Are job satisfaction and role breadth self-efficacy the links to proactive work behavior?

This research is based on a phenomenon that occurs in State Detention Centers in Indonesia. It attempts to test the relation among proactive personality (PP), proactive work behavior (PWB), job satisfaction (JS) and role breadth self-efficacy (RBSE) variables. With a quantitative approach using AMOS, this study took data from 455 respondents from Detention Centers in Indonesia. The results show that PP, JS and RBSE have direct effect on PWB. Further findings will be discussed. The results are expected to increase the understanding of PWB and can be the basis for the human resource management team to decide better approach to build PWB in the organization and eventually implement appropriate policy.

Introduction

A State Detention Center is a facility where a suspect is held while the case is investigated, prosecuted, and examined in court. It is an institution that provides services to community members who are incarcerated in cells while also preserving security and order. Currently, the average population of State Detention Center throughout Indonesia is overcrowded which increases the urgency of PP in the workplace since it is an integral part of the maintaining order and security process. This process includes providing protection, prevention and prosecution against any threats and interference from outside the State Detention Center.

Increased order and security disturbances and the lack of quality of detention officers have widened the gap in State Detention Centers throughout Indonesia. In order to answer these challenges, an effective human resource management system is needed to ensure that the organization can carry out its duties through human resources who are motivated, proactive, professional and high-performing towards the tasks being carried out. In order to build and sustain PWB in Detention Centers, integrative, collaborative, and motivating action toward tasks by adopting a larger role is required (Parker et al., 2006). Additionally, Parker (1998) stated that employees with RBSE appear to possess these kinds of actions and motivations.

Moreover, based on Judge (1993), individuals also need an affective disposition related to JS to increase proactive action in the workplace. As a result, the management of State Detention Centers must encourage those with PP to carry out responsibilities at State Detention Centers, either directly to PWB or mediated by RBSE and JS.

Literature review

Proactive personality

Individuals with PP are capable of considering all potential risks and chances (Parker and Collins, 2010). This personality tends to have a stable position, and is not easily affected by situational challenges and environmental changes (Bateman and Crant, 1993). According to Akgunduz et al. (2018) those with PP are usually goals-oriented and have initiative to seize potential chances. Thus, they are able to trigger changes within themselves as well as their environment (Presbitero, 2015).

Proactive work behavior

This behavior generally involves challenging the status quo (Crant, 2000), by taking the initiative to improve current circumstances or create new things, rather than passively adapt to current conditions. It entails actions which are self-directed and future-oriented to change or improve oneself and situations (Unsworth and Parker, 2008). Individuals that exhibit this behavior can go above and beyond the specified work, establish targets, and take a long-term approach to avoid conflicts (Frese and Fay, 2001). According to Bateman & Crant (1993) and Buss (1987), PWB is able to transform conditions, restructure the mindset, and change the existing state, both social and non-social. To highlight, proactive idea enactment and problem solving are two traits that receive the most attention (Parker et al., 2006).

Job satisfaction

Employees who are satisfied in their work are the source of job happiness (Luthans, 2011 in Diana et al., 2020). Methodologically, JS is employees ‘affective response to their work, which is a contrast between real and intended results, according to Mosadegh Rad A.M (2003). As a result, there may be both satisfaction and dissatisfaction with what is done (Eliyana and Sridadi, 2020). Employment satisfaction may also be the outcome of a positive evaluation of one's job or work experience (Locke, 1969).

Role breadth self-efficacy

People's opinion of their own capacity to do specific tasks is referred to as self-efficacy (Gist and Mitchell, 1992). It is particularly the perceived capacity to engage in a variety of proactive, social, and blended activities, beyond assigned duties (Parker, 1998). In other terms, self-efficacy that is significant as a job motivation factor is referred to as RBSE (Gist and Mitchell, 1992; Parker, 1998). RBSE refers to people's confidence in their ability to take on a larger and more proactive role (Parker et al., 2006). They are confident in their ability to complete a set of activities and are driven to do so whether or not they are permitted (Parker, 1998; Bandura, 1982, 2986).

