Effect Modification by Coping Strategies on the Association of Organizational Justice with Psychological Distress in Japanese Workers

Abstract

Objectives: Several previous studies showed that the lack of organizational justice was associated with poor mental health. However, no study examined the effect modification by internal factor, such as coping strategies, on the association of organizational justice with mental health. The purpose of the present study was to investigate the effect modification by coping strategies on the association of organizational justice with psychological distress. Methods: A total of 471 men and 764 women from a manufacturing company in Japan completed self-administered questionnaires, including the Organizational Justice Questionnaire, K6 scale (i.e., psychological distress scale), Brief Scales for Coping Profile, and demographic characteristics. Multiple logistic regression analyses were conducted for each coping strategy. Results: After adjusting for demographic characteristics, the association of the lack of procedural justice with psychological distress was greater among the low changing a point of view group than among their counterparts. Furthermore, the interaction term of procedural justice with changing a point of view was significant. The association of the lack of procedural justice and interactional justice with psychological distress was also greater among the high emotional expression involving others group than among their counterparts, while the interaction terms of procedural justice and interactional justice with emotional expression involving others were marginally significant. Conclusions: Positive emotion-focused coping strategies, such as changing a point of view, may effectively prevent psychological distress when there is a lack of organizational justice, while problem-focused coping strategies may have no effects, and negative emotion-focused coping strategies, such as emotional expression involving others, may have harmful effects on the association.

(J Occup Health 2014; 56: 111-123)

Introduction

Organizational justice has recently been introduced as a psychosocial determinant of employee health¹⁾. Epidemiological research on the association of organizational justice with employee health has increased in the past decade²⁾. Organizational justice consists of three components: procedural justice, interactional justice, and distributive justice^3-6). Procedural justice refers to the capability to influence a decision- making process, influence the outcome³⁾, or adhere to fair process criteria, i.e., consistency, lack of bias, correctability, representation, accuracy, and ethicality⁴⁾. Interactional justice refers to whether managers or supervisors treat their subordinates with respect and dignity and provide rationales for their decisions⁵⁾. Distributive justice refers to outcomes that are consistent with implicit norms for allocation, such as equity or equality⁶⁾, which is considered to be a product of fair decision processes through procedural justice and interactional justice⁷⁾. Thus, procedural justice and interactional justice are the primary characteristics of organizational justice in a workplace or organization. In fact, previous research on occupational health has focused mainly on procedural justice and interactional justice⁸⁾.

Previous prospective studies in Europe (mainly in Finland, the UK, and the Netherlands) reported that lack of organizational justice was associated with poor mental health (i.e., major depressive disorders and self-reported psychiatric morbidity)^9-14). However, research on the factors that modify the association of organizational justice with mental health is still limited. Some previous studies in Europe and Japan examined the effect modification by external factors, such as job insecurity and employment contract, on the association of organizational justice with mental health^15-17). However, the effect modification by internal or individual factors, such as coping strategies, on the association of organizational justice with mental health has not been fully investigated yet, while some previous studies on the association of traditional psychosocial factors at work (e.g., job demands, job control, and interpersonal conflict) with mental health have examined the effect modification by coping strategies^18-20).

Coping strategies can be theoretically categorized into problem-focused coping strategies and emotion-focused coping strategies²¹⁾. Problem-focused coping strategies are directed at changing or eliminating a stressor by planning, taking direct action, or seeking assistance, whereas emotion-focused coping strategies are directed towards changing one's own emotional reaction to a stressor. In addition, emotion-focused coping strategies include two different types of strategies: positive (or adaptive) coping strategies and negative (or maladaptive) coping strategies. Positive coping strategies are venting one's emotional distress or cognitively reframing a stressor's influence by changing mood or changing a point of view. On the other hand, negative coping strategies are used to avoid stressful situations, suppress one's emotional distress or vent one's emotional distress by expressing one's negative feelings to others.

Previous studies have suggested that problem-focused coping strategies are effective in improving psychological and physical health^{22, 23)}. They appear to play a major role in effective coping, particularly if some degree of control over the situation is possible, whereas such efforts may be ineffective or harmful in the face of an uncontrollable stressor²⁴⁾. On the other hand, emotion-focused coping strategies may be effective in dealing with stressors in uncontrollable situations, although these effects might differ by different types of strategies²⁴⁾. Previous studies have reported that positive emotion-focused coping strategies are effective in reducing distress^{20, 25)}, while negative emotion-focused coping strategies are not necessarily effective in improving mental or physical health²⁶⁾. Given the effect of each coping strategy on mental health, positive emotion-focused coping strategies may have an effect on reducing psychological distress in relation to organizational (in)justice, while problem-focused and negative emotion-focused coping strategies may have no effects or harmful effect because organizational (in)justice captures work-related social context or fundamental social structure²⁷⁾, which can be considered single-handedly uncontrollable. However, no previous study has estimated the effect modification by coping strategies on the association of organizational justice with mental health.

The purpose of the present study was to investigate the effect modification by coping strategies on the association of organizational justice with psychological distress in Japanese employees. It was hypothesized that higher levels of positive emotion-focused coping strategies would buffer the association of lack of organizational justice with psychological distress and that higher levels of problem-focused and negative emotion-focused coping strategies would have no effects or harmful effect on the association.

Methods

Participants

A cross-sectional study of employees from five branches of a manufacturing listed company located in Japan was conducted in August 2009. All employees (N=1,279) were invited to participate in this study. A total of 1,277 questionnaires were returned in sealed envelopes. The data were collected using a self-administered questionnaire, which included scales on organizational justice, psychological distress, coping profile, and demographic characteristics. After excluding 42 employees who had at least one missing response for variables relevant to this study, 1,235 employees (471 men and 764 women) were analyzed. Study purpose and procedures were explained to the employees at the beginning, and written informed consent was obtained from the employees prior to initiation of the study. The Ethic Committee of the Graduate School of Medicine/Faculty of Medicine, The University of Tokyo reviewed and approved the aims and procedures of the study (No. 2580). Detailed characteristics of participants are shown in Table 1.

