Cognitive job crafting as mediator between behavioral job crafting and quality of care in residential homes for the elderly.

Extending previous studies on job crafting, the aim of the present study is to analyze the effect of job crafting on quality of care in residential homes for elderly people in two European countries (Spain and Sweden). We hypothesize that cognitive crafting could be a consequence of behavioral crafting and that it will mediate the relationship between behavioral crafting and the perception of quality of care. A correlational design was used, with two-waves approximately 12 months apart (n = 226). Our results indicate that behavioral job crafting at T1 had an effect on cognitive job crafting at T2, relational job crafting at T1 increases quality of care at T2, and the mediation effect of cognitive job crafting. These results indicate that we must differentiate between the two forms of crafting (behavioral and cognitive), not as indicators of the same latent construct, but as aggregates. Additionally, we point out two main implications for managerial practice. First, as relational job crafting has a direct effect on quality of care, it is important to assure an organizational culture oriented towards employees. Secondly, due to the mediation effect of cognitive job crafting, managers should facilitate meaningful work environments. To do so, jobs should be re-designed, increasing skills variety, identity and significance.

Introduction

The recent European population projections indicate that the total population will increase from 511 million in 2016 to 520 million in 2070. Nevertheless, in the same period the working-age population (15–64 years old) will decrease significantly from 333 million in 2016 to 292 million [1]. The percentage of people over 65 years old will increase by 41.3%, and 12.5% of the total population will be represented by people over 80 years of age. In this light, the European Commission report draw two possible scenarios related to high life expectation; the positive one assumes that “all future gains in life expectancy are spent in good health” [1]. The other scenario, the negative one, associates the increasing years to health problems, and consequently with additional cost devoted to healthcare. In this scenario, the European Commission foresees that long-term care and health care costs are expected to contribute the most to the rise in age-related spending.

Facing this situation, countries in Europe are concerned about how to make the welfare system sustainable, and additionally how to guarantee healthcare quality. In this sense, professionals are vital to the provision of health services and goods.

In Spain, the Catalan Association for Health Care Resources (Asociació Catalana de Recursos Asistencials, ACRA), which is formed by most enterprises and entities devoted to the assistance of elder people, indicated that this sector has more than 50,000 employees [2], and provide services for taxes equating 1,121.809,000 € in 2014 year’s currency rate. In Sweden, according to the Council for the Promotion of Municipal Analysis [3], there is about 250,000 employees working in municipal eldercare and the corresponding total costs are 126,318.315,000 SEK in 2018.

Due to the high impact of this sector and its employees, from an economic and social point of view, it is important to point out that the last few years, the healthcare sector has been characterized by slow growth of costs, but healthcare professionals need an adequate amount of resources to reach their work-related goals and offer high-quality of care [4]. In this context, several authors [5] recognize the need for healthcare professionals to take a proactive role in shaping their future jobs to improve healthcare systems, to assure improvements on the quality of care they provide [6], and on other hand, this proactive role “can be learned and effectively transferred from training to organizational practice”[7 p321].

In this sense, the scientific literature has pointed out that the work environment can be proactively modified by the employees, who are not passive subjects, but can influence their workplace by redesigning their work experience from various processes. These processes have been called job crafting [8, 9].

The present study aims to analyze the effect of job crafting on quality of care in residential homes for the elderly people.

Theoretical framework and hypotheses

Job crafting has been defined by Wrzesniewski and Dutton [9] as the “physical and cognitive changes individuals make in the task or relational boundaries of their work” (p179). Following Berg, Grant and Johnson [10] this proposal is the understanding that employees are often interested in customizing their jobs to fit optimally their motivations, competences and desires.

Organization’s management should not unilaterally decide how its employees spend their time and energy. Rather, the employees themselves should be allowed to decide what to do, creating and searching a comfortable and enjoyable context, over and beyond the job descriptions provided by the management, particularly in complex and uncertain situations [9-12]. In this sense it is possible to consider that the job crafting emerges as a strategy for the leverage of work meaning and identity.

From the perspective of Wrzesniewski and Dutton [9], the concept of job crafting is based on three dimensions: a) the scope or amount of tasks that the employee develops (for example, the introduction of new tasks more suited to their abilities or interests; task job crafting), b) the relationships with other people and the activities that it carries out with the co-workers (for example, making friends with people with similar abilities or interests, relational job crafting), and c) the modification of the meaning of the labor and social environment (for example, recognize the effect and importance of work activity for users, patients, clients, etc., cognitive job crafting).

More recently, authors such as Zhang and Parker [13] grouped the three dimensions into two, called behavioral crafting (task and relational crafting) and cognitive crafting. The differentiation between behavioral and cognitive crafting arises from the debate between what should be considered job crafting or not. In this sense, from the perspective of job demands and resources theory (JD-R), it is considered that cognitive crafting is a mere passive adaptation to work, so it cannot be conceived as crafting since it is not a proactive behavior of change [8, 14]. In contrast, Zhang and Parker [13], based on Wrzesniewski and Dutton [9] perspective, include cognitive job crafting since “it involves altering how one frames or views their tasks or job, which is self‐initiated, self‐targeted, intentional, and represents meaningful changes to the job aspects” [13 p5]. In this sense, it is important to differentiate between the two forms of crafting (behavioral and cognitive), not as indicators of the same latent construct, but as aggregates.

Preliminary research has indicated that cognitive crafting has only moderate relationships with behavioral crafting [15-17]. These results also reveal different relationship dynamics for task and relational job crafting, on the one hand, and cognitive job crafting, on the other. Additionally, recent studies also suggest that behavioral crafting (task and relational) can cause changes at the cognitive level [13,18]. Research within the field of social psychology demonstrates that individuals actively shape their reactions through cognitive reframing. Reframing is defined by Ashforth and Kreiner [19] as the transformation of the meaning assigned to an occupation by instilling it with positive value and neutralizing the negative values previously assigned.

In this sense, Unsworth, Mason, and Jones [18] and Zhang and Parker [13], based on the studies of Ashforth and Kreiner [19], point out that employees could change their perceptions about their tasks and relations at work (behavioral crafting) involving cognitive reframing (cognitive crafting). The authors [18] explained that employees reframe their cognitions in order to compare themselves with others in terms of their strengths. For example, if a group has productivity as a strength their members will tend to assess other groups or employees on the basis on this characteristic. In our research, following this suggestion, we consider that employees could cognitive reframe the meaning of their jobs by means of the tasks (Task crafting) they develop and the relationships they have at work (Relational crafting).

Related to the effects of job crafting, several empirical studies have demonstrated its influence on individual and organizational variables. On the individual level, job crafting has been positively related with employees’ wellbeing [10, 11, 20, 21], engagement [22, 23], commitment [11] and job satisfaction [9]. On the other hand, on the organizational level, job crafting facilitates organizational change [24], improves performance [6, 25–28] and reduces turnover [11, 12, 21, 28].

In the healthcare sector, employees’ job crafting contributes to employees’ performance, analyzed as quality of care they give to patients [29]. During the past decades, performance indicators have become increasingly sophisticated [30], and the quality of care has been included as an indicator of performance [31]. Nevertheless, some authors [32] consider that it is difficult to compare quality of care and to transfer performance indicators directly between different health systems and cultures, because each country has established different indicators. However, the quality of care is used as an indicator of patients’ outcome in studies focused in hospital care [33] and in the geriatric care [34].

Considering the complexity of the healthcare sector Teoh, Hassard, and Cox [35] developed a systematic review on the relationship between the working conditions and the quality of patient care. Results showed several investigations where an improved work environment exerted a positive effect on nurses assessed quality of care. For instance, promotion prospects, perceived salary and job security [36-38] were positively related with quality of care. Nevertheless, these authors concluded that the relationships between quality of care and work environment reported divergent findings, showing the complexity of these relationships [39, 40]. They recommended, among other aspects, to use multilevel or longitudinal designs, as well as mediating and moderating variables, in order to present a more realistic interpretation of these relationships.

