Understanding Followers’ Critical Thinking Behaviors: A GLOB(E)al Perspective of Mid-Level Managers

Abstract

The main purpose of this exploratory study was to examine the association between mid-level leaders’ perceptions of their followers’ critical thinking skills and leaders’ geographic location. Studies such as the GLOBE study provide the foundation for recognizing variation in perspectives on leadership behaviors. Yet there is a lack of insight into how the effect of global location on followers’ perceived effectiveness varies across areas like critical thinking skills. Using a survey to collect data, mid-level leaders at a global oil and gas company participating in a leadership development program were asked to evaluate their followers against the QUEEN Model of Critical Thinking. Analyses were conducted at both the complete QUEEN model score level and within each critical thinking skill category, using one-way ANOVA and post hoc testing. Significant differences were found in relation to where mid-level leaders were based, as defined by the GLOBE clusters. This suggests a need to provide mid-level leader training tailored to global locations on understanding critical thinking and developing this skill in followers.

1. Introduction

As Riggio et al. (2023) postulated, followers are often invisible in organizational research. Another forgotten population is mid-level leaders (Spieler, 2024; Williams et al., 2010). Both positions are often seen as simple cogs in a perpetual-motion wheel, but these two populations contribute significantly to organizational success. To add to this cacophony, followership research has not yet focused on the impact of international organizations on effective followership. In (Tolstikov-Mast, 2016) noted, “Global followership requires new ways of thinking, theorizing, engaging in diverse epistemological and ontological approaches to formulate empirically supported global followership framework” (p. 144). Collinson (2008) stated, “The multiple identities, values and cultures of followers (and leaders) in various diverse regions, societies and continents are likely to have a significant impact on the possibilities and limits of leadership practices” (p. 251). This is reiterated by Thomas (2014), who found significant differences in followership types between Rwandan and American samples, but no significant differences in the factors of independent critical thinking and active engagement. Malakyan (2023), in a comprehensive study of 65 leadership and followership studies across the global North, West, South, and East, found that fundamental differences exist in cultural understandings and practices of leadership and followership between these regions. Too often, organizations take a follower with a proven track record and transplant that person into a global arena, expecting the same level of performance abroad as they had at home. Unfortunately, the skills needed to be a successful follower and leader in one’s country of origin are not necessarily the same as those needed to be successful abroad (Javidan et al., 2010). Exploratory research is needed to fill the gaps in the literature.

As the followership sub-discipline continues to expand its research, scholars must seek empirical connections and develop behaviorally anchored followership competencies (Uhl-Bien et al., 2009). Critical thinking is a standard followership competency that cuts across most major followership models (Linville & Rennaker, 2024). While Kelley (1992) explicitly uses that term, other models allude to higher-order thinking skills to describe followership effectiveness. Critical thinking has been identified as an essential skill for success, yet many global employers find it lacking in their followers (Dada et al., 2023). Defining the concept of critical thinking behaviors is also multifaceted. Manalo et al. (2015) found that college students had difficulty defining the term, and this cut across all cultures surveyed.

The purpose of this study was to explore mid-level leaders’ perceptions of their followers’ critical thinking behaviors in an international (global) oil and gas company. The research question guiding this study was as follows: Is there a statistically significant difference in mid-level leaders’ perceptions of their followers’ critical thinking skills across leaders from different parts of the world? Based on the findings of this study, it can be concluded that perceptions of followers’ critical thinking skills varied significantly across geographic regions, as defined by the GLOBE clusters, for this population.

