Master of International Tourism and Hospitality Management -journal article and write one paragraph that explains what it is about and what evidence is used to support the argument.

Journal of Leadership & Organizational Studies 2014, Vol. 21(2) 141 –149 © The Authors 2013 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1548051813515516 jlo.sagepub.com

Article

Since 2005 research on what has been called positive psy- chological capital (or simply PsyCap) has, to use a term from positive psychology, “flourished.” This multidimen- sional construct with antecedents in positive psychology is defined as

an individual’s positive psychological state of development characterized by: (1) having confidence (self-efficacy) to take on and put in the necessary effort to succeed at challenging tasks; (2) making a positive attribution (optimism) about succeeding now and in the future; (3) persevering toward the goals, and when necessary, redirecting paths to goals (hope) in order to succeed; and (4) when beset by problems and adversity, sustaining and bouncing back and even beyond (resilience) to attain success. (Luthans, Youssef, & Avolio, 2007, p. 3)

A recent meta-analysis (Avey, Reichard, Luthans, & Mhatre, 2011) found dozens of empirical studies which yielded sig- nificant predictive validity of PsyCap in individual out- comes. However, this meta-analysis also revealed a major omission in the theoretical development and empirical research on PsyCap. Avey, Reichard, and colleagues (2011) note they “found very few studies that measured anything pertaining to the formation of PsyCap. In other words, few have considered what is ‘to the left’ of PsyCap (i.e., the ante- cedents in a theoretical model).” After an exhaustive litera- ture review they also note “there has been no systematic method of examining antecedents to PsyCap, which sug- gests this may be a fruitful area of future research” (p. 148).

The purpose of this article is to begin to offer a system- atic consideration of the antecedents of PsyCap. In other words, at this point there is ample evidence to suggest PsyCap is a positive construct in organizations with useful predictive validity. However, this article addresses the need to understand what systems and structures within persons and organizational life are predictors of PsyCap. While this has been partially considered by the well-known work of Bandura (1997) on self-efficacy, there has been no overall consideration of these beginnings for overall PsyCap as this work was limited to self-efficacy specifically. Furthermore, an understanding of how PsyCap forms and the correspond- ing antecedents can offer insight into organizational poli- cies, human resource management systems, management structures, and leadership practices that enhance overall employee PsyCap for the benefit of the person and the firm.

Positive Psychological Capital

Beyond the definition of PsyCap, previous research has found at least seven boundary conditions and characteristics of the construct that are useful in an operational

515516 JLOXXX10.1177/1548051813515516Journal of Leadership & Organizational StudiesAvey research-article2013

1Central Washington University, Ellensburg, WA, USA

Corresponding Author: James B. Avey, College of Business, Central Washington University, 400 E. University Way, Ellensburg, WA 98926, USA. Email: [email protected]

The Left Side of Psychological Capital: New Evidence on the Antecedents of PsyCap

James B. Avey1

Abstract A recent meta-analysis suggests since 2005 there have been dozens of studies with positive psychological capital (PsyCap) predicting optimal individual performance, behaviors (e.g., citizenship, deviance), and attitudes (e.g., satisfaction, commitment, well-being, turnover intentions). However, in reviewing this literature there is an obvious silence as to the antecedents of PsyCap. Outside the examination of developmental interventions, very little is known about how and why an individual reaches and stabilizes at a given level of PsyCap. To address this gap, this study seeks to better understand the antecedents of individual-level PsyCap. In a field study of 1,264 engineers and technicians, a categorical perspective on PsyCap antecedents is tested. In a second study of 529 Chinese technology employees, the results from Study 1 are replicated with less overall variance explained raising measurement and cultural boundary conditions of current tools employed for empirical research on PsyCap. Implications for managerial practices conclude the article.

Keywords PsyCap, psychological capital, organizational behavior

at PENNSYLVANIA STATE UNIV on March 5, 2016jlo.sagepub.comDownloaded from

142 Journal of Leadership & Organizational Studies 21(2)

understanding. First, PsyCap is what Law, Wong, and Mobley (1998) refer to as a multidimensional construct. PsyCap is not a single dimension alone (e.g., simply opti- mism) but rather the shared variance of the four dimensions. Evidence for such constructs can be found in Hobfoll’s (2002) psychological resource theory, which suggests psy- chological constructs are members of a broader domain. In this case, the four dimensions of hope, efficacy, optimism, and resilience are members of a broader construct called PsyCap. This was demonstrated empirically by Luthans, Avolio, Avey, and Norman (2007) and in predictive validity by means of a usefulness analysis. A second characteristic of PsyCap is that it is domain specific, usually operational- ized at the work domain. This characteristic implies an indi- vidual may be high in work PsyCap but low in, for example, family or job searching PsyCap. In other words, an indi- vidual may have higher PsyCap to accomplish work goals and be resilient to setbacks at work, but in another domain (e.g., family) an individual may be low in hope and struggle to be resilient to such personal setbacks.

