Manager-Led Group Meetings

Meeting Relevance and the Psychological Conditions for Engagement

Meeting relevance refers to the degree to which workgroup meetings called by the manager are perceived as relevant to the employees who attend the meeting (Allen et al., 2012). In terms of psychological meaningfulness, relevant meetings are experienced as valuable and good uses of employee time (Allen et al., 2012). Furthermore, consistent with organizational support theory, meetings perceived as relevant serve to demonstrate respect and support for the employee’s efforts on the job, as they more readily promote achievement and goal accomplishment. Irrelevant meetings can demonstrate a lack of appreciation for an employee’s workload, responsibilities, and goals that should serve to decrease psychological meaningfulness.

Relevant meetings also promote psychological safety. By ensuring the topics discussed are related to the employees’ work-related activities, employees are in position to contribute actively to the discussion on hand (Allen et al., 2012; Sonnentag, 2001). Alternatively, if the topics are not perceived as relevant, but the employee is still present, he or she is not able to actively participate to the same extent. In fact, broaching tangential topics and/or topics not central to the employee may serve to threaten self-image, given that the employee is then not well-positioned to present himself or herself in a positive light to others or to the supervisor.

As for psychological availability, the connection to meeting relevance would appear fairly direct. Relevant meetings, by definition, should provide employees with the information and knowledge resources they see as needed to effectively carry out their role. As a result, by their very nature, relevant meetings should promote psychological availability. This assumption is consistent with goal setting theory, which suggests that one way to increase worker motivation is by connecting work processes and projects to overt goals of the organization as well as to personal goals of the employee (Klein, Wesson, Hollenbeck, & Alge, 1999).

Thus, the following are hypothesized:

Voice and the Psychological Conditions for Engagement

Voice in meetings refers to the degree to which managers encourage employees to speak up in workgroup meetings and provide them with adequate time to express their thoughts and ideas in that setting (Appelbaum, Hebert, & Leroux, 1999; Gordon & Infante, 1980). Instead of simply asking for feedback on particular decisions relevant to each employee’s job (i.e., participation in decision making), managers also promote the free flow of ideas and opinions generally about all topics discussed during the meeting. Consistent with organizational support theory, employees who feel they have voice in meetings are likely to be willing to bring up (i.e., reciprocate) issues, concerns, or problems they are facing, rather than simply responding to decision points presented by the manager. Encouraging voice in meetings may affect all three psychological conditions for engagement.

In terms of psychological meaningfulness, employees are likely to feel like a valued member of the workgroup when they observe that managers are encouraging them to share their ideas in meetings. These feelings of freedom of expression also help employees develop psychological safety by enabling them to feel safe to share ideas without experiencing negative consequences to their self-image. After all, it is hard to imagine employees perceiving that they have voice in meetings if they did not truly feel safe to contribute and participate. Furthermore, when employees’ ideas are shared openly, answers to questions raised in the meeting may be answered more fully, which should provide for increased intellectual resources needed for them to engage in their work—thus, promoting psychological availability. Therefore, the following are hypothesized:

Meeting Time Management and the Psychological Conditions for Engagement

Meeting time management is the extent to which managers start meetings on time, end when scheduled to end, and schedule meetings with adequate time for employees to arrange their other activities for the day. In terms of the psychological conditions for engagement, effective time management behavior may facilitate the experience of psychological meaningfulness for the employees by demonstrating that the manager cares and respects their time (Kahn, 1990). Furthermore, workgroup meeting time management may promote psychological safety by ensuring predictability (e.g., respect for the attendees’ schedule) in the work environment so that the employee is able to effectively meet his or her other commitments (Kahn, 1990). Finally, in terms of psychological availability, managers who schedule workgroup meetings in an appropriate manner (e.g., not over-scheduling meetings) do not overly infringe upon employees’ time, which is a scarce resource that is likely related to employees’ availability to engage. Employees may reciprocate these efforts by the manager with increased engagement in their work activities. Thus, the following are hypothesized:

According to Kahn (1990), as discussed earlier, psychological meaningfulness, safety, and availability define “the experiential conditions whose presence influenced people to personally engage and whose absence influenced them to personally disengage” (p. 703). In other words, these psychological conditions are necessary for engagement to occur and without them individuals may not incorporate themselves in their work. In addition, previous research shows the connection between these psychological conditions and overall engagement (see May et al., 2004) and this study seeks to confirm those relationships once more. Thus, the following are proposed:

