The Academic Environment and Faculty Well-Being: The Role of Psychological Needs

Abstract

Recent research has examined the well-being of higher education faculty, but it has typically lacked a theoretical model. The present study used self-determination theory to model the well-being of 581 tenured and tenure-eligible faculty members at a large mid-Western university. Volitional autonomy, perceived competence, and perceived relatedness were hypothesized to partially mediate the relationships between several environmental factors (e.g., administrative support, research support, promotion and tenure support) and faculty well-being (i.e., teaching/service satisfaction and global satisfaction). Results of path analysis indicated that all relations between the environment and teaching/service satisfaction were fully mediated by volitional autonomy and perceived competence, whereas all relations between the environment and global satisfaction were partially mediated by perceived relatedness. These findings highlight the centrality of psychological needs in understanding the relations between the environment and faculty well-being. Additional implications and future directions for research are discussed.

Keywords

faculty well-being self-determination theory academic environmental factors volitional autonomy perceived competence perceived relatedness

Faculty members constitute the core asset of every academic institution, taking on the primary tasks of teaching, research, and service activities that contribute to the livelihood of the university. Moreover, they mentor the next generation of scholars who allow the institution to thrive into the near future. Given faculty members’ central role in the sustainability and growth of the institution, consideration of the factors that contribute to faculty satisfaction are critical, especially based on meta-analyses that link job satisfaction with job performance for complex jobs like tenure-track faculty roles (Judge, Thoresen, Bono, & Patton, 2001).

One key index of well-being in the work domain is job satisfaction (e.g., Horn, Taris, Schaufeli, & Schreurs, 2004). Past faculty satisfaction research has shown that faculty satisfaction is affected by institutional factors such as leadership; relationships with colleagues, students, and administrators; and perceptions of climate and culture of the university (e.g., Eagan, Jaeger, & Grantham, 2015; Hagedorn, 2000; Moors, Malley, & Stewart, 2014), highlighting the power that institutions have in shaping the experiences of their faculty. In general, this body of research has identified particular environmental factors that may negatively or positively relate to faculty satisfaction. What is generally missing in this body of research is a conceptual model or framework that places the faculty member’s psychological needs as a central figure in the model.

One lens through which faculty well-being can be viewed is through self-determination theory (SDT; Deci & Ryan, 1985; Ryan & Deci, 2000). The theory posits three central psychological needs including the need to make one’s own choices (volitional autonomy), the need to experience mastery of the environment (perceived competence), and the need to feel a sense of belongingness and attachment with others (perceived relatedness). A subtheory of SDT, labeled cognitive evaluation theory (CET; Deci & Ryan, 1985), is focused on volitional autonomy and perceived competence as mediating the relationship between the work environment and one’s well-being (e.g., faculty satisfaction), while a second SDT subtheory, organismic integration theory (OIT; Deci & Ryan, 1985), is focused on relatedness as mediating the relationship between the work environment and well-being.

The present study applied SDT as a conceptual framework in predicting faculty well-being. The authors used a preexisting database of tenured and tenure-eligible (T/TE) faculty who had completed the Collaborative on Academic Careers in Higher Education (COACHE) faculty job satisfaction survey (Benson, Mathews, & Trower, 2014), a measure that has been used widely to improve the academic workplace for faculty (e.g., Lawrence, Celis, & Ott, 2014; Ott & Cisneros, 2015; Seipel & Larson, 2016). In doing so, we hypothesized that the three basic psychological needs would partially mediate the relationship between environmental factors and faculty well-being (operationalized as teaching/service satisfaction and global satisfaction). Figure 1 presents the conceptual model.

Figure 1.

The hypothesized path model. Correlations were specified among the exogenous, mediator, and outcome variables respectively, but are omitted for visual clarity.

The Relation of Environmental Factors with Faculty Well-Being

The present study conceptualized environmental factors as encompassing academic area, chair support, administrative support, research support, department feedback/support, promotion and tenure support, and faculty teaching quality. Hypothesized positive links between each of the environmental supports and faculty well-being are outlined below on the basis of extant empirical literature.

Chair support

Chair support encompassed factors such as decision-making, priority setting, and communication by department chairs. Positive relations between chair behaviors and faculty satisfaction have been found (Brown & Moshavi, 2002), and relations with the chair have also been found to be a significant predictor of overall faculty satisfaction (August & Waltman, 2004). Meta-analytic findings (Ng & Sorensen, 2008) further support the relationship between immediate supervisor support and employee well-being in the general population.

Administrative support

Administrative support in the present study was conceptualized in a manner similar to that of chair support. Several studies have shown positive relationships between administrators’ support and faculty satisfaction (e.g., Chen, Beck, & Amos, 2005; Verhaegen, 2005).

Research support

Research support encompassed multiple institutional (e.g., grant money and course release time) and instrumental (e.g., classrooms and clerical support) aspects of the job. Verhaegen (2005) found that factors related to research support were significant predictors of business faculty satisfaction. In one study, those faculty members who spent the most rather than the least time on research had the lowest scores for burnout and the highest for engagement and intrinsic satisfaction (Vera, Salanova, & Martín, 2010). Thus, support for research activities can be seen as intimately connected to faculty well-being, in that conditions that allow for effective engagement with research are associated with positive outcomes.

Department feedback/support

This construct was conceptualized as being related to department feedback through the annual evaluation and recognition for teaching and scholarly efforts, and related to departmental support around work-life balance. With regard to evaluation procedures, new faculty members who received information about research, teaching, and service expectations from mentors felt more connectedness and ownership in their work environments (Schrodt, Cawyer, & Sanders, 2003). These findings highlight the importance of department feedback in faculty satisfaction, in that faculty members who are given continuous and accurate feedback around their performance and expectations are more likely to experience higher levels of satisfaction. Other research showed positive linkages between colleagues’ recognition and faculty satisfaction (e.g., Grawitch, Trares, & Kohler, 2007; Huggett et al., 2012; Leung, Siu, & Spector, 2000). Additionally, support of work-life balance has been shown to relate positively to employee well-being in meta-analyses (e.g., Butts, Casper, & Yang, 2013) and to relate to faculty well-being in individual studies (e.g., Chung et al., 2010; Grawitch et al., 2007).