Table 1. Demographic and occupational characteristics, organizational justice, coping profile, and psychological distress of participating employees by gender

Demographic characteristics	Men (N=471)		Women (N=764)
	Average (SD)	n (%)	Average (SD)	n (%)
Age	41.7 (8.3)		34.0 (6.5)
50 years old or more		71 (15.1)		14 (1.8)
40–49 years old		237 (50.3)		130 (17.0)
30–39 years old		113 (24.0)		408 (53.4)
29 years old or less		50 (10.6)		212 (27.7)
Education	15.7 (1.7)		14.9 (1.5)
More than 12 years		384 (81.5)		695 (91.0)
12 years or less		87 (18.5)		69 (9.0)
Marital status
Currently married		369 (78.3)		263 (34.4)
Never married		82 (17.4)		446 (58.4)
Divorced/widowed		20 (4.2)		55 (7.2)
Occupation
Administrator/clerk		128 (27.2)		142 (18.6)
Quality assurance/after service		33 (7.0)		6 (0.8)
Sales support staff		2 (0.4)		274 (35.9)
Sales/sales engineer		272 (57.7)		53 (6.9)
Call talker		31 (6.6)		276 (36.1)
Others		5 (1.1)		13 (1.7)
Employment contract
Manager		203 (43.1)		18 (2.4)
Regular employee		216 (45.9)		166 (21.7)
Group businesses employee		23 (4.9)		12 (1.6)
Contract employee		19 (4.0)		252 (33.0)
Temporary employee		2 (0.4)		315 (41.2)
Others		8 (1.7)		1 (0.1)
Working hours in the past month	194.8 (27.2)		172.0 (25.9)
191 hours or more		210 (44.6)		115 (15.1)
173–190 hours		182 (38.6)		308 (40.3)
172 hours or less		79 (16.8)		341 (44.6)
Scale scores†	Average (SD)	Cronbach's α	Average (SD)	Cronbach's α
Organizational justice (OJQ)
Procedural justice (1–5)	3.56 (0.65)	0.86	3.40 (0.71)	0.88
Interactional justice (1–5)	4.09 (0.72)	0.93	3.87 (0.80)	0.94
Coping profile (BSCP)
Active solution (3–12)	9.90 (1.88)	0.77	9.52 (2.04)	0.76
Seeking help for solution (3–12)	8.65 (2.34)	0.84	8.94 (2.43)	0.81
Changing mood (3–12)	7.82 (2.39)	0.76	8.39 (2.52)	0.79
Emotional expression involving others (3–12)	4.09 (1.32)	0.63	4.17 (1.44)	0.68
Avoidance and suppression (3–12)	5.82 (2.04)	0.76	5.96 (2.01)	0.71
Changing a point of view (3–12)	8.04 (2.05)	0.72	8.29 (2.21)	0.75
Psychological distress (K6) (0–24)	4.73 (4.55)	0.89	5.30 (4.86)	0.89

^† OJQ, Organizational Justice Questionnaire; BSCP, Brief Scales for Coping Profile.

Measures

1) Organizational justice

Organizational justice was measured with the Japanese version of the Organizational Justice Scale^{7, 28, 29)}, sometimes known as the Organizational Justice Questionnaire (OJQ)³⁰⁾. The original OJQ was developed by Moorman⁷⁾, and its modified version²⁸⁾ has been widely used in research on organizational justice and health^{10-12, 14, 17, 31, 32)}. The OJQ consists of a seven-item scale assessing procedural justice and a six-item scale assessing interactional justice, both measured on a five-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). The former scale measures the degree of provision of relevant information to employees, the consistency of relevant information to employees, and the consistency of decision-making policy in the workplace. The latter scale measures the supervisor's degree of fairness and consideration of subordinates' needs and feelings. A total score for each OJQ subscale was calculated by averaging item scores. The modified English version of the OJQ was translated into Japanese, and the internal consistency reliability as well as factor and construct validities have been reported to be acceptable for this version²⁹⁾. In this sample, the Cronbach's alpha coefficient was 0.86 for procedural justice and 0.93 for interactional justice for male participants and 0.88 for procedural justice and 0.94 for interactional justice for female participants (Table 1). According to some previous studies^{9, 14, 17)}, the participants were classified into tertiles (high, moderate, and low) based on each OJQ subscale score.

2) Psychological distress

Psychological distress was measured with the Japanese version of the K6 scale^{33, 34)}. The K6 scale, developed by Kessler et al.³³⁾, consists of six items measured on a five-point Likert-type scale ranging from 0 (none of the time) to 4 (all of the time) (the range of the scale score, 0–24). In this sample, the Cronbach's alpha coefficients were 0.89 for both male and female participants. Participants were dichotomized into those with psychological distress (a total K6 score of 5 or more) and those without psychological distress (0–4 score) using a recommended cutoff point³⁵⁾.

3) Coping profile

Coping profile was measured with the Brief Scales for Coping Profile (BSCP). The BSCP, developed by Kageyama et al.³⁶⁾, comprises six subscales assessing coping strategies: active solution, seeking help for solution, changing mood, emotional expression involving others, avoidance and suppression, and changing a point of view. The active solution and seeking help for solution subscales refer to problem-focused coping strategies; the changing mood and changing a point of view subscales refer to positive (or adaptive) emotion-focused coping strategies; and the emotional expression involving others and avoidance and suppression subscales refer to negative (or maladaptive) emotion-focused coping strategies. Each subscale has three items measured on a four-point Likert-type scale ranging from 1 (never) to 4 (often) (the range of each subscale score, 3–12). The participants were dichotomized (high and low) based on median score for each subscale.

