl Psychol 2003, 88(5):879–903.
52. Van der Heijden BIJM, Nijhof AHJ: The value of subjectivity: Problems and
prospects for 36-degree appraisal systems. Int J Human Resour Manage
2004, 15(3):493–511.
53. Coffeng JK, Hendriksen IJM, Duijts SF, Twisk JWR, Van Mechelen W, Boot CRL:
Effectiveness of a combined social and physical environmental
intervention on presenteeism, absenteeism, work performance, and work
engagement in office employees. J Occup Environ Med 2014, 56(3):258–265.
54. Abma FI, Van der Klink JJL, Terwee CB, Amick BC III, Bultmann U:
Evaluation of the measurement properties of self-reported
health-related work-functioning instruments among workers with
common mental disorders. Scand J Work Environ Health 2012, 38(1):5–18.
55. Abma FI, van der Klink JJ, Bültmann U: The work role functioning questionnaire
2.0 (dutch version): Examination of its reliability, validity and responsiveness
in the general working population. J Occup Rehabil 2013, 23(1):135–147.
56. Demerouti E, Bakker AB, Demerouti E, Bakker AB: Employee well-being and
job performance: Where we stand and where we should go. In
Occupational Health Psychology: European Perspectives on Research, Education Submit your next manuscript to BioMed Central
and Practice. 1st edition. Edited by Houdmont J, McIntyre S. Maia: ISMAI and take full advantage of:
Publications; 2006.
57. Mokkink LB, Terwee CB, Gibbons E, Stratford PW, Alonso J, Patrick DL, Knol DL, • Convenient online submission
Bouter LM, De Vet HCW: Inter-rater agreement and reliability of the COSMIN
• Thorough peer review
(COnsensus-based standards for the selection of health status measurement
instruments) checklist. BMC Med Res Methodol 2010, 10(82):1–11. • No space constraints or color figure charges
• Immediate publication on acceptance
doi:10.1186/1471-2458-14-513
• Inclusion in PubMed, CAS, Scopus and Google Scholar
Cite this article as: Koopmans et al.: Responsiveness of the individual work
performance questionnaire. BMC Public Health 2014 14:513. • Research which is freely available for redistribution

Submit your manuscript at

www.biomedcentral.com/submit