Proactive personality on proactive work behavior

It is a personality that sees possibilities and seizes them, takes initiative, acts, and perseveres until significant changes take place (Crant, 2000). This personality is also not limited by situational forces (Parker and Collins, 2010), meaning that this internal control may lead to behavior. Some literature in psychology as well as organizational behavior have also stated that PWB can be controlled both internally or externally (Schneider, 1983 in Bateman and Crant, 1993). Therefore, it can be concluded that PP can influence PWB (Parker et al., 2006). Similar researches were also conducted by McCormick et al. (2019) and Wu et al. (2018) which stated that PP has a significant effect on PWB. The nature of the PP will create change and control that can support PWB in the workplace regardless of the work context because of its natural tendency to be a self-starter and take initiative (McCormick et al., 2019).

Proactive Personality significantly influences Proactive Work Behavior

Proactive personality on job satisfaction

Individuals with proactive personalities have a steady stance that is not impacted by environmental changes and is not constrained by situational pressures (Bateman and Crant, 1993). This is one aspect of personality traits that leads to JS (N. Li and Crant, 2010; Wang, 2010). An individual with a PP will be more likely to take action to alter and attain an ideal self or circumstance, and will potentially lead to better JS over time due to three key characteristics: self-initiative, change orientation, and future focus (Kuo et al., 2019). Because proactive people prefer to create situations that are favourable to personal work achievement, PP is linked to JS (N. Li and Crant, 2010).

Proactive Personality significantly influences Job Satisfaction

Proactive personality on role breadth self-efficacy

PP is a character that can initiate and act on it, in other terms, it has a more flexible role-taking style. It also has a strong drive to change, that is comparable to mastery or control (Bateman and Crant, 1993). Moreover, individuals with RBSE believe they are able to perform a larger role in the workplace (Parker, 1998). This notion, in addition to the source of motivation, must be impacted by one's talents (Parker et al., 2006). For example, the ability to have a more adaptable role orientation as well as control (Parker and Sprigg, 1999). According to Andri et al. (2020), PP is connected to RBSE in a substantial way. When people with PP feel they can effectively start an organization, they are more inclined to do so (Travis and Freeman, 2017).

Proactive Personality significantly influences Role Breadth Self-Efficacy

Job satisfaction on proactive work behavior

According to Strauss et al. (2013), JS is a resource that allows individuals to continue the efforts needed to maintain proactive action. However, based on (Parker et al., 2006), it is a construction that is more closely related to compliance than with proactivity. In particular, through proactive goal setting and achievement, individuals tend to fulfil their needs when they are satisfied with their work (Weigelt et al., 2019). Individuals who experience positive affective states associated with JS tend to change their situation proactively (Judge, 1993) and to exhibit higher levels of innovative behavior (George and George, 1990).

Job Satisfaction significantly influences Proactive Work Behavior

Role breadth self-efficacy on proactive work behavior

Individuals who believe they are capable of performing a task are more likely to do it efficiently (Barling and Beattie, 1983). RBSE refers to employees’ belief on their ability to engage in proactive, social, and integrative activities beyond their standardized tasks (Parker, 1998). RBSE can inspire each individual to believe that they can perform a broader and more proactive role, one that goes beyond the usually specified technical criteria, resulting in PWB (Peariasamy et al., 2020). As indicated earlier by Parker et al. (2006) that there are two crucial attributes of Proactive Behavior in the workplace: proactive enactment of ideas and problem solving.

Role Breadth Self-efficacy significantly influences Proactive Work Behavior

Mediating role of job satisfaction

According to Bateman and Crant (1993), individuals who have PP have control from within and can influence their environment in the workplace. Likewise, individuals who experience positive affective states associated with JS tend to change their situation proactively (Judge, 1993). By establishing a positive cycle, JS may be a key concern in organizational behavior and occupational health (Kuo et al., 2019). In short, individuals with PP who have control through JS or a situation where they have an affective disposition will better influence the PWB.

Job Satisfaction significantly mediates Proactive Personality and Personality Work Behavior

Mediating role of role breadth self-efficacy

Employees' perceived ability to carry out different proactive, social, and integrative actions that go beyond specific technical responsibilities is referred to as RBSE (Parker, 1998). RBSE, on the other hand, is not as stable as PP since it evolves with experience and organizational situation (Crant, 2000). PP rather emphasizes more on future changes (Unsworth and Parker, 2008). As a result, in order to engage in PWB, PP requires RBSE, which indicates the ability to start, play a greater role, and have social and integrative qualities (Crant, 2000).