Promotion and tenure support

This construct encompasses the clarity of the promotion and tenure process, such as having knowledge about the criteria and standards that will be used in evaluation of faculty members. Schrodt, Cawyer, and Sanders (2003) found that mentoring that provides promotion support is positively associated with a protégé’s sense of ownership in their department, while Chung et al. (2010) similarly found that career advancement variables (e.g., understanding the process for promotion) emerged as a significant predictor of job satisfaction for clinical track faculty members at a medical school. These findings lend support to the importance of promotion and tenure processes in predicting faculty member well-being such that those who perceive more clarity around promotion and tenure expectations will be more satisfied overall.

Perceived Autonomy, Competence, and Relatedness as Mediators

The three psychological SDT needs have rarely been examined as mediators in the faculty well-being literature, although there is extensive support in the broader literature. In the mental health and physical health literature, a recent meta-analysis showed all three SDT needs positively mediated the relation between autonomy support (an environmental factor) and mental health and physical health (Ng et al., 2012). The three SDT needs have also been shown to fully mediate or partially mediate the relationship of environmental factors in the workplace with well-being in samples of: (a) secondary teachers and school personnel in Spain and the United States (Doménech-Betoret, Lloret-Sequra, & Gómez-Artiga, 2015; Fernet, Austin, Trépanier, & Dussault, 2013), (b) nurses in France (Gillet, Colombat, Michinov, Pronost, & Fouquereau, 2013), and (c) masters’ students and graduates in business in Canada and the United States (Gillet, Forest, Benabour, & Bentein, 2015; Graves & Luciano, 2013). All but two of those studies combined the three SDT needs, so their individual contributions could not be examined. Fernet and colleagues (2013) examined the three psychological needs as mediators of four job environment factors (e.g., role overload) and three burnout indices (e.g., emotional exhaustion). They showed that volitional autonomy partially or fully mediated six of the environment/burnout relationships. Perceived competence and relatedness each partially or fully mediated two of the environment/burnout relationships. Graves and Luciano’s (2013) focus was on the environmental factor of leader–member exchange, which overlaps the most with chair support in the current study. They showed that leader–member exchange is directly and indirectly related to job satisfaction through perceived competence and volitional autonomy.

While the use of SDT as a conceptual framework for faculty well-being is an emerging area that is still in its infancy, a study by Seipel and Larson (2016) examined the SDT needs of autonomy and relatedness as mediators between environmental factors and the well-being of non-tenure-track (NTT) faculty members. Perceived relatedness fully mediated all relationships between environmental supports (i.e., departmental support, administrative support, personal/family support) and faculty well-being (i.e., teaching/service satisfaction and global satisfaction), while volitional autonomy marginally mediated the relation between administrative support and global satisfaction. Although Seipel and Larson did not include competence as one of the SDT needs, it does provide preliminary evidence to suggest the usefulness of SDT in conceptualizing the links between environmental factors and faculty well-being.

The above mentioned studies highlight the mediating role of SDT needs between the environment and employee well-being, in that environmental supports may affect well-being indirectly through one or more of the psychological needs. Thus, the current study attempted to extend these findings by applying the SDT framework to a T/TE faculty sample, such that the relations between environmental factors and faculty well-being would be partially mediated by the SDT needs.

Hypotheses

The primary hypothesis for this study was that the partial mediation model shown in Figure 1 would yield a significantly better fit to this faculty sample than a fully mediated model. Specifically, we hypothesized that the environmental factors would have significant, positive, and direct associations with two indices of well-being (i.e., faculty teaching/service satisfaction and global satisfaction) and that perceived autonomy, competence, and relatedness would also have direct relations with the two indices of well-being. As for mediation hypotheses, we predicted that the environmental factors would also have significant, positive, and indirect relations with the two indices of well-being through volitional autonomy, perceived competence, and perceived relatedness.

Method

Participants

The participants included 581 full-time T/TE faculty at a large Midwestern university who responded to the 2013–2014 administrative of the COACHE faculty satisfaction survey (Benson et al., 2014). Twenty-three participants were excluded from data analysis due to excessive missing responses, making the final sample N = 558. Women constituted 37% of the sample. A preponderance of respondents was Caucasian (81.9%), and other racial/ethnic groups included Asian, Asian American, or Pacific Islander (11.8%); Hispanic or Latino/a (2.6%); Black or African American (2.3%); American Indian or Native Alaskan (0.6%); Multiracial (0.6%); and Other (0.3%). The largest proportion of respondents were full professors (45.0%), followed by associate professors (31.7%), and assistant professors (23.3%). The mean age was 51.3 years (SD = 10.6). Science, technology, engineering, and math (STEM) faculty represented 58.1% of the participants. There was sufficient power for the analyses, since a sample size of 345 (5 participants for each of the 69 parameters) was recommended to provide adequate statistical power for path analysis of the model with the most parameters (i.e., the trimmed model; Hatcher, 1994).

The COACHE Survey

The COACHE faculty satisfaction survey was developed based on peer-reviewed research related to faculty job satisfaction, interviews with faculty across ranks, and focus groups with higher education administrators. Its validity is bolstered by evidence across multiple types of institutions. It is a secure, web-based instrument that takes approximately 25 minutes to complete. The COACHE survey has been administered to more than 250 academic institutions around the United States since 2003 (COACHE, 2017) for the purposes of institutional improvement and higher education research (e.g., Lawrence et al., 2014; Ott & Cisneros, 2015). It contains 170 five-point Likert-type items pertaining to specific aspects of participants’ work (e.g., teaching, service, research, climate, satisfaction), with higher scores indicating greater agreement, satisfaction, clarity, and so on. It also contains various demographic/background questions. COACHE groups certain questions into scales called “benchmarks.” They typically were not conceptually consistent with SDT and thus were not included in our model, but the benchmarks are referenced where appropriate in the subsequent sections to provide estimates of concurrent validity.