4) Other covariates

Other covariates included demographic characteristics (i.e., age, gender, education, and marital status) and occupational characteristics (i.e., occupation, employment contract, and working hours in the past month). Demographic characteristics were assessed using the self-administered questionnaire. Age was classified into four groups: 50 years old or more, 40–49 years old, 30–39 years old, and 29 years old or less. Education was dichotomized into some college or higher (i.e., more than 12 years) and senior high school or less (i.e., 12 years or less). Marital status was classified into three groups: currently married, never married, and divorced or widowed. Information on occupational characteristics was obtained from the personnel data of the company in which the survey was performed. Occupation was classified into six groups: administrator/clerk, quality assurance/after service worker, sales support staff (showroom sales staff), sales/sales engineer, call talker, and others. Employment contract was also classified into six groups: manager, regular employee, group businesses employee, contract employee, temporary employee, and others. Furthermore, participants were classified into three groups based on tertiles of working hours in the past month (i.e., 191 hours or more, 173–190 hours, and 172 hours or less).

Statistical analysis

Firstly, we calculated Pearson's correlation coefficients among gender, age, each justice dimension, each coping strategy, and psychological distress to determine if these variables are associated with each other. Secondly, using the high procedural justice or interactional justice group as a reference, a series of multiple logistic regression analyses were conducted to estimate the odds ratios (ORs) and 95% confidence intervals (CIs) of psychological distress (defined as having a score of five or more on the K6 scale) for the moderate or low procedural justice and interactional justice group according to the frequency (i.e., high and low) of each coping strategy. The analyses were adjusted first for demographic characteristics (i.e., age, education, and marital status) (Model 1) and then additionally for occupational characteristics (i.e., occupation, employment contract, and working hours in the past month) (Model 2). In a series of analyses, a linear trend test was also conducted to examine the dose-response relationship of procedural justice and interactional justice with psychological distress. Furthermore, we tested to determine the statistical significance of the interaction of procedural justice and interactional justice with each coping strategy if the association of procedural justice and interactional justice with psychological distress differed by frequency of each coping strategy. In addition, to examine gender difference in the interaction of each justice dimension with coping strategy, we tested the gender-interaction effect (i.e., three-way interaction term of each justice dimension × coping strategy × gender was included in the model). We also conducted alternative analyses including procedural justice or interactional justice as a continuous variable instead of a categorical variable. The level of significance was 0.05 (two-tailed). The statistical analyses were conducted using SPSS 17 for Windows.

Results

Pearson's correlation coefficients among the study variables are shown in Table 2. Active solution, seeking help for solution, changing mood, and changing a point of view were positively and significantly associated with procedural justice and interactional justice; and negatively and significantly associated with psychological distress. In contrast, emotional expression involving others and avoidance and suppression were negatively associated with procedural justice and interactional justice; and positively associated with psychological distress, while the association of avoidance and suppression with interactional justice was exceptionally insignificant.

Table 2. Pearson's correlation coefficients among the study variables (471 men and 764 women)

Variables†	1	2	3	4	5	6	7	8	9	10
1. Gender‡
2. Age	−0.46**
3. Procedural justice (OJQ) (1–5)	−0.11**	0.06*
4. Interactional justice (OJQ) (1–5)	−0.14**	−0.01	0.61**
5. Active solution (BSCP) (3–12)	−0.09**	0.02	0.13**	0.14**
6. Seeking help for solution (BSCP) (3–12)	0.06*	−0.12**	0.21**	0.18*	0.40**
7. Changing mood (BSCP) (3–12)	0.11**	−0.05	0.12**	0.06**	0.08**	0.17**
8. Emotional expression involving others (BSCP) (3–12)	0.03	−0.02	−0.13**	−0.13**	−0.08**	0.07*	0.07*
9. Avoidance and suppression (BSCP) (3–12)	0.03	−0.03	−0.10**	−0.05	−0.20**	−0.15**	0.07*	0.32**
10. Changing a point of view (BSCP) (3–12)	0.06*	0.02	0.20**	0.19**	0.25**	0.18**	0.34**	−0.03	0.12**
11. Psychological distress (K6) (0–24)	0.06*	−0.15**	−0.31**	−0.26**	−0.15**	−0.15**	−0.13**	0.19**	0.23**	−0.30**

* p<0.05; ** p<0.01. ^† OJQ, Organizational Justice Questionnaire; BSCP, Brief Scales for Coping Profile. ^‡ Men=0, women=1.

For procedural justice, after adjusting for demographic and occupational characteristics (Models 1 and 2), the low procedural justice group had significantly higher prevalence ORs of psychological distress than the high procedural justice group, regardless of any coping strategies (Table 3). For the interaction term, significant interactions of procedural justice with active solution and changing a point of view were observed (p for interaction <0.05). More specifically, among the moderate procedural justice group, the high active solution group had higher prevalence ORs of psychological distress than the low active solution group. Also, among the low procedural justice group, the low changing a point of view group had higher prevalence ORs of psychological distress than the high changing a point of view group. Furthermore, interaction of procedural justice with emotional expression involving others was also observed, although it was only marginally significant (p for interaction <0.10). More specifically, among the low procedural justice group, the high emotional expression involving others group had higher prevalence ORs of psychological distress than the low emotional expression involving others group.

Table 3. Association of procedural justice with psychological distress by coping strategies: logistic regression analysis^†