Ultimately, we are proposing a mediated model, whereby the three psychological conditions mediate the relationship between manager usage/facilitation of meetings and employees engagement at work (see Figure 1). Previous research often neglected to examine these psychological conditions (e.g., Saks, 2006) that Kahn originally theorized must be met before attitudinal and behavioral engagement could occur for individuals (Kahn, 1990). Furthermore, akin to what is seen in the organizational support framework, it is likely that manager usage/facilitation of meetings first impacts employees psychologically before their attitudes and behaviors change (i.e., engagement). In this study, the focus is on the development of the employee engagement attitudes (see Macey & Schneider, 2008, for a discussion of attitudinal vs. behavioral engagement), though the general assumption is that the attitude precedes behavioral engagement. Thus, the following mediation hypotheses are proposed.

We also propose that this mediated model persists even while controlling for important attitudinal (i.e., satisfaction with supervisor and work satisfaction) and meeting-related (i.e., meeting load) variables (Luong & Rogelberg, 2005).

Participants and Procedure

Participants for this study were recruited from among the alumni of a large university in the Southeast United States. A prenotification email was sent to potential participants (n = 11,552). This prenotification email served two purposes: (a) to screen out nondeliverable email addresses (n = 3,142) and (b) to provide notification that a survey would be arriving via email soon. Two surveys were administered in an effort to test the forgoing hypotheses. The surveys were administered using an online survey tool (i.e., SurveyMonkey). After sending the prenotification email, an email invitation was sent that included the link to the first survey. The first survey assessed demographics, meeting-related variables, and psychological conditions. One week later, a second survey was emailed to those who completed the first survey. This second survey assessed employee engagement and satisfaction with the supervisor.

Through the development and administration of the surveys, two major steps were taken to mitigate common-method bias concerns. Most substantively, measurement of the outcome variable was separated in time from measurement of the predictor variables. Another procedural remedy for common-method bias was counterbalancing question order on the survey instrument (Conway & Lance, 2010; Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). By rearranging the order of the measures on survey one, we were able to better control for item-context-induced mood states, priming effects, and other biases related to question context or item location on the survey. For this study, five different versions of the first survey were created. Each survey had a different ordering of variables/scales for participants to assess.

Of the potential participants who received the link to the first survey in a subsequent email (n = 8,410), 673 completed the survey for a response rate of 8%. Because the response rate was so low, the email list administrator (i.e., director of the university’s alumni association) was contacted for feedback on why so few people participated. The email list administrators indicated that at least 50% of the emails are not checked frequently. Thus, the actual response rate is likely higher than 8%, perhaps even as high as 16%.

Given the study’s focus, the population of interest was working adults who attend meetings with their supervisor on a regular basis. Participants in this sample who did not meet these criteria were removed and were not asked to participate in the second survey (n = 86). Thus, a total of 587 individuals were sent an invitation to complete the second survey. Of those individuals, 63.2% (n = 370) completed the second survey. Following recommendations from current Structural Equation Modeling (SEM) researchers (Schumacker & Lomax, 2004), respondents with more than 5% of their data missing or who had more than two items missing from the focal scales were dropped (n = 51). Thus, the final usable sample included 319 respondents. Item-mean substitution was used to replace all missing values before proceeding with data analysis. The sample was 52.7% female with an average age of 43 years. The average tenure with the current work organization was 9.5 years with an average tenure of 3.7 years with the current supervisor. Ninety-seven percent were college graduates and worked more than 20 hr per week. About half the sample respondents indicated they supervise others (49%). The sample also represented a variety of organizational types: 32% publicly traded firms, 19% privately held firms, 16% nonprofit firms, and 33% public sector (e.g., government).

