Abstract
This study investigates Tolerance of Ambiguity in project managers, focusing on how their ability to navigate ambiguous situations varies over time and influences their performance. In a four-week study of 275 project managers, we tracked changes in Tolerance of Ambiguity in project managers and found that these temporary fluctuations were significantly associated with positive mood, adaptive performance, and project progress. Factors such as self-control, coping skills, creativity, and supportive leadership emerged as key drivers of these changes. This study highlights Tolerance of Ambiguity as a dynamic, context-dependent competency in project managers, offering novel insights for enhancing project managers’ ability to navigate ambiguous situations.
Introduction
The project management literature is increasingly recognizing the role of soft competencies in explaining the successes and failures of project managers. For example, project managers’ personality traits (Creasy & Anantatmula, 2013), intellectual ability, managerial style, and emotional competencies (Müller & Turner, 2010; Rezvani et al., 2016; Stevenson & Starkweather, 2010) have all been linked to project success. Furthermore, such soft competencies have been found to contribute as much, or even more, to project success than technical skills such as planning or scheduling (Gray & Ulbrich, 2017; Lloyd-Walker & Walker, 2011). Tolerance of Ambiguity, sometimes referred to as ambiguity acceptance, has recently emerged as a critical factor in driving project success, enabling project managers to navigate uncertainty, adapt to challenges, and foster innovative problem-solving. Project success, in this context, refers to meeting key objectives such as completing projects on time, within budget, and to the desired quality standards. Managers with higher Tolerance of Ambiguity are better positioned to achieve these outcomes (Gray & Ulbrich, 2017; Hagen & Park, 2013; Wiewiora & O’Connor, 2022).
However, while the existing literature highlights Tolerance of Ambiguity as a critical soft competency for project managers, it offers limited guidance for individuals seeking to improve their own performance. This is because most studies rely on cross-sectional designs, which compare different individuals at a single point in time (see Hecht et al., 2023; Sosnowska et al., 2021). While these studies show that project managers with higher Tolerance of Ambiguity tend to outperform those with lower Tolerance of Ambiguity, they cannot determine whether increasing Tolerance of Ambiguity within an individual will lead to improved outcomes. Similarly, research identifying factors that contribute to Tolerance of Ambiguity in different managers does not establish that these factors will increase Tolerance of Ambiguity within a single person. We identify this as a significant research gap, which can be addressed through within-person studies that measure the same individuals over time (see Collins et al., 2017). Given the accelerated pace of global changes, technological advancements, and increasing project complexity (Lovett et al., 2023), understanding and enhancing Tolerance of Ambiguity among project managers is increasingly important.
By investigating the dynamics of Tolerance of Ambiguity within project managers, this study directly addresses the practical challenge of managing ambiguity and uncertainty in projects, an issue magnified by the increasing complexity of the modern world (Lovett et al., 2023). Our study examines how changes in Tolerance of Ambiguity correlate with variations in key outcomes such as positive mood, project manager performance, and project progress (i.e., the extent to which a project meets its scheduled milestones and objectives within the planned timeframe). This approach provides valuable insights into the relationship between managing ambiguity and achieving project success. Furthermore, by identifying the individual and organizational factors that influence Tolerance of Ambiguity, we offer a nuanced understanding of how to create an environment that enhances project managers’ ability to navigate uncertainty. These insights are important for developing targeted interventions to strengthen project managers’ capacity to manage the inherent uncertainties of their projects, ultimately improving overall project outcomes.
Therefore, the purpose of this article is to is to investigate the dynamic relationships between fluctuations in project managers’ Tolerance of Ambiguity and key outcome variables such as positive mood, performance, and project progress. By examining the underlying drivers of Tolerance of Ambiguity within individuals, this study aims to reveal how and why Tolerance of Ambiguity varies over time, contributing to a deeper understanding of its role in managing ambiguity. Our findings aim to fill existing gaps in the literature and offer a clearer understanding of how Tolerance of Ambiguity influences project management effectiveness.
Recognizing the need for project management research to draw from psychology and behavioral studies (Geraldi & Soderland, 2018), this project applies insights from Whole Trait Theory (Fleeson & Jayawickreme, 2015) and personality literature that distinguishes between personality traits and personality states (e.g., Cervone, 2005; Collins et al., 2017). Our focus is on exploring how Tolerance of Ambiguity changes within individuals over time. Guided by this literature and the limitations of existing studies, we pose two research questions:
Do changes in Tolerance of Ambiguity within project managers over time/situations predict changes in performance and affect within project managers and in project progress over time? Do state emotional control, state creativity, situational leader support, and state problem-focused coping influence Tolerance of Ambiguity in project managers?
To explore these questions, we conducted a longitudinal study with 275 project managers from the United States who completed up to five weekly surveys. Conducting a longitudinal survey allowed us to track fluctuations in our focal variables across participants over a four-week period. Each week, we asked project managers to describe an ambiguous situation they had encountered in the previous seven days and respond to a set of questionnaires relating to this situation. We adapted established questionnaires to measure project managers’ personality states in these ambiguous situations. Personality states refer to the temporary emotional and behavioral responses to specific situations, distinct from enduring personality traits. We measured state Tolerance of Ambiguity (O’Connor et al., 2018), state creativity (Hogan Personality Inventory (HPI), Hogan & Hogan, 1995) and state problem-focused coping (Matthews & Campbell, 1998). We also measured project managers’ well-being in terms of positive affect (Positive and Negative Affect Scale [PANAS]; Watson et al., 1988), their weekly work performance (Griffin et al., 2007) project progress, and the support they receive from their supervisors.