	n	No. of cases (%)	Odds ratio (95% confidence interval)
			Model 1‡	Model 2§
Procedural justice
High active solution group
High	156	32 (20.5)	1.00	1.00
Moderate	217	86 (39.6)	2.55 (1.56–4.16)	2.39 (1.45–3.96)
Low	124	62 (50.0)	3.82 (2.21–6.59)	3.76 (2.15–6.58)
Test for linear trend			p<0.001	p<0.001
Low active solution group
High	172	63 (36.6)	1.00	1.00
Moderate	341	145 (42.5)	1.26 (0.86–1.84)	1.23 (0.83–1.82)
Low	225	154 (68.4)	3.84 (2.50–5.90)	3.74 (2.41–5.80)
Test for linear trend			p<0.001	p<0.001
p for interaction (procedural justice × active solution)			p=0.023	p=0.037
High seeking help for solution group
High	176	50 (28.4)	1.00	1.00
Moderate	231	89 (38.5)	1.65 (1.06–2.56)	1.54 (0.98–2.42)
Low	117	64 (54.7)	3.15 (1.88–5.29)	3.08 (1.81–5.24)
Test for linear trend			p<0.001	p<0.001
Low seeking help for solution group
High	152	45 (29.6)	1.00	1.00
Moderate	327	142 (43.4)	1.73 (1.14–2.62)	1.70 (1.10–2.60)
Low	232	152 (65.5)	4.44 (2.83–6.97)	4.32 (2.73–6.85)
Test for linear trend			p<0.001	p<0.001
p for interaction (procedural justice × seeking help for solution)			p=0.425	p=0.441
High changing mood group
High	181	46 (25.4)	1.00	1.00
Moderate	259	98 (37.8)	1.80 (1.17–2.75)	1.71 (1.10–2.65)
Low	143	78 (54.5)	3.54 (2.19–5.73)	3.38 (2.07–5.52)
Test for linear trend			p<0.001	p<0.001
Low changing mood group
High	147	49 (33.3)	1.00	1.00
Moderate	299	133 (44.5)	1.54 (1.01–2.35)	1.52 (0.98–2.35)
Low	206	138 (67.0)	3.99 (2.50–6.37)	4.05 (2.50–6.55)
Test for linear trend			p<0.001	p<0.001
p for interaction (procedural justice × changing mood)			p=0.738	p=0.755
High emotional expression involving others group
High	88	33 (37.5)	1.00	1.00
Moderate	157	71 (45.2)	1.29 (0.74–2.24)	1.26 (0.71–2.23)
Low	135	99 (73.3)	4.82 (2.65–8.75)	4.66 (2.52–8.62)
Test for linear trend			p<0.001	p<0.001
Low emotional expression involving others group
High	240	62 (25.8)	1.00	1.00
Moderate	401	160 (39.9)	1.87 (1.31–2.67)	1.83 (1.27–2.63)
Low	214	117 (54.7)	3.35 (2.23–5.04)	3.34 (2.19–5.07)
Test for linear trend			p<0.001	p<0.001
p for interaction (interactional justice × emotional expression involving others)			p=0.063	p=0.096
High avoidance and suppression group
High	84	32 (38.1)	1.00	1.00
Moderate	182	94 (51.6)	1.89 (1.09–3.28)	1.84 (1.04–3.25)
Low	130	96 (73.8)	4.97 (2.69–9.16)	4.88 (2.61–9.12)
Test for linear trend			p<0.001	p<0.001
Low avoidance and suppression group
High	244	63 (25.8)	1.00	1.00
Moderate	376	137 (36.4)	1.59 (1.11–2.28)	1.53 (1.05–2.21)
Low	219	120 (54.8)	3.45 (2.30–5.17)	3.39 (2.24–5.11)
Test for linear trend			p<0.001	p<0.001
p for interaction (procedural justice × avoidance and suppression)			p=0.733	p=0.710
High changing a point of view group
High	181	45 (24.9)	1.00	1.00
Moderate	266	95 (35.7)	1.76 (1.14–2.72)	1.77 (1.14–2.76)
Low	129	58 (45.0)	2.57 (1.55–4.26)	2.65 (1.58–4.43)
Test for linear trend			p<0.001	p<0.001
Low changing a point of view group
High	147	50 (34.0)	1.00	1.00
Moderate	292	136 (46.6)	1.66 (1.09–2.53)	1.60 (1.04–2.47)
Low	220	158 (71.8)	5.06 (3.17–8.05)	5.00 (3.09–8.09)
Test for linear trend			p<0.001	p<0.001
p for interaction (procedural justice × changing a point of view)			p=0.034	p=0.041

^† Psychological distress was defined as having a score of five or more on the K6 scale. ^‡ Adjusted for age, gender, education, and marital status. ^§ Additionally adjusted for occupation, occupational class, and working hours in the past month.

As observed for procedural justice, the low interactional justice group had also significantly higher prevalence ORs of psychological distress than the high interactional justice group after adjusting for demographic and occupational characteristics (Models 1 and 2), regardless of any coping strategies (Table 4). For the interaction term, significant interaction of interactional justice with emotional expression involving others was observed after adjusting for demographic characteristics (Model 1) (p for interaction=0.038), while this interaction was attenuated and marginally significant after additionally adjusting for occupational characteristics (Model 2) (p for interaction=0.053). More specifically, among the low interactional justice group, the high emotional expression involving others group had higher prevalence ORs of psychological distress than the low emotional expression involving others group.

Table 4. Association of interactional justice with psychological distress by coping strategies: logistic regression analysis^†