Role Breadth Self-efficacy significantly mediates Proactive Personality and Personality Work Behavior

All hypotheses are conceptualized in Figure 1 (Research Model)

Figure 1

Research model.

Methods

The data collection was done on a total of 455 respondents at the State Detention Centers in Indonesia, an organization that plays a crucial role in society that is ensuring security and order. The number of officers from the State Detention Center (RUTAN) in Indonesia is very limited, especially compared to the number of inmates which are over capacity. This situation increases the need for a form of personality from State Detention officers that can affect their support for institutional effectiveness through efficacy and satisfaction (Whiteacre, 2019). For this reason, this study uses officers from the State Detention Centers in Indonesia to be the population in measuring the variables used, because they have equality with the phenomena that occur. In this study, Akgunduz et al. modified 10 items to measure PP (2018). Following that, 8 items from Parker et al. were developed to measure PWB (2006). There were 7 items developed from Parker et al. for RBSE (2006). Finally, JS was measured using 10 items modified from Jaiswal and Fit (2017). Responses were then gathered using a five-point Likert scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree). The study's control variables included gender, education, status, length of work, and age. To acquire an assessment of the research variables, this study conducted a data collecting method using a developed questionnaire that was distributed to respondents using Google Form. All employees and managers who were sampled in this study were given questionnaires. The data was then evaluated with Amos Software's SEM (Structural Equation Modeling). For examining diverse phenomena of society, corporations, organizations, and other groupings that incorporate humans as direct research objects, a mix of quantitative methodologies is considered appropriate (Saebani and Sutisna, 2018).

Results

Table 1 shows the demographics of the respondents in this survey. A total of 383 male responses out of 455 total respondents gains the majority. Then, 398 respondents (or 87.3 percent) with a high school degree become the majority in this survey. 307 respondents who are married make up the bulk of this study's participants. In this survey, the majority of individuals who have worked at Indonesian Detention Centers (RUTAN) for more than 8 years were 229 respondents, accounting for 50.2 percent of the total. Finally, respondents ranging from 20 and 30 years old cover 231 respondents.

Table 1

Demographics.

	N	Percentage
Gender
Man	383	84,0
Women	73	16,0
Education
Senior high school	398	87,3
Diploma	24	5,3
Bachelor degree	34	7,5
Master degree	0	0
Doctorate degree	0	0
Status
Married Single	307	67,3
Widower	139	30,5
Widow	10	2,2
Length of work
<1 year	2	0,4
1–2 year	159	34,9
3–4 year	41	9,0
5–6 year	3	0,7
7–8 year	22	4,8
>8 year	229	50,2
Age
<20 years old	0	0
20–30 years old	231	50,7
31–40 years old	142	31,1
41–50 years old	39	8,6
>50 years old	44	9,6

An overview of the validity and reliability

The validity test is used to establish the extent to which the statement items may measure each variable, and the results of the validity test are reported in Table 2. Validity measurement uses corrected item-total correlation (r corrected), for the correlation value greater than 0.30 or significant at the real level α 5%, then the statement item is declared valid. Validity testing was carried out with the help of the SPSS version 24 program. The results of the validity test can be seen in Table 2:

Table 2

Validity statistics.

Variable	Indicator	Corrected item-total correlation	Decision
Proactive Personality	PP1	0.657	Valid
	PP2	0.683	Valid
	PP3	0.660	Valid
	PP4	0.703	Valid
	PP5	0.680	Valid
	PP6	0.653	Valid
	PP7	0.691	Valid
	PP8	0.641	Valid
	PP9	0.659	Valid
	PP10	0.737	Valid
Job Satisfaction	JS1	0.618	Valid
	JS2	0.677	Valid
	JS3	0.729	Valid
	JS4	0.775	Valid
	JS5	0.763	Valid
	JS6	0.774	Valid
	JS7	0.734	Valid
	JS8	0.503	Valid
	JS9	0.722	Valid
	JS10	0.647	Valid
Role Breadth Self-efficacy	RBSE1	0.698	Valid
	RBSE2	0.786	Valid
	RBSE3	0.802	Valid
	RBSE4	0.787	Valid
	RBSE5	0.811	Valid
	RBSE6	0.788	Valid

As shown on Table 2, the correlation value for each item on all variables has a range between 0.503-0.823, so that all statement items have a correlation value greater than 0.30 and are also significant at the real level α 5%. Thus, it can be concluded that all statement items used to measure all variables are valid and can be used for further analysis.

