Types of workplace social support in the prediction of job satisfaction.

Research on social support and job satisfaction has yielded mixed results, partly because studies have rarely examined different types of workplace social support, such as collegial support, task support, coaching, and career mentoring. This study identified the relative contributions of different types of social support to job satisfaction and explored the relationship between social support and job tenure. Overall, social support accounted for approximately 17% of the variance in job satisfaction and 9% of the variance in job tenure. Career mentoring and task support were the types of social support most predictive of job satisfaction. Coaching and task support were the types of social support most predictive of job tenure.

**********

Workplace social support has been defined as the "actions of others that are either helpful or intended to be helpful" (Deelstra et al., 2003, p. 324). It includes a variety of interpersonal behaviors among workers that enhance individuals' psychological or behavioral functioning. These may include mentoring, providing emotional support, assisting others with assigned tasks, and teaching about social power structures (Hill, Bahniuk, Dobos, & Rouner, 1989). Beginning with the earliest need-fulfillment theories of job satisfaction, workplace social support has been identified as a predictor of job satisfaction (Orpen & Pinshaw, 1975; Smither, 1988; Stamps, 1997; Vroom, 1964). Most research has found workplace social support to be positively predictive of job satisfaction and other positive outcomes (Harris, Moritzen, Robitschek, Imhoff, & Lynch, 2001; Smith & Tziner, 1998; Winstead, Derlega, Montgomery, & Pilkington, 1995). However, there are exceptions to this rule in the research literature (Ducharme & Martin, 2000). Workplace social support also predicts a variety of negative outcomes, including absenteeism and turnover (Winstead et al., 1995), burnout (Myung-Yong & Harrison, 1998), and depression and anxiety (El-Bassel, Guterman, Bargal, & Su, 1998; Olson & Shultz, 1994).

Findings may be mixed because the construct of workplace social support is multifaceted (Bahniuk, Dobos, & Hill, 1990). The source of support may be a supervisor, mentor, or colleague; the content of the support may include information, appraisal, assistance with tasks, or emotional support (Bahniuk et al., 1990; Deelstra et al., 2003). Hill et al. (1989) defined four types of workplace social support: Task support focuses on sharing and exchanging work assignments and ideas. Career mentoring refers to parentlike or adviser relationships with other individuals who have more experience. Coaching involves teaching organizational/professional rules and goals, including organizational politics. Collegial social support includes sharing friendships, personal problems, and confidences. Bahniuk et al. found that, among business managers, instrumental support from colleagues and mentoring both predicted higher levels of job satisfaction, along with perceived success, managerial level, and income. However, coaching and collegial support had no effect on job satisfaction.

The relationship between workplace social support and the length of time an employee chooses to remain at the same job (i.e., job tenure) has recently been explored. Positive relationships with supervisors have been reported to strongly predict job tenure (Buckingham & Coffman, 1999; Van Breukelen, Van Der Vlist, & Steensma, 2004; Vecchio & Boatwright, 2002). Positive organizational social climates and the presence of friends or family at a particular work site have also been reported to predict employee retention (Van Breukelen et al., 2004; Milman, 2003; Pizam & Ellis, 1999). Given the importance of supervisor support and access to friends or family at work, it seems reasonable to hypothesize that career mentoring (i.e., support from the supervisor) and collegial support and task support (both of which may be more readily available from friends and family) may be types of workplace social support that strongly predict job tenure.

In this study, we attempted to determine types of workplace social support that best predict job satisfaction and job tenure. On the basis of Bahniuk et al.'s (1990) previous findings, we hypothesized that career mentoring and task support would be the strongest predictors of job satisfaction and that career mentoring, collegial support, and task support would be the strongest predictors of job tenure.