	n	No. of cases (%)	Odds ratio (95% confidence interval)
			Model 1‡	Model 2§
Interactional justice
High active solution group
High	215	60 (27.9)	1.00	1.00
Moderate	144	49 (34.0)	1.35 (0.84–2.15)	1.26 (0.77–2.04)
Low	138	71 (51.4)	2.77 (1.73–4.43)	2.53 (1.56–4.11)
Test for linear trend			p<0.001	p<0.001
Low active solution group
High	210	78 (37.1)	1.00	1.00
Moderate	286	136 (47.6)	1.59 (1.10–2.30)	1.52 (1.05–2.22)
Low	242	148 (61.2)	2.79 (1.88–4.12)	2.69 (1.80–4.00)
Test for linear trend			p<0.001	p<0.001
p for interaction (interactional justice × active solution)			p=0.882	p=0.891
High seeking help for solution group
High	226	69 (30.5)	1.00	1.00
Moderate	166	61 (36.7)	1.38 (0.89–2.15)	1.30 (0.83–2.05)
Low	132	73 (55.3)	2.93 (1.83–4.67)	2.96 (1.82–4.81)
Test for linear trend			p<0.001	p<0.001
Low seeking help for solution group
High	199	69 (34.7)	1.00	1.00
Moderate	264	124 (47.0)	1.67 (1.14–2.45)	1.65 (1.11–2.45)
Low	248	146 (58.9)	2.69 (1.81–4.01)	2.54 (1.69–3.81)
Test for linear trend			p<0.001	p<0.001
p for interaction (interactional justice × seeking help for solution)			p=0.709	p=0.656
High changing mood group
High	220	66 (30.0)	1.00	1.00
Moderate	191	74 (38.7)	1.47 (0.97–2.23)	1.41 (0.92–2.17)
Low	172	82 (47.7)	2.13 (1.39–3.26)	2.01 (1.30–3.11)
Test for linear trend			p<0.001	p=0.002
Low changing mood group
High	205	72 (35.1)	1.00	1.00
Moderate	239	111 (46.4)	1.68 (1.12–2.50)	1.72 (1.14–2.59)
Low	208	137 (65.9)	3.83 (2.49–5.89)	3.94 (2.53–6.14)
Test for linear trend			p<0.001	p<0.001
p for interaction (interactional justice × changing mood)			p=0.157	p=0.155
High emotional expression involving others group
High	114	44 (38.6)	1.00	1.00
Moderate	129	61 (47.3)	1.49 (0.88–2.52)	1.48 (0.86–2.55)
Low	137	98 (71.5)	4.43 (2.55–7.71)	4.25 (2.40–7.52)
Test for linear trend			p<0.001	p<0.001
Low emotional expression involving others group
High	311	94 (30.2)	1.00	1.00
Moderate	301	124 (41.2)	1.64 (1.17–2.31)	1.63 (1.15–2.31)
Low	243	121 (49.8)	2.25 (1.57–3.24)	2.19 (1.51–3.18)
Test for linear trend			p<0.001	p<0.001
p for interaction (interactional justice × emotional expression involving others)			p=0.038	p=0.053
High avoidance and suppression group
High	130	55 (42.3)	1.00	1.00
Moderate	130	72 (55.4)	1.88 (1.12–3.16)	1.98 (1.17–3.37)
Low	136	95 (69.9)	3.73 (2.18–6.36)	3.74 (2.15–6.50)
Test for linear trend			p<0.001	p<0.001
Low avoidance and suppression group
High	295	83 (28.1)	1.00	1.00
Moderate	300	113 (37.7)	1.55 (1.09–2.20)	1.48 (1.03–2.12)
Low	244	124 (50.8)	2.59 (1.79–3.74)	2.47 (1.70–3.60)
Test for linear trend			p<0.001	p<0.001
p for interaction (interactional justice × avoidance and suppression)			p=0.710	p=0.610
High changing a point of view group
High	248	70 (28.2)	1.00	1.00
Moderate	172	60 (34.9)	1.37 (0.89–2.12)	1.36 (0.88–2.12)
Low	156	68 (43.6)	1.96 (1.26–3.04)	1.89 (1.20–2.97)
Test for linear trend			p=0.003	p=0.006
Low changing a point of view group
High	177	68 (38.4)	1.00	1.00
Moderate	258	125 (48.4)	1.59 (1.07–2.38)	1.57 (1.04–2.36)
Low	224	151 (67.4)	3.51 (2.28–5.39)	3.39 (2.18–5.25)
Test for linear trend			p<0.001	p<0.001
p for interaction (interactional justice × changing a point of view)			p=0.162	p=0.178

^† Psychological distress was defined as having a score of five or more on the K6 scale. ^‡ Adjusted for age, gender, education, and marital status. ^§ Additionally adjusted for occupation, occupational class, and working hours in the past month.

The results revealed non-significant interaction of procedural justice or interactional justice with seeking help for solution, changing mood, or avoidance and suppression (p for interaction>0.05).

When we tested the gender-interaction effect, only the interaction term of interactional justice × changing a point of view × gender was significant (p for interaction=0.014 and 0.021 for Model 1 and 2, respectively). Post-hoc gender-stratified analyses revealed significant interaction of interactional justice with changing a point of view only among women (p for interaction=0.015 and 0.024 for Model 1 and 2, respectively). More specifically, among the low interactional justice female group, the low changing a point of view group had higher prevalence ORs of psychological distress than the high changing a point of view group (data available upon request).

When we conducted alternative analyses including procedural justice or interactional justice as a continuous variable instead of as a categorical variable, a quite similar pattern was observed. However, the interaction of procedural justice with active solution was not significant (p for interaction=0.699 and 0.725 for Model 1 and 2, respectively), and the interaction of interactional justice with changing a point of view was significant or marginally significant (p for interaction=0.048 and 0.057 for Model 1 and 2, respectively) (data available upon request).

Discussion

In the present study, a significant buffering effect of changing a point of view on the association of the lack of procedural justice with psychological distress was observed. Gender-stratified analyses showed a significant buffering effect of changing a point of view on the association of the lack of interactional justice with psychological distress only among women. Furthermore, there was a marginally significant harmful effect of emotional expression involving others on the association of the lack of procedural justice and interactional justice with psychological distress. However, there was no buffering or harmful effect of seeking help for solution, changing mood, or avoidance and suppression on the association of the lack of procedural justice or interactional justice with psychological distress.

Pearson's correlation coefficients showed negative associations of problem-focused coping strategies (i.e., active solution and seeking help for solution) and positive (or adaptive) emotion-focused coping strategies (i.e., changing mood and changing a point of view) and positive associations of negative (or maladaptive) emotion-focused coping strategies (i.e., emotional expression involving others and avoidance and suppression) with psychological distress. These findings are consistent with the previous studies that found that problem-focused and positive (or adaptive) emotion-focused coping strategies are effective in improving psychological health or reducing distress^{22, 23, 25)} and that negative (or maladaptive) emotion-focused coping strategies are not necessarily effective in improving mental health²⁶⁾. Furthermore, the present study showed positive associations of problem-focused and positive (or adaptive) emotion-focused coping strategies and negative associations of negative (or maladaptive) emotion-focused coping strategies with both justice dimensions. These findings are also consistent with a previous study that found that problem-focused and positive (or adaptive) emotion-focused coping strategies are negatively associated with job stressors, which were defined by high job overload and poor human relations, and that negative (or maladaptive) emotion-focused coping strategies are positively associated with job stressors²⁰⁾. The present findings expanded this evidence into justice perception among a Japanese working population and suggest that frequent use of problem-focused and/or positive (or adaptive) emotion-focused coping strategies is effective but that of negative (or maladaptive) emotion-focused coping strategies are not necessarily effective in improving justice perception or mental health.