2. Theoretical Framework

The swift pace of globalization and internationalization of organizations led to the focus on leader development within international contexts (Hocagil, 2024). The most renowned model of leadership within a global context is the GLOBE study. House et al.’s (2004) Global Leadership and Organizational Behavior Effectiveness model is a cultural classification system consisting of nine dimensions: (1) uncertainty avoidance, (2) power distance, (3) collectivism–institutional, (4) collectivism–group, (5) gender egalitarianism, (6) assertiveness, (7) future orientation, (8) performance orientation, and (9) humane orientation. House et al. (2004) also divided the globe into regions to correlate with these dimensions and describe the culture. The regions are Eastern Europe, Nordic Europe, Middle East, Confucian Asia, Southern Asia, Latin America, Anglo, Germanic Europe, Latin Europe, and sub-Saharan Africa. These researchers also identified six global leadership behaviors and correlated them to the regions that showed this type of leadership. These global leadership behaviors are charismatic/values-based leadership, team-oriented leadership, participative leadership, humane-oriented leadership, autonomous leadership, and self-protective leadership (Hackman & Johnson, 2009). The GLOBE study has yielded substantial research on effective leadership and culture, yet Linville and Rennaker (2024) note that no research has examined the GLOBE study in relation to followership. Hocagil (2024) also observed the need for continued development of followership competencies within a global organizational context.

Critical thinking has been identified as a key effective behavior for followers. Kelley’s (1992) questionnaire measures independent/critical thinking. It is descriptive and situationally based, making it difficult to improve one’s score. Kelly is not alone in this plight. Historically, measures of critical thinking have focused on dispositions or propensities rather than on actual critical thinking behaviors and skills (Lamm et al., 2011; Possin, 2008). The QUEEN Critical Thinking Model was developed to fill this gap and to help practitioners, leaders, followers, and those seeking to increase their critical-thinking behaviors by focusing on behaviorally anchored critical-thinking competencies. The QUEEN Critical Thinking Model used a grounded theory approach and a modified Delphi technique (Patton, 2015) to develop the five behaviorally anchored constructs: (1) Questioning, (2) Understanding, (3) Evaluating, (4) Explaining, and (5) Neoteric Thinking (Boyd & Strong, 2025). The Questioning construct seeks relevant and accurate information about a situation while asking inquisitive questions about the “why” and “what” of an issue. The Understanding construct describes behaviors that deepen understanding of others’ backgrounds, purposes, and assumptions, prompting individuals to seek alternative perspectives to form a holistic picture. For the QUEEN Critical Thinking Model, evaluating means judging the relevance and credibility of sources. This construct also examines the potential consequences of actions, examines assumptions, and distinguishes between fact and value claims. Explaining is not only discussing the what of an issue but also the why of how you came to view a situation in that way. Using examples to connect main points and explaining the reasoning process with theoretical models or frameworks increases this critical-thinking behavior. Neoteric thinking involves behaviors that promote the emergence of new concepts and the synthesis of new solutions or ideas (Boyd & Strong, 2025).

Organizational governance conditions, such as accountability structures, decision rights, transparency mechanisms, and norms for participation, play a central role in shaping whether followers’ critical-thinking behaviors are recognized, evaluated, and developed. Governance systems that emphasize procedural justice, voice, and psychological safety signal that questioning assumptions and engaging in reasoned dissent are legitimate and valued behaviors, increasing the likelihood that such behaviors will be noticed and positively appraised rather than sanctioned (Detert & Burris, 2007; Edmondson, 1999). In contrast, highly centralized or compliance-oriented governance frameworks tend to privilege adherence to rules and risk avoidance, which can suppress evaluative judgment and limit opportunities for followers to practice and refine critical thinking skills (Argyris & Schön, 1978; Yukl, 2013). Moreover, governance arrangements influence developmental pathways by determining how performance is assessed and rewarded; when evaluation systems explicitly incorporate learning, reflection, and evidence-based reasoning, organizations create institutional incentives for leaders to coach and cultivate critical thinking among followers (Hannah et al., 2008). Thus, governance conditions function not merely as structural constraints but as interpretive frames that define what counts as competent thinking and whether followers’ critical engagement is developed as an organizational capability rather than treated as deviance. Given the complex cacophony of organizational culture and dynamics, and the emphasis on exploring this phenomenon, this study focused on followers’ critical-thinking behaviors as perceived by their immediate supervisor (leader).