Following these suggestions and taking into account that job crafting is a modification of working conditions initiated by the employees, the present research propose job crafting as an antecedent of quality of care and a longitudinal design. Secondly, we propose cognitive crafting as a mediator between behavioral crafting (tasks and relationships), which modify working conditions, and quality of care.

Quality of care could be analyzed from the nurses’ perception as an indicator of patient outcome [29, 33]. The present study uses this perspective to analyze this variable and, additionally, even though it is not unusual to assess quality of care with a single-item [41], we follow Westerberg and Tafvelin [31] recommendations, adding information about satisfaction “with the way in which the clients were treated, kept informed and their wishes respected” [p464], and use a 5-items scale.

The results of previous studies [9-11] reveal different relationship dynamics for task and relational job crafting, on the one hand, and cognitive job crafting, on the other. In light of the revisions of Unsworth et al. [18] and Zhang and Parker [13], there exist several empirical and theoretical reasons to hypothesize that cognitive crafting could be a consequence of behavioral crafting and that it will mediate the relationship between behavioral crafting and the perception of quality of care. Compared to previous studies, the main novelty of our proposal lies in the fact that we analyze job crafting as a dynamic process and differentiate between behavioral job crafting (tasks and relational crafting), as an antecedent, and cognitive job crafting as a consequence, and as a mediator between it and the quality of care.

Consequently, the following hypothesis will be tested:

H1. Behavioral Job Crafting (JC) (T1) is positively related with Cognitive JC (T2)

H1.1. Task JC (T1) is positively related with Cognitive JC (T2)

H1.2. Relational JC (T1) is positively related with Cognitive JC (T2)

H2. Behavioral JC (T1) is positively related with Quality of Care (T2)

H2.1. Task JC (T1) is positively related with Quality of Care (T2)

H2.2. Relational JC (T1) is positively related with Quality of Care (T2)

H3. Cognitive JC (T2) mediates the relationship between Behavioral JC (T1) and Quality of Care (T2)

H3.1. Cognitive JC (T2) mediates the relationship between Task JC (T1) and Quality of Care (T2)

H3.2. Cognitive JC (T2) mediates the relationship between Relational JC (T1) and Quality of Care (T2)

Fig 1 illustrates the variables and hypotheses that constitute our research model.

Fig 1

Research model.

Materials and methods

Participants and procedure

Data was collected in 18 residential homes for elderly people in two European countries (Spain and Sweden). A correlational design was used, with two-waves approximately 12 months apart.

The protocol for the research project has been approved by the managers and research ethics committees of the participating centers, in Spain (due to confidentiality agreements, the complete names of the participating centers cannot be disclosed), and the regional ethics committee in Sweden (Regional Board of Ethics, EPN, Umeå [ref 2015-62-31Ö]). All participants gave informed consent for the research, and that their anonymity was preserved.

At time 1 (2016), participants received information about the study aim and ethical considerations, and the written questionnaire to be completed. Those who returned usable surveys were contacted again at time 2 (2017).

At T1 and T2, participants provided data about their age, gender, if they have a permanent contract, a managerial position, and worked night shift. The questionnaire included self-reported measures of behavioral job crafting (task and relational job crafting), cognitive job crafting and perceived quality of care. Participants were asked to add a secret code that allowed us to match responses from T1 and T2 without identifying individual employees. The application format was “paper and pencil” in all centers.

Of the 928 contacted employees at T1, 628 returned complete questionnaires (68% response rate), and at T2 890 employees were contacted and 501 answered (56% response rate). A total of 226 employees answered at both T1 and T2. Employees in the final sample (n = 226) were predominantly female (92.5%), with a mean age of 45.76 years (SD = 11.42), 92.5% of them had permanent contracts, 10.2% had managerial positions and 10.2% worked on shifts with night shifts. Most respondents worked in Spain (65.93%)

Materials

Two instruments were used in this study to evaluate the variables that constitute our research model. To measure job crafting, the Job Crafting Questionnaire (JCQ) [16] was used, both for behavioral and cognitive crafting. Recently, a new Spanish adaptation of this questionnaire has been developed [42]. The Spanish versions of the instrument used in the present study [6] and in the newest one [42] are very similar. Only small changes on translation have been observed. The internal consistency for the Spanish version of this scale was .85 [6] (.88 in the newest version [42]), and .83 for the Swedish version [6]).

On the other hand, for measuring quality of care the instrument developed by Westerberg and Tafvelin [31] was used.

Behavioral job crafting (10 items). Task (TJC) and Relational (RJC) Job Crafting dimensions of the Job Crafting Questionnaire (JCQ) [16] were used to assess Behavioral Job Crafting (BJC). Responses were measured on a five-point Likert scale ranging from 1, “hardly ever”, to 5, “very often”. An example of item for TJC is “How often do you change the scope or types of tasks that you complete at work?” and for RJC “How often do you make friends with people at work who have similar skills or interests?” Task (TJC) Job Crafting subscale had a Cronbach’s alpha of .87 and Relational (RJC) Job Crafting .83 [16].

Cognitive job crafting (5 items). Cognitive Job Crafting dimension of the Job Crafting Questionnaire (JCQ) [16] was used to assess Cognitive Job Crafting (CJC). Responses were measured on a five-point Likert scale ranging from 1, “hardly ever”, to 5, “very often”. An example of item is “How often do you remind yourself of the importance of your work for the broader community?” Cognitive (CJC) Job Crafting subscale had a Cronbach’s alpha of .89 [16].

Quality of care (QoC). The instrument developed by Westerberg and Tafvelin [31] was used. It includes three statements about how often the respondent felt satisfied with the way in which the clients were treated, kept informed, and their wishes respected, one statement about how well the help and support provided met the client’s needs, and another item related to the overall satisfaction with care work. An example of item is “At my workplace I experience that enough consideration is taken to the users’/clients’ opinions and wishes”. Responses were measured on a five-point Likert scale ranging from 1, “very seldom or never”, to 5, “very often or always”. The internal consistency of the scale was .86 [31].

Control variables. We controlled for the possible effects of participants’ age, gender, contract, managerial position, and shift.

Results

Descriptive statistics

Descriptive statistics for all the study variables in both waves are included in Table 1. No significant differences were found between T1 and T2 variables’ scores, indicating the stability of the measures. In all cases, the scores for job crafting and quality of care variables were above 3 in a 5-points scale. Following the scale range these results indicated that employees sometimes job craft, and often considered that quality of care in their workplaces was appropriate.

Table 1

Descriptive statistics at time 1 (T1) and time 2 (T2).

	N	Minimum	Maximum	Mean	SD	Paired t-test
QoC T1	226	1.80	5.00	4.04	.60	.131 (ns)
QoC T2	225	2.00	5.00	4.03	.61
TJC T1	224	1.00	5.00	3.22	.75	.633 (ns)
TJC T2	223	1.00	5.00	3.20	.73
RJC T1	226	1.00	5.00	3.36	.80	.646 (ns)
RJC T2	224	1.00	5.00	3.32	.82
CJC T1	226	1.00	5.00	3.19	.82	-.948 (ns)
CJC T2	223	1.00	5.00	3.23	.80

Note: QoC, Quality of care; TJC, Task Job Crafting; RJC, Relational Job Crafting; CJC, Cognitive Job Crafting; SD, Standard Deviation; ns, no significant relationship.

Regarding correlations among the variables, task job crafting (T2) was not associated with quality of care, neither at T1 nor at T2. All other variables had significant relationships among them. Additionally, all scales showed good reliability values, above .70 (Table 2).

Table 2

Correlations and alpha coefficients.

