Person-centered insights into organizational change: Identifying and analyzing profiles using latent profile analysis

Abstract

Organizational change is well-studied, yet remains a fragmented field. While existing theory and research identify various organizational change factors, empirical studies tend to examine these factors in isolation. In this respect, the change field has been ineffective in capturing meaningful profiles or configurations informed by the complex interdependencies among these change factors. This limit contributes to a narrower understanding of organizational change phenomena and how change is studied. To address this gap, we propose a person-centered approach as an accessible and effective approach for studying the underlying profiles that characterizes the complexity of organizational change. This article introduces person-centered research and provides a step-by-step guide to latent profile analysis (LPA), a flagship technique used to analyze profiles. After explaining LPA, we outline essential steps for applying this technique in the context of organizational change, illustrating the value of a person-centered approach in conducting this type of analysis. Offering practical insights for researchers and practitioners, we demonstrate how LPA can uncover hidden profiles of subgroups, providing a more nuanced understanding of organizational change. By making person-centered research more accessible, we promote its use to capture the underlying complexity and diversity of organizational change and its impact on the success of change initiatives.

JEL Classification: M50

Keywords

Latent profile analysis organizational change research person-centered research

1. Introduction

Organizational change is a rich, complex field that encompasses a broad range of research topics and change factors. As organizational change has wide-reaching impacts on work, people, and processes, the past several decades have seen a notable growth in research and practice in this area (Armenakis and Bedeian, 1999; Bouckenooghe et al., 2023; Oreg et al., 2011). This growth has led to a broad array of research designs and statistical methods aimed at better understanding change dynamics, including how individuals and groups perceive and respond to change initiatives (e.g. Anderson and Lenz, 2001; Bartunek and Moch, 1987). Yet, despite these advancements, the field remains dominated by variable-centered approaches, focused on the relationships between variables across a population. While valuable, this often leads to fragmented knowledge (Bouckenooghe et al., 2021), limiting our understanding of the complex and interdependent nature of change processes.

The traditional, variable-centered approach has undeniably contributed important insights to organizational change research, particularly by highlighting how certain variables (e.g. communication, participation) are associated with change outcomes (e.g. resistance, commitment). This approach, however, tends to examine the effects of one variable on another in isolation, overlooking interdependencies among them that can inform broader profiles and dynamic patterns that characterize organizational responses to change (Bouckenooghe et al., 2021). In this respect, as change researchers rely on variable-centered approach, the field has left important themes largely ignored. For example, while the variable-centered approach might explore how employee participation in change-related decision-making is associated with resistance to change, it rarely considers how multiple change-related variables interrelate to form unique profiles that could more accurately capture an employee’s holistic response to change. In addition, while change processes characterize specific actions or interventions undertaken by management to implement change (see Hagl et al., 2024; Stouten et al., 2018), we do not know if these actions work in isolation or come together to represent profiles of organizational units that employ high- and low-quality change processes. Moreover, we lack studies that uncover the climate or context profiles that are favorable (or unfavorable) for the implementation of a newly proposed change initiative.

In response to these limitations, we advocate for the integration of a person-centered approach in organizational change research; a methodology that identifies theoretically meaningful profiles or configurations within a population based on multiple variables (Howard and Hoffman, 2018). Taking this outlook we recognize that the person-centered approach focuses on uncovering why and how variables of interest constitute complex, within-group patterns and subpopulation dynamics. This method has been successfully applied in other areas of organizational research, including studies of commitment (Hofmans et al., 2020; Meyer and Morin, 2016), emotional labor (Woo et al., 2018), and career development (Nguyen and Stinglhamber, 2020). Despite its demonstrated utility, the person-centered approach remains underutilized in organizational change research, limiting our understanding of how diverse change-related variables may combine to create meaningful profiles within different organizational contexts (Bouckenooghe et al., 2021).

Although the field has been strengthened by traditional variance-based methods, there is a tendency among scholars to gravitate toward familiar methods and data analysis techniques. Alternative approaches, particularly those that fall outside the mainstream, often encounter resistance, partly because they require us to move beyond established routines and comfort zones (Aguinis and Edwards, 2014; Harley and Cornelissen, 2022). This tendency can inadvertently narrow our research scope and constrain methodological diversity. As scholars, we must sometimes push beyond the familiar to prevent the “hammer and nail” problem—where the same methodological tools are repeatedly applied, potentially oversimplifying complex phenomena. Addressing this issue, with this article, we intend to expand the methodological toolkit for change research to include both variance-based and person-centered approaches. This diversification enables us to better analyze the nuanced configurations that exist within change-related data, facilitating a more comprehensive understanding of organizational change.