This research makes several key contributions to the field of project management and the understanding of soft competencies. First, by employing a longitudinal within-person design, we demonstrate how fluctuations in Tolerance of Ambiguity within individual project managers are associated with changes in positive mood, performance, and project progress. This provides a more nuanced understanding of how Tolerance of Ambiguity influences project success at the individual level. Second, by identifying both individual and organizational factors that affect Tolerance of Ambiguity, we offer insights into how project managers can enhance their capacity to navigate ambiguity and uncertainty in dynamic project environments. Third, by integrating concepts from psychology and personality research—specifically Whole Trait Theory and the distinction between personality traits and states—we contribute to an interdisciplinary framework that enriches the theoretical foundations of project management. Overall, our research addresses significant gaps in the existing literature and provides practical implications for developing strategies to improve project managers’ ability to manage uncertainty, ultimately enhancing project outcomes.
Literature Review
Tolerance of Ambiguity and Project Management
Ambiguity, as defined by Wiewiora and O'Connor (2022), is largely characterized by inadequate information and reflects an agent's perception of the availability of important information. In the context of project management, situational ambiguity occurs when information is vague, inaccessible, poorly communicated, insufficient, or comes from unreliable sources. This type of ambiguity is particularly important for project managers, as it directly impacts project managers’ decision-making and their ability to allocate resources effectively (Wiewiora & O’Connor, 2022).
Tolerance of Ambiguity can be defined as the tendency to perceive ambiguous situations as desirable (Budner, 1962) rather than aversive (McLain, 2009). Those who score high on common measures of Tolerance of Ambiguity, tend to be comfortable in the context of ambiguity, have a desire for complexity and novelty, and tend to be low in moral absolutism (i.e., they tend not to have strong and absolute opinions about what constitutes right/wrong; Lauriola et al., 2016). On the contrary, those low in Tolerance of Ambiguity tend to experience discomfort and anxiety when confronted with ambiguity, do not seek out complexity, tend to be high in moral absolutism, and have difficulty managing ambiguity when it occurs (Lauriola et al., 2016; O’Connor et al., 2018).
Given that project management is characterized by high levels of ambiguity, it is not surprising that researchers have explored—and ultimately confirmed—the importance of Tolerance of Ambiguity in project managers (Gray & Ulbrich, 2017; Hagen & Park, 2013). Based on a comprehensive review of the Tolerance of Ambiguity literature, Gray and Ulbrich (2017) concluded that project managers with high Tolerance of Ambiguity perform well due to their confidence in handling unknowns, managing change, and leading others through uncertainty. Indeed, the benefits of Tolerance of Ambiguity are not limited to project managers. Broader research has confirmed that Tolerance of Ambiguity differentiates effective from ineffective workers operating in the context of high ambiguity such as those occupying leadership positions (O’Connor et al., 2017).
Although initially regarded as a single trait-like dimension (Budner, 1962), growing research on Tolerance of Ambiguity has demonstrated that it is not actually a single competency, but instead is best conceptualized as a set of related attitudes toward ambiguity (Lauriola et al., 2016; O’Connor et al., 2018; O’Connor et al., 2022). In a study focusing on Tolerance of Ambiguity in employees, O’Connor and colleagues (2018) found evidence for Tolerance of Ambiguity facets termed comfort with ambiguity, desire for challenging work, and managing uncertainty and found that each was associated with worker performance. Indeed, this view of a broad, multifaceted nature of Tolerance of Ambiguity is consistent with Gray and Ulbrich (2017) who identified multiple Tolerance of Ambiguity facets from the project management literature, including what they termed flexibility, investigating, integrative thinking, and creative thinking, among others. In the current research, therefore, we adopt a broad measure of Tolerance of Ambiguity that captures the multifaceted dimensions assessed in recent measures. We also introduce a nuanced view of Tolerance of Ambiguity, acknowledging it as a personality characteristic with both trait-like and state-like properties (a distinction we will explore further in the following section).
Between Versus Within-Person Variation in Tolerance of Ambiguity
An important distinction in the current research is the difference between personality traits and personality states. Personality traits can be defined as patterns in thoughts, feelings, behaviors, and attitudes that remain stable over time (Anglim & O’Connor, 2019). On the contrary, personality states can be defined as more temporary behaviors, thoughts, and feelings regarding specific situations or time points (see Horstmann & Ziegler, 2020). As outlined extensively in the personality psychology literature (e.g., Cervone, 2005; Collins et al., 2017) but largely unrecognized elsewhere, traits are constructs that vary between people (i.e., some individuals have more or less of a trait than others), whereas states are constructs that can vary between and within people (i.e., over time, individuals experience increases and decreases in states). For example, if someone is high in trait positive mood, it means they tend to experience more positive affect in general compared to others. In contrast if someone is high in state positive mood/affect, it means they are feeling positive at a certain point in time, which may be more or less positive than how they feel at other times.