A third characteristic of PsyCap refers to the stability of the construct. Test–retest validity work by Luthans, Avolio, and colleagues (2007) shows PsyCap to be more stable than emotions but more open to change than personality. Developmental interventions (e.g., Luthans, Avey, Avolio, & Peterson, 2010) that show an increase in the PsyCap of participants support this notion that PsyCap is developable, although the sustaining nature of this developmental change has yet to be tested over time. This state-like characteristic is important to this study given the investigation into how PsyCap develops. A fourth characteristic of this construct is the self-opinion operationalization. Although some studies have considered rating others’ (e.g., the leader’s) PsyCap (e.g., Norman, Avolio, & Luthans, 2010), the primary oper- ationalization of PsyCap has been from the self. Given that by definition PsyCap is an individual’s state of develop- ment, ratings of what Person A believes Person B’s PsyCap to be have been less frequent.

A fifth characteristic of PsyCap, and one originally stated in Luthans’s positive organizational behavior manifesto (Luthans, 2002), is that PsyCap is measureable. There are several instruments currently in use. The primary instru- ment for PsyCap is the PCQ-24 (Luthans, Avolio, et al., 2007), which contains 24 items (six items for each of the four components) that were adapted from existing measures of each individual construct and includes three reverse- coded items. A second instrument is a 12-item reduced ver- sion of the PCQ-24 (Avey, Avolio, & Luthans, 2010). A third and more recent instrument measuring PsyCap moves away from the standard self-report survey methodology and includes an implicit measure of PsyCap (Harms & Luthans, 2012).

A sixth characteristic of PsyCap and also originally noted by Luthans (2002) and supported by several studies is

that PsyCap is predictive of performance. One of the pri- mary reasons for conducting this study is to understand what firms can do to develop employee PsyCap. This prac- tice matters only to the extent that PsyCap is a desirable construct to increase in employees. Meta-analysis (Avey, Reichard, et al., 2011) suggests that this is the case as PsyCap is consistently, positively, and significantly related to employee performance.

A seventh and final boundary condition or characteristic of PsyCap is the level of analysis. Although current research underway is considering the role of multiple levels of PsyCap, of all the published studies empirically examining PsyCap only two have considered anything other than the individual level. Both Walumbwa, Luthans, Avey, and Oke (2011) and Clapp-Smith, Vogelgesang, and Avey (2009) considered the role of team-level PsyCap using a referent shift model. The more than 50 other unique studies with more than 13,000 participants include the individual level of analysis proposed in the original conceptualization.

Categories of Antecedents

Although much has been learned about PsyCap since 2002, a review of the literature suggests that outside of develop- mental interventions, very little is known about the anteced- ents of PsyCap. To address this systematically, I reviewed extant research on PsyCap and found at least four possible categories of antecedents. As a starting point I first consid- ered the role of trait-like individual differences. Cognitive mediation theory (Lazarus, 1991, 1993) suggests that cog- nitions are a foundation of emotional arousal. People are presented with a situation in organizational life and after assessing the environment against their cognitions, there is an emotional response. Given the strong correlations between PsyCap and cognitive dispositions, in an explor- atory process I determined stable individual differences may be a source of PsyCap. For example, if generalized self-efficacy is a possible (not the only) predictor of task specific PsyCap, then perhaps there are individual differ- ences that serve as a foundation for PsyCap. Second, research suggests leaders may have an impact on individual PsyCap (e.g., Avey et al., 2010; Gooty, Gavin, Johnson, Frazier, & Snow, 2009; Walumbwa & Schaubroeck, 2009). This position, that supervision and leadership can influence a person’s PsyCap on the job is also found in Eden and Shani’s (1982) work on the Pygmalion effect, which sug- gests leadership actions and communications can build self- efficacy in their followers. Although this work includes only self-efficacy, it is a starting point to explore the role of supervision in influencing employees’ overall PsyCap.

A third consideration of a domain for antecedents for PsyCap is job design. The well-known job characteristic model offered by Hackman and Oldham (1980) suggests that certain job characteristics influence self-opinions

at PENNSYLVANIA STATE UNIV on March 5, 2016jlo.sagepub.comDownloaded from

Avey 143

regarding task motivation. In a simple example, a person with an extremely difficult task at work who commonly experiences setbacks and failures may be more likely over time to report lower levels of PsyCap given that a founda- tion of, for example, self-efficacy is task mastery, some- thing often not achieved. If task mastery is rarely or never obtained, this would likely reduce employee’s level of PsyCap. Furthermore, in a well-designed challenging job where an individual experiences success and overcoming obstacles at work, levels of PsyCap are more likely to increase. Therefore, job characteristics were considered as a potential predictor of PsyCap.