The following constructs as shown in Figure 1 were operationalized from the 2013–2014 COACHE survey. First, item content unique to NTT faculty (e.g., contract renewal) was excluded. Second, scales were adapted from similar scales developed from COACHE data sets that had included SDT or SDT-related constructs (i.e., Lawrence et al., 2014; Ott & Cisneros, 2015; Seipel & Larson, 2016). Third, subscales were derived based on the results of an exploratory factor analysis in which each set of scale items loaded uniquely on one factor. Fourth, internal consistency analyses yielded satisfactory estimates.

Chair support

This scale consists of five 5-point Likert-type items, with higher scores indicating perception of more support from the department chair. It includes items from Seipel and Larson (2016) departmental support scale for NTT faculty that are specific to the department chair. The item content concerns the following aspects of the department chair: pace of decision-making, stated priorities, communication of priorities to faculty, fairness in evaluating work, and recognition received from the department chair. An example item stem was: “My department head’s or chair’s fairness in evaluating my work.” The items loaded on a unitary factor with factor loadings ranging from .90 to .96. The value of Cronbach’s coefficient of reliability α in this sample was .94. This scale correlated with the other environmental support scales in this sample at levels ranging from .07 to .42, suggesting that they are distinct environmental factors. Scores on this scale also correlated strongly with the COACHE benchmark for departmental leadership (r = .97, p < .01) in this sample. An additional initial estimate of convergent validity shows that Seipel and Larson (2016) measure of departmental support, which contains items pertaining to chair support, correlated .74 with the COACHE benchmark for departmental leadership.

Administrative support

This scale includes nine 5-point Likert-type items, with higher scores indicating more perceived administrative support, and was used in the Seipel and Larson (2016) study. The item content concerns: whether institutional priorities are acted upon consistently across all levels of leadership, as well as the stated priorities, communication of priorities, and pace of decision-making of the president/chancellor, chief academic officer, and dean/division head. An example item stem was: “My institution’s president’s/chancellor’s communication of priorities to faculty.” The items loaded on a unitary factor, with factor loadings ranging from .62 to .81. The value of Cronbach’s coefficient of reliability α was .92 in this sample. This scale correlated with the other environmental support scales in this sample at levels ranging from .16 to .42. Scores on this scale also correlated strongly with the COACHE benchmarks for senior leadership (r = .94, p < .01) and division leadership (r = .74, p < .01). Additional initial estimates of convergent validity show that Seipel and Larson (2016) administrative support scale correlated .92 and .89 with the COACHE benchmarks for senior leadership and division leadership, respectively.

Research support

This scale expanded upon the resource scales used in the COACHE studies conducted by Ott and Cisneros (2015) and Lawrence et al. (2014). It contained thirteen 5-point Likert-type items with higher scores indicating more perceived research support. The item content concerns institutional support for scholarly work, engaging undergraduates in research, obtaining external funding, securing graduate student assistance, managing leadership responsibilities, improving teaching, and traveling to present research; availability of course release time for research; and classrooms, office space, equipment, computing/technical support, and clerical support. An example item stem was: “The support your institution provides you for engaging undergraduates in your research/scholarly/creative work.” The items loaded on a unitary factor, with factor loadings ranging from .42 to .65. The value of Cronbach’s coefficient of reliability α was .86 in this sample. This scale correlated with the other environmental support scales in this sample ranging from .26 to .56. Scores on this scale also correlated strongly with the COACHE benchmarks for research (r = .86, p < .01) and facilities/resources (r = .82, p < .01). The comparable resource scales displayed moderate correlations with indices of well-being in two faculty samples: job satisfaction (labeled organizational commitment: r = .42, p < .001; Ott & Cisneros, 2015) and turnover intentions (r = .30, p < .001; Lawrence et al., 2014). Resources had a value of Cronbach’s coefficient of reliability α of .69 in Ott and Cisneros and .80 in Lawrence et al.

Department feedback/support

This measure is a composite that includes components of Seipel and Larson (2016) departmental support and personal/family support scales. It contained eleven 5-point Likert-type items with higher scores indicating more perceived feedback and support. The item content concerns departmental annual review process, course-level student evaluations, and consistency of evaluative standards; recognition received from peers; recognition received (in general) for teaching, scholarly work, and service; and perceived ability to find work–life balance. An example item stem was: “The course-level student teaching evaluations at my institution provide a meaningful evaluation of my teaching performance.” The items loaded on a unitary factor with factor loadings ranging from .49 to .72. The value of Cronbach’s coefficient of reliability α was .90 in this sample. This scale correlated with the other environmental support scales in this sample at values ranging from .31 to .56. Scores on this scale also correlated strongly with the COACHE benchmark for appreciation/recognition (r = .82, p < .01).

Promotion and tenure support

This measure is adapted from the tenure clarity scale used by Lawrence et al. (2014) to apply to clarity in both tenure and promotion. It incorporated items pertaining to tenure that were asked of assistant professors and pertaining to promotion that were asked of associate and full professors. The final scale contained four 5-point Likert-type items, with higher scores indicating more perceived support in pursuing promotion and tenure. The item content concerns the clarity of the process, criteria, standards, and dossier contents needed to obtain tenure/promotion. An example item stem was “The tenure standards (the performance thresholds) in my department.” The items loaded on a unitary factor with factor loadings ranging from .87 to .95. The value of Cronbach’s coefficient of reliability α was .94 in this sample. This scale correlated with the other environmental support scales in this sample at levels ranging from .07 to .50. Scores on this scale also correlated strongly with the COACHE benchmarks for tenure policies (r = .97, p < .01) and promotion (r = .95, p < .01).

Volitional autonomy

Volitional autonomy was operationalized by the autonomy scale used in two other COACHE studies (Lawrence et al., 2014; Ott & Cisneros, 2015) and identical to the Seipel and Larson (2016) volitional autonomy scale with the addition of one tenure-track item not relevant to their NTT sample. This scale contained seven 5-point Likert-type items, with higher scores indicating more volitional autonomy. The item content concerns discretion over committee work, equity of committee assignments, discretion over courses taught, equity of teaching workload, influence over research focus, and input regarding departmental policies and institutional priorities. An example item stem was: “The discretion you have over the content of the courses you teach.” The items in this sample loaded on a unitary factor with factor loadings ranging from .35 to .70. The value of Cronbach’s coefficient of reliability α was .76 in this sample and .79 in the Seipel and Larson NTT sample. Volitional autonomy displayed moderate correlations with indices of well-being in three faculty samples: global satisfaction (r = .51, p < .01; Seipel & Larson, 2016), job satisfaction (r = .39, p < .001; Ott & Cisneros, 2015) and turnover intentions (r = .26, p < .001; Lawrence et al., 2014). Autonomy had a value of Cronbach’s coefficient of reliability α of .62 in Ott and Cisneros and .72 in Lawrence et al. Faculty who intended to leave their institutions reported less autonomy than those who intended to stay (Ott & Cisneros, 2015).