A common mistake in applied psychology and management research is the tendency for researchers to generalize findings from between-person studies (e.g., studies that link traits to performance outcomes in different people) to within-person relationships. In the current context, for example, it would be incorrect to assume that because trait Tolerance of Ambiguity predicts project manager performance between individuals, that state Tolerance of Ambiguity will predict strong performance within individuals. A commonly cited example illustrating how trait and state associations can differ was provided by Corr et al. (2013), who discussed the association between conscientiousness (i.e., the tendency to be organized) and neuroticism (i.e., the tendency to be emotionally unstable and anxious). When the associations between trait consciousness and trait neuroticism are examined between people, they are reliably negative (Mount et al., 2005), meaning that conscientious people tend to be emotionally stable. However, when the associations between these constructs are examined within people, they are positively associated (Minbashian et al., 2010), meaning that people are more likely to behave conscientiously when experiencing negative emotions such as fear and anxiety. Researchers and practitioners must therefore use caution when drawing from between-person studies to make claims about personality states and individual performance.
In the context of Tolerance of Ambiguity and project manager outcomes, we believe that Tolerance of Ambiguity states within project managers will generally drive positive outcomes; in other words, in situations whereby project managers accept and tolerate the ambiguity, they will tend to have better outcomes (and vice versa). This is based on the known benefits of trait Tolerance of Ambiguity in project managers (Gray & Ulbrich, 2017) and the tendency for traits to influence states (Fleeson & Jayawickreme, 2015). However, we are cognizant of the possibility that this may not be the case, and, in the absence of within-person empirical research, we cannot assume such positive relationships exist. For example, it is plausible that in highly ambiguous contexts (e.g., situations in projects whereby the best course of action cannot be determined), that being too tolerant of ambiguity will actually lead to negative outcomes. In fact, it may be more adaptive to be less tolerant of ambiguity in highly ambiguous situations (see O’Connor et al., 2022). Additionally, it is possible that those high in trait Tolerance of Ambiguity are more successful in the context of ambiguity because they have a set of other traits and abilities that allow them to effectively address the challenges of ambiguity (e.g., emotional intelligence; see Rezvani et al., 2016). If this is the case, then it may be counterproductive to attempt to boost Tolerance of Ambiguity within individuals, but not boost the various competences that allow individuals to effectively navigate ambiguity. The current study, which uses a within-person design and several control variables, will allow us to investigate these different possibilities and determine the unique impact of state Tolerance of Ambiguity on project manager performance within individuals over time.
Development of Hypotheses
In formulating our hypotheses regarding the effects of state/situational Tolerance of Ambiguity on project manager outcomes, we draw first from Whole Trait Theory (Fleeson & Jayawickreme, 2015, 2021), which is an influential personality theory that recognizes the dual state- and trait-like properties of personality constructs. It proposes that traits are averages of states, conceptualized as density distributions of states measured over time within individuals. According to Whole Trait Theory, people with high levels of certain personality traits are more likely to engage in respective personality states, because traits influence the likelihood that individuals will engage in respective states in any given situation. The theory therefore recognizes that traits are an important driver of states; however, it also recognizes that states vary substantially within people and are affected by a range of contextual variables in addition to traits. Importantly therefore, it follows from Whole Trait Theory that, in any one situation, people are capable of states that are both consistent and inconsistent with their underlying traits. In the context of our study, we are particularly interested in exploring how state Tolerance of Ambiguity, regardless of underlying traits, affects diverse project manager outcomes.
Based on Whole Trait Theory and the project management literature on Tolerance of Ambiguity, we propose that the known benefits of trait Tolerance of Ambiguity on project manager performance are partially attributable to key Tolerance of Ambiguity states. That is, project managers high in trait Tolerance of Ambiguity, are more likely to experience positive outcomes because they are more likely to enact state Tolerance of Ambiguity than those low in trait Tolerance of Ambiguity. We also draw from research on mechanisms linking trait Tolerance of Ambiguity to performance, that make theoretical sense at the state level. Specifically, consistent with Wiewiora and O’Connor (2022) and Furnham and Marks (2013) we suggest that tolerating ambiguity in specific situations will allow project managers to view ambiguity as an opportunity rather than a threat and cause them to utilize internal resources to make good decisions. Additionally, consistent with Gray and Ulbrich (2017) we suggest that tolerating ambiguity within specific situations, will mean that project managers will have more confidence in their abilities to remedy, deal with, and control the situation.
In hypothesizing the effects of Tolerance of Ambiguity on project outcomes, we identify three key categories. The first category is project manager well-being, with a focus on state positive affect, which has been studied in various contexts, including by Hancock and Mattick (2020) as an example of how Tolerance of Ambiguity influences well-being. The second category, project manager performance, includes indices of core job performance, adaptive performance, and proactive performance which have been explored in broader management contexts by O’Connor et al. (2022). The third category is project progress, specifically linked to Tolerance of Ambiguity in project managers by Wiewiora and O’Connor (2022). Based on this foundational research, we hypothesize that state Tolerance of Ambiguity will positively impact all three outcomes for project managers. Consistent with theoretical perspectives of within-person variation (Fleeson & Jayawickreme, 2015, 2021), we expect that the effects of state Tolerance of Ambiguity will be significant, even when controlling for the effects of trait Tolerance of Ambiguity (i.e., we believe that variation in Tolerance of Ambiguity within individuals will account for performance outcomes beyond variation in trait Tolerance of Ambiguity between individuals). Thus, we hypothesize: State Tolerance of Ambiguity will be associated with positive affect (H1a), project manager performance (H1b), and project progress (H1c) within project managers.