Method

Participants

The study included 122 female and 57 male full-time paid employees in two training hospitals in the southwestern part of the United States. The sample was 57% Caucasian, 10% African American, 17% Asian American, 9% Hispanic, and 7% "other." Mean age was 41.8 years (SD = 10.8), average education level was 15.3 years (SD = 2.6), and mean annual income was $28,500 (SD = $14,500). To derive a range of occupations, we approached multiple hospital departments, including patient care areas and departments such as accounting, building maintenance, and laundry; 50% of the sample was in direct patient care. After announcing the opportunity to participate in the study, an investigator (first author) visited the department to distribute a survey that participants completed and that was retrieved later on the same day. A total of 237 surveys were distributed; 184 were returned, yielding a return rate of 77.6%. The institutional research board for the hospital administering the study reviewed and approved the protocol.

Instruments

Participants completed a demographic questionnaire that elicited data on their age, gender, ethnicity, education level, job tenure, job title and job description, salary, relationship status, and number of children. Participants also completed the Job in General scale (Ironson, Smith, Brannick, Gibson, & Paul, 1989), a global index of job satisfaction. Participants were asked to indicate if each adjective-phrase item described their jobs by answering "yes," "no," or "?" to each item. Scores ranged from 0 to 54. Scores on the instrument yielded alpha coefficients that ranged from .91 to .95 (Ironson et al., 1989), indicating adequate reliability. Correlations with scores on similar job satisfaction measures ranged from .67 to .80 (Ironson et al., 1989), indicating scale validity.

Workplace social support was measured using the Mentoring and Communication Support Scale (Hill et al., 1989), a 15-item measure that yields subscale scores for Career Mentoring, Coaching, Collegial Social Support, and Task Support. Examples of items measuring career mentoring include "Someone of higher rank has shown a parental-like interest in me and my career" and "Someone of higher rank has placed me in important assignments or positions." Examples of items measuring coaching include "I have been coached about office politics" and "I have had an associate teach me the informal rules of my organization." Examples of items measuring collegial social support include "My associates and I are friends as well as coworkers" and "My associates and I share confidences with each other." Examples of items measuring task support include "My associates and I assist each other in accomplishing assigned tasks" and "I work jointly in major projects or cases with my associates." Items are rated on a Likert-type scale of 1 (strongly disagree) to 5 (strongly agree); therefore, scores ranged from 15 to 75. Cronbach's alphas for scores on the subscales ranged from .75 to .89 (Downs, Hill, Bahniuk, & Rouner, 1994), indicating adequate reliability. Scores had positive correlations with upward career mobility and satisfaction with promotions (Hill et al., 1989), indicating validity.

Analysis

Variable distributions were normal, except for the Job in General score, which was negatively skewed. Such a negative skew, or ceiling effect, is a common observation in job satisfaction scores regardless of the measure used (Spokane, Meir, & Catalano, 2000). Evidently, research participants' ability to place themselves in satisfying jobs exceeds researchers' present ability to measure the upper limits of job satisfaction. A reflect log transformation normalized the distribution. It was inverted to ease interpretation of findings. Means and standard deviations are presented in Table 1, as are intercorrelations. Results of t tests indicated no gender differences in age, number of children, education, job tenure, income, job satisfaction, or social support in the workplace. Career Mentoring, Coaching, and Task Support all had significant positive relationships with job satisfaction. Coaching and job tenure were negatively correlated, whereas Task Support and job tenure were positively correlated.

Simultaneous multiple regression analysis using the four types of social support to predict job satisfaction (see Table 2) yielded an adjusted [R.sup.2] of .176 (F = 10.22, p < .001), indicating that these combined types of social support were significant predictors of job satisfaction. Significant individual predictors of job satisfaction were Career Mentoring ([beta] = .38, p < .001) and Task Support ([beta] = .17, p < .05). Coaching and Collegial Support were not significant individual predictors of job satisfaction. Simultaneous multiple regression analysis using the four types of social support to predict job tenure (see Table 3) yielded an adjusted [R.sup.2] of .091 (F = 5.32, p < .001), indicating that these combined types of social support were significant predictors of job tenure. Individual predictors of job tenure were Coaching ([beta] = -.31, p < .001) and Task Support ([beta] = .22, p < .02).

Discussion