Perceived competence

Perceived competence was operationalized in terms of faculty members’ perceptions of the time spent in multiple broad domains of competence (i.e., teaching, research, and service) and included 3 of 5 items from Ott and Cisneros (2015) labeled skill variety. Items concerning administration and outreach was not relevant in this sample. This scale contained three 5-point Likert-type items, with higher scores indicating more perceived competence in the domains mentioned above. An example item stem was: “The portion of your time spent on service (e.g., department/program administrative, faculty governance, committee work, advising/mentoring students, speaking to alumni or prospective students/parents).” The items in this sample loaded on a unitary factor with factor loadings ranging from .57 to .75 with a value of Cronbach’s coefficient of reliability α of .66. Perceived competence previously displayed a moderate correlation with job satisfaction (r = .36, p < .001; Ott & Cisneros, 2015).

Perceived relatedness

Perceived relatedness was operationalized by using the Seipel and Larson (2016) relatedness scale. The authors expanded the shorter collegiality subscales reported by Ott and Cisneros (2015) and Lawrence et al. (2014) to include dimensions of relatedness content beyond collegiality (e.g., intellectual vitality and interaction). This scale included twenty 5-point Likert-type items, with higher scores indicating more perceived relatedness. The item content included perceived collegiality of the department, amount of professional and personal contact with peers, intellectual vitality and productivity of peers, how much peers “pitch in” when needed, promotion of diversity in the department, overall perceived fit in the department, communication, conflict resolution, and engagement in dialogue with peers concerning student learning, teaching practices, and effective use of technology. An example item stem was: “How well you fit in your department (e.g., your sense of belonging in your department)?” The items loaded on a unitary factor, with factor loadings ranging from .38 to .79. The value of Cronbach’s coefficient of reliability α was .92 in this sample and .93 in the Seipel and Larson NTT sample. Volitional autonomy correlated strongly with global satisfaction (r = .51, p < .01) in Seipel and Larson (2016). In their faculty samples, collegiality displayed modest correlations with job satisfaction (r = .28, p < .001; Ott & Cisneros, 2015) and turnover intentions (r = .24, p < .001; Lawrence et al., 2014). Collegiality had a value of Cronbach’s coefficient of reliability α of .88 in both Ott and Cisneros and Lawrence et al. It also differentiated Asian international faculty who intended to stay from those who intended to leave (Lawrence et al., 2014).

Faculty well-being

Two indicators of well-being were used, teaching/service satisfaction and global satisfaction, previously used by Seipel and Larson (2016). The teaching/service satisfaction scale included seven 5-point Likert-type items with higher scores indicating more satisfaction with the teaching and service roles. The item content included number of committees served, students advised, and courses taught; level of courses taught, class sizes, quality of students taught, and amount of external funding faculty were expected to find. An example item was “Please rate your level of satisfaction or dissatisfaction with the following: The level of courses you teach.” In this sample, the items loaded on a unitary factor with factor loadings ranging from .40 to .66. The value of Cronbach’s coefficient of reliability α was .75 in this sample and was .72 in the Seipel and Larson NTT sample. Teaching/service satisfaction correlated .45 with global satisfaction in their study, suggesting they are measures of distinct constructs. Initial estimates of convergent validity show that the teaching/service satisfaction scale correlated .81 and .88 with the COACHE benchmark for nature of work—teaching in this sample and the Seipel and Larson sample, respectively.

The global satisfaction scale was the same scale used by Seipel and Larson (2016) with 2 additional items added. It included the 1-item job satisfaction measure used by Lawrence et al. (2014) and by Ott and Cisneros (2015). It consisted of six 5-point Likert-type items, with higher scores indicating greater overall job satisfaction. The item content included overall satisfaction with the position, department, and institution. An example item was: “If I had it to do all over, I would again choose to work at this institution.” The items loaded on a unitary factor, with factor loadings ranging from .42 to .90. The value of Cronbach’s coefficient of reliability α was .88 in this sample and was .93 in the Larson and Seipel NTT sample. No COACHE benchmarks pertained to global faculty satisfaction; however, the global satisfaction scale correlated with the 1-item measure used by Lawrence et al. and Ott and Cisneros at .92 and .91 in the present study and Seipel and Larson, respectively.

Procedure

The data set was an archival faculty satisfaction survey that had been collected at one Midwestern university in the fall and early spring of 2013–2014. Full-time T/TE faculty were invited by the Provost’s office to complete the survey. They then received an e-mail from COACHE containing a unique link to the survey. COACHE subsequently sent three reminder e-mails, and the Provost’s office sent a final e-mail to participants who had not yet responded. All full-time T/TE faculty at the university not hired in the same academic year as the survey (n = 1,171) were invited to complete the survey, and 49.6% did so. Because all data received from COACHE were deidentified and the authors had no contact with participants, approval was not required from the authors’ institutional review board.

Results

Preliminary Analyses and Descriptive Statistics

Missing data were dealt with systematically. First, before sending the data to the authors, COACHE had deleted respondents who completed the survey in an inordinately short time (i.e., more than two standard deviations below the mode) or used an identical response pattern throughout the survey. This resulted in an initial sample size of 581. The authors then removed items in which more than 20% of the data were missing and removed participants (n = 23) with extensive incomplete responses. Finally, participants’ responses on any construct in which more than 50% of the items were missing were omitted from the analysis. Table 1 presents the means, standard deviations, and correlations of the variables included in our analyses.