Our second set of hypotheses relate to the question: Why do individual project managers tolerate ambiguity in some situations but not others? We therefore explore whether a range of situational and psychological factors increase and/or reduce the likelihood that project managers will tolerate ambiguity in a given situation. In formulating our second set of hypotheses, we draw from the broader literature on correlates of Tolerance of Ambiguity in employees and also from the stress literature. We focus on stress theories because ambiguity is widely recognized as a potent stressor. These theories offer perspectives on the within-person processes that facilitate positive responses to ambiguity.
One of the most influential theories of stress and coping was proposed by Lazarus and Folkman (1984, 1987), who argued that individuals’ responses to stressors could be explained in terms of appraisal processes. According to their transactional model of stress and coping (Lazarus & Folman, 1984, 1987), when individuals are confronted with a stressor (e.g., ambiguity, threat, challenge), they appraise whether they have sufficient resources to manage and/or overcome the stressor. If individuals believe they have sufficient resources, the stressor will not elicit a strong stress response (e.g., worry, negative affect, etc.). These resources could theoretically be internal to the individual (e.g., emotional control, willpower) and/or external to the individual (e.g., social support). Following the appraisal process, the model then stipulates that individuals engage in a coping strategy (e.g., adopting problem- solving strategies to manage stressors), which, depending on its efficacy, may or may not lead to positive outcomes (Lazarus & Folkman, 1984, 1987).
Applying the transactional model of stress and coping to the project manager context, we argue that, in an ambiguous situation, project managers will be more likely to accept the ambiguity (i.e., a potential source of stress) when they believe they have sufficient internal and external resources to positively appraise and manage the stressor. Specifically, we suggest that when confronted with ambiguity, project managers who believe they have the capacity to control their emotions and adopt a creative mindset in that situation (internal resources), who perceive they have support from their leader (external resource), and who engage in problem-focused coping, will be well-positioned to tolerate and accept the ambiguity. Indeed, these proposed resources and coping strategies align with the various internal and external resources that have been studied in the context of the transactional model of stress and coping broadly (e.g., Rafferty & Griffin, 2006; Reif et al., 2021). Furthermore, these resources have also been specifically linked to employee Tolerance of Ambiguity in a recent between-person study exploring predictors of Tolerance of Ambiguity in employees (O’Connor et al., 2018). We therefore expect these factors to drive fluctuations in state Tolerance of Ambiguity within project managers over time. Thus, we hypothesize: State emotional control (H2a), state creativity (H2b), situational leader support (H2c), and state problem-focused coping (H2d) will predict state Tolerance of Ambiguity within project managers.
In selecting these four factors—state emotional control, state creativity, situational leader support, and state problem-focused coping—as predictors of state Tolerance of Ambiguity within project managers, we grounded our choice in both the transactional model of stress and coping and Whole Trait Theory. These factors have been empirically linked to effective ambiguity management and performance outcomes in ambiguous situations, as noted by O’Connor et al. (2018) and supported by the work of Furnham and Marks (2013) and Lazarus and Folkman (1984, 1987). Specifically, state emotional control and state creativity represent internal resources that enable project managers to adaptively respond to ambiguity by harnessing personal resilience and innovative problem-solving. Situational leader support acts as an external resource, bolstering the project manager's capacity to navigate uncertainty through perceived organizational backing. Lastly, state problem-focused coping reflects a proactive strategy in managing the demands of ambiguous project situations, aligning with effective stress and ambiguity resolution tactics. While there are undoubtedly other factors that could influence state Tolerance of Ambiguity, these four were chosen for their direct relevance to the appraisal and coping processes central to the transactional model of stress and coping. Figure 1 provides a summary of the hypotheses.
Proposed research model.
Methods
Participants
Participants were 275 U.S. project managers recruited using the Qualtrics Panel, which has access to several thousand project managers. They were from a broad range of industries, including professional, scientific and technical, manufacturing, construction, education and training, and public administration and safety, among others. Participants were invited to complete five surveys over the course of four weeks, which included one baseline and four weekly surveys. In total, we collected data from 686 observations across these five waves. Surveys were completed online using the Qualtrics survey software, and participants were given access to the weekly survey on Friday afternoons (i.e., the end of the work week) and were given until the following Monday morning to complete each survey. Up to one reminder was administered to participants each week, approximately 24 hours after they were sent the initial link. Participants received a small reward in exchange for completing each survey (i.e., points that could be exchanged for store vouchers), and the value of this reward increased each subsequent week to serve as extra motivation for participants to respond. The average age of participants was 47.97; the sample included slightly more females (n = 155) than males (n = 120). The four-week survey period was chosen based on prior research, indicating that meaningful changes in psychological states and behaviors can occur within this timeframe (e.g., Fleeson, 2001). The dynamic nature of project management, where project phases and challenges can shift rapidly, makes this period appropriate for capturing variations in Tolerance of Ambiguity and related outcomes.