The fourth and final domain of PsyCap antecedents was included for pragmatic reasons. Most individuals will not change demographics such as age, tenure (at a given point in time), ethnicity, or gender. Furthermore, if PsyCap is to be used in selection batteries, a correlation between PsyCap and a legally protected class status would be problematic considering disparate impact. Thus, to determine if there are roots of PsyCap in demographic variables, this category was included in the analysis to address questions such as if men have higher PsyCap than women and if senior employ- ees have higher PsyCap than junior employees. No a priori hypotheses or expectations are considered in this explor- atory analysis. Although one could speculate that, for exam- ple, senior employees have had more time to experience successes and thus have higher PsyCap, one could also speculate they have had more time for failures. Furthermore, one may speculate that in a male-dominated industry or firm men will have higher PsyCap. Thus, one could specu- late that entrance into this firm or industry by women would require higher PsyCap, and thus women in such firms would be higher on PsyCap. Therefore, the category of demo- graphics is exploratory. Overall, the purpose of a categori- cal analysis is to provide a foundation for a systematic review of the antecedents of PsyCap. By systematic I mean the consideration of these four categories.

Study 1

Method To test hypotheses on the antecedents of PsyCap, 1,264 electrical, mechanical, and aeronautical engineers from a large aerospace firm participated in Study 1. These partici- pants were located in the northwestern United States. All participants had at least a bachelor’s or technical degree in an engineering-related field. There were 512 women and 752 men, with an average age of 44 years (SD = 14). Data were collected at two time points separated by 1 week with predictors of PsyCap at Time 1 and PsyCap collected at Time 2 to minimize effects of common method variance (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). All par- ticipants completed instruments electronically through the Internet following a link emailed from the researcher.

All instruments used in the study were previously pub- lished in academic journals undergoing the psychometric evaluation process. Unless otherwise noted, all instruments were on a scale of 1 (strongly disagree) to 6 (strongly agree). Furthermore, all instruments yielded a Cronbach’s reliability coefficient greater than .70, as can be seen on the diagonals of Table 1. In both studies the 24-item PCQ (Luthans, Avolio, et al., 2007) was used to measure PsyCap. For individual differences, participants completed instru- ments on self-esteem and proactive personality. Self-esteem was measured with the three items from Judge, Erez, Bono, and Thoresen’s (2003) core self-evaluation trait scale. An example item from this scale is “Overall, I am satisfied with myself.” In this study proactive personality was measured using Bateman and Crant’s (1993) measure, which com- prises 17 items. These items are averaged to arrive at a pro- active personality score. Example items are “I excel at identifying opportunities” and “No matter what the odds, if I believe in something I will make it happen.”

To assess the role of leadership and supervision in PsyCap three leadership styles were measured: authentic

Table 1. Bivariate Correlations for Study 1.

M SD 1 2 3 4 5 6 7 8 9

1. PsyCap 4.79 0.58 (.93) 2. Self-esteem 3.68 0.56 .50 (.75) 3. Proactive personality 3.76 0.55 .45 .31 (.77) 4. Authentic leadership 4.12 0.49 .51 .30 .40 (.89) 5. Ethical leadership 3.68 0.79 .31 .22 .16 .24 (.94) 6. Abusive supervision 1.50 0.68 −.27 −.19 −.03 −.21 −.61 (.90) 7. Task complexity 3.69 0.89 .28 .15 .20 .25 .17 .01 (.85) 8. Tenure 8.62 9.41 .13 .05 −.09 .04 −.03 −.03 .15 9. Age 44 14 .21 .06 −.10 .12 .01 −.01 .20 .62 10. Gender 0.43 0.49 −.02 .08 .20 −.03 .05 .02 .08 .04 −.08

Alpha reliabilities are on the diagonal. r > .06 = p < .05. r > .09 = p < .01. r > .11 = p < .001.

at PENNSYLVANIA STATE UNIV on March 5, 2016jlo.sagepub.comDownloaded from

144 Journal of Leadership & Organizational Studies 21(2)

leadership, ethical leadership, and abusive supervision. Authentic leadership was measured with Walumbwa, Avolio, Gardner, Wernsing, and Peterson’s (2008) instru- ment on a behavioral frequency scale from hardly ever if ever to frequently if not always. An example item from this scale is “My manager makes decisions based on his/her core values.” Ethical leadership was measured with Brown, Treviño, and Harrison’s (2005) ethical leadership scale. An example item from this scale is “My manager disciplines employees who violate ethical standards.” Abusive supervi- sion was measured with Tepper’s (2000) abusive supervi- sion scale. Items begin with the phrase “My supervisor” and include completed statements such as “ridicules me.” Given existing survey length, only one instrument was used to assess job characteristics, which was task complexity. This was measured with relevant items for this study from Wood, Mento, and Locke’s (1987) instrument. Specifically, par- ticipants were asked to rate the following items:

In my opinion, my work is generally: 1 = very easy, 2 = easy, 3 = somewhat easy, 4 = somewhat difficult, 5 = difficult, 6 = very difficult. The various tasks I perform during a typical work day relate to each other to the following degree: 1 = very similar, 2 = similar, 3 = somewhat similar, 4 = somewhat variable, 5 = variable, 6 = very variable. I need a job that is more challenging to me (reverse coded): 1 = strongly disagree, 2 = disagree, 3 = somewhat disagree, 4 = somewhat agree, 5 = agree, 6 = strongly agree.

In addition to these items, demographic details, such as age, gender, years in current position, and tenure with the firm, were also collected.