Table 1.
Means, Standard Deviations, and Intercorrelations of Environmental Factors, Psychological Needs, and Indices of Faculty Well-Being.

1 2 3 4 5 6 7 8 9 10

1. Chair support —

2. Administrative support .16 —

3. Research support .30 .42 —

4. Department feedback/support .31 .35 .56 —

5. Promotion and tenure support .21 .30 .38 .50 —

6. Volitional autonomy .41 .44 .64 .66 .46 —

7. Perceived competence .13 .21 .42 .41 .24 .47 —

8. Perceived relatedness .24 .38 .49 .67 .48 .58 .27 —

9. Teaching/service satisfaction .15 .17 .37 .28 .15 .43 .36 .23 —

10. Global satisfaction .30 .43 .60 .68 .48 .62 .35 .66 .25 —

Mean 3.01 3.09 3.26 3.32 3.58 3.52 3.52 3.63 3.40 3.62

Standard deviation 1.57 1.08 0.70 0.86 1.26 0.78 0.83 0.70 0.96 1.00

Note. N = 558. All variables are scored 1 to 5 on Likert-type scales, with higher scores indicating a greater degree of the variable. All correlations greater than .11 are significant at p < .05. Items 1–5 are environmental factors, items 6–8 are psychological needs, and items 9 and 10 are indices of faculty well-being.

Path Analysis

Path models were estimated in MPlus version 7.4 (Muthén & Muthén, 2012) using robust maximum likelihood (MLR) to estimate case wise parameters for missed items. MLR is robust to nonnormality and was used because multiple independent variables violated assumptions of normality, which were tested using a Kolmogorov–Smirnov test of univariate normality. Goodness of fit for the models was assessed using the guidelines of Hu and Bentler (1999): comparative fit index (CFI) of .95 or greater, root mean square error of approximation (RMSEA) of .06 or less, and standardized root mean square residual (SRMR) of .08 or less.

The hypothesized partial mediation model was a saturated model, so its fit was perfect by all indices. A trimmed model was then created in which the nonsignificant paths from the hypothesized model were fixed to zero. This model was created to examine if omitting paths with less empirical support would fit the data equally well while being more parsimonious. This model yielded a good fit, χ²(13, N = 558) = 16.51, p = .22, CFI = 1.00, RMSEA = .02, SRMR = .01. The Satorra–Bentler corrected chi-square statistic for this model was χ²(13, N = 558) = 17.71, p = .17. The nonsignificant chi-square difference test, Δχ²(13, N = 558) = 17.71, p = .17, indicated that the saturated and trimmed models fit the data equally well. Figure 2 presents the more parsimonious trimmed model with the standardized β coefficients for all the significant direct and indirect effects.

Figure 2.
Trimmed model results. Correlations were specified among the exogenous, mediator, and outcome variables, respectively, but are omitted for visual clarity. p < .05. p < .01. p < .001.

As seen in Figure 2, none of the environmental factors displayed significant direct relationships with teaching/service satisfaction. Administrative support, research support, department feedback/support, and promotion and tenure support had significant direct relationships with global satisfaction. All five environmental factors had significant direct relationships with volitional autonomy. Two environmental factors (research support and department feedback/support) had significant direct relationships with perceived competence. Four environmental factors (administrative support, research support, department feedback/support, and promotion and tenure support) had significant direct relationships with perceived relatedness. Volitional autonomy and perceived competence had significant direct relationships with only teaching/service satisfaction, whereas perceived relatedness had a significant direct relationship with only global satisfaction.

A fully mediated model was also tested in which no direct paths were allowed from the environmental factors to the outcome variables. This model did not yield a good fit, χ²(10, N = 558) = 99.69, p < .001, CFI = .91, RMSEA = .13, SRMR = .03. The Satorra–Bentler corrected chi-square statistic (Satorra & Bentler, 2001) for this model was χ²(10, N = 558) = 111.21, p < .001.

Bootstrap tests using bias-corrected 95% confidence intervals were used to evaluate the statistical significance of the mean indirect effects. The trimmed model was used for this analysis because it demonstrated good fit and was more parsimonious than the hypothesized model. The calculation was repeated with 1,000 samples to yield parameter estimates for total and specific indirect effects. A confidence interval not containing 0 indicated that the mean indirect effect across the samples was significant (p < .05). Bootstrap analysis provided greater statistical power and did not make any assumptions regarding multivariate normality (Preacher & Hayes, 2008). Table 2 presents the magnitude and statistical significance of the specific and total indirect effects of the environmental factors on faculty well-being through psychological needs using the bootstrapping procedure. Table 2 is organized so that the first five specific mean indirect effects reported concern the first mediator, volitional autonomy; the next two specific mean indirect effect reported concerns the second mediator, perceived competence. The final four specific mean indirect effects reported concern the third mediator, perceived relatedness.

Table 2.
Bootstrap Analysis of Magnitude and Statistical Significance of Indirect Effects of Environmental Supports on Faculty Well-being (Satisfaction) through the Psychological Needs.

Indirect Effects β and Product Mean Indirect Effect (b)^a SE of Mean^a 95% BC CI [Lower, Upper]^a

1a. Chair support Volitional autonomy Teaching/service satisfaction (.18) × (.34) = .061 .04 .01 [.015, .058]*

1b. Administrative support Volitional autonomy Teaching/service satisfaction (.14) × (.34) = .048 .04 .01 [.020, .061]*

1c. Research support Volitional autonomy Teaching/service satisfaction (.30) × (.34) = .102 .14 .03 [.087, .192]*

1d. Department feedback/support Volitional autonomy Teaching/service satisfaction (.34) × (.34) = .116 .13 .03 [.059, .194]*

1e. Promotion and tenure support Volitional autonomy Teaching/service satisfaction (.10) × (.34) = .034 .03 .01 [.009, .042]*

2a. Research support Perceived competence Teaching/service satisfaction (.27) × (.20) = .054 .07 .02 [.026, .120]*

2b. Department feedback/support Perceived competence Teaching/service satisfaction (.26) × (.20) = .052 .06 .02 [.024, .092]*

3a. Administrative support Perceived relatedness Global satisfaction (.11) × (.28) = .031 .03 .01 [.007, .049]*

3b. Research support Perceived relatedness Global satisfaction (.12) × (.28) = .034 .05 .02 [.006, .087]*

3c. Department feedback/support Perceived relatedness Global satisfaction (.49) × (.28) = .137 .16 .03 [.094, .226]*

3d. Promotion and tenure support Perceived relatedness Global satisfaction (.16) × (.28) = .045 .04 .01 [.013, .059]*

Note. N = 558. BC CI = bias-corrected confidence interval.