Because these response rates were low, a number of steps were taken to check for nonresponse bias following current guidelines from survey research methodologists (e.g., Rogelberg & Stanton, 2007). First, a wave analysis was conducted. Early respondents did not differ from later respondents (submitted after the imposed deadline) on the variables assessed in the first survey. Second, an interest-level analysis was conducted comparing those who indicated they wanted to receive a summary of the results with those who did not indicate an interest in seeing a results summary. Results indicate that the means and standard deviations on the focal variables were nearly identical across these groups, providing further evidence that nonresponse bias was not present in these data. Third, sample demographic parameters (e.g., education, gender, and age) were nearly identical to what was known about the overall population. Fourth, split-group mean comparison analyses were used to verify that those who completed both surveys did not differ substantially from those who completed only the first survey (and received the second survey invitation) on the predictor variables. The analyses showed no significant mean differences. Based on these analyses, nonresponse bias did not appear to be present.

Measures

As the focus of this study is on manager usage/facilitation of meetings, all the meetings variables focused on only meetings led by the manager (e.g., Baran, Rhoades Shanock, Rogelberg, & Scott, 2012). Survey directions asked participants only to think about meetings led by their manager.

Voice was assessed using a seven-item measure adapted from Gordon and Infante (1980) focusing on the degree to which employees felt they had voice and freedom to discuss concerns in the meeting context. Participants were asked to “Think of the meetings with your supervisor that he or she leads. Regarding only these meetings, how frequently does he or she do the following.” Similar instructions are used for all meeting-related measures. Sample items include “Give employees time to express concerns about company policies” and “Provide time for employees to express disagreements with management practices.” Ratings were made using a 5-point scale ranging from 1 being “never” to 5 being “always.”

Meeting time management was assessed using a five-item measure adapted from Baran et al. (2012) focusing on how the manager schedules and uses meeting time. A sample item is “Start meetings on time.” Ratings were made using a 5-point scale ranging from 1 being “never” to 5 being “always.”

Meeting relevance was assessed using seven items adapted from Sawyer’s (1992) goal and process clarity scale. The items were modified to assess whether supervisor-led meetings are relevant to the accomplishment of work goals. A sample item is “Meetings led by my supervisor are relevant to my job.” Items were assessed using a 5-point scale ranging from 1 being “strongly disagree” to 5 being “strongly agree.”

Psychological conditions for engagement were assessed using items developed by May et al. (2004). All scales were rated using a 5-point scale ranging from 1 being “strongly disagree” to 5 being “strongly agree.” Instructions for each scale state “Think about the work that you do. Indicate the extent to which you agree or disagree with the following statements.” Meaningfulness was assessed using six items (e.g., “the work I do on this job is very important to me”), psychological safety was assessed using three items (e.g., “My work environment is nonthreatening”), and psychological availability was assessed using five items (e.g., “I am confident in my ability to handle competing demands at work”).

Employee engagement was assessed using the Utrecht Work Engagement Scale designed to assess overall employee engagement (Schaufeli & Bakker, 2003). This is a 16-item measure designed to assess three facets of employee engagement: vigor, dedication, and absorption. The instructions read “The following statements are about how you feel at work. Please read each statement carefully and decide if you ever feel this way about your job.” Sample items include “At my work, I feel bursting with energy,” “I find the work that I do full of meaning and purpose,” and “Time flies when I am working.” Ratings were made on a 5-point scale, ranging from 1 being “never” to 5 being “always.” Research evidence indicates that the three dimensions of work engagement are highly correlated (e.g., r > .65; Schaufeli & Salanova, 2007) and often examined as one overall factor (Halbesleben & Wheeler, 2008; Mauno, Kinnunen, Makikangas, & Natti, 2005). Given parsimony and that a confirmatory factor analysis (CFA) showed that the one-factor model fit as well as the three-factor model for the current sample, an overall score for employee engagement was computed for each respondent (CFA results available upon request from the first author).

Descriptive Statistics

Means, standard deviations, internal reliabilities, and intercorrelations among the variables used in this study are reported in Table 1. Although the psychological conditions for engagement were only modestly correlated (i.e., .26-.31), the correlations between the manager meeting usage/facilitation variables were fairly substantial (i.e., .53-.68). Thus, the discriminant validity of the constructs is tested next before proceeding with hypothesis testing.

OPEN IN VIEWER

Table 1.

Means, Standard Deviations, and Intercorrelations of All Measures (N = 319).

