Within-Person Dynamics of Job Boredom and Counterproductive Work Behavior: A Latent Change Score Modeling Approach

Job boredom is one of the most common negative affective states experienced in the workplace, yet also among the least well-understood. One stream of research suggests that employees frequently react to job boredom by engaging in counterproductive work behaviors (CWB). However, recent studies show the converse—that engaging in CWB relates to job boredom. As studies on the job boredom-CWB relationship primarily have been cross-sectional and at the between-person level of analysis, the directionality between these constructs remains in question. Therefore, research examining the within-person dynamics of job boredom and CWB within a short timeframe is needed. In the current study, we explore whether job boredom in�uences subsequent changes in CWB and vice versa. We examined these relationships using latent change score (LCS) modeling with 10-day experience sampling data (N = 120 individuals providing 1,161 observations). Findings supported a reciprocal relationship. Employees’ level of job boredom on a given day was associated with a subsequent increase in CWB on the next day, and the level of CWB on a given day was associated with a subsequent increase in job boredom on the next day. We discuss the implications of our �ndings and future research directions.


Introduction
Job boredom can be de ned as an "unpleasant, transient affective state in which the individual feels a pervasive lack of interest in and di culty concentrating on the current activity" (Fisher, 1993, p. 397).Job boredom is one of the most prevalent affective states in the workplace and relates to various harmful outcomes such as job dissatisfaction, greater turnover intentions, and mental health challenges (Fisher, 1993;Mael & Jex, 2015).Yet, job boredom also remains among the least understood forms of affect at work (Fisher, 1993;Loukidou et al., 2009).
One stream of research links job boredom to increased counterproductive work behavior (CWB; e.g., Bruursema et al., 2011;Kim et al., 2021).CWBs are voluntary behaviors detrimental to the well-being of employees and the organization (Dalal et al., 2009).However, these studies primarily have been crosssectional and at the between-person level (for an exception, see Spanouli et al., 2023), failing to yield insight into within-person dynamics of job boredom and CWB.Consequently, research is unclear about the directionality of the job boredom-CWB relationship at the within-person level.Here, we address this issue by exploring within-person reciprocal relationships between job boredom and CWB.
Research demonstrating a positive linkage between job boredom and CWB (e.g., Bruursema et al., 2011;Kim et al., 2021) mainly is based on the notion that job boredom precedes CWB.Spector and Fox's (2002) model of voluntary work behaviors, for instance, posits that negative affect leads to CWB.Consistent with this notion, ndings from a daily diary study (Spanouli et al., 2023) demonstrated the lagged effect of job boredom on CWB at the within-person level.Thus, we expect job boredom to relate positively to next-day increases in CWB (Hypothesis 1).
However, the reverse directionality may be plausible too: CWB may foster job boredom.Indirectly supporting this notion, researchers have found that CWB tends to trigger other aversive states, including anger, frustration, guilt, and shame (e.g., Ilies et al., 2013;Spector & Fox, 2010).This said, boredom may not function in the same manner as these other emotions.Functional theories of boredom posit that it serves a regulatory mechanism, signaling to individuals that they should pursue other tasks/activities that are more interesting or enjoyable than the current one (Westgate & Wilson, 2018).According to these accounts, engaging in CWB (e.g., cyberloa ng; Pindek et al., 2018) to reduce boredom should be functional, resulting in a lower level of subsequent boredom.In sum, given the con icting notions about the sign of this relationship, we do not offer a hypothesis, examining it in an exploratory manner.