^aThese values are based on the unstandardized path coefficients.

*95% Confidence interval does not include zero and therefore is significant at p < .05.

Volitional autonomy significantly fully mediated the relationships between five environmental factors (chair support, administrative support, research support, department feedback/support, and promotion and tenure support) and teaching/service satisfaction. Perceived competence significantly fully mediated the relationships between two environmental factors (research support and department feedback/support) and teaching/service satisfaction. Perceived relatedness significantly partially mediated the relationships between four environmental factors (administrative support, research support, department feedback/support, and promotion and tenure support) and global satisfaction.

Discussion

The overarching purpose of this research is to provide an SDT conceptual framework to the research literature on the well-being of T/TE faculty. Prior to the Seipel and Larson (2016) study that focused on NTT faculty, the faculty satisfaction literature was predominantly aimed at identifying environmental factors and demographic factors that may help or hinder faculty well-being (e.g., Hagedorn, 2000). Our aim was to examine if SDT could be used as a lens to inform scholars and administrators as to what may be mediating the relationship between environmental supports and faculty well-being.

These findings suggest, as we suspected, that the three psychological needs proposed by SDT acted as partial or full mediators between environmental factors and faculty well-being. In comparing the trimmed partial mediation model with the original model, the trimmed model presented in Figure 2 yielded a good fit, was more parsimonious, and was not significantly different than the original model. Moreover, the fully mediated model did not yield a good fit to the data. The results support the findings of other SDT research conducted with other populations in concluding that the three SDT needs mediate the relationships between environmental factors and well-being in employees (Doménech-Betoret et al., 2015; Fernet et al., 2013; Gillet et al., 2013, 2015; Graves & Luciano, 2013). These findings also support research that shows the three SDT needs do not act as one construct but rather three related constructs (e.g., Graves & Luciano, 2013; Fernet et al., 2013).

Volitional autonomy fully explained the relationships between all five of the environmental factors and teaching/service satisfaction. As anticipated by SDT, T/TE faculty members’ linkage between their perceptions of the work environment and their satisfaction with their teaching/service roles was completely mediated by the extent to which they felt they were making their own choices. This finding is consistent with SDT research with employees other than faculty that have shown volitional autonomy mediates the relationship of work environment factors and positive indicators of well-being (job satisfaction; Graves & Luciano, 2013) as well as work environment factors and negative indicators of well-being (emotional exhaustion and depersonalization; Fernet et al., 2013).

The findings of this T/TE faculty sample are also partially supportive of Seipel and Larson (2016) NTT faculty sample. Volitional autonomy fully mediated the relation of administrative support in both studies. Perceived relatedness fully mediated the relation of all three Seipel and Larson’s environmental factors; in this sample, perceived relatedness fully or partially mediated four of the five environmental factors. However, some findings were inconsistent across the two studies. Seipel and Larson provided evidence that volitional autonomy was unrelated to departmental support (which included some item overlap with our environmental factors of chair support and department feedback/support), while the findings presented here showed chair support and department feedback/support to be related to volitional autonomy. Most notably, volitional autonomy was only a modest mediator of one environmental factors and one index of NTT faculty well-being. In contrast, volitional autonomy fully mediated the relationships of five environmental factors and teaching/service satisfaction in our T/TE faculty sample. Volitional autonomy’s meditational role may be dependent on the faculty’s assignment as either TT or NTT.

Perceived competence also fully explained the relationships of two of the five environmental factors with teaching/service satisfaction. Consistent with prior research, perceived competence mediated the relationships between the environmental work factors and indices related to well-being (Fernet et al., 2013). Moreover, the relationship of research support and department feedback/support with satisfaction with teaching/service was mediated by the extent to which faculty perceived they were spending an adequate time on research and not too much time on service and teaching. This is consistent with the Vera, Salanova, and Martín (2010) discussion of how faculty having the right balance of time spent on required tasks will feel more competent and experience greater well-being. One prior study (Graves & Luciano, 2013) yielded findings different from ours; namely, leader–member exchange, similar to chair support, was indirectly related to job satisfaction through perceived competence, whereas in our study it was unrelated to perceived competence.

Perceived relatedness was the only SDT need to partially mediate the relation between four of the five environmental factors and global satisfaction. These results support the assertion of SDT that a sense of caring and mutual concern for others in the work place mediates between perceptions of the work environment and one’s overall satisfaction. Perceived relatedness as a partial mediator in this T/TE faculty sample supports Seipel and Larson's (2016) findings of NTT faculty that perceived relatedness fully mediated the relationship of three environmental factors (departmental support, administrative support, and personal and family support) with two indices of faculty well-being. Other research has also shown this SDT need to mediate the relationship between environmental work factors (e.g., social support and leader–member exchange) and well-being (Fernet et al., 2013; Graves & Luciano, 2013).

Four environmental supports were directly related to global satisfaction, one of the two indices of faculty well-being, as can be seen in Figure 2. This is consistent with related work linking faculty satisfaction to research support (e.g., Vera et al., 2010), administrative support (e.g., Chen et al., 2005; Verhaegen, 2005), department feedback/support (e.g., Grawitch et al., 2007; Huggett et al., 2012), and promotion and tenure support (e.g., Chung et al., 2010; Schrodt et al., 2003).

Limitations

One limitation in this sample is the cross-sectional nature of the data collection, which does not facilitate interpretation of causality. Using SDT as a framework drove the researchers to present the three SDT needs as mediators between the work environment and faculty well-being. Future researchers could apply experimental and longitudinal designs to extend our findings. Another limitation concerns the inability to generalize our findings beyond predominantly Caucasian T/TE faculty. A third limitation of this sample was the archival nature of the data. Researchers were not able to select other measures or include other constructs beyond what was in the data set. The benefit of the archival data set was the wide applicability of COACHE data sets across the country to be re-envisioned from an SDT conceptual lens.