The next test is the reliability test presented in Table 3, which is used to determine the reliability or consistency of variable measurements. Reliability testing was carried out using the cronbach's alpha technique, according to Malholtra, the questionnaire was declared reliable if it produced a cronbach's alpha value greater than 0.60 (Solimun et al., 2017: 40).

Table 3

Reliability statistics.

Variable	Cronbach's α	No. of items	Decision
Proactive Personality	0,911	10	Reliable
Job Satisfaction	0,911	10	Reliable
Role Breadth Self-efficacy	0,926	7	Reliable
Proactive Work Behaviour	0,929	8	Reliable

Table 3 shows that the Cronbach's alpha value for all variables has a value greater than 0.60, meaning all statement items used to measure all variables can be declared reliable and believed to be a consistent measuring tool.

Construct validity

Construct Validity determines the extent to which indicators measure constructs. Convergent validity is used in the construct validity test in SEM. If the indicators in the construct have a standardized regression weight (lambda/factor loading) value larger than 0.50 and an Average Variance Extracted (AVE) value greater than 0.50, the construct has convergent validity. Table 4 shows the findings of the concept validity evaluation:

Table 4

Construct validity.

Variable	Indicator	Factor Loadings	Decision
Proactive Personality	PP1	0.698	Valid
	PP2	0.722	Valid
	PP3	0.687	Valid
	PP4	0.728	Valid
	PP5	0.710	Valid
	PP6	0.687	Valid
	PP7	0.740	Valid
	PP8	0.677	Valid
	PP9	0.696	Valid
	PP10	0.775	Valid
Job Satisfaction	JS1	0.640	Valid
	JS2	0.718	Valid
	JS3	0.763	Valid
	JS4	0.828	Valid
	JS5	0.812	Valid
	JS6	0.819	Valid
	JS7	0.760	Valid
	JS8	0.529	Valid
	JS9	0.770	Valid
	JS10	0.676	Valid
Role Breadth Self-efficacy	RBSE1	0.754	Valid
	RBSE2	0.836	Valid
	RBSE3	0.844	Valid
	RBSE4	0.801	Valid
	RBSE5	0.853	Valid
	RBSE6	0.807	Valid
	RBSE7	0.746	Valid
Proactive Work Behaviour	PWB1	0.712	Valid
	PWB2	0.798	Valid
	PWB3	0.796	Valid
	PWB4	0.815	Valid
	PWB5	0.856	Valid
	PWB6	0.800	Valid
	PWB7	0.749	Valid
	PWB8	0.793	Valid

Table 4 shows that each indicator in each construct yields an AVE value more than 0.50 and has a factor loading value higher than 0.50. As a result, the indicators are valid in building constructs and may be utilized to develop models (Table 5).

Table 5

Construct reliability.

Variable	AVE	Construct Reliability	Decision
Proactive Personality	0.508	0.911	Reliable
Job Satisfaction	0.543	0.921	Reliable
Role Breadth Self-efficacy	0.651	0.929	Reliable
Proactive Work Behaviour	0.626	0.930	Reliable

Construct reliability

The construct reliability value in Table 5 is used to assess the construct reliability test; if the construct reliability value is more than 0.70, the construct is considered to be reliable (Solimun et al., 2017: 78). According to Hair et al. (2014: 605), the construct value reliability must be larger than 0.70 as a rule of thumb, and a construct reliability value greater than 0.60 is acceptable as long as each indication meets the convergent validity. Table 5 shows the findings of evaluating construct reliability for each construct:

Table 5 shows that each construct produces a construct reliability value greater than 0.70, so it can be concluded that these indicators are reliable in reflecting the constructs of all variables.

Structural model fit

The structural model analysis stage follows after the measurement model analysis stage is completed. This stage begins with an assessment of the structural model fit (goodness of fit), which ensures that the model created is accurate and consistent with the data (fit). Figure 2 shows the structural model's estimation results as well as the value of the goodness of fit criteria:

Figure 2

Assessing the structural model.