Measures
Data were collected over a four-week period using five nearly identical surveys. The first week’s survey (the baseline survey) included demographic information and trait-level control variables, which were not repeated in subsequent weeks. For both the baseline and weekly surveys, participants were asked to describe an ambiguous situation and complete various questions designed to measure personality states and outcome variables: tolerance of ambiguity, positive and negative affect, weekly work performance, project progress, emotional control, creativity, and problem-focused coping. These variables are outlined in detail in the following paragraphs.
Demographics
Prior to completing our focal questionnaires, participants responded to a series of demographic questions regarding their age, gender (male, female, other), years in current role, industry, and years managing projects.
Description of Ambiguous Situation
Prior to completing each of the weekly surveys, project managers were instructed to respond to the following question: “In regard to the project you are currently most invested in, please think of an ambiguous situation you encountered in the last seven days, which was unclear, such that the best way to manage the situation was not obvious. The situation should not be trivial, but important to the outcomes of the project…. Please briefly describe the situation in one sentence.”
State Tolerance of Ambiguity
Participants were asked to complete 17 items regarding what they did, and how they felt about the ambiguous situation they described. The items addressed how comfortable they were with the ambiguous situation: “the situation made me feel anxious”; whether they perceived the ambiguity as a challenge: “I liked engaging with the situation because it was complex”; and how effectively they managed the ambiguity: “I was open to new ideas when trying to resolve the problem.” Participants responded to items on a 5-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). The items were adapted from the trait Tolerance of Ambiguity measure developed by O’Connor et al. (2018). The scale was internally reliable in the current study (omega total, ω = .88).
State Positive and Negative Affect
Participants completed the 20-item PANAS (Watson et al., 1988). This is a widely used scale that consists of words describing feelings and emotions (e.g., attentive, alert, upset). Participants were asked to rate the extent to which they agreed with each item from 1 (very slightly or not at all) to 5 (extremely/very much) over the previous week. Measures of both positive affect (ω = .96) and negative affect (ω = .95) showed good omega total scores in the current study.
Weekly Work Performance
Job performance was measured using Griffin et al.’s (2007) work role performance measure. Although it can be used as a measure of overall job performance, it contains three facets including proficiency (“I carried out the core parts of my job well”), adaptivity (“I adapted well to changes in core tasks”), and proactivity (“I initiated better ways of doing my core tasks). In the current study we explored how Tolerance of Ambiguity impacted these facets, rather than overall job performance. All scales showed good omega total scores (ω > .72).
Project Progress
To assess project progress, participants responded on a 5-point scale (ranging from strongly disagree to strongly agree) to the items “The project is currently meeting its time goals,” “The project is currently meeting its budget goals,” “The project is currently meeting scope and requirements goals,” “The client is satisfied with the progress of the project,” “The project team members are satisfied with the progress of the project,” and “My manager is satisfied with the progress of the project.” The scale had a good omega total score (ω = .93).
State Emotional Control
This variable was measured using the self-control facet of the trait emotional intelligence questionnaire (Petrides, 2009), which was modified to measure situational rather than trait emotional control. Participants were asked to respond to five items designed to measure emotional control in terms of what they “did and felt in the past seven days” when they “approached the ambiguous situation they described.” An example item was “I was able to find ways to control my emotions when I wanted to.” The scale has established validity (O’Connor et al., 2017) and was found to have a good omega total score in this study (ω = .73).
State Creativity
To measure state creativity, we adapted items from the creativity measure from the HPI (Hogan & Hogan, 1995). Participants were again asked to indicate how they approached the ambiguous situation they described in terms of five items designed to measure creativity. Example items include “I used my vivid imagination when seeking to resolve the situation,” and “I was interested in exploring abstract ideas to solve the situation.” The scale had a good omega total score (ω = .93).
State Problem-Focused Coping
State problem-focused coping was measured using a modified version of the task-focused coping scale (Matthews & Campbell, 1998). Participants were asked to respond to the five-item measure in terms of how they approached the ambiguous situation they described. An example item is: “I focused my attention on the most important parts of the task.” The scale had a good omega total score (ω = .81).
Situational Leadership Support
To measure situational leadership support, project managers completed three items adapted from Madjar et al.’s (2002) measure of supportive leadership. Participants responded to these items based on their interactions with their manager over the previous seven days. An example item is: “My supervisor was always ready to support me even if I introduced an unpopular idea or solution at work.” The scale demonstrated good reliability, with an omega score (ω) of .80.
Control Variables
In this study, we incorporated several control variables to mitigate potential confounding effects. These controls included gender, age, trait positive affect, trait negative affect, trait self-control, trait sociability, trait self-efficacy, and trait Tolerance of Ambiguity, as detailed in Table 1. We selected these specific variables as controls due to their established relationships with one or more of the outcome variables under investigation (see, for example, Furnham & Marks, 2013; Wiewiora & O’Connor, 2022; O’Connor et al., 2022). Additionally, there is a plausible expectation for correlations between these traits and state Tolerance of Ambiguity.