Implications and Future Directions

Higher education administrators may want to consider SDT as a way to conceptually frame future COACHE faculty satisfaction survey results. The benefit for administrators is that they could envision ways of supporting and structuring policies, practices, and structures in the campus environment that would enhance faculty members’ volitional autonomy, perceived competence, and perceived relatedness. Administrators may provide means to support faculty competence in their roles as instructors and researchers but may not be thinking in terms of ensuring faculty volitional decision-making in the various aspects of their roles. Moreover, administrators may not be considering the crucial role that belonging or connectedness to other colleagues plays in recruitment, retention, and job satisfaction.

Future researchers will want to extend this research to specific racial and ethnic minority faculty as well as to other geographic regions and other institutions besides doctoral universities: highest research activity (Research I) institutions. Other environmental factors (e.g., mentoring support) could be included in future investigations. Access to data across a variety of institutions is imperative, as it would provide sample sizes needed to investigate additional demographic and environmental variables and use more informative statistical analyses such as hierarchical linear modeling and structural invariance analysis. There is ample opportunity for future researchers to expand the scope of scholarship on faculty well-being through the lens of SDT, and it will serve to improve the academy and the society that it serves.

	1	2	3	4	5	6	7	8	9	10
1. Chair support	—
2. Administrative support	.16	—
3. Research support	.30	.42	—
4. Department feedback/support	.31	.35	.56	—
5. Promotion and tenure support	.21	.30	.38	.50	—
6. Volitional autonomy	.41	.44	.64	.66	.46	—
7. Perceived competence	.13	.21	.42	.41	.24	.47	—
8. Perceived relatedness	.24	.38	.49	.67	.48	.58	.27	—
9. Teaching/service satisfaction	.15	.17	.37	.28	.15	.43	.36	.23	—
10. Global satisfaction	.30	.43	.60	.68	.48	.62	.35	.66	.25	—
Mean	3.01	3.09	3.26	3.32	3.58	3.52	3.52	3.63	3.40	3.62
Standard deviation	1.57	1.08	0.70	0.86	1.26	0.78	0.83	0.70	0.96	1.00

Indirect Effects	β and Product	Mean Indirect Effect (b)^a	SE of Mean^a	95% BC CI [Lower, Upper]^a
1a. Chair support	Volitional autonomy	Teaching/service satisfaction	(.18) × (.34) = .061	.04	.01	[.015, .058]*
1b. Administrative support	Volitional autonomy	Teaching/service satisfaction	(.14) × (.34) = .048	.04	.01	[.020, .061]*
1c. Research support	Volitional autonomy	Teaching/service satisfaction	(.30) × (.34) = .102	.14	.03	[.087, .192]*
1d. Department feedback/support	Volitional autonomy	Teaching/service satisfaction	(.34) × (.34) = .116	.13	.03	[.059, .194]*
1e. Promotion and tenure support	Volitional autonomy	Teaching/service satisfaction	(.10) × (.34) = .034	.03	.01	[.009, .042]*
2a. Research support	Perceived competence	Teaching/service satisfaction	(.27) × (.20) = .054	.07	.02	[.026, .120]*
2b. Department feedback/support	Perceived competence	Teaching/service satisfaction	(.26) × (.20) = .052	.06	.02	[.024, .092]*
3a. Administrative support	Perceived relatedness	Global satisfaction	(.11) × (.28) = .031	.03	.01	[.007, .049]*
3b. Research support	Perceived relatedness	Global satisfaction	(.12) × (.28) = .034	.05	.02	[.006, .087]*
3c. Department feedback/support	Perceived relatedness	Global satisfaction	(.49) × (.28) = .137	.16	.03	[.094, .226]*
3d. Promotion and tenure support	Perceived relatedness	Global satisfaction	(.16) × (.28) = .045	.04	.01	[.013, .059]*

Footnotes

Acknowledgments

We would like to thank Dr. Meifen Wei for assistance in statistical consulting and Drs. Bonnie Bowen, Nancy Franz, Susan Renoe, and Mariko Chang who served as grant consultants. A special thanks to Dr. Kiernan Mathews, Executive Director of the Collaborative on Academic Careers in Higher Education (COACHE) for giving us access to the data.

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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This manuscript was supported by an EAGER Grant from the National Science Foundation (Award # IIA-1449187).

References

August

Waltman

(2004). Culture, climate, and contribution: Career satisfaction among female faculty. Research in Higher Education, 45, 177–192. doi:10.1023/b: rihe.0000015694.14358.ed

Benson

R. T.

Mathews

Trower

(2014). The collaborative on academic careers in higher education: Faculty Job Satisfaction Survey, 2011-2014 (Dataverse version) [data file and codebook]. Cambridge, MA: Harvard University.

Brown

F. W.

Moshavi

(2002). Herding academic cats: Faculty reactions to transformational and contingent reward leadership by department chairs. Journal of Leadership Studies, 8, 79–93. doi:10.1177/107179190200800307

Butts

M. M.

Casper

W. J.

Yang

T. S.

(2013). How important are work–family support policies? A meta-analytic investigation of their effects on employee outcomes. Journal of Applied Psychology, 98, 1–25. doi:10.1037/a0030389

Chen

Beck

S. L.

Amos

L. K.

(2005). Leadership styles and nursing faculty job satisfaction in Taiwan. Journal of Nursing Scholarship, 37, 374–380. doi:10.1111/j.1547-5069.2005.00064.x

Chung

K. C.

Song

J. W.

Kim

H. M.

Woolliscroft

J. O.

Quint

E. H.

Lukacs

N. W.

Gyetko

M. R.

(2010). Predictors of job satisfaction among academic faculty members: Do instructional and clinical staff differ? Medical Education, 44, 985–995. doi:10.1111/j.1365-2923.2010.03766.x

Collaborative on Academic Careers in Higher Education website. (2017). Retrieved from http://coache.gse.harvard.edu/

Deci

E. L.

Ryan

R. M.

(1985). The general causality orientations scale: Self-determination in personality. Journal of Research in Personality, 19, 109–134. doi:10.1016/00926566(85)90023-6

Doménech-Betoret

Lloret-Segura

Gómez-Artiga

(2015). Teacher support resources, need satisfaction and well-being. The Spanish Journal of Psychology, 18, 12. doi:10.1017/sjp.2015.8

10.