In addition to reliability, I conducted a confirmatory fac- tor analysis using maximum likelihood techniques with Mplus software. Each item was fit to its latent variable (e.g., self-esteem Item 1 fit to the overall self-esteem variable). However, in the case of multidimensional constructs (e.g., PsyCap, authentic leadership), each item was fit to its first- order factor and each first-order factor was fit to the overall latent construct (e.g., four PsyCap dimensions loaded to one PsyCap variable). Without employing modification indices the model demonstrated a strong fit (comparative fit index [CFI] = .93, root mean square error of approximation [RMSEA] = .04, standardized root mean square residual [SRMR] = .04) with both RMSEA and SRMR being within the traditional strong fit range. Thus, the hypothesized model showed acceptable psychometric fit. After allowing four within-construct items to covary (e.g., Efficacy 1 and Efficacy 5), the CFI increased to .95. Overall, the evidence suggests the model provided adequate psychometric fit.

Results

Bivariate correlations can be seen in Table 1. In addition to bivariate correlation, regression was used to understand the degree to which each category (e.g., leadership, individual

differences, job characteristics, and demographics) as well as each individual construct (e.g., self-esteem) predicted PsyCap. Results from the regression analysis can be seen in Table 2.

Individual Differences. As seen in Table 2, individual differ- ences as a category was the strongest predictor of PsyCap predicting 45% of the variance. Both proactive personality and self-esteem were independently significant suggesting they explain unique variance in PsyCap from each other. Self-esteem was the highest predictor of PsyCap in this category.

Supervision. The supervision category was the next overall best predictor of PsyCap, explaining 32% of the variance in PsyCap with authentic and ethical leadership predicting unique variance in PsyCap. The significant relationship between abusive supervision and PsyCap, however, was washed out in the presence of the other two predictors with authentic leadership being the strongest predictor of PsyCap overall in this category.

Job Characteristics. With only one variable in the job charac- teristics category task complexity was a significant predic- tor of PsyCap, and this category predicted 12% of the variance in overall PsyCap.

Demographics. The three demographics in this study included age, which was a significant predictor of PsyCap and tenure, and gender, which were not. Overall, this cate- gory was the weakest predictor of PsyCap, with all three variables explaining only 2% of the variance in PsyCap.

As a final step in the regression analysis it was important to determine the uniqueness of each overall predictor. In other words, did each category predict unique variance in

Table 2. Regression Results With PsyCap as the Criterion for Study 1.

Predictor Standardized

beta Significance R2 of

category

Individual differences .45 Proactive personality .36 <.001 Self-esteem .46 <.001 Supervision .32 Authentic leadership .47 <.001 Ethical leadership .21 <.001 Abusive supervision −.02 .58 Job characteristics .12 Task complexity .34 <.001 Demographics .02 Age .13 <.01 Tenure −.01 .94 Gender −.03 .43

at PENNSYLVANIA STATE UNIV on March 5, 2016jlo.sagepub.comDownloaded from

Avey 145

PsyCap or simply overlapping variance in the outcome. To test this question the best predictors from each category (self-esteem, authentic leadership, task complexity, and age) were set as predictors of PsyCap. Results demonstrated that self-esteem, (b = .43, p < .001), authentic leadership (b = .36, p < .001), and task complexity (b = .16, p < .001) each predicted unique overall variance in PsyCap, while age was not a significant predictor (b = .03, p = .31), with the regression model explaining 49% of the variance in overall PsyCap.

Study 2

Method

To replicate and converge results of Study 1, I conducted a second study in mainland China with 524 technology employees of a large telecommunications firm. Participants completed translated instruments using Brislin’s (1980) back-translation method. Participants included 179 women and 345 men, with an average age of 32 (SD = 5.67) and approximately 6 years average tenure with the firm. Similar to in Study 1, all instruments used in the study were previ- ously published in academic journals undergoing the psy- chometric evaluation process. Furthermore, all instruments yielded a Cronbach’s reliability coefficient greater than .70, as can be seen on the diagonals of Table 3.

For individual differences participants completed instru- ments on core self-evaluations and cultural dimensions of power distance, uncertainty avoidance, and collectivism. Core self-evaluations were measured with the instrument from Judge and colleagues (2003). Example items from this scale are “I am confident I get the success I deserve in life,” “Sometimes I feel depressed” (reverse scored), and “I deter- mine what will happen in my life.” Power distance, uncer- tainty avoidance, and collectivism were all measured with three-item scales from the GLOBE studies (House, Hanges, Javidan, Dorfman, & Gupta, 2004). Example items are,

In this organization, subordinates are expected to . . . 1 = obey their boss without question 2, 3, 4, 5, 6 = question their boss when in disagreement; In this organization, orderliness and consistency are stressed, even at the expense of experimentation and innovation; and In this organization, group members take pride in the individual accomplishments of their group.