The structural model suitability test in Table 6 shows that all the criteria for absolute fit indices, incremental fit indices, and parsimony fit indices have met the requirements (marginal fit and good fit), so that the structural model is acceptable, and then testing the significance of the interplay between variables, both direct influence and indirect effect. The explanation of each fit indices is presented below Table 6.

Table 6

Fit measure for the structural model.

Fit Measure		Structural Model	Critical Value	Decision
Absolute Fit Indices	Probability	0.000	≤0.05	Good fit
	Cmin/DF	2.362	≤3.00	Good fit
	GFI	0.839	≥0.90	Marginal fit
	RMSEA	0.057	≤0.08	Good fit
	SRMR	0.046	≤0.05	Good fit
Incremental Fit Indices	CFI	0.934	≥0.95	Marginal fit
	TLI	0.930	≥0.95	Marginal fit
	NFI	0.892	≥0.90	Marginal fit
	RFI	0.884	≥0.90	Marginal fit
Parsimony Fit Indices	AGFI	0.817	≥0.90	Marginal fit
	PNFI	0.832	≥0.90	Marginal fit

Chi-square

The chi-square statistic (χ2) is the simplest basic test tool for determining model fit, and it is quite sensitive to sample size. According to Hair et al. (2014: 584), in a model with a sample size of <250 and the number of indicators <30, the expected chi-square criterion is to produce a probability> 0.05 (insignificant p-values expected), while in models with a sample size of> 250 or the number of indicators >30, then the expected chi-square criterion is to produce a probability≤0.05 (significant p-values expected). The model in this study, the number of samples is 427 (>250) and the number of indicators is 35 (>30), so that the good fit model if the chi-square criteria yield a probability≤0.05. The structural model's estimate result yields a chi-square probability of 0.000, which is less than the real level of 5%, indicating that the structural model is a good fit.

Normed Chi-square (Cmin/Df)

The recommended values are 1.0 for the lower limit and 2.0 or 3.0 for the higher limit. The structural model's estimation result is a cmin/df of 2.362, which is less than 3, indicating that the structural model is good-fit.

Goodness of fit index (GFI)

The GFI is a suitability indicator for estimating a systemized population covariance matrix's weighted proportion of variance. GFI ratings range from 0 to 1, with higher scores indicating better performance. A GFI of more than 0.90 denotes a good fit, whereas a GFI of 0.80–0.90 denotes a moderate fit (marginal fit). The GFI value obtained from the structural model estimate is 0.839, which is within the allowed range of 0.80–0.90, showing that the structural model is still acceptable (marginal fit).

Root mean square error of approximation (RMSEA)

RMSEA calculates the difference between a model's parameter values and its population covariance matrix. If the RMSEA value is less than or equal to 0.05, the model is considered to be close fit, and if the RMSEA value is between 0.05-0.08, the model is said to be good fit. The structural model's estimate results provide an RMSEA value of 0.057, which is within the range of 0.50–0.08, indicating that the structural model is determined to be a good fit.

Standardized root mean residual (SRMR)

SRMR is the residual average between the observed covariance/correlation matrices and the estimated results. The model is considered good fit if the SRMR value is less than 0.05. The estimation result of the structural model produces an SRMR value of 0.046, this value is less than 0.05, which indicates that the structural model is good (good fit).

Comparative fit index (CFI)

The CFI value varies from 0 to 1. If a model's CFI value is more than or equal to 0.95, it is considered to be good fit, and if it is between 0.80-0.95, it is said to be marginal fit. The structural model's estimate findings provide a CFI value of 0.934, which is within the allowed range of 0.80–0.95, indicating that the structural model is still acceptable (marginal fit).

Tucker Lewis index (TLI)

The TLI value ranges from 0 to 1, and it is also known as a non-normed fit index. A model is deemed good fit if its TLI value is more than or equal to 0.95, and marginal fit if it is between 0.80-0.95. The structural model's estimation result yields a TLI value of 0.930, which is within the allowed range of 0.80–0.95, indicating that the structural model is still accepted (marginal fit).

