BUS603: Managing People

Dispositional Mindfulness

Leaders' dispositional mindfulness was measured with the short-version of the Freiburg Mindfulness Inventory. The scale consists of 14 items assessing the frequency of mindful states. A sample item is "I am open to the experience of the present moment". The items were answered on a 6-point frequency scale (ranging from 1 = never to 6 = almost always). Cronbach's alpha was 0.90.

Followers' Satisfaction With Leader–Follower Communication

In order to measure followers' satisfaction with leader–follower communication, we used two items from the questionnaire for communication in organizations developed by Sperka. The two items were "I am content about how the communication with my leader takes place" and "I would like to have a better communication with my leader" (reverse coded). Each follower was asked to rate their own level of satisfaction. Again, a 6-point response scale (ranging from 1 = strongly disagree to 6 = strongly agree) was employed. The reliability of this measure was estimated by using the Spearman–Brown formula, and was 0.66.

Followers' General Satisfaction With Their Leaders

Followers' general satisfaction with their leaders was measured with two items taken from Felfe. The two items were: "My leader uses methods of leadership that are satisfying" and "My leaders works with me in a satisfactory way". Responses were given on a 6-point scale, ranging from 1 = strongly disagree to 6 = strongly agree. The reliability of this measure was again estimated by using the Spearman–Brown formula, and was r = 0.84.

Mindfulness in Communication

Since there was no existing scale for what we call mindfulness in communication, we developed a new scale for this study. The purpose of the scale was to assess "behavioral correlates" of leaders' mindfulness when communicating with followers. Followers were explicitly asked to rate their leaders' behavior in communication situations. Below, we describe in more detail how the measure was constructed.

Item development and exploratory factor analysis

We first generated 14 items based on a review of the literature addressing mindfulness in leadership. The items addressed the following three facets of mindfulness in communication:  (1) being present and paying attention in conversations, (2) showing an open, non-judging attitude during a conversation, and (3) being calm and non-impulsive during conversations, not becoming overwhelmed by emotional reactions. Second, the content validity was assessed by asking four experts (i.e., experts on mindfulness practice) to rate the items in terms of their conceptual fit. As a result of the expert rating, four items were omitted. The remaining 10 items were included in the questionnaire described above and were answered in total by 204 followers (including followers that could be matched with a leader and followers that could not be matched with a leader and were therefore not considered in the main analyses). For the analyses of the scale and the items, all followers (N = 204) were included. One item showed a low level of communality (communality = 0.17) and was therefore excluded. With the remaining nine items, an Exploratory Factor Analysis using a principal–axis analysis with Promax rotation was performed. The results suggested one factor with an Eigenvalue > 1 explaining 59% of the variance. The factor loadings, communalities, and standardized item-scale-correlations were satisfactory (Table 1). Therefore, the final scale consists of nine items and shows adequate reliability (α = 0.91) which could not be improved by deleting further items.

TABLE 1

Table 1. Items of the scale "mindfulness in communication" including their factor loadings, communalities, and corrected item-scale-correlations.

Confirmatory factor analysis

Because a new measure of mindfulness in communication was created for this study, we collected data from a separate sample to confirm the construct validity using confirmatory factor analysis (CFA). Specifically, 214 employees from various organizations in Germany completed the newly developed measure. The mean age of the participants was 33.36 (SD = 8.89); 47% were male and 65% had a university degree. The majority of the participants worked in the for-profit sector (69%) and the average tenure in the current position was 3.92 years (SD = 3.25).

We conducted a CFA using the Lavaan package in R and compared the fit of two nested models. The first one was a single-factor model with all nine items loading on the same factor. The second one was a second-order factor model in which items loaded on their respective factors (i.e., presence, openness, and calmness) and the three factors loaded on a second-order latent mindfulness in communication factor. This second-order factor model showed a reasonable fit with χ² = 58.65, df = 24, p < 0.001, comparative fit index (CFI) = 0.96, root mean square error of approximation (RMSEA) = 0.08, SRMR = 0.04 and was clearly preferable over the single-factor model (χ² = 219.82, df = 27, p < 0.001, CFI = 0.80, RMSEA = 0.18, SRMR = 0.85; Δχ² = 161.17, df = 3, p < 0.001, ΔCFI = 0.16). It should be noted that with three latent factors, the second-order model is mathematically equivalent to a first-order model (i.e., a model in which items load on their respective factors and the factors are allowed to correlate) and thus, both solutions produce identical fit statistics. Yet, since we assumed that a common latent mindfulness in communication factor accounts for the relation between the three subscales (i.e., presence, openness, and calmness), the second-order model represents a more parsimonious and meaningful approach. This supports the use of the combined mindfulness in communication scale (comprising the three sub-facets) in the analysis presented below.

Discriminant validity

To examine the discriminant validity among the follower-related outcome measures that we used in our main study (i.e., perceived mindfulness in communication, satisfaction with the communication with the leader, and general satisfaction with the leader), we again used CFA. For mindfulness in communication, we used parcels as indicators (i.e., the three sub-dimensions, which is in line with the content-based algorithm of parcel building. For the two satisfaction constructs, items were used as indicators. Table 2 reports the models we tested. To compare the fit for different models, we used the chi-square difference test. However, given that chi-square tests are very sensitive to sample size and non-normality, even small differences may become statistically significant. Thus, in line with the recommendations of Chen, we also relied on the change in CFI and RMSEA. Specifically, for small samples (i.e., N < 300) a change of 0.005 in CFI, supplemented by a change of 0.010 in RMSEA indicates that the models are significantly different. As shown in Table 2, the proposed three-factor model (Model 1) fitted the data reasonably well and was preferable over all alternative models (Models 2, 3, and 4). Taken together, these results provide evidence that our follower reported measures captured distinct constructs.

TABLE 2

Table 2. Test of measurement models.