Among the moderate procedural justice group, the high active solution group had higher prevalence ORs of psychological distress than the low active solution group. When the decision-making procedure is moderately unfair or unclear, the high active solution group may try to improve or solve this unfair or unclear situation because it seems to be single-handedly controllable on the surface. However, as mentioned earlier, procedural justice captures work-related social context or fundamental social structure²⁷⁾, and it may be difficult for one person to change it even if the levels of unfairness or unclarity are moderate, which may cause psychological distress, especially among the high active solution group. On the other hand, among the low procedural justice group, the high active solution group had slightly higher but almost the same prevalence ORs of psychological distress as the low active solution group (i.e., ORs of 3.76 and 3.74 in Model 2, respectively), while these ORs were higher than those obtained for the moderate procedural justice group. When decision-making procedure becomes worst, employees may consider it to be absolutely uncontrollable; thus, they may become more psychologically distressed, regardless of their tendency to choose an active solution. However, it should be noted that the prevalence of psychological distress cases differed greatly between the high active solution group (20.5%) and low active solution group (36.6%) within the high procedural justice group, while quite similar prevalence of psychological distress cases emerged between the high active solution group (39.6%) and low active solution group (42.5%) within the moderate procedural justice group (see Table 3). These findings may lead to an apparently higher prevalence OR of psychological distress among the high active solution group than the low active solution group within the moderate procedural justice group. Furthermore, when we conducted alternative analyses including procedural justice as a continuous variable instead of as a categorical variable, the interaction of procedural justice with active solution was not significant. Therefore, considering the present findings comprehensively, the effect of active solution on the association of the lack of procedural justice with psychological distress may not necessarily be harmful, and interpretation of the present findings should be done with caution.

Among the low procedural justice group, the low changing a point of view group had higher prevalence ORs of psychological distress than the high changing a point of view group. Those who had a higher tendency to change a point of view may be likely to seek positive aspects of their own company and/or organization even if the decision-making procedure is unfair, which may have a preventive role in psychological distress in relation to the lack of procedural justice. Furthermore, when we conducted gender-stratified analyses, among the low interactional justice group, the low changing a point of view group had higher prevalence ORs of psychological distress than the high changing a point of view group only for women. A previous study reported that the effect of cognitive behavioral therapy, which is aimed at cognitive restructuring such as changing a point of view, on depressive symptoms was greater for women than for men³⁷⁾. Although further studies on gender differences in cognitive patterns are needed, women may be better at cognitive restructuring than men when they experience unfair treatment by their supervisors, which is more likely to cause chronic distress in the daily working life for employees than lack of procedural justice.

On the other hand, among the low procedural justice and interactional justice groups, the high emotional expression involving others group had higher prevalence ORs of psychological distress than the low emotional expression involving others group, while their interaction terms were only marginally significant. These findings support the literature indicating that negative (or maladaptive) emotion-focused coping strategies do not necessarily improve mental health²⁶⁾. Because the methods of emotional expression involving others include complaining to people who have nothing to do with the problem and blaming the person who caused the situation³⁶⁾, they may lead to interpersonal conflicts, which are associated with psychological distress at work³⁸⁾. The present study suggests that emotional expression involving others has a harmful effect on the association of organizational justice with psychological distress.

The results revealed no significant interaction of procedural justice or interactional justice with seeking help for solution. Even if employees have a higher tendency to seek help for a solution, it may be difficult for them to consult with their supervisor about organizational injustice because some injustice may be attributable to the supervisor him/herself, which may explain the present findings. In a similar way, the results showed no significant interaction of procedural justice or interactional justice with avoidance and suppression. On the other hand, among the low procedural and interactional justice groups, the high avoidance and suppression group had higher prevalence ORs of psychological distress than the low avoidance and suppression group. Since avoidance and suppression are thought to be negative (or maladaptive) coping strategies, they may have a harmful effect on the association of the lack of organizational justice with psychological distress, while this effect may be only modest compared with other kinds of negative coping strategy, such as emotional expression involving others. In addition, the results showed no significant interaction of procedural justice or interactional justice with changing mood. This finding does not support the literature indicating that positive emotion-focused coping strategies are effective in reducing distress^{20, 25)}. When employees experience unfair decision-making in their workplace and/or unfair treatment by their supervisors, changing mood, such as traveling and spending their free time on hobbies or entertainment, may not be effective for reducing psychological distress. With regard to emotion-focused coping strategies, the present findings rather suggest that a cognitive approach, such as changing a point of view, is more effective for reducing psychological distress among employees when they are facing organizational injustice.

Since the present study was conducted with Japanese employees, some parts of the present findings may reflect typical Japanese culture, such as vertical collectivism³⁹⁾. For example, from a viewpoint of collectivism, emotional expression involving others may disturb the workplace harmony, which may lead to greater harmful effect of emotional expression involving others than in Western countries. Furthermore, from a viewpoint of vertical society, Japanese employees may tend to hesitate to seek help for a solution from their supervisor, which may have lead to the non-significant buffering effect of seeking help for a solution in the present study. To investigate cultural differences in the effect modification by coping strategies on the association of organizational justice with psychological distress, similar studies should be performed in Western countries.