Normed fit index (NFI)

The NFI value varies from 0 to 1. If the NFI value of a model is more than or equal to 0.90, it is considered to be good fit, and if the value is between 0.80-0.90, it is said to be marginal fit. The structural model's estimation results generate an NFI value of 0.892, which is within the permitted range of 0.80–0.90, indicating that the structural model is still acceptable (marginal fit).

Relative fit index (RFI)

If the RFI value of a model is more than or equal to 0.90, it is considered to be good fit, and if the RFI value is between 0.80-0.90, it is said to be marginal fit. The structural model's estimation result yields an RFI value of 0.884. This number falls between 0.80-0.90, indicating that the structural model is still appropriate (marginal fit).

Adjusted goodness of fit index (AGFI)

For the degree of freedom (df) in the model, AGFI is a variation of GFI. When an AGFI is more than or equal to 0.90, a model is considered to be good fit, and when it is between 0.80-0.90, it is said to be marginal fit. The structural model's estimation results generate an AGFI value of 0.817, which is within the range of 0.80–0.90, indicating that the structural model is still acceptable (marginal fit).

Parsimony normed fit index (PNFI)

PNFI values greater than 0.50 indicate parsimony which indicates a better fit, and is only used for comparisons between alternative models. The results of the structural model estimate yield a PNFI value of 0.832, which indicates that the structural model is good fit.

Hypothesis testing

Testing structural relationships (direct effect)

Furthermore, in Table 7, it can be seen that the next stage of structural model analysis is the testing of structural relationships in the direct effect path, namely examining the estimated parameters of the relationship between variables that represent each theoretical hypothesis. The hypothesis can be accepted if the path parameters are statistically significant with the direction of influence predicted, meaning that the path parameters must be greater than zero for the positive direction and less than zero for the negative direction (Hair et al., 2014: 589).

Table 7

Summary of the direct effect hypothesis.

Hip.	Structural relationship	Std.Estimate	C.R.	P value
H1	Proactive Personality → Proactive Work Behaviour	0.122	2.025	0.043∗
H2	Proactive Personality →Job Satisfaction	0.675	10.963	0.000∗∗
H3	Proactive Personality → Role Breadth Self-efficacy	0.807	13.298	0.000∗∗
H4	Job Satisfaction →Proactive Work Behaviour	0.311	7.046	0.000∗∗
H5	Role Breadth Self-efficacy → Proactive Work Behaviour	0.575	9.472	0.000∗∗

∗Significant at the 0.05 level

∗∗Significant at the 0.01 level

In testing structural relationships, a hypothesis is tested to test the significance of the influence between variables, using the critical ratio (CR) and probability values (p-value). If the CR value is ≥1.96 or the p-value ≤ 5% significance level is significant or not, it is decided that there is a significant influence between these variables.

Following are the results of testing structural relationships in order to test each research hypothesis based on the SEM output:

The estimate findings of the influence of PP on PWB demonstrate a substantial effect with a CR value of 2.025 (more than 1.96) and a significance value (p-value) of 0.043 (smaller than the 5 percent real level), as shown in Table 7. The resultant coefficient of effect is 0.122 (positive), indicating that the more proactive one's personality is, the more proactive one's work behavior is. As a result, the first hypothesis is accepted.

With a CR value of 10,963 (more than 1.96) and a significance value (p-value) of 0,000, the estimate findings of the PP influence parameter on JS likewise demonstrated a significant effect (less than the 5 percent real level). The resultant coefficient of influence is 0.675 (positive), indicating that the more proactive a person is, the more satisfied they are at work. As a result, the second hypothesis is accepted.

The influence of PP on RBSE was also shown to have a significant effect, with a CR value of 13,298 (higher than 1.96) and a significance value (p-value) of 0.000 in the estimation results (less than the 5 percent real level). The resulting coefficient of effect is 0.807 (positive), indicating that the more proactive a person is, the greater their RBSE. As a result, the third hypothesis is accepted.

With a CR value of 7,046 (higher than 1.96) and a significance value (p-value) of 0,000 (less than the 5% actual level), the estimate result of the influence of JS on PWB demonstrates a significant effect. The estimated coefficient of effect is 0.311 (positive), indicating that JS correlates with PWB. As a result, the fourth hypothesis is accepted.