Correlations, Means, Standard Deviations, and Omega Total Estimates
Note: Means and standard deviations are at the observation level. Between-person correlations (nj = 275 after listwise deletion) are under the diagonal and within-person correlations (ni = 686) are above the diagonal. Omega total estimates are shown in bold italics on the diagonal. Correlations at .13 or higher are significant at .05, .16 or higher significant at .01, and .19 or higher significant at .001. State = situational. Gender is coded as 1 = male, 2 = female. TOA = Tolerance of Ambiguity.
Estimation Method
We constructed three two-level multilevel models for each dependent variable with two levels where level one was the repeated measures (wave) and level two was the person (i.e., between-person level). The first model tested for random intercepts via the intraclass correlation coefficient (ICC) to assess the need for a multilevel model, whereby a higher ICC indicates that there is sufficient variance in the outcome variable at the between-person level (i.e., project managers experience different levels of the outcome, on average, and must therefore be modeled individually within a multilevel model approach). The second model added a random effect with respect to the wave, thus allowing the trajectory of the outcome variable to vary by person over time, which is essential to understanding individual differences in response to project management challenges. The third added the fixed effects, which are necessary for testing our hypotheses on the direct effects of state Tolerance of Ambiguity and other predictors on project outcomes. All analyses were conducted in R (R Core Team, 2018), where the multilevel modeling used the “lme4” package (Bates et al., 2015). The lme4 package handles missing data by using Full Information Maximum Likelihood estimation. This approach utilizes all available data points to provide robust and unbiased parameter estimates, ensuring that the analysis is comprehensive and makes the best use of the collected data without requiring explicit imputation. Thus, our analysis method allowed us to include participants with partial data.
Results
Means, standard deviations, correlations, and omega total scores are shown in Table 1.
We first report on the analyses regarding the hypothesized outcomes of situational Tolerance of Ambiguity (H1a–H1c). As indicated in the model 1 column in Tables 2 through 7, the ICCs for every model, which ranged from .38 to .67, justified the use of multilevel modeling without exception (i.e., there was sufficient between-person variance in the respective outcome variables). After modeling for random effects shown in the model 2 column (which were present for some outcome variables), incremental variance was accounted for in the final model (see the model 3 column) for every one of our proposed outcomes of situational Tolerance of Ambiguity. Consistent with our hypotheses, we found that across our fixed effect models (shown in the model 3 column), state Tolerance of Ambiguity was a unique predictor of state positive affect (H1a) (b = .32, SE = .06, p < .001), situational adaptive performance (H1b) (b = .24, SE = .07, p < .01), and situational project progress (H1c) (b = .15, SE = .06, p < .05). This indicates that as state Tolerance of Ambiguity increases within project managers, so too does state positive affect and state adaptive performance. However, state Tolerance of Ambiguity within project managers did not predict job proficiency or proactive performance, which were additional project manager performance variables relevant to H1c. While not hypothesized, it is worth noting various relationships among control variables and outcome variables. Specifically, state negative affect was negatively predicted by state self-control (b = -.17, SE = .05, p < .01). Job proficiency was positively predicted by state self-control (b = .15, SE = .07, p < .05), state task-focused coping (b = .21, SE = .07, p < .01), and supportive leadership (b = .09, SE = .04, p < .05). Our only significant predictor of proactive performance was situational creativity (b = .22, SE = .08, p < .01).
Multilevel Model Output Where the Dependent Variable is Situational Positive Affect
Note: ni = 686, nj = 275. Full information maximum likelihood estimation.
Values for fixed effects are unstandardized betas with standard errors shown in parentheses. State = situational.
***p < .001. **p < .01. *p < .05.
Multilevel Model Output Where the Dependent Variable is Situational Negative Affect
Note: ni = 686, nj = 275. Full information maximum likelihood estimation. Values for fixed effects are unstandardized betas with standard errors shown in parentheses. State = situational.
***p < .001. **p < .01. *p < .05.
Multilevel Model Output Where the Dependent Variable is Situational Adaptive Performance
Note: ni = 686, nj = 275. Full information maximum likelihood estimation. Values for fixed effects are unstandardized betas with standard errors shown in parentheses. State = situational.
***p < .001. **p < .01. *p < .05.
Multilevel Model Output Where the Dependent Variable is Situational Proactive Performance
Note: ni = 686, nj = 275. Full information maximum likelihood estimation. Values for fixed effects are unstandardized betas with standard errors shown in parentheses. State = situational.
***p < .001. **p < .01. *p < .05.
Multilevel Model Output Where the Dependent Variable is Situational Project Progress
Note: ni = 686, nj = 275. Full information maximum likelihood estimation. Values for fixed effects are unstandardized betas with standard errors shown in parentheses. State = situational.
***p < .001. **p < .01. *p < .05.
We next report the analysis for state Tolerance of Ambiguity as the outcome variable (H2a – H2d). As shown in Table 7, model 1 column, the ICC justified use of a multilevel model (ICC = .52). In the second model (model 2 column), we tested for the effect of random slopes of the wave, which was significantly different from the first model (χ2 = 13.80, Δdf = 2, p < .01), thus indicating that allowing the wave to vary within person meaningfully accounted for variance in the model. The third model (model 3 column) added the main effects and covariates, which were significantly different compared to the second model (χ2 = 349.52, Δdf = 18, p < .001). All hypothesized main effects were significant, including state self-control (b = .22, SE = .05, p < .001), state task-focused coping (b = .17, SE = .05, p < .001), state creativity (b = .16, SE = .04, p < .001), and supportive leadership (b = .08, SE = .03, p < .01). The positive values of these estimates indicate that increases in each predictor (state self-control, state task-focused coping, state creativity, and supportive leadership) were uniquely associated with increases in state Tolerance of Ambiguity within project managers over time. We therefore received strong support for H2a through H2d.