Some possible limitations of this study must be considered. First, the response rate was quite high compared with previous studies in Japan as well as in Western countries. Because supervisors collected each questionnaire in a sealed envelope, some respondents might have felt forced to participate in the study even though all employees were told that their participation was voluntary and that supervisors could not open the sealed envelopes. This process of collection of questionnaires may provide imprecise information or lack due consideration, which could result in underestimation of the association of organizational justice (especially of interactional justice) with psychological distress. Second, the present sample came from one particular manufacturing company in Japan, and all participants were white-collar workers, who could be considered a higher socioeconomic group. If we conduct a similar study with blue-collar workers or a lower socioeconomic group, who are reported to be more likely to face adverse psychosocial work environments⁴⁰⁾ and to use maladaptive coping strategies⁴¹⁾, different patterns in effect modification by coping strategies may be observed. Therefore, generalization of the present findings should be done with caution. Third, organizational justice, psychological distress, and coping profile were measured by self-report. Therefore, associations among the study variables may be overestimated due to a common response style, although several studies have shown that these influences are not as much as one would expect⁴²⁾. Fourth, low reliability was observed for each coping strategy subscale of the BSCP, which may have affected the present findings, while the Cronbach's alpha coefficients in the present study were similar to or higher compared with those in a previous study³⁶⁾. Fifth, those who had higher levels of psychological distress might be likely to choose negative (or maladaptive) coping strategies, which may mask the true association, as a previous study has reported that it is difficult for depressed patients to use positive (or adaptive) coping strategies⁴³⁾. Finally, causal inferences are limited due to the cross-sectional nature of the study. Regarding the association of organizational justice with psychological distress in particular, the present findings can also be interpreted as indicating that those who perceived higher levels of psychological distress might assess organizational justice as low. A prospective study is needed to investigate more precise effect modification by coping strategies on the association of organizational justice with psychological distress.

In conclusion, the present study suggests that positive (or adaptive) emotion-focused coping strategies, especially changing a point of view, have a buffering effect on the association of the lack of organizational justice with psychological distress and that problem-focused and negative (or maladaptive) emotion-focused coping strategies have no effects or harmful effect on the association. Our findings suggest that the choice of positive emotion-focused coping is effective for prevention of psychological distress when organizational justice is low. Furthermore, for employees who experience prolonged situations of organizational injustice, information about the choice of coping according to situations and training designed to improve coping skills may be important in an intervention program. However, it should be noted that an individual's effort might not solve the problems with lack of organization justice directly. Therefore, it may be important to promote improvement of the work environment, focusing on the enhancement of organizational justice, as well as to improve the coping skills of individuals. Further research should investigate other indicators that may affect the coping-stress relationship.

Acknowledgment: The present study was supported by a Grant-in-Aid for Scientific Research (A) 2009 (No. 20240062) from the Ministry of Education, Culture, Sports, Science and Technology, Japan. Preparation of the manuscript was partially supported by a Grant-in-Aid for Young Scientists (B) 2013 (No. 24790599) from the Ministry of Education, Culture, Sports, Science and Technology, Japan.