3. Materials and Methods

Mid-level leaders at a global oil and gas company, selected to participate in a leadership development program, were asked about their experiences leading and following in a global organization. During this program, conversations began about the phenomenon of followership, which most had not heard of. It was quickly noted that there was a need to explore followership further in this context. To better understand the complexities of followership within an international organization, a Qualtrics (2025) survey was created to study this organizational phenomenon. The survey included the QUEEN Model of Critical Thinking (Boyd & Strong, 2025) and three open-ended questions on followership. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Institutional Review Board of Texas A&M University (2025-1195) on 7 January 2025. This paper focuses on quantitative data collected via the QUEEN Model of Critical Thinking.

The sample for this study initially included 101 mid-level leaders who work for an international oil and gas company. After eliminating incomplete survey responses, 88 participants were included in the analysis. Groups were assigned based on the reported leader location using the GLOBE clusters (Gupta et al., 2002). Self-reported locations indicated that 22.7% of participants were from Latin America, 40.9% from Anglo, 20.5% from the Middle East, 1.1% from Confucian Asia, 4.5% from Nordic Europe, 5.7% from Sub-Saharan Africa, 2.3% from Southern Asia, and 2.3% from Eastern Europe. Responses were bootstrapped, and ANOVA results were compared; no statistically significant variance was found.

A modified Delphi technique was utilized to develop the QUEEN Critical Thinking Model (Patton, 2015). Twenty experts in critical thinking were surveyed and asked to list specific behaviors identified by critical thinkers. After the third round of the Delphi, 23 behaviors were identified and categorized into five constructs: Questioning, Understanding, Evaluating, Explaining, and Neoteric Thinking. The QUEEN scale is: Never (1), Rarely (2), Sometimes (3), Often (4), and Always (5). Reliability coefficients were computed for each construct and for the entire instrument (Bryman, 2012). Cronbach’s Alpha resulted in high internal reliability for all constructs (Questioning, 0.83; Understanding, 0.78; Evaluating, 0.86; Explaining, 0.82; Neoteric Thinking, 0.82; and Total, 0.921). Followership researchers stationed across the GLOBE clusters provided construct reliability (Bryman, 2012).

The following analyses were used to determine whether there was a statistically significant difference in leaders’ perceptions of their followers’ critical thinking skills based on follower summated observations. Data analysis for all descriptive statistics, hypothesis testing, and statistical measures was conducted utilizing RStudio (2025.05, Mariposa Orchid).

For the research question, data were analyzed in two manners—by each critical thinking skill category and as a complete QUEEN model score. The goal of this analysis was to explore the nuances within each critical thinking skill category with greater depth. Testing each null hypothesis began with the statistical program, which generated descriptive statistics, including the mean, standard deviation, skewness, and kurtosis. The null hypothesis was then tested using a one-way ANOVA. The one-way ANOVA was used to compare the means of the independent groups to assess the impact of a single factor on our critical thinking score by analyzing variance within and between groups (Kabacoff, 2015). If differences occurred, then a post hoc test was required. The one-way ANOVA has multiple assumptions. First, each group had participants categorized into only one group per statistical test, thereby meeting the assumption of independence of observations. Normality was assessed using the Shapiro–Wilk test. The Assumption of Homogeneity of Variance was tested using Levene’s test of homogeneity. If violations of assumptions occurred, the Kruskal–Wallis test was used as an alternative to ANOVA to ensure accurate representation. The effect size was determined using Eta Squared (η²). The confidence level was set at 95%.

4. Results

Initial data analysis and statistical testing were conducted to address the research question: Is there a statistically significant difference in mid-level leaders’ perceptions of their followers’ critical thinking skills across leaders from different parts of the world? The null hypothesis states that there is no statistically significant difference in perceptions of follower critical thinking skills across leadership locations. Table 1 displays descriptive statistics for the average ratings of each of the five components of the QUEEN model and the total average for critical thinking across the GLOBE clusters. In total, 88 participants were included in the descriptive statistics analysis. When looking at the total QUEEN average score means, Confucian Asia rated their followers highest with a mean score of 3.62, while the lowest mean score of 2.96 came from Sub-Saharan Africa.