To assess the role of leadership and supervision in PsyCap, two leadership styles were measured: empowering leader- ship behaviors and ethical leadership. Empowering leader- ship behaviors were measured by Zhang and Bartol’s (2010) instrument. Example items from this instrument are “My manager helps me understand how my job fits into the big- ger picture,” “My manager believes that I can handle demanding tasks,” “My manager often consults me on stra- tegic decisions,” and “My manager allows me to do my job my way.” Similar to in Study 1, ethical leadership was mea- sured with Brown and colleagues’ (2005) ethical leadership scale. An example item from this scale is “My manager dis- ciplines employees who violate ethical standards.” While job characteristics were not available in Study 2 given time constraints and instrument length, demographic variables for age, tenure, and gender were collected in conjunction with the above instruments.

Results

Bivariate correlations for Study 2 can be seen in Table 3 with reliabilities in the diagonals. In addition to bivariate correlations, and similar to Study 1, regression was used to understand the degree to which each group (e.g., leadership, individual differences, and demographics) as well as each individual construct (e.g., core self-evaluations) predicted PsyCap. Results from the regression analysis for Study 2 can be seen in Table 4.

Individual Differences. As seen in Table 4, consistent with the results and Study 1, individual differences as a category was

Table 3. Bivariate Correlations for Study 2.

M SD 1 2 3 4 5 6 7 8 9

1. PsyCap 3.97 0.91 (.92) 2. Core self-evaluations 3.49 0.94 .45 (.73) 3. Power distance 2.86 0.88 −.10 −.18 (.76) 4. Uncertainty avoidance 2.75 1.09 −.24 −.32 .04 (.71) 5. Collectivism 3.24 0.94 .35 .35 −.19 −.15 (.78) 6. Ethical leadership 3.46 0.64 .38 .34 −.10 −.25 .23 (.80) 7. Empowering leadership 3.49 0.79 .42 .40 −.02 −.30 .24 .42 (.93) 8. Age 32.48 5.62 .02 −.04 −.04 −.02 .03 −.02 .04 9. Tenure 5.71 5.35 .04 −.05 −.05 .01 −.03 −.02 .02 .60 10. Gender 0.66 0.47 .04 −.08 −.08 −.04 .01 .05 .03 .16 .11

Alpha reliabilities are on the diagonal. r > .09 = p < .05. r > .11 = p < .01. r > .14 = p < .001.

at PENNSYLVANIA STATE UNIV on March 5, 2016jlo.sagepub.comDownloaded from

146 Journal of Leadership & Organizational Studies 21(2)

the strongest predictor with 24% of the variance in PsyCap explained. In this case, both core self-evaluations and col- lectivism were independently significant, suggesting they explain unique variance in PsyCap from each other. Power distance and uncertainty avoidance were not significant predictors of PsyCap. Overall, in this category core self- evaluations were the best predictor of PsyCap.

Supervision. Similar to Study 1, the supervision category was the next overall best predictor of PsyCap, explaining 23% of the variance in PsyCap with empowering leadership behav- iors and ethical leadership predicting unique variance in PsyCap. The strongest predictor of PsyCap in this category was empowering leadership behaviors (b = .33, p < .001).

Demographics. Replicating the analysis from Study 1, the three demographics in this study included age, tenure, and gender. In this case, none of the demographic predictors were significant predictors of PsyCap. Overall, this category was the weakest predictor of PsyCap with none of the three variables explaining any variance in PsyCap and the regres- sion model overall predicting no significant variance.

I replicated the final analysis from Study 1 to determine the uniqueness of each overall predictor on PsyCap. It was important to determine if each category predicted unique variance in PsyCap or simply overlapping variance in the outcome. To test this question, the best predictor from each significant category (in this case only core self-evaluations and empowering leadership behaviors) was again set as a predictor of PsyCap. Results demonstrate that core self- evaluations (b = .33, p < .001) and empowering leadership behaviors (b = .33, p < .001) each predicted unique overall

variance in PsyCap, with the regression model explaining 27% of the variance in overall PsyCap.

Discussion

Overall, the purpose of this study was to investigate the extent to which there are fixed antecedents of PsyCap. More than 10 years has passed since Luthans and his colleagues introduced positive organizational behavior and PsyCap. Since that time there have been a number of published arti- cles dealing with the utility and predictive power of PsyCap, including a few longitudinal research designs and many cross-sectional works linking PsyCap to performance, behaviors, and attitudes and a meta-analysis. However, until now research on PsyCap has not undertaken an empiri- cal investigation of where PsyCap begins, the antecedents. This study was intended to begin to fill that void in knowl- edge. Results from two field studies suggest there may be fixed constructs in the categories of individual differences, supervision, job characteristics, and demographics that pre- dict levels of PsyCap at work. With some replication and convergence, overall results from the two samples suggest PsyCap may have some (although not exclusive) anteced- ents in these categorical domains. Based on this work there are several points of discussion that add to existing knowl- edge on PsyCap.