Variables	1	2	3	4	5	6	7	8
1. QoC T1	.811
2. QoC T2	.575**	.838
3. TJC T1	.181**	.174**	.777
4. TJC T2	.056	.115	.519**	.753
5. RJC T1	.353**	.240**	.386**	.224**	.761
6. RJC T2	.265**	.319**	.274**	.359**	.595**	.717
7. CJC T1	.267**	.155*	.465**	.358**	.438**	.338**	.842
8. CJC T2	.265**	.226**	.378**	.421**	.330**	.445**	.700**	.866

**p < 0.01. Scale alpha coefficient in the diagonal.

Note: T1, Time 1; T2, Time 2; QoC, Quality of care; TJC, Task Job Crafting; RJC, Relational Job Crafting; CJC, Cognitive Job Crafting.

Test of hypotheses

To test H1 and H2 hierarchical linear regression analyses were performed. As a first step, control variables were included. Secondly, task (H1.1) and relational (H1.2) job crafting in T1 were introduced as antecedents of cognitive job crafting in T2. Similarly, task (H2.1) and relational (H2.2) job crafting in T1 were introduced as antecedents of quality of care in T2.

To test H3, PROCESS macro [43] was used, as it generates total, direct and indirect effects in simple mediation models. We estimated model 4 in PROCESS, with 10,000 bootstrap samples and 95% bias-corrected bootstrap confidence intervals for all indirect effects, controlling for all control variables. We conducted separate analyses for task job crafting (T1) and relational job crafting (T1). Cognitive job crafting in T2 was used as mediator on the relationship between behavioral crafting in T1 and quality of care in T2.

Our results supported H1, as task (βTJC1 = .373, p < .001; Corrected R2 = .138. F = 4.069, p = .001) and relational job crafting (βRJC1 = .406, p < .001; Corrected R2 = .16. F = 4.718, p < .001) in T1 were positively related with cognitive job crafting in T2. No significant effect of the control variables (gender, age, permanent contract, managerial position, night shift) was found.

H2 was partially supported, as only relational job crafting at T1 influenced quality of care at T2 (H2.2) (ΒRJC1 = .286, p = .002; Corrected R2 = .071. F = 2.501, p = .026). H2.1 was not supported (βTJC1 = .166, ns; Corrected R2 = .021. F = 1.405, p = .219), and in both cases no significant effect of the control variables (gender, age, permanent contract, managerial position, night shift) was found.

Finally, both mediation models were supported (H3). We observed a significant indirect effect of task job crafting at T1 (indirect effect = .0601, p < 0.01; SE = .0259; 95% CI: .015 to .1156) (Fig 2) and relational job crafting in T1 (indirect effect = .0404, p < 0.001; SE = .0203; 95% CI: .0056 to .0866) (Fig 3) on quality of care at T2 through cognitive job crafting at T2.

Fig 2

Simple mediation model of CJC (T2) between RJC (T1) and QoC (T2).

Scores are not standardized. *p < 0.05; **p < 0.01; ***p < 0.001.

Fig 3

Simple mediation model of CJC (T2) between TJC (T1) and QoC (T2).

Scores are not standardized. **p < 0.01; ***p < 0.001; ns: non-significant.

Discussion

Implications for theory and research

The purpose of the present study was to analyze the effect of job crafting on quality of care in residential homes for elderly people in two European countries (Spain and Sweden). Our research extends previous studies on job crafting from the perspective of Wrzesniewski and Dutton [9], considering task and relational crafting as behavioral crafting [13], and as antecedents of cognitive crafting.

Additionally, and contrary to the perspective of job demands and resources theory (JD-R) [8, 14], the present research on cognitive job crafting is not considered as a mere passive adaptation to work, but as a proactive behavior of change, as pointed out by Wrzesniewski and Dutton [9] and Slemp and Vella-Brodick [16]. It is considered that employees could change their perceptions about their tasks and relations at work (behavioral crafting) by involving cognitive reframing (cognitive crafting) [13, 18]. Consequently, this process could affect performance [9, 12, 29, 33, 34].

As an added value of this research, we take measures in two waves, to account for the dynamic nature of the phenomenon. Some studies have previously used longitudinal designs but the majority of them used the JD-R model of job crafting [14], as it is the most common model for studying job crafting [6, 44], and none of them introduced cognitive job crafting as consequence of behavioral job crafting. Our results supported the hypothesis that behavioral job crafting at T1 had an effect on cognitive job crafting in T2, as demonstrated by correlations and regression analyses performed.

This result is in line with other studies [15-17], which used general employed population samples and cross-sectional designs. However, a second longitudinal study by Niessen et al. [15], found no relationship between behavioral job crafting at T1 and cognitive job crafting at T2. The authors indicated that the time lag of two weeks may be too short, while in the present study we lasted a year between measures, following the recommendations of other researchers [45].

Secondly, our results supported partially the hypothesis that behavioral job crafting increases quality of care. Specifically, correlations between task job crafting and relational job crafting in T1 with quality of care in T2 were positive and significant, having relational job crafting in T1 the highest relation with quality of care in T2. Nevertheless, regression analyses have shown that only relational job crafting at T1 had a significant effect on quality of care at T2.

Research has shown that individuals who have a strong need for relatedness tend to have collectivist tendencies [46] and to help group members [47]. Leana et al. [12] consider that job crafting can also be a collaborative activity, as it “involves joint effort among employees in the service of changing work process” (p1173). It “is not the work of an individual agent, as described by Wrzesniewski and Dutton, but instead is the work of a dyad or group of employees who together make physical and cognitive changes in the task or relational boundaries of their work" [12 p1173).

Finally, we tested the mediation effect of cognitive job crafting at T2 on the relationship between behavioral job crafting in T1 and quality of care at T2. We observe different results depending on the independent variable (task or relational crafting). In the case of relational job crafting, the total and the direct effects are positive and significant, while in the case of task job crafting only the total effect is significant. Nevertheless, in both cases the results confirmed our hypotheses related to mediation effect of cognitive job crafting (indirect effect).

This result is in relation with our correlational and regression analyses, where only relational job crafting had a significant effect on quality of care. Previous cross-sectional studies [29] indicated that task job crafting had the lowest correlation coefficients with quality of care, even though they were statistically significant. The same study indicated that task and relational job crafting explained similar amounts of variance of quality of care (around 3.9%).

The current research has both, theoretical and practical implications, for understanding the psychological mechanisms underlying the association between behavioral job crafting and quality of care on residential homes for elderly people. To be more specific, the present study develops the perspective of Wrzesniewski and Dutton [9] on job crafting by conceptualizing it as a dynamic process where the three components of the model have different roles when explaining quality of care perceptions.

In this sense, our results suggest that we must differentiate between the two forms of crafting (behavioral and cognitive), not as indicators of the same latent construct, but as aggregates [48], unlike the original model from Wrzesniewski and Dutton [9] and Slemp and Vella-Brodrick [16], who conceive them as indicators of the same supraordinal construct.

In relation to Tim’s and Bakker’s model [28], our study provides evidence on the role that cognitive job crafting plays. Cognitive job crafting is not included in the mentioned model but, according to our study, it mediates the relationship between behavioral crafting, which could be equated with increasing social and structural job resources, and even with the increasing challenging job demands [13], and quality of care.

Our research assesses the effects of a positive organizational intervention. Specifically, our findings indicate that organizations may foster quality of care facilitating job crafting. In this sense, a practical implication of our research is that promoting job crafting in the healthcare sector might be worthwhile. The possibility of allowing workers to job craft should be highly considered by managers and organizations within the healthcare sector, especially if economic contractions are experienced, and considering that the healthcare professionals are working in a demanding environment. Following a psychosocial perspective, we considered important to point out the context key aspects for human resource management. In this sense, it is important to guarantee the working conditions that allow employees turn the job they have into the job they want [49].

Additionally, it is possible to point out the implications for job design, because “recognizing that the strength of individual needs varies across employees and allowing them the opportunity to adjust tasks, relationships, and skills in ways that enable need-fulfillment at work is important” [50 p624].