	M	SD	1	2	3	4	5	6	7	8	9	10
1. Voice	3.47	.95	(.93)
2. Meeting time management	3.83	.76	.56*	(.86)
3. Meeting relevance	3.66	.90	.68*	.53*	(.95)
4. Meaningfulness	4.14	.77	.26*	.31*	.32*	(.95)
5. Safety	3.79	.64	.56*	.40*	.42*	.32*	(.82)
6. Availability	4.37	.50	.11*	.18*	.05	.26*	.31*	(.87)
7. Employee engagement	3.70	.59	.29*	.32*	.32*	.74*	.42*	.30*	(.93)
8. Supervisor satisfaction	2.31	.87	.58*	.53*	.64*	.22*	.52*	.01	.30*	(.78)
9. Work satisfaction	2.55	.84	.29*	.27*	.28*	.65*	.39*	.10	.64*	.37*	(.87)
10. Meeting load a	.00	.94	.05	−.07	.09	.01	.01	.07	.02	.01	.04	(.88)

Note. Diagonal values are the internal consistency estimates for each scale.

a

Variable computed using z-scores.

*

p < .05 (two-tailed).

Discriminant Validity of the Constructs

A CFA was conducted to examine the distinctiveness of all 10 focal variables. The model fit for each of the nine nested models was compared, ranging from a single-factor model with a 10-factor model (Lance & Vandenberg, 2002; Rahim & Magner, 1995). Table 2 shows the results of these analyses. Specifically, the one-factor model includes all focal measures combined. Each subsequent model separates each measure out (i.e., voice, meeting time management, meeting relevance, meaningfulness, safety, availability, employee engagement, supervisor satisfaction, work satisfaction, and meeting load) one by one, until the 10-factor model, which separates each measure into a distinct factor, is presented.

OPEN IN VIEWER

Table 2.

Confirmatory Factor Analyses for All Focal Measures (N = 319).

Model	CFI	TLI	χ2	df	Difference	RMSEA
One-factor	.86	.86	19,934.04*	1,890		.17
Two-factor	.89	.88	13,418.57*	1,889	6,515.47*	.14
Three-factor	.90	.90	11,363.65*	1,887	2,054.92*	.13
Four-factor	.91	.91	10,614.23*	1,884	749.42*	.12
Five-factor	.92	.92	9,540.36*	1,880	1,073.87*	.11
Six-factor	.93	.93	8,206.86*	1,875	1,333.50*	.10
Seven-factor	.93	.93	7,702.52*	1,869	504.34*	.10
Eight-factor	.94	.94	6,400.38*	1,862	1,302.14*	.09
Nine-factor	.95	.95	5,311.63*	1,854	1,088.75*	.08
Ten-factor	.97	.97	3,910.31*	1,845	1,401.32*	.06

Note. The 1-factor model includes all focal measures combined. Each subsequent model separates each measure out, step-by-step, until the 10-factor model which separates each measure into distinct factors. CFI = comparative fit index; TLI = Tucker–Lewis index; Difference = difference in chi-square from the next model; RMSEA = root mean square error of approximation.

*

p < .05.

Considering several fit statistics, the 10-factor model showed the best overall fit. Although each more differentiated model showed a significantly better Chi-square statistic (James, Mulaik, & Brett, 1982) in comparison with the previous models, the 10-factor model showed a better root mean square error of approximation (RMSEA; Browne & Cudeck, 1993) and had both comparative fit index (CFI; Bentler, 1990) and Tucker–Lewis Index (TLI; Tucker & Lewis, 1973) values above their recommended cutoff of .90. All items in the 10-factor model loaded reliably on their predicted factors; the lowest loading was .35.

Proposed Model and Hypotheses

Figure 1 shows the standardized path coefficients estimated by LISREL 8.80 for the proposed full-mediation model. Two other models were tested as comparison points for assessing the efficacy of the proposed model in explaining the relationships hypothesized: a direct effects model; and a partial-mediation model (see Table 3).

OPEN IN VIEWER

Table 3.

Fit Indices and Standardized Path Coefficients for Theoretical Models (N = 319).