First, the categories were set a priori based on theoretical rationale. At first glance it seems these categories were set fairly well. Rationale for this conclusion is in the unique variance predicted by each category. A challenge with empirical work using surveys and correlations may be a mysterious and unmeasured third variable. Researchers may find a correlation between, for example, authentic leadership and task complexity on PsyCap. However, when subject to regression analysis they may find that only authentic leadership is a significant predictor implying the correlation between task complexity and PsyCap was shar- ing variance with authentic leadership. In other words, task complexity was correlated primarily with authentic leader- ship and not the actual outcome PsyCap. Results in these two studies however suggest the categorical differences were meaningful in predicting PsyCap. Specifically, the summary analysis from Study 1 suggest that self-esteem, task complexity, and authentic leadership all uniquely pre- dicted variance in PsyCap, as each was a significant predic- tor in the regression model. For research purposes this suggests levels of PsyCap do not have merely one anteced- ent but several, with at least three (self-esteem, task com- plexity, and authentic partnership) potentially explained in this study. Therefore, the findings suggest not only that PsyCap is a multidimensional construct but also that it is a multiestablished construct (i.e., established first in multiple other domains).

Table 4. Regression Results With PsyCap as the Criterion for Study 2.

Predictor Standardized

beta Significance R2 of

category

Individual differences .24 Core self-evaluations .35 <.001 Power distance .01 .99 Uncertainty avoidance −.05 .21 Collectivism .22 <.001 Supervision .23 Empowering leadership .33 <.001 Ethical leadership .24 <.001 Job characteristics N/A Demographics <.01 Age .02 .73 Tenure .01 .93 Gender .03 .53

at PENNSYLVANIA STATE UNIV on March 5, 2016jlo.sagepub.comDownloaded from

Avey 147

A second point of discussion is the nature of the replicated results. Study 1 found that (in order of predictive power) indi- vidual differences, supervision, and job characteristics were important in predicting PsyCap, with demographics being less or not important. Study 2 replicated these results in another geographic location, industry, and sample, showing (in order of predictive power) individual differences and supervision were most important in predicting PsyCap (job characteristics were not measured in Study 2) and, also like Study 1, demographics were not important. Although these are only two studies, the replication adds validity to the idea that among these domains, individual differences are the most powerful predictor of PsyCap, followed by leadership and job characteristics, with a substantial amount of variance explained in PsyCap. As a practical implication, this suggests organizations have significant influence on employee levels of PsyCap through selection and supervision.

A third and final point of discussion includes a consider- ation of the boundaries of PsyCap theory and measurement. The results in Study 1 (U.S.-based sample) were substantially more robust than the results in Study 2 (Chinese sample). I outline several possibilities for this, but it is important to note that I am not suggesting PsyCap research is not useful in a Chinese context. Rather, I am suggesting contextual consid- eration must be made when researching PsyCap outside the United States for the following reasons. One possibility for the results being more robust in the United States has to do with measurement. The 24-item PCQ was developed, tested, validated, and revalidated by U.S. research participants.

Although several PsyCap studies outside the United States have revealed meaningful results, the original valida- tion was in the United States. Word choice and meaning were developed by U.S. researchers and validated in U.S. labs and field studies. Given this, even with a back-transla- tion method, words and phrases may have different mean- ing across cultures. Hence, each phrase and word that means one thing in Culture A may mean another in Culture B. It is possible this lack of isomorphism between cultures partially explains why results were less robust in Study 2. A second explanation for results being more robust in the U.S. sample and less in the Chinese sample that has implications for PsyCap research is the subtle self-promotion in the con- struct itself. In the United States, a primarily individualist culture, there may be social desirability in the instrument artificially inflating PsyCap. In other words, a seasoned engineer who is recently struggling on his or her job may not want to say he or she is lower in PsyCap. In contract, in the Chinese context, historically higher in collectivism, it may be less socially desirable to decree one’s confidence and resilience in one’s self versus one’s work unit or organi- zation. This negative social desirability may lead the Chinese participant higher in PsyCap to artificially score lower. This is speculation, however, as these premises were not tested in the current study.

A final point on why the results were much more robust in the U.S. sample is that results here may have little to do with a macro-level U.S. context compared to the Chinese context but have to do with the unique nature of each sam- ple in these studies. In Study 1, highly educated, well-paid, professional talented employees were studied. In Study 2, there was a cross-section of education, job level, and expe- rience (e.g., tenure in Study 1 was significantly higher than in Study 2). Therefore, it is possible that the differences in predicting PsyCap are explained better by sample unique- ness (e.g., highly educated aeronautical engineers vs. vari- ous employees in a telecommunications firm) rather than measurement or cultural boundary conditions of PsyCap and its operationalization. More research comparing sam- ples within and between cultures is necessary to make more confident assertions on these propositions.

Practical Implications

Several studies including a meta-analysis (Avey, Reichard, et al., 2011) suggests PsyCap is a useful predictor of perfor- mance and job attitudes and behaviors. Furthermore, recent research has also suggested that developing employee PsyCap can lead to positive individual and organization outcomes (Luthans et al., 2010). Given this, it seems that it is in an organization’s best interest to design systems and structures that would enhance overall employee PsyCap. In consideration of the findings of this study, there are at least four practical implications for managers. First, select and hire employees for high PsyCap based on individual differ- ences. Although PsyCap is not a fixed trait, it is not as prone to fluctuation as are emotions (see Luthans, Avolio, et al., 2007, for test–retest reliability statistics) and thus can be used in selection batteries. Furthermore, by selecting in individual differences such as self-esteem, core self-evalua- tions, and proactive personality, organizations can used trait-based selection criteria to acquire a potentially more sustainable high level of PsyCap.