Multilevel Model Output Where the Dependent Variable is Situational Tolerance of Ambiguity
Note: ni = 686, nj = 275. Full information maximum likelihood estimation. Values for fixed effects are unstandardized betas with standard errors shown in parentheses. State = situational.
***p < .001. **p < .01. *p < .05.
Discussion
There is a growing recognition in the project management literature that soft competencies play an important role in the successes and failures of project managers. One such competency, Tolerance of Ambiguity, is receiving increased attention in project management research (e.g., Gray & Ulbrich, 2017: Hagen & Park, 2013; Wiewiora & O’Connor, 2022), with existing studies demonstrating that Tolerance of Ambiguity and related constructs differentiate between effective and less effective project managers. Our study adds to this literature by addressing how Tolerance of Ambiguity fluctuates within individuals over time and how this variability impacts outcomes. In our study, we sought to broaden the research on Tolerance of Ambiguity in the project management context, by exploring two related research questions: (1) Do changes in Tolerance of Ambiguity over time/situations, predict changes in performance, project success, and affect in project managers? (2) Do factors within project managers (e.g., emotional control) and organizational factors (e.g., leadership) foster Tolerance of Ambiguity within project managers? Drawing from Whole Trait Theory (Fleeson & Jayawickreme, 2015), and Lazarus and Folkman’s (1984, 1987), transactional model of stress and coping we formulated a range of hypotheses regarding drivers and outcomes of Tolerance of Ambiguity within individual project managers.
Using a longitudinal study across five time points, we found that when project managers accepted and tolerated ambiguity in specific situations, they were more likely to experience positive outcomes in these situations, in terms of their positive affect, adaptive performance, and project progress. We also found that project managers were more likely to accept and tolerate ambiguity, when they reported high levels of state emotional control, state creativity, state problem-focused coping, and when they felt they had a supportive leader/supervisor. This dynamic relationship between state-level factors and situational Tolerance of Ambiguity contributes to a more nuanced understanding of how ambiguity tolerance operates in real-world project environments. Our overall set of results therefore supports the model illustrated in Figure 1 and collectively allow us to better understand when, how, and why project managers effectively manage ambiguity in their projects.
Theoretical Implications
As we outlined earlier, a limitation of existing research on soft competencies in project managers is the near sole focus on factors (e.g., traits) that discriminate between project managers rather than on states that affect performance within project managers. While research on the traits of effective project managers is useful in developing theory on why different project managers have different outcomes, it has limited value in addressing how and why individual project managers differ in how they tolerate ambiguity across different situations and points in time. In other words, research on the traits of effective leaders speaks to phenomena that occur at the between-person level (operationalized as level 2 in this study), whereas research on the states of effective leaders speak to phenomena that occur at the within-person level (operationalized as level 1 in this study). Therefore, in exploring the drivers and outcomes of situational Tolerance of Ambiguity within project managers in this study, we aimed to enrich the understanding of effective project management behavior by focusing on the underexplored phenomenon of situational Tolerance of Ambiguity.
In seeking to explain how and why individual project managers differ in their Tolerance of Ambiguity across different situations, and why state Tolerance of Ambiguity leads to positive outcomes, we drew from aspects of Whole Trait Theory (Fleeson & Jayawickreme, 2015) and the transactional model of stress and coping (Lazarus & Folkman, 1984, 1987). We have explicitly linked our findings to these theories by demonstrating that known trait benefits of Tolerance of Ambiguity can be attributed to Tolerance of Ambiguity states. Adopting Tolerance of Ambiguity states allows project managers to view ambiguity as an opportunity rather than a threat, which is consistent with literature on Tolerance of Ambiguity outside of project management (Furnham & Marks, 2013).
We also theorized that state Tolerance of Ambiguity could be attributed to a range of within person and situational factors that, collectively, resulted in project managers forming the belief that they had sufficient resources in that particular situation to manage and/or overcome the ambiguity (i.e., a stressor). Consistent with our model, we found that state Tolerance of Ambiguity did drive positive outcomes in project managers, and that project managers were most likely to accept ambiguity in situations where they could maintain emotional control, adopt a creative and problem-focused mindset, and felt they had good leader/supervisor support. These insights advance the understanding of Tolerance of Ambiguity as a flexible, situation-dependent competence, rather than a fixed trait.
Our study builds on the work of Gray and Ulbrich (2017), Hagen and Park (2013), and Wiewiora and O’Connor (2022), who highlighted the importance of Tolerance of Ambiguity in differentiating effective from less effective project managers. Gray and Ulbrich (2017) demonstrated that project managers with higher Tolerance of Ambiguity were better able to navigate complex project environments, leading to superior performance. Hagen and Park (2013) found that Tolerance of Ambiguity was crucial for adapting to unforeseen project challenges, supporting the idea that ambiguity tolerance is a critical soft skill in project management. Wiewiora and O’Connor (2022) expanded on these findings by linking Tolerance of Ambiguity to project outcomes and team dynamics. Our research extends these studies by focusing on the situational aspect of Tolerance of Ambiguity, showing that it can fluctuate and be influenced by specific factors within the project environment. This distinction is important as it suggests that interventions aimed at enhancing Tolerance of Ambiguity can be situation-specific rather than solely focusing on long-term personality traits. Our findings align with and extend the work of Furnham and Marks (2013) by showing that Tolerance of Ambiguity states can be developed and harnessed in specific project scenarios, leading to positive outcomes such as improved adaptive performance and project progress.