Managers must acknowledge and understand employees’ perspectives, encourage self-initiative and minimize control wherever possible. In this sense, our research adds empirical evidence at the self-determination theory [51], suggesting how job crafting can facilitate self-satisfaction of employees’ needs. Managers should be aware about their employees’ needs and try to encompass employees’ needs into those behaviors that are most desirable for the organization [52]. In order to achieve this objective, it is important to assure an organizational culture oriented towards employees [53, 54], as relational job crafting has a direct effect on quality of care. In this sense, as pointed out by Berdicchia, Nicolli, and Masino [55] from the JD-R perspective, “managerial interventions aimed at changing the organizational culture may be useful to lower perceived social costs and increase the perceived instrumental value of social exchanges” (p 327).

Secondly, due to the mediation effect of cognitive job crafting, managers should facilitate meaningful work environments. The implications related this result for job design are clear, because to do so, jobs should be re-designed, increasing skills variety, identity and significance [56, 57].

Limitations and future research

Despite the interesting findings we have obtained, our study has some limitations to take into account. First, this research has limitations related to the psychosocial and individual factors considered. Several authors have pointed out as antecedents of job crafting proactive personality [8, 58–62], knowledge, skills, and abilities [62], self-efficacy [58], social support [58], and situational features of accountability, ambiguity, and autonomy [59]. Future research should include some of these variables, either as possible predictors or as control variables.

Secondly, sample characteristics (size and organizational context). This limits the external validity of our findings, as we collected data from twelve organizations from two European countries, even though they represent the north and south of Europe with their contextual particularities. Although the sample obtained cannot be considered as representative due to the non-probability selection system, the demographic data for this sample is similar to that available in Spain (ACRA) [2] and Sweden [3] in terms of gender (mostly women), age, and type of work. In future research other countries should be included, as well as measures about the possible effects of organizational context and/or culture. Additionally, separate analyses by country should be done, but the small Swedish sample (n = 77; 34.1%) in the present research did not allow it.

Thirdly, we collected data on two waves. It would be interesting to have a third wave to confirm the mediation results strongly and to test new hypothesis. In this sense, as a dynamic process, we understand that there can be a reciprocal effect between variables, so the behavioral job crafting would affect the cognitive job crafting and vice versa. Having a third wave would allow to test this hypothesis. Nevertheless, as we have explained previously, the response rate in T2 was 33.3%, decreasing in almost 45%, and the participant organizations declined to collect a third wave.

Fourtly, we focus on the perception of employees’ job crafting and quality of care. Related to job crafting, future research must consider the quality and sustainability of the job crafting developed by employees, in order to tackle the phenomenon in greater depth. Related quality of care, although it is acceptable to collect employees’ perception as an indicator of patient outcome [33, 34] using self-reports, some authors recommend to include alternative external assessment as well as measures about relational and functional aspects of quality [63], providing greater levels of objectivity. In this sense, future research should include employees, managers, users and relatives’ perceptions of quality of care, as well as objective measures as survival (mortality), incidence of disease (morbidity), health-related quality of life issues (ulcer rates, number of users falls), or ratio of residents per employee.

Finally, a limitation of PROCESS is that only one exogenous variable can be entered in a single analysis. That is the reason why we used 2 separate mediation models. Additionally, as we aforementioned, previous research has arisen that the 3 components of job crafting have different relationship dynamics for task and relational job crafting, on the one hand, and cognitive job crafting, on the other. Additionally, as pointed out by Hayes et al. [64], SEM is more suitable with large samples, as it relays on large sample asymptotic theory. For this reason, as we have a sample of 226 participants, we decided to use PROCESS. Nevertheless, in future research with larger samples, the data would be more parsimoniously handled with SEM models.

Data set.

(SAV)

Click here for additional data file.

1 Apr 2020

PONE-D-20-01554

Cognitive job crafting as mediator between behavioral job crafting and quality of care in residential homes for the elderly

PLOS ONE

Dear Dr. Yepes-Baldó,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

We would appreciate receiving your revised manuscript by May 16 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Anthony Montgomery

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: PONE-D-20-01554: Cognitive job crafting as mediator between behavioral job crafting and quality of care in residential homes for the elderly

The present study was a prospective study of job crafting in the residential care industry, examining whether cognitive crafting mediated the relation between behavioral crafting and quality of care. Authors concluded that cognitive crafting was a mediator between these other variables. A strength of the study was the use of prospective data and the unique sample. Notwithstanding these strengths, I have some concerns about the paper that I outline below in case the authors might use the feedback to improve their work:

1. The introduction was too brief and did not adequately build a case for why a job crafting study was needed in this population (residential care staff), and similarly, why cognitive crafting would act as a mediator between the more behavioral aspects of crafting and quality of care. These needs to be unpacked quite substantially, as far too much is left to the reader currently. Questions I had were:

a. Why is cognitive crafting the mediating variable? Why not another dimensions of crafting?

b. Could cognitive crafting precede the more behavioral aspects of crafting?

c. Why wouldn’t all three aspects of job crafting be correlated predictors that are theorized to occur simultaneously?

d. Why residential care staff? I realise they are an important population and the statistcs presented at the beginning of the manuscript illustrate that point, but my question was more around why job crafting would be a particularly helpful strategy with this population? Can you link the two in some way? I feel this would be instrumental in helping to build a more persuasive case for the contribution of the study

2. Regarding the method, please add the inclusion and exclusion critieria for the participants. Also, how do the characteristics of this sample match the population of residential care staff in either Spain or Sweden? To they approximate these groups?

3. There were some aspects of the measures that were not well described. For example, I was confused by this sentence: "The internal consistency for the different Spanish versions of this scale was .85 [32] and .88 [31], and .83 for the Swedish version [32]." Is that in the present study, or in the original publications? What about the validity of the measures?

4. For the analyses, it was odd that the models were tested as separated mediation models using PROCESS. I think the data would be more parsimoniously handled with SEM models, where you could a) include all variables simultaneously, and b) account for attenuation caused by measurement error when you model with latent variables, and c) test competing positions of the variables so that you could better establish which one is better position as a mediator, predictor, etc.

5. On that point above, I would like to see these models compared against a model in which all three job crafting dimensions were treated as the predictor of quality of care. If this mediation model was better, then it at least provides some empirical support behind the authors’ reasoning/hypotheses, but currently there is not much. I recommend SEM as the approach to do this.

6. There is a bit too much describing of the results in the discussion. I was more interested in how this all fits with the existing literature, and a deeper discussion of the practical and theoretical implications. Especially the practical limitations feel very light, with very few details provided. I suggest you expand both of these sections considerably.

All the best with your research!

Reviewer #2: Dear Authors,

Thank you for submitting your manuscript and for the opportunity to review it. I certainly think it has merit but there are a number of recommendations I believe you should consider in order to improve it.

1. The Introduction is somewhat descriptive and could be more critical in reviewing the existing literature. First and foremost, there needs to a clear explanation into what Job Crafting is and why it is important. In particular, the manuscript alludes to a difference between the JDR perspective of crafting and that of Wrzesniewski and Dutton which is used here – however unless a reader is familiar with the job crafting literature many will not know the difference. There isn’t a clear explanation on why job crafting is important.

2. Similarly, how does this fit in with existing theoretical models. Much of the Introduction describes existing relationships (particularly from Line 102 onwards on Page 5), but why do these relationships happen? What is the process behind it? This is particularly the case between crafting and performance (Tims et al., 2015). It would also be useful to reflect on how quality of care is different/ same to existing performance measures, and how this links in with the wider psychosocial working environment (see Teoh et al., 2019).

3. Could you please expand and clarify the sentence “It is considered that cognitive crafting is a mere passive adaptation to work, so it cannot be conceived as crafting since it is not a proactive behavior of change” please? I’m not familiar with this perspective. From what I understand it appears that Zhang and Parker are saying that it isn’t job crafting, but then in the next sentence argues that is still is?

4. Page 6, 113. What are the recommendations from Westerberg and Tafvelin and how is it relevant to the point being made? When I first read this paragraph I understood it that the intention was to use a single item measure of quality of care, but this is not the case as the study uses a 5 item measure instead.