Measures	Direct model	Full-mediation model	Partial-mediation model
Fit indices
χ2	3,172.45	2,964.43	2,963.63
df	1,209	1,209	1,206
CFI	.96	.97	.96
TLI	.96	.97	.96
RMSEA	.07	.06	.07
Direct effects on engagement
Meeting relevance	.16*	—	.01
Voice	.02	—	−.04
Meeting time management	.33*	—	.05
Meaningfulness	—	.76*	.75*
Safety	—	.24*	.24*
Availability	—	.16*	.16*
Direct effects on meaningfulness
Meeting relevance	.23*	.23*	.22*
Voice	−.07	−.03	−.03
Meeting time management	.30*	.24*	.23*
Direct effects on safety
Meeting relevance	.00	.01	.00
Voice	.51*	.52*	.52*
Meeting time management	.20*	.18*	.17*
Direct effects on availability
Meeting relevance	−.14	−.14	−.14
Voice	.02	.04	.04
Meeting time management	.28*	.25*	.25*

Note. Difference = difference in chi-square from the next model; CFI = comparative fit index; TLI = Tucker–Lewis index; RMSEA = root mean square error of approximation.

*

p < .05.

Hypothesis 1a indicated that meeting relevance would positively relate to employees’ psychological meaningfulness. Consistent with this hypothesis, the path coefficient for the relationship between meeting relevance and meaningfulness was significant with the expected sign (β = .23, p < .05). However, meeting relevance was not significantly related to safety (Hypothesis 1b) or availability (Hypothesis 1c).

Hypotheses 2a, 2b, and 2c stated that voice would positively relate to employees’ psychological meaningfulness, safety, and availability, respectively. Consistent with Hypothesis 2b, voice significantly predicted psychological safety (β = .64, p < .05), but was unrelated to both meaningfulness (Hypothesis 2a) and availability (Hypothesis 2c).

Hypotheses 3a, 3b, and 3c suggested that meeting time management would be positively related to employees’ psychological meaningfulness, safety, and availability, respectively. Consistent with these hypotheses, the path coefficient for the relationship between meeting time management and meaningfulness, safety, and availability were significant and in the expected direction (β = .24, .19, .25, respectively, p < .05).

Hypotheses 4a, 4b, and 4c suggested that psychological meaningfulness, safety, and availability, respectively, would be positively related to overall employee engagement. Consistent with these hypotheses, psychological meaningfulness, safety, and availability had significant positive relations with engagement (β = .76, .24, and .16, respectively, p < .05).

Structural equation modeling with LISREL 8.80 was used to test the hypothesized model presented in Figure 1 as well as several additional models (see Table 4). The proposed full-mediation model showed good fit, χ²(1209) = 2964.43, p < .05; RMSEA = .06, TLI = .97, CFI = .97. Thus, no direct effects of meeting relevance, voice, or meeting time management on engagement were tested in this model. To test the meditational hypothesis, two processes were followed given current conventions concerning testing mediation hypotheses using SEM (Mackinnon, Coxe, & Baraldi, 2012). First, the steps described by Kenny, Kashy, and Bolger (1998) were followed and tested simultaneously using SEM (see Figure 1). Second, the indirect effects of the main predictors (i.e., meeting relevance, meeting time management, and voice) on the outcome (i.e., employee engagement) through the mediators (i.e., psychological conditions for engagement) were tested using bootstrapping methods developed by Preacher and Hayes (2008).

OPEN IN VIEWER

Table 4.

Mediation of the Effects of Manager Usage/Facilitation of Meetings on Employee Engagement Through the Psychological Conditions for Engagement (N = 319).

				Bootstrapping
		Product of coefficients		Percentile 95% CI		BC 95% CI		BCa 95% CI
	β	SE	Z	Lower	Upper	Lower	Upper	Lower	Upper
MR → M → EE	.15*	.026	5.76	.093	.216	.095	.221	.096	.224
V → S → EE	.13*	.024	5.47	.081	.183	.082	.184	.080	.181
TM → M → EE	.16*	.029	5.56	.089	.242	.093	.248	.097	.253
TM → S → EE	.05*	.014	3.79	.030	.084	.031	.085	.031	.086
TM → A → EE	.01	.006	1.62	−.002	.029	−.001	.030	−.002	.029

Note. V = voice; TM = meeting time management; MR = meeting relevance; S = safety; M = meaningfulness; A = availability; EE = employee engagement; CI = confidence interval; BC = bias corrected; BCa = bias corrected and accelerated; 5,000 bootstrap samples.