A further issue relevant to our theoretical contribution relates to the question of causation, and specifically, whether our data can speak to causation. First, we want to point out that our theoretical model does imply causation, in that we believe the factors specified in H2 have causal effects on state Tolerance of Ambiguity, and that state Tolerance of Ambiguity then has a causal effect on our various project manager outcome variables. We also point out that our design had a number of features that are consistent with a causal interpretation. Specifically, (1) our proposed Independent Variables (IVs) covaried in expected directions with our proposed Dependent Variables (DVs); (2) the observed covariances between IVs and DVs occurred at the appropriate level of analysis (i.e., within person); (3) our measures were time-ordered and specific to certain points in time, such that participants completed questions regarding how they approached and responded to discrete ambiguous situations respectively; and (4) we took steps to rule out a range of competing explanations by controlling for constructs that could theoretically covary with relevant IV–DV pairs (see Antonakis et al., 2010 for a comprehensive discussion on the various criteria for causation). Taking these design features into account, we believe that our results provide some evidence for the proposed causal effects we imply. Nevertheless, given the methodological challenges in providing clear evidence for causation (e.g., in particular the necessity for experimental methods and full experimental control), we do not claim our results conclusively indicate causation and acknowledge that experimental work is required to conclusively make this claim.
Practical Contributions
Our findings offer specific, actionable insights for both project managers and organizations seeking to improve project outcomes. The demonstrated link between increased Tolerance of Ambiguity and improved performance and well-being, grounded in likely causal relationships as outlined previously, highlights the importance of creating environments where ambiguity is effectively managed. For project managers, targeted training programs focusing on enhancing emotional control, creativity, and problem-focused coping skills are crucial. These programs can equip project managers with practical strategies for adapting to ambiguity in real-time, allowing them to respond more effectively to dynamic and uncertain project demands. For organizations, understanding the role of supportive leadership in fostering Tolerance of Ambiguity provides a clear pathway to developing leadership frameworks aligned with the uncertain and often evolving nature of project environments. By integrating leadership strategies that promote resilience and adaptability, organizations can cultivate a workforce capable of navigating complex project challenges. These actions can ultimately drive improved project performance and contribute to sustained organizational success.
Limitations and Future Research
As indicated above, the primary limitation of this study relates to our inability to conclusively demonstrate our findings reflect causal mechanisms. Although we conducted longitudinal, within-person, time-ordered research, we acknowledge that conclusive claims regarding causation require experimental work (Antonakis et al., 2010). A further limitation in this study relates to our treatment of Tolerance of Ambiguity. Although we used a broad measure of Tolerance of Ambiguity, some researchers argue that Tolerance of Ambiguity reflects several lower order facets such as novelty seeking, desire for challenge, and dichotomous thinking (e.g., O’Connor et al., 2018). In this study, we chose to focus on the higher-order Tolerance of Ambiguity construct to limit the scope of our already substantive study; however, we believe that exploring determinants and outcomes of Tolerance of Ambiguity facets in the project management context will likely reveal a set of informative findings not uncovered by our study.
Future research could therefore explore the lower-order facets of Tolerance of Ambiguity, such as novelty seeking and desire for challenging work, to identify specific aspects beneficial for different project types (O’Connor et al., 2018). Extending the duration of longitudinal studies could provide insights into the long-term stability and variability of Tolerance of Ambiguity. Experimental studies with randomized control trials are recommended to establish stronger causal evidence for Tolerance of Ambiguity’s impact on project outcomes. Investigating the role of different leadership styles in fostering Tolerance of Ambiguity among project managers could offer valuable insights for organizational training. Additionally, examining the interplay between Tolerance of Ambiguity and other competencies like emotional intelligence and coping strategies can inform comprehensive training programs (Müller & Turner, 2010). Finally, cross-cultural studies could determine if the benefits of Tolerance of Ambiguity are universally applicable or vary based on cultural differences.
Conclusion
This study explored the within-person drivers and outcomes of situational Tolerance of Ambiguity. Using a longitudinal study across five time points, we explored whether (1) changes in situational Tolerance of Ambiguity over time predict changes in project manager performance and well-being (in terms of positive affect) over time, and (2) whether changes in a range of situational factors enhanced situational Tolerance of Ambiguity within project managers. Consistent with our predictions, we found that state emotional control, state creativity, situational leader support, and state problem-focused coping are likely drivers of situational Tolerance of Ambiguity, and that situational Tolerance of Ambiguity is a success factor within project managers. Our results have implications for theory and practice, notably, they provide direction for project managers seeking to enhance their performance in the context of ambiguous situations. These findings also highlight potential pathways for organizational leaders to foster environments that encourage ambiguity tolerance, further contributing to project success.