Table 1. Descriptive Statistics for QUEEN Components and Total Ratings by GLOBE Clusters.

Cluster Group	n	M	SD	Sk	Ku
Questioning
Latin America	20	3.68	0.59	0.05	−1.11
Anglo	36	3.5	0.36	0.25	−0.11
Middle East	18	3.54	0.39	−0.41	−0.19
Confucian Asia	1	3.6	NA	NA	NA
Nordic Europe	4	3.25	0.3	−0.75	−1.69
Sub-Saharan Africa	5	3.48	0.64	0.07	−1.45
Southern Asia	2	3.3	0.14	0	−2.75
Eastern Europe	2	3.6	0.28	0	−2.75
Understanding
Latin America	20	3.61	0.63	−0.78	−0.71
Anglo	36	3.19	0.43	−0.09	−0.82
Middle East	18	3.56	0.43	−0.25	0.93
Confucian Asia	1	3.25	NA	NA	NA
Nordic Europe	4	2.81	0.31	0.42	−1.82
Sub-Saharan Africa	5	2.8	0.74	0.69	−1.47
Southern Asia	2	3.25	0.35	0	−2.75
Eastern Europe	2	3.12	0.18	0	−2.75
Evaluating
Latin America	20	3.58	0.58	−0.17	−0.57
Anglo	36	3.37	0.31	−0.53	−0.74
Middle East	18	3.57	−0.34	0.37	−0.1
Confucian Asia	1	3.67	NA	NA	NA
Nordic Europe	4	3.12	0.55	−0.58	−1.82
Sub-Saharan Africa	5	3.3	0.48	0.16	−1.45
Southern Asia	2	3.33	0	NA	NA
Eastern Europe	2	3	0.24	0	−2.75
Explaining
Latin America	20	3.5	0.64	0.25	−0.79
Anglo	36	3.37	0.4	0.46	−0.7
Middle East	18	3.43	0.42	0.05	−0.44
Confucian Asia	1	3.8	NA	NA	NA
Nordic Europe	4	3	0.43	0.45	−1.87
Sub-Saharan Africa	5	2.84	0.71	0.66	−1.43
Southern Asia	2	2.9	0.14	0	−2.75
Eastern Europe	2	3.3	0.14	0	−2.75
Neoteric Thinking
Latin America	20	3.54	0.56	−0.25	−1.21
Anglo	36	3.17	0.42	0.5	−0.13
Middle East	18	3.29	0.49	0.48	−0.55
Confucian Asia	1	3.8	NA	NA	NA
Nordic Europe	4	3.35	0.34	0.28	−1.96
Sub-Saharan Africa	5	2.36	0.62	0.74	−1.29
Southern Asia	2	2.9	0.14	0	−2.75
Eastern Europe	2	3.4	0	NA	NA
Total QUEEN
Latin America	20	3.58	0.48	−0.2	−1.18
Anglo	36	3.52	0.27	−0.06	−0.41
Middle East	18	3.48	0.29	0.68	0.01
Confucian Asia	1	3.62	NA	NA	NA
Nordic Europe	4	3.11	0.27	0.66	−1.75
Sub-Saharan Africa	5	2.96	0.58	0.88	−1.12
Southern Asia	2	3.14	0.1	0	−2.75
Eastern Europe	2	3.29	0.07	0.1	0

n = 88.