With a CR value of 9,472 (more than 1.96) and a significance value (p-value) of 0,000, the effect of RBSE on PWB is similarly significant (less than the 5 percent real level). The resulting coefficient of effect is 0.575 (positive), indicating that the greater the role breadth of self-efficacy, the more PWB is displayed. As a result, the fifth hypothesis is accepted.

PP has more dominant influence on role breadth of self-efficacy, then on JS, and finally on PWB. PWB is more dominantly influenced by RBSE, then JS, and finally PP. This shows that PP affects PWB more through indirect channels.

Testing structural relationships (indirect effect)

Then in Table 8, it can be seen that the next stage of structural model analysis is the testing of structural relationships in the path of the indirect effect. Hypothesis testing to test the significance of this indirect effect is carried out in the same way, namely using the critical ratio (CR) value and the probability value (p-value). If the CR value is ≥1.96 or the p-value ≤ 5% is significant, it is decided that there is a significant effect.

Table 8

Summary of the indirect effect hypothesis.

Hip.	Structural relationship	Std.Estimate	SE	C.R.	P-value	Type of mediator
H6	Proactive Personality → Job satisfaction → proactive work behavior	0.210	0.033	5.891	0.000∗∗	Partial mediation
H7	Proactive Personality →Role Breadth Self-efficacy → Proactive Work Behavior	0.464	0.057	7.680	0.000∗∗	Partial mediation

∗Significant at the 0.05 level

∗∗Significant at the 0.01 level

The nature of the mediation must be determined once the importance of the mediation effect has been determined. According to Ghozali (2011: 248), detecting the nature of mediation can be seen in the effect of the mediation; if the direct effect of exogenous variables on endogenous variables is significant, and the indirect effect through intervening variables is also significant, then it is partially mediation. In contrast, fully mediation or perfect mediation occurs when the direct influence of exogenous variables on endogenous variables is minor, but the indirect effect through intervening variables is significant.

Following are the results of testing structural relationships in order to test each research hypothesis of the indirect effect based on the SEM output:

Based on Table 8 above, it can be explained as follows: The results of the indirect path significance test PP → JS → PWB showed a significant effect with a CR value of 5,891 (greater than 1.96) and a significance value (p-value) of 0.000 (less than the 5% real level). Thus, the sixth hypothesis is accepted. The nature of the mediator is known to be partially mediation, meaning that increasing the PWB of employees can only be done by increasing the employee's PP, but if it is also supported by high JS, then the employee's PWB can be even more improved.

The results of the indirect path significance test PP → RBSE → PWB also showed a significant effect with a CR value of 7,689 (greater than 1.96) and a significance value (p-value) of 0,000 (less than the 5% real level). Thus, the seventh hypothesis is accepted. The nature of the mediator is known to be partially mediation, meaning that increasing the PWB of employees can only be done by increasing the employee's PP, but if it is also supported by high role breadth of self-efficacy, then the employee's PWB can be even more improved. The mediation level of RBSE is stronger than JS, in mediating the influence of PP on PWB.

Discussion

PP has a considerable beneficial influence on PWB, according to the findings, which is consistent with several previous studies (Crant, 2000; Parker et al., 2006). PP is found to influence PWB, which focuses on enacting ideas and solving problems at the workplace. PP is described as a personality trait that contributes to proactive behaviors such as taking initiative, taking action, and not depending on external factors.

Furthermore, there is evidence that PP has a beneficial impact on JS. This is consistent with past researches that have shown comparable results (M. Li et al., 2017; N. Li and Crant, 2010). As can be observed, a proactive individual tends to establish conditions in the workplace that are beneficial to personal achievement as well as an atmosphere that leads to JS (N. Li and Crant, 2010).

Similarly, PP has an impact on RBSE, which is similar to some previous studies (Parker, 1998; Parker et al., 2006). It suggests that employees with PP are more likely to be motivated or confident in expanding their responsibilities. Confidence in oneself when doing a task extends beyond the allotted technical core and is integrated and coordinated (Parker, 1998). This notion will be advantageous in positions that demand collaboration and disclosure of information, such as in Indonesian state detention centers.