5. Page 6, line 115. It would be useful to consider reiterating these reasons, or then making them clearer earlier on, as it is not really evident.

6. Its great, a real strength, to be able to carry out a longitudinal study. Nevertheless there needs to be a rationale and description of this in the Introduction. In particularly, within the hypotheses need to make clear whether the measures are at T1 or T2 because as they currently are this appears to be a cross-sectional study.

7. Within the material section, it is not clear if the surveys administered were done so in Swedish or in English. For the JCQ in Spanish, I don’t understand the two different Spanish versions. There is a Spanish adaption [31] that wasn’t used which I presume because it came out after the study began? Bu then the version that was used in this study, was also used in a previous study [4], so why is there a need to mention the new version develop in [31]?

8. It isn’t clear if the Cronbach Alpha’s in the material section refers to that from previous studies or from the current study. I understand it as the former but I am not sure.

9. Page 9, Line 197 – what does medium to high mean? Are there established thresholds these are compared against?

10. I believe that the data from Sweden and Spain were mixed-together which is fine, but I think it would be important to control for this within the analysis. Considering the longitudinal design, the relationships between cognitive crafting and quality of care still are cross-sectional. At the least, cognitive crafting at T1 should be tested as a predictor of quality of care at T2. Building on the Introduction, there needs to be some rationale as to why cognitive at crafting at T2 was used as the mediator (opposed to at T1).

11. Considering that job crafting is strongly influenced by psychosocial and individual factors (Rudolph et al., 2016), I am left to wonder if these could/should have been included within this study. Either as possible predictors or then as control variables. At the very least, they should be acknowledged within the manuscript and discussed.

I hope you find the comments above constructive and I wish you all the best.

References

Rudolph CW, Katz IM, Lavigne KN, Zacher H. Job crafting: A meta-analysis of relationships with individual differences, job characteristics, and work outcomes. J Vocat Behav. 2017;102(314):112–38.

Teoh, K., Hassard, J., & Cox, T. (2019). Doctors’ perceived working conditions and the quality of patient care: a systematic review. Work & Stress, 33(4), 385-413.

Tims M, Bakker AB, Derks D. Job crafting and job performance: A longitudinal study. Eur J Work Organ Psychol. 2015;24(6):914–28

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

7 May 2020

As indicated in the Response to reviewers, we include here our answers:

Reviewer #1: PONE-D-20-01554: Cognitive job crafting as mediator between behavioral job crafting and quality of care in residential homes for the elderly

The present study was a prospective study of job crafting in the residential care industry, examining whether cognitive crafting mediated the relation between behavioral crafting and quality of care. Authors concluded that cognitive crafting was a mediator between these other variables. A strength of the study was the use of prospective data and the unique sample.

Thank you for your comments!

Notwithstanding these strengths, I have some concerns about the paper that I outline below in case the authors might use the feedback to improve their work:

1. The introduction was too brief and did not adequately build a case for why a job crafting study was needed in this population (residential care staff), and similarly, why cognitive crafting would act as a mediator between the more behavioral aspects of crafting and quality of care. These needs to be unpacked quite substantially, as far too much is left to the reader currently. Questions I had were:

a. Why is cognitive crafting the mediating variable? Why not another dimensions of crafting?

Zhang and Parker (2018) analyze Bruning and Campion's (2018) proposal, showing that behavioral and cognitive crafting are closely related and are not mutually exclusive dimensions.

The differentiation between behavioral and cognitive crafting arises from the existing debate between what should or should not be considered job crafting. Precisely, from the perspective of demands and resources, cognitive crafting is considered a mere passive adaptation to work, so it cannot be conceived as crafting since it is not a proactive behavior of change (Bakker, Tims, & Derks, 2012; Tims & Bakker, 2010).

In contrast, the model by Zhang and Parker (2018), following Wrzesniewski and Dutton (2001), includes cognitive crafting since “it involves altering how one frames or views their tasks or job, which is self-initiated , self-targeted, intentional, and represents meaningful changes to the job aspects ”(Zhang and Parker 2018, p.5).

In this sense, it is important to differentiate between the two forms of crafting (behavioral and cognitive), not as indicators of the same latent construct, but as aggregates.

The results of previous studies [9-11] reveal different relationship dynamics for task and relational job crafting, on the one hand, and cognitive job crafting, on the other. Additionally, recent studies also suggest that behavioral crafting (task and relational) can cause changes at the cognitive level (Unsworth, Mason, & Jones, 2004; Zhang & Parker, 2018).

b. Could cognitive crafting precede the more behavioral aspects of crafting?

Yes, it could. As we indicated in Limitations and future research section, we collected data on two waves. It would be interesting to have a third wave to confirm the mediation results strongly and to test new hypothesis. In this sense, as a dynamic process, we understand that there can be a reciprocal effect between variables, so the behavioral job crafting would affect the cognitive job crafting and vice versa. Having a third wave would allow to test this hypothesis. Nevertheless, as we have explained, the response rate in T2 was 33.3%, decreasing in almost 45%, and the participant organizations declined to collect a third wave.

c. Why wouldn’t all three aspects of job crafting be correlated predictors that are theorized to occur simultaneously?

From the Wrzesniewski and Dutton (2001) perspective, this is the original relationships model. The novelty of our research is theoretically based on Unsworth, Mason, & Jones (2004) and Zhang & Parker (2018) revisions.

Various studies have linked job crafting with performance, including the quality of service offered by employees (Lichtehthaler and Fischbach, 2019; Rofcanin, Bakker, Berber, Gölgeci and las Heras, 2018; Rudolph et al., 2017; Tims et al., 2015; Yepes-Baldó et al., 2018). Compared to previous studies, the main novelty of our proposal lies in the fact that we analyze job crafting as a dynamic process and differentiate between behavioral job crafting, as an antecedent, and cognitive job crafting as a consequence, and as a mediator between it and the quality of care .The results of previous studies [9-11] reveal different relationship dynamics for task and relational job crafting, on the one hand, and cognitive job crafting, on the other.

Additionally, recent studies also suggest that behavioral crafting (task and relational) can cause changes at the cognitive level (Unsworth, Mason, & Jones, 2004; Zhang & Parker, 2018).

d. Why residential care staff? I realize they are an important population and the statistics presented at the beginning of the manuscript illustrate that point, but my question was more around why job crafting would be a particularly helpful strategy with this population? Can you link the two in some way? I feel this would be instrumental in helping to build a more persuasive case for the contribution of the study

The last few years, the healthcare sector has been characterized by reducing and slow growth of costs, but healthcare professionals need an adequate amount of resources to reach their work-related goals and offer high-quality care (French, Ikenwilo, & Scott, 2007). In this context, several authors, as Gordon, Demerouti, Le Blanc, & Bipp (2015), recognize the need for healthcare professionals to take a proactive role in shaping their future jobs to improve healthcare systems, and “job crafting is particularly interesting for health care organizations because, on one hand it can be to assure improvements on the quality of care they provide [4], and on other hand, this proactive role “can be learned and effectively transferred from training to organizational practice” (Bakker 2018, 321).

2. Regarding the method, please add the inclusion and exclusion critieria for the participants. Also, how do the characteristics of this sample match the population of residential care staff in either Spain or Sweden? To they approximate these groups?

All the employees who were effectively working (not on leave) and who did so voluntarily participate at T1 and T2. For the final database, only the participants who participated at both T1 and T2 were included.

Although the sample obtained cannot be considered as representative due to the non-probability selection system, the demographic data for this sample is similar to that available in Spain (ACRA) and Sweden (FORTE) in terms of gender (mostly women), age, and type of work.

3. There were some aspects of the measures that were not well described. For example, I was confused by this sentence: "The internal consistency for the different Spanish versions of this scale was .85 [32] and .88 [31], and .83 for the Swedish version [32]." Is that in the present study, or in the original publications? What about the validity of the measures?