Method
Participants and Procedure Initially, we recruited 165 individuals who satis ed our eligibility criteria in the initial baseline survey.Of these, we excluded 12 participants who failed one or more of the three attention check items.We also omitted 33 participants who did not complete a minimum of two full days of daily surveys (i.e., four daily surveys).The nal sample size was 120 individuals providing 1,161 observations (response rate = 92.49%).On average, participants were 32.93 years old (SD = 3.27), 32.03% of whom were female.They worked for an average of 3.39 years in their current jobs (SD = 0.58).A majority of participants had obtained a bachelor's degree (55.56%) and were either Black/African American (58.17%) or White (40.52%).
We recruited full-time employees by posting an advertisement on Craigslist, an online community website (see https://www.craigslist.org/about) in multiple large cities in the United States (see Judge et al., 2014;Vogel et al., 2020, for examples using Craigslist).The advertisement contained information about the study procedures, eligibility criteria, and compensation, and was posted for approximately two weeks in June 2022.
Interested participants were requested to email the researchers and were subsequently invited to complete an initial baseline survey, which included informed consent and measures of demographic variables.In this survey, participants were rst screened based on three eligibility criteria: (a) age (i.e., 18 + years old), (b) full-time employment status (i.e., 30 + work hours per week), and (c) work schedules (i.e., whether they had a traditional work schedule from 9:00 AM to 5:00 PM), the last of which was employed to ease scheduling of daily surveys.At the end of the baseline survey, we employed a commitment device that asked participants to explicitly commit to completing a minimum of 75% of the daily surveys (Gabriel et al., 2019).If respondents failed to commit to completing 75% or more of daily surveys, we kindly asked them to consider increasing their commitment to maximize the utility of their effort in participating in the research study.As a result, all participants chose to commit 75% or more of the daily surveys.
For 10 consecutive workdays (i.e., two consecutive workweeks), participants received two brief surveys per day, the rst survey at 10:00 a.m.(Time 1 or T1) and the second survey at 1:00 p.m. (Time 2 or T2).In both surveys, we assessed momentary measures of job boredom and CWB.In exchange for participation, we provided a maximum possible amount of 55.00U.S. dollars in the form of an Amazon gift card.
To ensure data quality, we followed best practice recommendations for convenience sampling.
Speci cally, we implemented such strategies as posting the advertisement in major metropolitan areas (Antoun et al., 2016), including attention check items in the initial survey (e.g., "Please select Never for this item.";Meade & Craig, 2012), and tracking unique identi ers to assure only one survey was completed per participant (Mason & Suri, 2012).Additionally, to ensure that our study had su cient statistical power, we followed Gabriel et al.'s (2019) sample size recommendations, which were based on their calculation of the mean sample size of the past experience sampling studies.Accordingly, we aimed to obtain a minimum Level 2 sample size of 83 and a minimum Level 1 sample size of 835 by administering three daily surveys for 10 consecutive workdays.This strategy of sampling for two weeks allowed us to capture a generally representative sample of a person's daily experiences (Wheeler & Reis, 1991).

Measures Job Boredom
Momentary job boredom was assessed using a 4-item measure (Park et al., 2019).On a 5-point Likert scale (1 = not at all to 5 = extremely), participants were asked to rate the extent to which they were feeling each of the following four emotion adjectives at the current moment: "bored," "sluggish," "dull," and "lethargic."We aggregated momentary job boredom assessed in T1 and T2 up to a morning measure of job boredom.Across the days of data collection, the average Cronbach's alpha coe cient was .99.

Counterproductive Work Behavior
CWB was measured with 2 items from Dalal et al. (2009).Although the full scale originally contained 6 items, we only included 2 of them that may be likely to have greater within-person variability during one's typical workday (Koopman et al., 2021).On a dichotomous scale (1 = yes, 0 = no), we asked participants to rate each item while thinking about the last two hours at work.Example items include "Did not work to the best of my ability" and "Spent time on tasks unrelated to work."Across the days of data collection, the average Cronbach's alpha coe cient was .91.