A one-way analysis of variance (ANOVA) was used to test the research questions and null hypotheses, examining differences in QUEEN model scoring across global location. A one-way ANOVA required that assumptions of independence, normality, and homogeneity of variance were met. The normality of the assumption was verified because the number of participants exceeded the minimum of 30, which satisfies the Central Limit Theorem (Le Cam, 1986). However, there could be expected error due to uneven group amounts, and with three group populations that are too small to run the Shapiro–Wilk formal test for normality. The skewness and kurtosis remain within −3 to 3, so normality can be assumed. The assumption of homogeneity was tested using Levene’s Test. The assumption of homogeneity was met in all tests, except for the total average QUEEN rating for participants, which yielded F (7, 80) = 2.89, p = 0.01. For this statistic, Welch’s Alternative Test was then utilized to ensure the reliability of the ANOVA analysis. See Table 2 for Levene’s Test of Equality of Error Variances.

Table 2. Levene’s Test of Equality of Error Variances in relation to GLOBE Clusters.

Group	Df1	Df2	F	p
Questioning	7	80	1.3	0.26
Understanding	7	80	1.27	0.28
Evaluating	7	80	2.1	0.053
Explaining	7	80	1.96	0.07
Neoteric Thinking	7	80	1.65	0.13
Total QUEEN	7	80	2.89	0.01 *

* violation of Levene’s with p < 0.05.

An ANOVA analysis was conducted for each of the QUEEN components and the total QUEEN individual score to test the null hypothesis that there would be no statistically significant difference between the mid-level leader’s location on the perception of followers’ critical thinking skills. Table 3 displays the results of the ANOVA analyses conducted. The resulting p-values showed significant differences for the component of understanding [F (7, 80) = 2.9, p = 0.01], neoteric thinking [F (7, 80) = 4.11, p = 0.0007], and the total QUEEN score [F (7, 80) = 2.9, N = 87, p = 0.01] using a predetermined Type 1 error rate of 0.05.

Table 3. ANOVA analyses in relation to GLOBE Clusters.

Group	Df	SS	MS	F	p
Questioning
Between Groups	7	0.29	0.13	0.55	0.8
Within Groups	80	19.35	0.24
Understanding
Between Groups	7	5.75	0.82	2.9	0.01 **
Within Groups	80	22.65	0.28
Evaluating
Between Groups	7	1.89	0.27	1.58	0.15
Within Groups	80	13.63	0.17
Explaining
Between Groups	7	2.98	0.43	1.81	0.10
Within Groups	80	18.89	0.24
Neoteric Thinking
Between Groups	7	6.52	0.93	4.11	0.0007 **
Within Groups	80	18.12	0.23
Total QUEEN
Between Groups	7	2.53	0.36	2.87	0.01 **
Within Groups	80	10.07	0.13

** indicates statistical significance with p < 0.05.

To further explore the three values with statistical significance following the ANOVA test, effect size and a Tukey HSD test were completed. The critical thinking component of Understanding yielded an effect size, η², of 0.2 at the 95% confidence level. This can be interpreted to mean that approximately 20% of the variation in Understanding average scores is attributable to the leader’s location. When a Tukey HSD was conducted to identify significant pairings, no pairings showed statistically significant differences. This could be explained by having unequal group sizes.

When analyzing the component of neoteric thinking, an effect size, η², of 0.26 with a 95% confidence level was produced, suggesting that 26% of the variation in Neoteric Thinking average scores can be attributed to the location of the mid-level leader. To further explore which groups differ from one another by a statistically significant amount, a Tukey HSD test was conducted on all pairwise comparisons. The mean differences and confidence intervals for the statistically significant results are shown in Table 4. The results indicate that Sub-Saharan Africa differs significantly from Latin America, Anglo, and Middle East in terms of average Neoteric Thinking rating for followers.

Table 4. Statistically Significant Pairwise Group Comparisons of Neoteric Thinking by Location.

Comparison	Raw Score Mean Difference	Raw Score 95% Tukey CI	p
Sub-Saharan Africa—Latin America	−1.18	−1.92 to −0.44	0.0001
Sub-Saharan Africa—Anglo	−0.81	−1.52 to −0.11	0.01
Sub-Saharan Africa—Middle East	−0.93	−1.68 to −0.18	0.01

The total QUEEN model average effect size, η², was 0.2 at the 95% confidence level. This suggests that 20% of the variation in average total QUEEN scores is attributable to the leader’s location. When using a Tukey HSD test to identify statistically significant pairs, only one pair was identified. This was between Sub-Saharan Africa and Latin America. Table 5 shows this test.