*

p < .05.

First, for Hypothesis 5a, regarding psychological meaningfulness as a potential mediator, meeting relevance and meeting time management were related to psychological meaningfulness and psychological meaningfulness was related to engagement. For Hypothesis 5b, regarding psychological safety as a mediator, voice and meeting time management were related to psychological safety and psychological safety was related to engagement. For Hypothesis 5c, regarding psychological availability as a mediator, meeting time management was related to psychological availability and psychological availability was related to engagement. Figure 1 provides an illustration of the proposed model with these significant paths shown.

Second, using 5,000 bootstrap samples, indirect effects estimates were computed along with 95% confidence intervals around the estimates. Results of these analyses are reported in Table 4. All the indirect effects were significant (p < .05) except the effect of meeting time management on employee engagement through psychological availability.

As an alternative test of the proposed model, a partial-mediation model was tested in which a direct path from each of the meetings variables to engagement was added. None of these direct paths were statistically significant. In addition, the approximate fit indices (i.e., CFI, TLI, and RMSEA) for this partial-mediation model were essentially the same as the full-mediation model. Furthermore, as the partial-mediation model is nested within the full-mediation model, the chi-square difference test is an appropriate statistic for comparing these two models. Interestingly, the test showed a nonsignificant reduction in the chi-square statistic, χ²(3) difference = .80, p > .05, suggesting that the partial-mediation model does not represent the data better than the full-mediation model. All these results, taken together, provide support for Hypotheses 5a and 5b, with partial support for Hypothesis 5c.

Finally, in an effort to verify that the three meeting variables were not just proxy variables for overall satisfaction with the supervisor, satisfaction with work in general, or a function of meeting load (which was shown to be related to employee well-being; Luong & Rogelberg, 2005), a revised model was tested controlling for these factors. The revised model allows supervisor satisfaction, work satisfaction, and meeting load to predict each of the psychological conditions for engagement and overall engagement. Although all the models appear to have adequate fit, the contribution here is the fact that all but one of the paths (i.e., the path from meeting time management to psychological safety) from the original proposed model (see Figure 1) remained significant in the revised model. Overall, these analyses suggest that employees’ satisfaction with their supervisor and work are not confounding factors. In addition, it suggests that the observed relationships exist across various levels of meeting load.

Psychological Meaningfulness and Engagement Through Meetings

First, meeting relevance and meeting time management related to psychological meaningfulness, whereas voice did not. This finding suggests that managers who attempt to make their meetings more relevant (i.e., focused on accomplishing organizational and employee goals; Reinig, 2002) to attendees as well as manage their meeting time effectively (e.g., start/end on time) may help employees feel more meaning in their work, thereby promoting engagement. In terms of meeting relevance, managers who make their meetings relevant to employees may be demonstrating a level of respect toward their employees, given the employees’ other obligations and goals. Meetings that are not relevant to an employee are likely viewed as a waste of time, energy, and effort by the employee. In fact, recent practical recommendations to managers suggest that providing an “opt-out” clause for meetings for employees that would allow them to excuse themselves from a meeting may be ideal (Rogelberg et al., 2007). Furthermore, in terms of meeting time management, these relevant meetings that start/end on time and use employee time effectively show care and respect for employees (Kahn, 1990). This is consistent with Kahn (1990), who suggests that job tasks that involve a “clear delineation of procedures and goals” are likely to positively influence the development of psychological meaningfulness (p. 705). Surprisingly, voice was not related to psychological meaningfulness. One possible explanation is that voice behaviors need to be validated by the group (i.e., acknowledged and supported) to experience feelings of value and worth (Detert & Burris, 2007).

Psychological Safety and Engagement Through Meetings

With regard to psychological safety, meeting time management and voice were significant correlates; however, meeting relevance was unrelated. These relationships suggest that employees appear to feel more psychologically safe when managers schedule meetings at appropriate times and start/end meetings according to schedule. As previously stated, meeting time management behaviors ensure predictability in the work environment (Kahn, 1990). Predictability from a time perspective allows employees to schedule their other work commitments around the meetings in a way they personally find effective and helps them feel safe to schedule other activities throughout the day. In terms of voice, it stands to reason that when employees feel safe to share ideas in meetings without personal attacks upon their self-image, they would also feel psychologically safe in other areas of their work. Thus, the meeting may become a location to promote feelings of safety that subsequently permeate the work environment, allowing employees to contribute their whole self and engage more broadly in their work efforts.