A second implication is for leadership development. By investing in leadership development interventions that train leaders in practices (e.g., authentic leadership, ethical lead- ership, empowering leadership behaviors) that enhance employee PsyCap, not only will organizations acquire more highly developed leaders, but they will also benefit from the possible outcome of higher employee PsyCap from those leadership behaviors and styles. Third, ever since Hackman and Oldham’s (1980) classic work on job design, it is known that job characteristics (such as task complexity) can influ- ence employees’ psychological states (such as PsyCap). Thus, where possible and feasible, organizations can rede- sign jobs to enhance employee PsyCap. Each of these areas may uniquely and positively influence employee PsyCap.

A fourth implication for managers stems from the lack of robust results on employee demographics. Given that these

at PENNSYLVANIA STATE UNIV on March 5, 2016jlo.sagepub.comDownloaded from

148 Journal of Leadership & Organizational Studies 21(2)

results were not statistically significant, organizations can rule out being concerned about selecting for age or gender to enhance employee PsyCap. In the United States, the Equal Pay Act, the Civil Rights Act, and the Age Discrimination in Employment Act prohibit employment discrimination on the basis of age and gender. Given that neither of these variables was a strong predictor of PsyCap, organizations should of course monitor them. However, they are not likely to have the challenges of discrimination when selecting for PsyCap as with other criterion (e.g., gen- eral mental ability tests). In other words, the lack of correla- tion with demographic variables from a legal standpoint is actually good news for organizations selecting on PsyCap for higher performance and better attitudes and behaviors.

Limitations and Conclusion

As with any empirical study, the results of these studies should be interpreted with caution based on several com- mon limitations. First, not all categories of the antecedents of PsyCap were measured. While the four categories of individual differences, supervision, job characteristics and demographics were relatively broad and encompassing, there may be a category that is uniquely powerful in pre- dicting PsyCap not measured in this study. Second, no one category was saturated. For example, I did not measure every leadership style. Future research may find the best predictor of PsyCap in a supervision category is transfor- mational leadership (see, e.g., Gooty et al., 2009). Future research should consider if there are better predictors within the given categories. Third, self-report questionnaires were the primary measures used in this study. Future research should utilize other methods (e.g., functional magnetic res- onance imaging brain scans, implicit measures such as the I-PCQ; Harms & Luthans, 2012), which may yield other categories or strong predictors of PsyCap. Fourth, although steps were taken to reduce common method variance, the exclusive use of surveys can artificially increase the corre- lations between predictors and PsyCap (e.g., Podsakoff et al., 2003). Fifth, all data were collected from the partici- pants as the point of reference. It was determined the most useful data on, for instance, individual differences were self-ratings rather than from another source (e.g., supervi- sor) through the personality of constituents. While this was necessary in this study, both common method and common source variance may influence the results. Sixth, as previ- ously mentioned there may be cultural and language con- straints influencing the results of Study 2. Finally, in Study 1 the predictive effect of abusive supervision were not sta- tistically significant in a regression model with authentic and ethical leadership as predictors. Given the strong cor- relation between abusive supervision and ethical leader- ship, it is possible that the two were collinear thus artificially suppressing the effect of abusive supervision on PsyCap.

While ethical leadership and abusive supervision are not two ends of the same continuum (Avey, Palanski, & Walumbwa, 2011), in this study there was a high correla- tion, suggesting a threat of collinearity.

In conclusion, research on PsyCap has flourished over the past decade. The empirical work has grown substan- tially to consider the role PsyCap may play in employee performance, behaviors, and attitudes. The results of this study considering the “left” of the theoretical model may give insights into the antecedents of PsyCap in employees. These results may be useful moving forward to understand where PsyCap begins and how it develops and changes and how to design more useful systems, structures, and inter- ventions to enhance overall employee PsyCap.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Avey, J. B., Avolio, B. J., & Luthans, F. (2010). Experimentally analyzing the impact of leader positivity on follower positiv- ity and performance. Leadership Quarterly, 22, 282-294.

Avey, J. B., Palanski, M. E., & Walumbwa, F. O. (2011). When leadership goes unnoticed: The moderating role of follower self-esteem on the relationship between ethical leadership and follower behavior. Journal of Business Ethics, 98, 573-582.

Avey, J. B., Reichard, R., Luthans, F., & Mhatre, K. (2011). Meta-analysis of the impact of positive psychological capital on employee attitudes, behaviors and performance. Human Resource Development Quarterly, 22, 127-152.

Bandura, A. (1997). Self-efficacy: The exercise of control. New York, NY: Freeman.

Bateman, T. S., & Crant, J. M. (1993). The proactive compo- nent of organizational behavior. Journal of Organizational Behavior, 14, 103-118.