Furthermore, JS shows significant positive results to PWB. Those who are satisfied with their work will influence their behavior in the workplace. It is supported by research by Strauss et al. (2013), which stated that JS is a resource that enables individuals to continue with the effort necessary to maintain proactive action in the workplace.

Then, when it comes to PWB, RBSE reveals a strong positive effect. Those who have a strong belief in their ability to carry out tasks with a larger scope will be more proactive. When no one feels capable or understands the position, it leads to taking over behavior (Parker et al., 2006). This conclusion is consistent with earlier research that shows RBSE has a large favorable impact on proactive work personality (Parker et al., 2006).

The importance of JS as a mediator between PP and PWB has shown to be beneficial. It appears that JS plays a role as a partial mediator. Previously, PP has a direct influence on PWB, and PP is also possessed by employees who have an affective work disposition, in other words, JS will support PWB (Judge, 1993; Parker et al., 2006).

In this study, the significance of RBSE as a mediator between PP and PWB is found to be important. These findings back with prior research that shows RBSE can considerably positively moderate the relationship between PP and PWB (Parker et al., 2006). Partial mediation is also stated in the findings. RSBE plays a function in behaving confidently while also optimizing their broader role in proactive and integrative action. PP had previously supported this function with initiative and steadiness in order to create future adjustments. Having a steady personality and the confidence to take on a larger role in the job leads to proactive professional activities, which are also advantageous for future changes.

Conclusion and suggestion

Conclusion

This study found that: (1) PP has a positive and significant effect on PWB with a p-value of 0.043 (<0.05), (2) PP has a positive and significant effect on JS with a p-value of 0.000 (<0.05), (3) PP has a positive and significant effect on RBSE with a p-value of 0.000 (<0.05), (4) JS has a positive and significant effect on PWB with a p-value of 0.000 (<0.05), (5) RBSE has a positive and significant effect to PWB with a p-value of 0.000 (<0.05), (6) There is a positive and significant effect of PP on PWB mediated by JS because of the significant direct effect of PP on JS and JS on PWB with a value positive value of 0.210. (7) There is a positive and significant effect of PP on PWB mediated by RBSE because of the significant direct effect of PP on RBSE and RBSE on PWB with a positive value of 0.464. This is consistent with the findings by Nurjaman et al. (2019), who discovered that PWB is important for organizations since it is likely to enhance present work situations and open up new chances. The State Detention Center (RUTAN) in Indonesia need something comparable. Individuals will be encouraged to participate in proactive activity when they feel in control of their activities, according to Hua et al. (2020).

Suggestion

Several suggestions for improving PWB may be offered based on the findings. Employees at Indonesian State Detention Centers should be encouraged to engage in PWB. This is especially true for employees that have a PP. Then, in order to increase PWB, employees want good sentiments associated to JS. Employees with PP must also be given the confidence to take on greater roles and be placed in appropriate positions, which will foster PWB. Similarly, developing or bringing forth a PP in other employees is critical. Preventative actions in state prisons will be easier to adopt this way. As a result, it is suggested that officers be willing to take initiative and persevere in completing job responsibilities in order to improve the working environment and better deal with problems.

For further research, it is suggested that PP be compared to PWB directly, as well as indirectly through JS and RBSE, using a variety of sampling methods and numbers of samples, or with the addition of several other antecedents. Because there is still a scarcity of study on this subject. As a result, variables with diverse types of responders will be developed. So that future research in human resource management and organizational research can focus on PP, JS, RBSE, and PWB. Furthermore, it is preferable to study broader factors in future studies in order to affect PWB, or to employ a new research object. Because proactive conduct may be developed in a variety of ways, including PP, work happiness, and RBSE, it is not limited to Detention Center officers.

Declarations

Author contribution statement

Nanank Syamsudin: Conceived and designed the experiments; Analyzed and interpreted the data; Wrote the paper.

Anis Eliyana: Conceived and designed the experiments; Analyzed and interpreted the data.

Nurliah Nurdin: Performed the experiments.

Agus Sudrajat: Performed the experiments.

Bambang Giyanto: Contributed reagents, materials, analysis tools or data.

Alvin Permana Emur: Analyzed and interpreted the data.

Marziah Binti: Contributed reagents, materials, analysis tools or data.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability statement

Data will be made available on request.

Declaration of interests statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.