Eagan

M. K.

Jaeger

A. J.

Grantham

(2015). Supporting the academic majority: Policies and practices related to part-time faculty’s job satisfaction. The Journal of Higher Education, 86, 448–483. doi:10.1353/jhe.2015.0012

11.

Fernet

Austin

Trépanier

S. G.

Dussault

(2013). How do job characteristics contribute to burnout? Exploring the distinct mediating roles of perceived autonomy, competence, and relatedness. European Journal of Work and Organizational Psychology, 22, 123–137. doi:10.1080/1359432X.2011.632161

12.

Gillet

Colombat

Michinov

Pronost

A. M.

Fouquereau

(2013). Procedural justice, supervisor autonomy support, work satisfaction, organizational identification and job performance: The mediating role of need satisfaction and perceived organizational support. Journal of Advanced Nursing, 69, 2560–2571. doi:10.1111/jan.12144

13.

Gillet

Forest

Benabou

Bentein

(2015). The effects of organizational factors, psychological need satisfaction and thwarting, and affective commitment on workers’ well-being and turnover intentions. Le Travail Humain: A Bilingual and Multi-Disciplinary Journal in Human Factors, 78, 119–140. doi:10.1017/sjp.2015.30

14.

Graves

L. M.

Luciano

M. M.

(2013). Self-determination at work: Understanding the role of leader-member exchange. Motivation and Emotion, 37, 518–536. doi:10.1007/s11031012-9336-z

15.

Grawitch

M. J.

Trares

Kohler

J. M.

(2007). Healthy workplace practices and employee outcomes. International Journal of Stress Management, 14, 275. doi:10.1037/10725245.14.3.275

16.

Hagedorn

L. S.

(2000). Conceptualizing faculty job satisfaction: Components, theories, and outcomes. New Directions for Institutional Research, 2000, 5–20. doi:10.1002/ir.10501

17.

Hatcher

(1994). A step-by-step approach to using the SAS system for factor analysis and structural equation modeling. Cary, NC: SAS Institute.

18.

Horn

J. E.

Taris

T. W.

Schaufeli

W. B.

Schreurs

P. J.

(2004). The structure of occupational well-being: A study among Dutch teachers. Journal of occupational and Organizational Psychology, 77, 365–375. doi:10.1348/0963179041752718

19.

L. T.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1–55. doi:10.1080/10705519909540118

20.

Huggett

K. N.

Greenberg

R. B.

Rao

Richards

Chauvin

S. W.

Fulton

T. B.

… Simpson

. (2012). The design and utility of institutional teaching awards: A literature review. Medical Teacher, 34, 907–919. doi:10.3109/0142159x.2012.731102

21.

Judge

T. A.

Thoresen

C. J.

Bono

J. E.

Patton

G. K.

(2001). The job satisfaction-job performance relationship: A qualitative and quantitative review. Psychological Bulletin, 127, 376–407. doi:10.1037/0033-2909.127.3.376

22.

Lawrence

J. H.

Celis

Ott

(2014). Is the tenure process fair? What faculty think. The Journal of Higher Education, 85, 155–192. doi:10.1353/jhe.2014.0010

23.

Leung

Siu

Spector

P. E.

(2000). Faculty stressors, job satisfaction, and psychological distress among university teachers in Hong Kong: The role of locus of control. International Journal of Stress Management, 7, 121–138.

24.

Moors

A. C.

Malley

J. E.

Stewart

A. J.

(2014). My family matters: Gender and perceived support for family commitments and satisfaction in academia among postdocs and faculty in STEMM and non-STEMM fields. Psychology of Women Quarterly, 38, 460–474.

25.

Muthén

L. K.

Muthén

B. O.

(2012). Mplus user’s guide (7th ed.). Los Angeles, CA: Muthén & Muthén.

26.

J. Y.

Ntoumanis

Thøgersen-Ntoumani

Deci

E. L.

Ryan

R. M.

Duda

J. L.

Williams

G. C.

(2012). Self-determination theory applied to health contexts a meta-analysis. Perspectives on Psychological Science, 7, 325–340. doi:10.1177/1745691612447309

27.

T. W.

Sorensen

K. L.

(2008). Toward a further understanding of the relationships between perceptions of support and work attitudes: A meta-analysis. Group & Organization Management, 33, 243–268. doi:10.1177/1059601107313307

28.

Ott

Cisneros

(2015). Understanding the changing faculty workforce in higher education: A comparison of full-time non-tenure track and tenure line experiences. Education Policy Analysis Archives, 23, 1–28. doi:10.14507/epaa.v23.1934

29.

Preacher

Hayes

(2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891. doi:10.3758/BRM.40.3.879

30.

Ryan

R. M.

Deci

E. L.

(2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55, 68–78. doi:10.1037//0003-066x.55.1.68

31.

Satorra

Bentler

P. M.

(2001). A scaled difference chi-square test statistic for moment structure analysis. Psychometrika, 66, 507–514. doi:10.2139/ssrn.199064

32.

Schrodt

Cawyer

C. S.

Sanders

(2003). An examination of academic mentoring behaviors and new faculty members’ satisfaction with socialization and tenure and promotion processes. Communication Education, 52, 17–29. doi:10.1080/03634520302461

33.

Seipel

M. T.

Larson

L. M

. (2016). Supporting non-tenure-track faculty well-being. Journal of Career Assessment, 26, 154–171.

34.

Vera

Salanova

Martin

(2010). University faculty and work-related well-being: The importance of the triple work profile. Electronic Journal of Research in Educational Psychology, 8, 581–602. doi:10.1037/e573012012-089

35.

Verhaegen

(2005). Academic talent: Quo vadis? Recruitment and retention of faculty in European business schools. Journal of Management Development, 24, 807–818. doi:10.1108/02621710510621312