There are different studies that use the scales in their Spanish and Swedish versions. Validity about the different scales can be seen in these studies. The studies to which we refer have been clarified in the text.

While redacting the present paper a new Spanish version was published. The Spanish versions of the instrument used in the present study [4] and in the new one are very similar. Only small changes on translation have been observed. The internal consistency for the Spanish version of this scale was .85 and .88, and .83 for the Swedish version. We considered important to mention the new published version of the instrument and its similarities with the version we used in our study.

4. For the analyses, it was odd that the models were tested as separated mediation models using PROCESS. I think the data would be more parsimoniously handled with SEM models, where you could a) include all variables simultaneously, and b) account for attenuation caused by measurement error when you model with latent variables, and c) test competing positions of the variables so that you could better establish which one is better position as a mediator, predictor, etc.

Hayes (2013) showed the similarity in results between PROCESS and an SEM program. Additionally, PROCESS estimates each equation included in a model to be tested separately (Hayes, Montoya, & Rockwood, 2017).

Secondly, as pointed out by Hayes et al. (2017), SEM is more suitable with large samples, as it relays on large sample asymptotic theory. For this reason, as we have a sample of 226 participants, we decided to use PROCESS.

Related to latent variables, we reduced them to observed variable proxies (averages of indicators), which make them, by definition, observed and not latent (Hayes et al, 2017).

In relation to measurement error, any analysis that can be expressed in the form of a linear regression model is not exempt of it, including SEM. On the other hand, latent variable mediation analysis may be more accurate in the estimation of effects than observed variable analysis, but less powerful in detecting them (Hayes et al, 2017).

A limitation of PROCESS is that only one exogenous variable can be entered in a single analysis. That is the reason why we used 2 separate mediation models. Additionally, as we aforementioned, previous research has arisen that the 3 components of job crafting have different relationship dynamics for task and relational job crafting, on the one hand, and cognitive job crafting, on the other. So, our decision is based on theoretical and empirical reasons.

5. On that point above, I would like to see these models compared against a model in which all three job crafting dimensions were treated as the predictor of quality of care. If this mediation model was better, then it at least provides some empirical support behind the authors’ reasoning/hypotheses, but currently there is not much. I recommend SEM as the approach to do this.

As we have different hypothesis to test, we selected PROCESS as analytical strategy to analyses the data on a more parsimonious way. The reviewer’s proposal is related to measures of fit of the model to the data, whereas PROCESS do not offer this kind of measures. However, the model built into PROCESS is saturated, so fit by some measures would be perfect when this model is estimated using SEM. Furthermore, as stated by MacCallum et al. (1993) any good fitting model typically has several minor variations with different interpretations that fit equally well (the equivalent models’ problem).

6. There is a bit too much describing of the results in the discussion. I was more interested in how this all fits with the existing literature, and a deeper discussion of the practical and theoretical implications. Especially the practical limitations feel very light, with very few details provided. I suggest you expand both of these sections considerably.

Relation with existing literature

Research has shown that individuals who have a strong need for relatedness tend to have collectivist tendencies (Baumeister & Leary, 1995) and to help group members (Den Hartog, De Hoogh, & Keegan, 2007). Leana et al., (2009) consider that job crafting can also be a collaborative activity, as it “involves joint effort among employees in the service of changing work process” (2009, p1173). It “is not the work of an individual agent, as described by Wrzesniewski and Dutton, but instead is the work of a dyad or group of employees who together make physical and cognitive changes in the task or relational boundaries of their work" (Leana et al, 2009, 1173).

Practical implications

Our research assesses the effects of a positive organizational intervention. Specifically, our findings indicate that organizations may foster quality of care facilitating job crafting. In this sense, a practical implication of our research is that promoting job crafting in the healthcare sector might be worthwhile. The possibility of allowing workers to job craft should be highly considered by managers and organizations within the healthcare sector, especially if economic contractions are experienced, and taking into account that the healthcare professionals are working in a demanding environment.

Additionally, following Bindl, Unsworth, Gibson, & Stride (2019), it is possible to point out the implications for job design, because “recognizing that the strength of individual needs varies across employees and allowing them the opportunity to adjust tasks, relationships, and skills in ways that enable need-fulfillment at work is important” (2019, 42-43).

Managers must acknowledge and understand employees’ perspectives, encourage self-initiative and minimize control wherever possible. In this sense, our research adds empirical evidence at the self-determination theory (Deci, Connell & Ryan, 1989), suggesting how job crafting can facilitate self-satisfaction of employees’ needs. Managers should be aware about their employees’ needs and try to encompass employees’ needs into those behaviors that are most desirable for the organization (Gagné 2003). In order to achieve this objective, it is important to assure an organizational culture oriented towards employees [35,36].

Limitation

Our research did not consider the quality and sustainability of the job crafting developed by employees, as we have focused on employees’ job crafting and quality of care perception.

Reviewer #2: Dear Authors,

Thank you for submitting your manuscript and for the opportunity to review it. I certainly think it has merit but there are a number of recommendations I believe you should consider in order to improve it.

Thank you for your comments!

1. The Introduction is somewhat descriptive and could be more critical in reviewing the existing literature. First and foremost, there needs to a clear explanation into what Job Crafting is and why it is important. In particular, the manuscript alludes to a difference between the JDR perspective of crafting and that of Wrzesniewski and Dutton which is used here – however unless a reader is familiar with the job crafting literature many will not know the difference. There isn’t a clear explanation on why job crafting is important.

Job crafting has been defined by Wrzesniewski and Dutton (2001) as the “physical and cognitive changes individuals make in the task or relational boundaries of their work” (p 179).

Following Berg, Grant and Johnson (2010) this proposal is the understanding that employees are often interested in customizing their jobs to fit optimally their motivations, competences and desires. Organization’s management should not unilaterally decide how its employees spend their time and energy. Rather, the employees themselves should be allowed to decide what to do, creating and searching a comfortable and enjoyable context, over and beyond the job descriptions provided by the management, particularly in complex and uncertain situations (Wrzesniewski & Dutton, 2001; Ghitulescu, 2006; Leana, Appelbaum, & Shevchuk, 2009; Berg, Grant & Johnson., 2010). In this sense it is possible to consider that the job crafting emerges as a strategy for the leverage of work meaning and identity.

2. Similarly, how does this fit in with existing theoretical models. Much of the Introduction describes existing relationships (particularly from Line 102 onwards on Page 5), but why do these relationships happen? What is the process behind it? This is particularly the case between crafting and performance (Tims et al., 2015).

Our main objective is to test a possible mechanism to explain the relationship between job crafting and performance (in this case, quality of care) using cognitive job crafting as a mediator.

Various studies have linked job crafting with performance, including the quality of service offered by employees (Lichtehthaler and Fischbach, 2019; Rofcanin, Bakker, Berber, Gölgeci and las Heras, 2018; Rudolph et al., 2017; Tims et al., 2015; Yepes-Baldó et al., 2018). Compared to previous studies, the main novelty of our proposal lies in the fact that we analyze job crafting as a dynamic process and differentiate between behavioral job crafting, as an antecedent, and cognitive job crafting as a consequence, and as a mediator between it and the quality of care .The results of previous studies [9-11] reveal different relationship dynamics for task and relational job crafting, on the one hand, and cognitive job crafting, on the other.

Additionally, recent studies also suggest that behavioral crafting (task and relational) can cause changes at the cognitive level (Unsworth, Mason, & Jones, 2004; Zhang & Parker, 2018).

2.1. It would also be useful to reflect on how quality of care is different/ same to existing performance measures, and how this links in with the wider psychosocial working environment (see Teoh et al., 2019).

During the past decades, performance indicators have become increasingly sophisticated (Majeed, Lester, Bindman, 2007), and the quality of care has been included as an indicator of performance (Westerberg and Tafvelin, 2014). Nevertheless, Marshall, Roland Brook, McGlynn and Shekelle (2003) consider that it is difficult to compare quality of care and to transfer performance indicators directly between different health systems and cultures, because each country has established different indicators. However, the quality of care is used as an indicator of patients’ outcome in studies focused in hospital care (Aiken, Clarke & Slone, 2002) and in the geriatric care (Lapointe McKenzie, Blandford, Menec, Boltz & Capezuti, 2011).