Brislin, R. W. (1980). Translation and content analysis of oral and written material. In H. C. Triandis & J. W. Berry (Eds.), Handbook of cross-cultural psychology (Vol. 2, pp. 349-444). Boston, MA: Allyn & Bacon.

Brown, M. E., Treviño, L. K., & Harrison, D. A. (2005). Ethical leadership: A social learning perspective for construct devel- opment and testing. Organizational Behavior and Human Decision Processes, 97, 117-134.

Clapp-Smith, R. O., Vogelgesang, G., & Avey, J. B. (2009). Authentic leadership and positive psychological capital: The mediating role of trust at the meso-level of analysis. Journal of Leadership and Organization Studies, 15, 227-240.

Eden, D., & Shani, A. B. (1982). Pygmalion goes to boot camp: Expectancy, leadership, and trainee performance. Journal of Applied Psychology, 67, 194-199.

at PENNSYLVANIA STATE UNIV on March 5, 2016jlo.sagepub.comDownloaded from

Avey 149

Gooty, J., Gavin, M., Johnson, P. D., Frazier, M. L., & Snow, D. B. (2009). In the eyes of the beholder: Transformational leader- ship, positive psychological capital, and performance. Journal of Leadership and Organizational Studies, 15, 353-367.

Hackman, J. R., & Oldham, G. R. (1980). Work redesign. Reading, MA: FT Press.

Harms, P. D., & Luthans, F. (2012). Measuring implicit psycho- logical constructs in organizational behavior: An example using psychological capital. Journal of Organizational Behavior, 33, 589-594.

Hobfoll, S.E. (2002). Social and psychological resources and adaptation. Review of General Psychology, 6, 307-324.

House, R. J., Hanges, P. J., Javidan, M., Dorfman, P. W., & Gupta, V. (2004). Culture, leadership, and organizations: The GLOBE study of 62 societies. Thousand Oaks, CA: Sage.

Judge, T. A., Erez, A., Bono, J. E., & Thoresen, C. J. (2003). The Core Self-Evaluations Scale: Development of a measure. Personnel Psychology, 56, 303-331.

Law, K. S., Wong, C., & Mobley, W. H. (1998). Toward a taxon- omy of multidimensional constructs. Academy of Management Review, 23, 741-755.

Lazarus, R. S. (1991). Emotion and adaptation. New York, NY: Oxford University Press.

Lazarus, R. S. (1993). From psychological stress to the emo- tions: A history of changing outlooks. Annual Review of Psychology, 44, 1-21.

Luthans, F. (2002). The need for and meaning of positive orga- nizational behavior. Journal of Organizational Behavior, 23, 695-706.

Luthans, F., Avey, J. B., Avolio, B. J., & Peterson, S. J. (2010). The development and resulting performance impact of posi- tive psychological capital. Human Resource Development Quarterly, 21, 41-67.

Luthans, F., Avolio, B. J., Avey, J. B., & Norman, S. M. (2007). Psychological capital: Measurement and relationship with perfor- mance and job satisfaction. Personnel Psychology, 60, 541-572.

Luthans, F., Youssef, C. M., & Avolio, B. J. (2007). Psychological capital: Developing the human competitive edge. Oxford, UK: Oxford University Press.

Norman, S. M., Avolio, B. J., & Luthans, F. (2010). The impact of positivity and transparency on trust in leaders and their per- ceived effectiveness. Leadership Quarterly, 21, 350-364.

Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879-903.

Tepper, B. J. (2000). Consequences of abusive supervision. Academy of Management Journal, 43, 178-190.

Walumbwa, F. O., Avolio, B. J., Gardner, W. L., Wernsing, T. S., & Peterson, S. J. (2008). Authentic leadership: Development and validation of a theory-based measure. Journal of Management, 34, 89-126.

Walumbwa, F. O., Luthans, F., Avey, J. B., & Oke, A. (2011). Authentically leading groups: The mediating role of collec- tive psychological capital and trust. Journal of Organizational Behavior, 32, 4-24.

Walumbwa, F. O., & Schaubroeck, J. (2009). Leader personality traits and employee voice behavior: Mediating roles of ethical leadership and work group psychological safety. Journal of Applied Psychology, 94, 1275-1286.

Wood, R. E., Mento, A. J., & Locke, E. A. (1987). Task complex- ity as a moderator of goal effects: A meta-analysis. Journal of Applied Psychology, 72, 416-425.

Zhang, X. M., & Bartol, K. M. (2010). Linking empowering lead- ership and employee creativity: The influence of psychologi- cal empowerment, intrinsic motivation, and creative process engagement. Academy of Management Journal, 53, 107-128.

Author Biography

James B. Avey, PhD, is an associate professor of Management at Central Washginton University, College of Business. He reiceved his PhD from the University of Nebraska-Lincoln and publishes on topics such as Positive Psychological Capital, Ethical Leadership and Psychological Ownership. His research has appeared in outlets such as Personell Psychoogy, Journal of Management, The Leadership Quarterly and Journal of Leadership and Organizational Studies.

at PENNSYLVANIA STATE UNIV on March 5, 2016jlo.sagepub.comDownloaded from