The lack of findings for meeting relevance and psychological safety may be due to the nature of the actual meeting purpose. Given the diversity of meeting purposes, it is not hard to imagine a meeting that is both relevant to employees and experienced as psychologically unsafe. For example, meetings concerning layoff decisions are relevant to employees affected by the layoffs. However, employees would probably not leave that meeting with increased feelings of safety to fully engage in their work.

Psychological Availability and Engagement Through Meetings

Finally, with regard to psychological availability, only meeting time management was a significant correlate. The other two manager usage/facilitation of meetings variables were unrelated to psychological availability. It appears that managing meetings effectively from a time perspective may help provide the resources employees need to engage in their work. Employees often view meetings as interruptions (Rogelberg et al., 2006) and meetings are always effortful events that require cognitive resources (Allen et al., 2012). Thus, anything a manager can do to reduce the resource impact of a given meeting is ideal.

It should be noticed that the mediated framework suggesting that meeting time management relates to engagement through psychological availability was not statistically significant (see Table 4). In looking at the results, clearly psychological meaningfulness is absorbing much of the relationship between the psychological conditions for engagement and overall engagement. The complexity of the model and the presence of meaningfulness in the model may actually be obscuring, to some extent, the strength/importance of psychological availability. Furthermore, the path coefficient between meeting time management and availability was the strongest of the three paths between meeting time management and the psychological conditions. Thus, the time management aspect is clearly important and should not be overlooked simply because the availability to engagement relationship is less robust. In sum, the findings suggest that how managers use/facilitate their group meetings promotes the psychological conditions in varied ways necessary for an engaged workforce.

Empirical Contributions

This study contributes to the literature concerning the various antecedents of employee engagement. Previous research tended to focus on more global job characteristics (e.g., autonomy, training, technology; Salanova et al., 2005) as well as employee attitudes about their job (e.g., perceived organizational support and procedural justice; Saks, 2006) as antecedents to engagement. In contrast, this study narrows the focus to a particular context, the meeting, and the types of supportive behaviors and processes that managers can follow to promote engagement. This study demonstrates that managers may be able to promote engagement by simply running their workgroup meetings more effectively in terms of allowing open communication, starting/ending on time, and calling relevant meetings. These relationships were shown to remain even after controlling for previously tested attitudinal antecedents to employee engagement, supervisor satisfaction and work satisfaction (May et al., 2004). In addition, these relationships remained after controlling for meeting load, suggesting that employees with few or many meetings are still affected by the way managers facilitate the meeting tool.

The current study contributes to the growing body of literature on workgroup meetings. The call to study meetings as an important social phenomenon was relatively recent (Rogelberg et al., 2006). As such, the literature base on workgroup meetings is nascent and many areas of inquiry exist for research. One area that seemed lacking in the meetings literature was a connection between workplace meetings and employee performance. Although recent research is beginning to bridge this gap (Rogelberg, Allen, Shanock, Scott, & Shuffler, 2010), no study has focused on how workgroup meetings affect employee and organizational performance. By showing a relationship between how managers use/facilitate meetings and employee engagement, this study connects the design and execution of meetings to important employee outcomes. Because employee engagement is an important predictor of employee performance (Crawford et al., 2010), showing that meetings can promote an engaged workforce illustrates their potential importance for achieving competitive advantage through improved performance. Therefore, this study adds to the legitimacy of researchers’ and practitioners’ growing focus on studying and improving workgroup meetings within organizations.

Practical Implications

The current findings illustrate a general need to maximize the quality and effectiveness of organizational meetings. To do so, a learning, feedback, and accountability approach will likely be needed (Rogelberg, Shanock, & Scott, 2012). This approach starts with the teaching and development of meeting skills in managers. This approach is not only relevant for current managers, but should be an important piece of the onboarding process for new leaders. Next, feedback and accountability systems for managers that target meetings should be developed. For example, a 360-degree appraisal system or employee survey could easily include a section evaluating managers’ performance in workgroup meetings. Surveys could also be used as a focused initiative looking at a series of meetings by a single manager. For example, employees who attend meetings with this manager would provide assessments over a given period of time (e.g., a week or month) for each meeting they attend. This survey approach would allow a more focused assessment of the manager’s meeting skills, highlight areas for improvement, and illustrate skills that organizational leaders may wish to propagate among managers in their organization. Ultimately, making leaders aware of the importance of their meeting activity, providing feedback on set activity, and creating a development plan to leverage strengths and mitigate weakness, when compounded across leaders, can serve to further strengthen employee engagement initiatives across the organization.