Considering the complexity of the healthcare sector Teoh, Hassard & Cox (2019) developed a systematic review on the relationship between the working conditions and the quality of patient care. The results show, on the one hand, several investigations where an improved work environment exerted a positive effect on nurses assessed quality of care. For instance, promotion prospects, perceived salary and job security (Loerbrokset, Weigl, & Angerer, 2016. Weigl, Schneider, Hoffmann, & Angerer, 2015, Purdy, Spence Laschinger, Finegan, Kerr & Olivera, 2010) were positively related with quality of care. Nevertheless, these authors concluded that the relationships between quality of care and work environment reported divergent findings, showing the complexity of these relationships (Hannan et al. 2001, Hasson & Arnetz 2009). They recommended, among other aspects, to use multilevel or longitudinal designs, as well as mediating and moderating variables, in order to present a more realistic interpretation of these relationships.

Quality of care could be analyzed from the nurses’ perception as an indicator of patient outcome [27,28]. The present study uses this perspective to analyze this variable

3. Could you please expand and clarify the sentence “It is considered that cognitive crafting is a mere passive adaptation to work, so it cannot be conceived as crafting since it is not a proactive behavior of change” please? I’m not familiar with this perspective. From what I understand it appears that Zhang and Parker are saying that it isn’t job crafting, but then in the next sentence argues that is still is?

The differentiation between behavioral and cognitive crafting arises from the existing debate between what should or should not be considered job crafting. Precisely, from the perspective of demands and resources, cognitive crafting is considered to be a mere passive adaptation to work, so it cannot be conceived as crafting since it is not a proactive behavior of change (Bakker, Tims, & Derks, 2012; Tims & Bakker, 2010).

In contrast, the model by Zhang and Parker (2018), following Wrzesniewski and Dutton (2001), includes cognitive crafting since “it involves altering how one frames or views their tasks or job, which is self-initiated , self-targeted, intentional, and represents meaningful changes to the job aspects ”(Zhang and Parker, 2018, p.5).

In this sense, it is important to differentiate between the two forms of crafting (behavioral and cognitive), not as indicators of the same latent construct, but as aggregates.

4. Page 6, 113. What are the recommendations from Westerberg and Tafvelin and how is it relevant to the point being made? When I first read this paragraph, I understood it that the intention was to use a single item measure of quality of care, but this is not the case as the study uses a 5-item measure instead.

Even though it is not unusual to assess quality of care with a single-item [29], we follow Westerberg and Tafvelin [30] recommendations, adding information about satisfaction “with the way in which the clients were treated, kept informed and their wishes respected” [30 p464]. That is the reason why we use a scale with 5 items.

5. Page 6, line 115. It would be useful to consider reiterating these reasons, or then making them clearer earlier on, as it is not really evident.

We hope that the changes made in the Theoretical framework and hypotheses section could clarify this point.

6. It’s great, a real strength, to be able to carry out a longitudinal study. Nevertheless, there needs to be a rationale and description of this in the Introduction. In particularly, within the hypotheses need to make clear whether the measures are at T1 or T2 because as they currently are this appears to be a cross-sectional study.

We agree. We changed our hypothesis to include time of each measure. Additionally, we justified our hypothesis theoretically

7. Within the material section, it is not clear if the surveys administered were done so in Swedish or in English. For the JCQ in Spanish, I don’t understand the two different Spanish versions. There is a Spanish adaption [31] that wasn’t used which I presume because it came out after the study began? Bu then the version that was used in this study, was also used in a previous study [4], so why is there a need to mention the new version develop in [31]?

Questionnaires were administered in Spanish and Swedish. While redacting the present paper a new Spanish version was published. The Spanish versions of the instrument used in the present study and in the new one are very similar. Only small changes on translation have been observed. The internal consistency for the Spanish versions of this scale was .85 and .88, and .83 for the Swedish version. We consider it is important to mention the new published version of the instrument and its similarities with the version we used in our study.

8. It isn’t clear if the Cronbach Alpha’s in the material section refers to that from previous studies or from the current study. I understand it as the former, but I am not sure.

We have clarified this in the text, including both, the former and the present study alpha’s

9. Page 9, Line 197 – what does medium to high mean? Are there established thresholds these are compared against?

As we used a 5-point Likert scale, we consider 3 or higher a medium to high score. We indicated this in the text.

10. I believe that the data from Sweden and Spain were mixed-together which is fine, but I think it would be important to control for this within the analysis.

It is correct, Sweden and Spain data were mixed-together. Unfortunately, the small number of cases in Sweden made it impossible to control for this. We include this in limitations section

10.1. Considering the longitudinal design, the relationships between cognitive crafting and quality of care still are cross-sectional. At the least, cognitive crafting at T1 should be tested as a predictor of quality of care at T2. Building on the Introduction, there needs to be some rationale as to why cognitive at crafting at T2 was used as the mediator (opposed to at T1).

To analyze the mediation effect of CJC it is more accurate to use T2, as the antecedent (behavioral JC) is measured in T1. Nevertheless, it would be interesting to have a third wave to confirm the mediation results strongly and to test new hypothesis. In this sense, as a dynamic process, we understand that there can be a reciprocal effect between variables, so the behavioral job crafting would affect the cognitive job crafting and vice versa. Having a third wave would allow to test this hypothesis. Unfortunately, as we have explained previously, the response rate in T2 was 33.3%, decreasing in almost 45%, and the participant organizations declined to collect a third wave.

11. Considering that job crafting is strongly influenced by psychosocial and individual factors (Rudolph et al., 2016), I am left to wonder if these could/should have been included within this study. Either as possible predictors or then as control variables. At the very least, they should be acknowledged within the manuscript and discussed.

Certainly, several author have pointed out as antecedents of job crafting proactive personality (Parker, Bindl, & Strauss, 2010; Crant, 2000, Tims, & Bakker, 2010; Grant & Ashford, 2008; Vermooten, Boonzaier, & Kidd, , 2019), knowledge, skills, and abilities (Plomp,Tims,Akkermans, Khapova, Jansen, & Bakker, 2016), self-efficacy (Crant, 2000), social support (Crant, 2000), and situational features of accountability, ambiguity, and autonomy (Grant and Ashford, 2008). Future research should include these variables, either as possible predictors or as control variables.

12. I hope you find the comments above constructive and I wish you all the best.

Your comments are really interesting, and they have allowed us to improve our paper. We hope our changes make our paper suitable for publication in PlosOne.

Submitted filename: Response to reviewers.docx

Click here for additional data file.

18 Sep 2020

PONE-D-20-01554R1

Cognitive job crafting as mediator between behavioral job crafting and quality of care in residential homes for the elderly

PLOS ONE

Dear Dr. Yepes-Baldó,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

The manuscript has improved, but the authors need to address the issues raised by both reviewers. In addition to the comments of the reviewers (at the bottom of this letter), I would like the authors to address the following points concerning their work:

1.     The authors have not yet adequately explained why  cognitive crafting would act as a mediator between the more behavioral aspects of crafting and quality of care.  Could cognitive crafting precede the more behavioral aspects of crafting?

2.    The authors have not yet answered the question of a previous reviewer.  For the analyses, why were the models tested as separated mediation models using PROCESS? As noted by a previous reviewer,  the data would be more parsimoniously handled with SEM models, where they could a) include all variables simultaneously, and b) account for attenuation caused by measurement error when you model with latent variables, and c) test competing positions of the variables so that you could better establish which one is better position as a mediator, predictor, etc. So, the authors should address this issue.

3.     What are the practical implications for the people involved in their research? How should they change the way they work based on the results of the research? We need to know the implications of the research for work practices.