Data Analysis
To study within-person dynamics between job boredom and CWB, we used latent change score (LCS) modeling (see Matusik et al., 2021, for a review).In an LCS model (see Fig. 1), we used 10 daily measurements of both job boredom and CWB and speci ed constant change effect (i.e., whether and to what extent there is an overall increasing or decreasing trend for a given variable), proportional change effect (i.e., whether and to what extent a level of the variable in uences subsequent change in that same variable), and autoregression of change scores for both boredom and CWB.Importantly, we speci ed a lagged coupling effect of job boredom on change in CWB, as well as lagged coupling effect of CWB on change in job boredom.The lagged coupling effects were the parameters of primary interest in this study.For instance, the lagged coupling effect of job boredom on change in CWB indicates whether and to what extent job boredom at Day i -1 predicts increases (indicated by a signi cant, positive effect) or decreases (indicated by a signi cant, negative effect) in CWB between Day i -1 and Day i.As is common when using LCS models, all estimates were assumed to be equal across time points (Matusik et al., 2021).The LCS model showed acceptable t (CFI = .90,TLI = .91,RMSEA = .09,SRMR = .10).

Results
We rst report results regarding constant change (i.e., µ) to examine whether there was a signi cant overall trend, and then ndings regarding lagged coupling effects (i.e., γ).As reported in Table 1, results from LCS modeling demonstrated that, regarding job boredom, the constant change estimate was not statistically signi cant (µ = − 1.36, p = .289).The nonsigni cant constant change estimate indicates that there was no general increasing or decreasing trend in job boredom over the course of 10 workdays.Regarding CWB, ndings demonstrated that the constant change effect was signi cant and positive (µ = 0.78, p = .030),indicating that there was a general increasing trend in CWB engagement over the course of 10 workdays.We next report ndings regarding lagged coupling effects.Results showed the lagged coupling effect of job boredom on subsequent changes in CWB was signi cant and positive (γ = 0.32, p = .039).That is, true score values for job boredom at Day i -1 were positively associated with change score values for CWB at Day i.Consistent with our prediction, this suggests that employees who experience greater job boredom on a given day are likely to engage in more CWB on the next day, compared to those who are less bored.Therefore, ndings provide support for Hypothesis 1.
Regarding the reverse-direction relationship, ndings also demonstrated that the lagged coupling effect of CWB on subsequent changes in job boredom was signi cant and positive (γ = 2.95, p = .020).That is, true score values for CWB at Day i -1 were positively associated with change score values for job boredom at Day i.This suggests that employees who engage in more CWB on a given day are likely to experience greater job boredom on the next day, compared to those who engage in less CWB.

Discussion
In this study, we examined within-person changes in job boredom and CWB over a short timeframe.
Using LCS modeling, we found that reciprocal within-person relationships exist between job boredom and CWB.Consistent with our prediction, we found that job boredom related positively to next-day CWB change.We also found a positive relationship between CWB and next-day boredom change, suggesting that CWB enactment may increase boredom rather than reducing it.

Consent to Participate
Informed consent was obtained from all participants.

Figures
Figures

Figure 1 A
Figure 1

Table 1
Parameter Estimates for a Latent Change Score Model with Job Boredom and Counterproductive Work Behavior CWB = counterproductive work behavior; γ JB.CWB = latent change in JB between Day i -1 and Day i determined by latent true score of CWB at Day i -1; γ CWB.JB = latent change in CWB between Day i -1 and Day i determined by latent true score of JB at Day i -1.
Although this latter nding is inconsistent with the theorized regulatory function of boredom, it is consistent with results showing that CWB engagement leads to other negative affective states (e.g., shame, guilt, frustration, anger; Ilies et al., 2013; Spector & Fox, 2010).Our results extend the literature on job boredom and CWB by providing empirical evidence for withinperson change in job boredom and CWB over a short timeframe.Most studies on job boredom have used cross-sectional designs and are at the between-person level of analysis.Using cross-lagged data in the experience sampling framework, the current study provides evidence clarifying the directionality of the within-person relationship between job boredom and CWB.Regarding practical implications, our ndings suggest that HR practitioners should be cognizant that CWB enactment may be generative, not reparative, of job boredom.As recent studies show bene cial effects of some CWB forms on employee well-being, such as cyberloa ng (e.g., Lim & Chen, 2012), a reasonable inference is that such activities may mitigate boredom.However, the current results suggest that the opportunity to engage in such activities could instead foster (more) boredom.In this sense, those activities may trigger the outcomes they partly are meant to avoid.In general, more research is needed on organizational policies and interventions meant to mitigate job boredom.