In relation to this, managers should consider specific ways in which they can make their workgroup meetings relevant to meeting attendees. In this study, meeting relevance was assessed as the perception of employees that their meetings are relevant. As such, the study does not identify key tasks that managers can perform to make their meetings relevant to employees, or how relevance for one participant may result in irrelevance for another participant. However, a manager might consider taking a goal setting approach to meetings. A meeting goal is “any need or want that an individual makes a conscious effort to fulfill” within the meeting context (Reinig, 2002, p. 2). When managers communicate the link between employee work goals and the goals of the organization and deliberately connect the meeting purposes to these shared goals, the meetings may take on more meaning for employees.

Limitations and Future Directions

Though this study is an important first step within the engagement and workgroup meetings literature, several limitations exist as well as opportunities for future research. An obvious methodological limitation of this study is the use of correlational analysis and the resulting inability to draw causal conclusions, despite the fact that a time-lag assessment of engagement was introduced. This limitation is particularly salient with the direction of causality between voice, the psychological conditions, and engagement. One could argue that voice may be a manifestation of employees’ engagement in their work as opposed to something that causes them to feel engaged. When employees incorporate their whole self in their work role, they may see relationships between work tasks and their other life experiences not otherwise acknowledged, and then feel encouraged to express those relationships. Thus, engagement may create a desire to voice opinions and ideas. Though theory supports the current causal inferences, future research can address this limitation by using experimental designs. For example, one could vary the levels of some of the meetings variables (e.g., meeting time management) across meeting contexts and assess the degree to which individuals leave the meeting experiencing more or less of the psychological conditions for engagement. This manipulation would allow for a clearer indication of the degree to which certain strategic meeting behaviors are important to the development of employee engagement.

Another limitation related to the sample is the possibility of range restriction on the education level variable, which may reduce generalizability. According to the U.S. Census Bureau (2000), only 24.4% of the population has completed a bachelor’s degree (or 4-year equivalent). As 97% of the study sample were college graduates (i.e., 4-year degree or more education), it does not accurately reflect the variance in education level within the workforce. This artifact of the data is a direct result of the sampling frame (i.e., university alumni), thus, future research can benefit from targeting a more diverse sample. Doing so will allow for greater generalizability to other areas of the workforce.

Another limitation of this study is the deliberately narrow content coverage. This study focused on meetings as a location for promoting engagement in the workplace. As such, the variables measured focused almost exclusively on characteristics of the meeting and the behaviors of managers associated with their workgroup meetings. However, previous research showed other job attitudes (e.g., organizational support, supervisor support, and procedural justice) are related to overall employee engagement (Crawford et al., 2010; Saks, 2006). Although the alternative model incorporates two such critical and theoretically the most relevant antecedents, given the topic area of meetings (satisfaction with one’s supervisor and work satisfaction), future research should consider modeling both the meeting variables and other job attitudes concurrently. This combined model will allow for a more nuanced understanding of the contribution of effectively run meetings to overall employee engagement.

Another potential future direction stems from the fact that the current study focused on global assessments of employees’ experiences in their workgroup meetings, rather than focusing on the nature of any single meeting experience. Although this focus is a necessary first step in understanding the importance of meetings generally, future research could begin to look at specific meetings that supervisors lead and how they affect employees from an engagement perspective. One way to do this research would be to perform a diary study. Employees would provide ratings of various meeting characteristics after each meeting over a given period of time. They would also provide assessments of their level of the psychological conditions for engagement and overall engagement on a meeting-by-meeting basis. This within-subjects multilevel (i.e., events nested within individuals) design would allow for an understanding of how meeting characteristics and process affect individuals, and also how individual characteristics may affect the evaluation of the meetings.