==============================

Please submit your revised manuscript by Nov 02 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Anthony Montgomery

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: (No Response)

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: No

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: Thank you for your revisisions and addressing my previous comments. These are mostly addressed and I have a few more clarification points:

The inclusion of the Wrzesniewski and Dutton definition should be integrated into one of the paragraphs and not be a standalone sentence.

I don’t feel the second original point on why do these relationships happen (from line 80 onwards) is evident. The revisions are still descriptive, although the work of Ashforth and Kreiner helps with this. But this could still be unpacked further, and especially in explaining how and why job crafting is associated with quality of care.

In terms of the Teoh et al. paper, the intention was not so much to summarise the findings of it, as actually the description and links between psychosocial working conditions and quality of care is not directly relevant. Instead, that paper highlights how quality of care can be operationsed in different ways and that it rather complex. What is needed here is for the paper to consider what is quality of care and how it is measured here relates to performance more generally and/or other measures of quality of care. The revised new paragraph (Line 131, During the past decades…) does a decent job at this, but it needs to better link in with the next paragraph on the review. This paper does highlight the review’s call for longitudinal and mediating research, which this study actually answers, and it would be good to use this as additional justification to strengthen the need for this study.

With regards to the point “As we used a 5-point Likert scale, we consider 3 or higher a medium to high score. We indicated this in the text”, I don’t think this is needed. What is the justification of the 3 and above as a medium to high score? Is this a cut-off determined as part of the questionnaire or something that was decided here? If there is no set comparison (normed) score than its not appropriate to say its high.

There are elements of duplication that can be removed from this paper. For example there is no need to describe the five-point QoC scale in the Introduction and then again in the Method section. Equally if Cronbach Alpha scores are available in the Table 2 and it is mentioned that they all have good reliability then there is no need to mention them also in the measures section.

For H1 and H2, was it hierarchical multiple regressions that were carried out or simple linear regressions? If it is the former, which it should be, then this should be made clearer and the control variables also acknowledged.

Cognitive job crafting T2 should be made clearer within the hypotheses. While ideally there would have been three time points (but two is really good as well – especially as it’s a year apart), as this is not the case here there should be some consideration as to why cognitive job crafting at Time 2 was used and not that from Time 1.

Page 4 line 66: “the healthcare sector has been characterized by reducing and slow growth of costs” – this sentence isn’t clear, is it reducing or is it having slow growth?

Reviewer #3: I consider this to be a very interesting job as it applies job crafting to a group of professionals who were traditionally analyzed from the wrong perspectives.

The design of the work is very appropriate as it combines the cross-cultural perspective with an exhaustive control over the variables of the study.

I have a question regarding the tests they have used:

a. In which languages ​​have the work questionnaires been applied? Spanish and Swedish? (page 10)

b. Did they apply the questionnaires in "paper and pencil" format? (page 9)

c. With an "online" application of the questionnaires could they have achieved greater success in answering? (page 11)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

2 Oct 2020

ANSWERS TO EDITOR

Following the instructions, we attached the answers to reviewers file to our submission, including our answer to the editor comments. Nevertheless, you can see here our answers to editor comments again.

COMMENT 1. The authors have not yet adequately explained why cognitive crafting would act as a mediator between the more behavioral aspects of crafting and quality of care. Could cognitive crafting precede the more behavioral aspects of crafting?

AUTHORS ANSWER: Recent studies suggest that behavioral crafting (task and relational) can cause changes at the cognitive level (Unsworth, Mason, & Jones, 2004; Zhang & Parker, 2018). In this sense, Unsworth et al. (2004) pointed out that “group members cognitively frame their comparisons such that they compare themselves with other groups on dimensions that represent in-group strengths (…). For example, members of a highly productive group would tend to evaluate groups in terms of their productivity, but a less productive group might choose to focus on other attributes, such as friendliness” (p.5). In our research, following this suggestion, we consider that employees could cognitive reframe the meaning of their jobs by means of the tasks (Task crafting) they develop and the relationships they have at work (Relational crafting) (Theoretical framework and hypotheses section)

Various studies have linked job crafting with performance, including the quality of service offered by employees (Lichtehthaler and Fischbach, 2019; Rofcanin, Bakker, Berber, Gölgeci and las Heras, 2018; Rudolph et al., 2017; Tims et al., 2015; Yepes-Baldó et al., 2018). Compared to previous studies, the main novelty of our proposal lies in the fact that we analyze job crafting as a dynamic process and differentiate between behavioral job crafting (tasks and relational crafting), as an antecedent, and cognitive job crafting as a consequence, and as a mediator between it and the quality of care .The results of previous studies [9-11] reveal different relationship dynamics for task and relational job crafting, on the one hand, and cognitive job crafting, on the other.

COMMENT 2. The authors have not yet answered the question of a previous reviewer. For the analyses, why were the models tested as separated mediation models using PROCESS? As noted by a previous reviewer, the data would be more parsimoniously handled with SEM models, where they could a) include all variables simultaneously, and b) account for attenuation caused by measurement error when you model with latent variables, and c) test competing positions of the variables so that you could better establish which one is better position as a mediator, predictor, etc. So, the authors should address this issue.

AUTHORS ANSWER: Hayes (2013) showed the similarity in results between PROCESS and an SEM program. Additionally, PROCESS estimates each equation included in a model to be tested separately (Hayes, Montoya, & Rockwood, 2017).

Secondly, as pointed out by Hayes et al. (2017), SEM is more suitable with large samples, as it relays on large sample asymptotic theory. For this reason, as we have a sample of 226 participants, we decided to use PROCESS.

Related to latent variables, we reduced them to observed variable proxies (averages of indicators), which make them, by definition, observed and not latent (Hayes et al, 2017).

In relation to measurement error, any analysis that can be expressed in the form of a linear regression model is not exempt of it, including SEM. On the other hand, latent variable mediation analysis may be more accurate in the estimation of effects than observed variable analysis, but less powerful in detecting them (Hayes et al, 2017).

A limitation of PROCESS is that only one exogenous variable can be entered in a single analysis. That is the reason why we used 2 separate mediation models. Additionally, as we aforementioned, previous research has arisen that the 3 components of job crafting have different relationship dynamics for task and relational job crafting, on the one hand, and cognitive job crafting, on the other. So, our decision is based on theoretical and empirical reasons.

COMMENT 3. What are the practical implications for the people involved in their research? How should they change the way they work based on the results of the research? We need to know the implications of the research for work practices.

AUTHORS ANSWER: Following a psychosocial perspective, we considered important to point out the context key aspects for human resource management. In order to allow people “to change the way their work” we focused on the organizational context and managerial practices. We consider that it is important to guarantee the working conditions that allow employees turn the job they have into the job they want (Wrzesniewski , Berg, and Dutton, 2010).

We added this consideration in the text (Implications for theory and research).

Submitted filename: Responses to reviewers.docx

Click here for additional data file.

30 Nov 2020

Cognitive job crafting as mediator between behavioral job crafting and quality of care in residential homes for the elderly

PONE-D-20-01554R2

Dear Dr. Yepes-Baldó,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Ting Ren

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: (No Response)

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

Reviewer #3: (No Response)

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: Dear Authors,

Thank you for addressing the comments from the last revision. I only have one minor point. The included sentence in line 120 with the quote by Unsworth et al. could perhaps be paraphrased and not kept as a direct quotation. Also, both sentence start of with point/pointed out.

Returning to the point on PROCESS mediation and SEM, your explanation is fine but I still think SEM has its advantages and not doing so remains a limitation that should be acknowledged.

Reviewer #3: The work responds to the quality that is demanded in the magazine. The authors' answers to my questions are correct and adequate

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

Reviewer #3: No

3 Dec 2020

PONE-D-20-01554R2

Cognitive job crafting as mediator between behavioral job crafting and quality of care in residential homes for the elderly

Dear Dr. Yepes-Baldó:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Ting Ren

Academic Editor

PLOS ONE