Table 5. Statistically Significant Pairwise Group Comparisons of Total QUEEN Score by Location.

Comparison	Raw Score Mean Difference	Raw Score 95% Tukey CI	p
Sub-Saharan Africa—Latin America	−0.63	−1.18 to −0.08	0.01

A total of 101 mid-level leaders were surveyed, and 88 completed responses were received regarding their perceptions and beliefs about their followers’ critical thinking skills. Data were analyzed, examining each individual pillar of the QUEEN Critical Thinking Model as well as the overall average of all pillar scores. Data were also segmented to examine differences by mid-level leaders’ global locations, tenure in the field, and tenure in a leadership position.

Significant differences were found in relation to where mid-level leaders were based worldwide. The resulting p-values showed significant differences for the component of understanding [F (7, 80) = 2.9, p = 0.01], neoteric thinking [F (7, 80) = 4.11, p = 0.0007], and the total QUEEN score [F (7, 80) = 2.9, N = 87, p = 0.01] using a predetermined Type 1 error rate of 0.05. The critical thinking component of Understanding yielded an effect size, η², of 0.2 at the 95% confidence level, indicating that 20% of the difference in means could be attributed to the location of the mid-level leader. There was not enough data to determine which global locations differed significantly. Statistical significance was found for Sub-Saharan Africa compared to Latin America, Anglo, and Middle East for Neoteric Thinking, with 26% of the variance being explained by location from the effect size.

It can be concluded that there are statistically significant differences in the perceptions of critical thinking skills across geographic regions, as defined by the GLOBE clusters. This finding correlates with the recognition of differing styles of leadership across global leadership behaviors (Hackman & Johnson, 2009) but asks the further question of how those leadership skills and styles may influence perceptions of critical thinking skills in followers. This is consistent with Linville and Rennaker’s (2024) observation that no research has yet examined followership within the GLOBE framework.

5. Discussion

These findings have important implications for leadership development, workforce training, and understanding of employees or followers. However, with its limitation of unbalanced sampling across geographic regions and limited participation size, it serves as an initial exploratory study into the connection between leaders’ perceptions of critical thinking of their followers across varying regions. Given the statistical significance of differences in critical thinking ratings worldwide, international organizations may need to provide tailored training, specific to the GLOBE clusters, to help followers understand critical thinking and develop critical thinking skills. This customization can help to provide the best opportunity for success among mid-level leaders and followers (Putera et al., 2026). The results also contribute to the theoretical understanding of critical thinking, reinforcing the notion that perceptions of it vary among mid-level leaders and warrant further study.

As stated previously, while this study provides valuable insights, it is not without limitations. The sample was drawn from mid-level leaders selected to participate in a leadership development program, but the sample and respondents were not equally distributed across the identified GLOBE clusters. Further data collection is advised to provide a more normalized distribution of the respondents across the identified GLOBE clusters. This could enable more thorough data analysis and further testing to identify significant differences between paired clusters. The study also does not account for other factors that could potentially influence a leader’s perception of their followers, including things like team size, proximity to workers (hybrid/in-person), and the training of supervisors to recognize critical thinking. There is also a limitation on self-perceived behaviors of followers.

Additionally, there are many additional areas for analysis utilizing the same methodology as this study. One factor that could impact perceptions of critical thinking is asking mid-level leaders to identify a numerical range of followers they are considering when providing the average rating. This could indicate whether there is a statistical difference in perceptions of critical thinking by the number of followers observed. Mid-level leaders could also be asked to rate and compare their best and worst followers to better understand the differences between what they view as positive and negative. Further research should also explore followers’ critical thinking behaviors across varying industries.