References

• 1)   Bourbonnais  R. Are job stress models capturing important dimensions of the psychosocial work environment? Occup Environ Med 2007; 64: 640-1.
• 2)   Fujishiro  K,  Heaney  CA. Justice at work, job stress, and employee health. Health Educ Behav 2009; 36: 487-504.
• 3)   Thibaut  J,  Walker  L. Procedural justice: a psychological analysis. Hillsdale (NJ): Erlbaum; 1975.
• 4)   Leventhal  GS. What should be done with equity theory? New approaches to the study of fairness in social relationships. In:  Gergen  K,  Greenberg  M,  Willis  R, editors. Social exchange: advances in theory and research. New York (NY): Plenum Press; 1980. p.27-55.
• 5)   Bies  RJ,  Moag  JS. Interactional justice: communication criteria of fairness. In:  Lewicki  RJ,  Sheppard  BH,  Bazerman  MH, editors. Research on negotiation in organizations, vol 1. Greenwich (CT): JAI Press; 1986. p.43-55.
• 6)   Adams  JS. Inequity in social exchange. In:  Berkowits  L, editor. Advances in experimental social psychology, vol 2. New York (NY): Academic Press; 1965. p.267-99.
• 7)   Moorman  RH. Relationship between organizational justice and organizational citizenship behaviors: Do fairness perceptions influence employee citizenship? J Appl Psychol 1991; 76: 845-55.
• 8)   Kawachi  I. Injustice at work and health: causation or correlation? Occup Environ Med 2006; 63: 578-9.
• 9)   Ferrie  JE,  Head  J,  Shipley  MJ,  Vahtera  J,  Marmot  MG,  Kivimäki  M. Injustice at work and incidence of psychiatric morbidity: the Whitehall II study. Occup Environ Med 2006; 63: 443-50.
• 10)   Kivimäki  M,  Elovainio  M,  Vahtera  J,  Ferrie  JE. Organizational justice and health of employees: prospective cohort study. Occup Environ Med 2003; 60: 27-34.
• 11)   Kivimäki  M,  Elovainio  M,  Vahtera  J,  Virtanen  M,  Stansfeld  SA. Association between organizational inequity and incidence of psychiatric disorders in female employees. Psychol Med 2003; 33: 319-26.
• 12)   Kivimäki  M,  Vahtera  J,  Elovainio  M,  Virtanen  M,  Siegrist  J. Effort-reward imbalance, procedural injustice and relational injustice as psychosocial predictors of health: complementary or redundant models? Occup Environ Med 2007; 64: 659-65.
• 13)   Ybema  JF,  van den Bos  K. Effects of organizational justice on depressive symptoms and sickness absence: a longitudinal perspective. Soc Sci Med 2010; 70: 1609-17.
• 14)   Ylipaavalniemi  J,  Kivimäki  M,  Elovainio  M,  Virtanen  M,  Keltikangas-Järvinen  L,  Vahtera  J. Psychosocial work characteristics and incidence of newly diagnosed depression: a prospective cohort study of three different models. Soc Sci Med 2005; 61: 111-22.
• 15)   Elovainio  M,  van den Bos  K,  Linna  A, et al. Combined effects of uncertainty and organizational justice on employee health: testing the uncertainty management model of fairness judgments among Finnish public sector employees. Soc Sci Med 2005; 61: 2501-12.
• 16)   Kausto  J,  Elo  AL,  Lipponen  J,  Elovainio  M. Moderating effects of job insecurity in the relationships between procedural justice and employee well-being: gender differences. Eur J Work Organ Psychol 2005; 14: 431-52.
• 17)   Inoue  A,  Kawakami  N,  Tsuno  K,  Tomioka  K,  Nakanishi  M. Organizational justice and psychological distress among permanent and non-permanent employees in Japan: a prospective cohort study. Int J Behav Med 2013; 20: 265-76.
• 18)   Parkes  KR. Coping, negative affectivity, and the work environment: additive and interactive predictors of mental health. J Appl Psychol 1990; 75: 399-409.
• 19)   Koeske  GF,  Kirk  SA,  Koeske  RD. Coping with job stress: which strategies work best? J Occup Organ Psychol 1993; 66: 319-35.
• 20)   Takaya  M,  Hasegawa  Y. An analysis of the relation between stress coping profile and job stress using Structural Equation Modeling (SEM). Sangyo Eiseigaku Zasshi 2010; 52: 209-15 (in Japanese).
• 21)   Lazarus  RS,  Folkman  S. Stress Appraisal and Coping. New York (NY): Springer; 1984.
• 22)   De Rijk  AE,  Le Blanc  PM,  Schaufeli  WB,  De Jonge  J. Active coping and need for control as moderators of the job demand-control model: effects on burnout. J Occup Organ Psychol 1998; 71: 1-18.
• 23)   Panley  JA,  Tomaka  J,  Wiebe  JS. The association of coping to physical and psychological health outcomes: a meta-analytic review. J Behav Med 2002; 25: 551-603.
• 24)   Collins  DL,  Baum  A,  Singer  JE. Coping with chronic stress at Three Mile Island: psychological and biochemical evidence. Health Psychol 1983; 2: 149-66.
• 25)   Fennell  MJ,  Teasdale  JD. Effects of distraction on thinking and affect in depressed patients. Br J Clin Psychol 1984; 23: 65-6.
• 26)   Eileen  KM,  Jeanne  CW. Expressing Emotion. New York (NY): the Guilford Press; 1999.
• 27)   van den Bos  K,  Lind  EA. Uncertainty management by means of fairness judgments. In:  Zanna  MP, editor. Advances in experimental social psychology. San Diego (CA): Academic Press; 2002. p. 1-60.
• 28)   Elovainio  M,  Kivimäki  M,  Vahtera  J. Organizational justice: evidence of a new psychosocial predictor of health. Am J Public Health 2002; 92: 105-8.
• 29)   Inoue  A,  Kawakami  N,  Tsutsumi  A, et al. Reliability and validity of the Japanese version of the Organizational Justice Questionnaire. J Occup Health 2009; 51: 74-83.
• 30)   Sutinen  R,  Kivimäki  M,  Elovainio  M,  Forma  P. Associations between stress at work and attitudes towards retirement in hospital physicians. Work Stress 2005; 19: 177-85.
• 31)   Inoue  A,  Kawakami  N,  Ishizaki  M, et al. Organizational justice, psychological distress, and work engagement in Japanese workers. Int Arch Occup Environ Health 2010; 83: 29-38.
• 32)   Inoue  A,  Kawakami  N,  Tsuno  K,  Tomioka  K,  Nakanishi  M. Organizational justice and major depressive episodes in Japanese employees: a cross-sectional study. J Occup Health 2013; 55: 47-55.
• 33)   Kessler  RC,  Andrews  G,  Colpe  LJ, et al. Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychol Med 2002; 32: 959-76.
• 34)   Furukawa  TA,  Kawakami  N,  Saitoh  M, et al. The performance of the Japanese version of the K6 and K10 in the World Mental Health Survey Japan. Int J Methods Psychiatr Res 2008; 17: 152-8.
• 35)   Sakurai  K,  Nishi  A,  Kondo  K,  Yanagida  K,  Kawakami  N. Screening performance of K6/K10 and other screening instruments for mood and anxiety disorders in Japan. Psychiatry Clin Neurosci 2011; 65: 434-41.
• 36)   Kageyama  T,  Kobayashi  T,  Kawashima  M,  Kanamaru  Y. Development of the Brief Scales for Coping Profile (BSCP) for workers: basic information about its reliability and validity. San Ei Shi 2004; 46: 103-14 (in Japanese).
• 37)   Pieh  C,  Altmeppen  J,  Neumeier  S,  Loew  T,  Angerer  M,  Lahmann  C. Gender differences in response to CBT-orientated multimodal treatment in depressed patients with chronic pain. Psychiatr Prax 2012; 39: 280-5 (in German).
• 38)   Tsuno  K,  Kawakami  N,  Inoue  A, et al. Intragroup and intergroup conflict at work, psychological distress, and work engagement in a sample of employees in Japan. Ind Health 2009; 47: 640-8.
• 39)   Nakane  C. Japanese society. Berkeley (CA): University of Calfornia Press; 1970.
• 40)   Siegrist  J,  Marmot  M. Health inequalities and the psychosocial environment-two scientific challenges. Soc Sci Med 2004; 58: 1463-73.
• 41)   Haan  N. Coping and defending: processes of self-environment organization. New York (NY): Academic Press; 1977.
• 42)   Lachman  ME,  Weaver  SL. Sociodemographic variations in the sense of control by domein: findings from the MacArthur studies of midlife. Psychol Aging 1998; 13: 553-62.
• 43)   Edwards  JR. To prosper, organizational psychology should…overcome methodological barriers to progress. J Organ Behav 2008; 29: 469-91.
• 44)   Maximos  MHF,  Ramadan  FH,  Naeem  MM. Relationship between coping styles and level of depression among depressed patients. J Am Sci 2011; 7: 322-35.