With this focus, we aim to demystify latent profile analysis (LPA), a core technique and flagship model in person-centered research (e.g. Meyer and Morin, 2016; Spurk et al., 2020; Vermunt and Magidson, 2002), by providing a detailed guide that highlights its utility in exploring complex change configurations. By illustrating how person-centered methods like LPA can be applied in change research, we hope to lower barriers to its adoption, fostering a more comprehensive understanding of the multifaceted nature of change. Expanding our methodological toolbox enriches the insights available to the field while ensuring that the diverse and interconnected realities of organizational change are captured and understood at more actionable levels.

2. Origins of person-centered research and its presence in management literature

Person-centered research, grounded in the holistic-interactionistic view of individual development psychology (Magnusson and Törestad, 1993), views individuals as integrated wholes, emphasizing the interplay between personal traits and environmental factors. Unlike traditional variable-centered methods that isolate variables, person-centered approaches identify meaningful patterns of interacting factors within individuals, rather than between individuals, offering a richer understanding of complex dynamics (Bergman and Magnusson, 1997). Magnusson (1988) compared this perspective to biological taxonomies, where species are defined not by a single characteristic but by configurations of traits that provide a complete understanding. Using data from the Individual Development and Adjustment (IDA) longitudinal program, he demonstrated that clustering childhood characteristics could predict adult outcomes like criminality or alcohol abuse, offering insights that might be missed in variable-centered approaches.

While person-centered methods have long been applied in developmental psychology, management research has largely relied on variable-centered methods, focused on relationships between isolated variables to derive generalizable conclusions (Bouckenooghe et al., 2021). Techniques like regression and structural equation modeling help researchers identify broad patterns across populations, assuming that these relationships apply uniformly to all individuals (Howard and Hoffman, 2018; Laursen and Hoff, 2006). While valuable, variable-centered methods often oversimplify the intricate interactions of personal and contextual factors that shape employee attitudes, behaviors, and outcomes. By contrast, the person-centered approach, with focusing on complex interdependencies among factors, identifies shared patterns within subsets of individuals, teams or organizations, allowing for classification into meaningful categories. For example, in organizational change contexts, a person-centered study might reveal different organizational types—such as “innovation-driven” or “stability-focused—or classify teams within an organization based on performance and support needs. Such groupings enable tailored strategies to address each group’s specific dynamics and needs (Marsh et al., 2009). Overall, by identifying distinct subgroups or profiles within organizational populations, person-centered approaches provide a more nuanced view of how individual, team or organizational characteristics and experiences configure into meaningful patterns.

In organizational behavior, person-centered methodologies have deepened theoretical development and practical application. For instance, Meyer and Morin (2016) used LPA to reveal employee subgroups based on commitment types, aiding in retention strategies, while Gellatly et al. (2014) examined commitment profiles’ effects on work unit relations and turnover intentions, and Gabriel et al. (2015) identified distinct emotional labor profiles linked to outcomes like emotional exhaustion and job satisfaction. Exploring Psychological Capital (PsyCap), Bouckenooghe et al. (2019) identify distinct profiles of self-efficacy, hope, optimism, and resilience with implications for engagement and performance. Similarly, Hirschi and Valero (2017) demonstrated how adaptability profiles inform tailored career support, and Chou et al. (2015) identified paternalistic leadership profiles that shape employees’ occupational commitment and turnover intentions. This adoption of LPA and similar person-centered methods has expanded significantly in applied psychology and management research, reflecting a trending shift toward capturing the complex profiles within organizational settings. Reviewing 37 studies using LPA, Woo et al. (2018) note its growing acceptance as a tool for identifying latent subgroups. Since their work, the application of person-centered research has accelerated, underscoring its value in revealing diverse ways individuals and groups experience and engage with organizational contexts.

Yet despite this increasing recognition, and as Bouckenooghe et al. (2021) illustrate in their systematic review of attitudes toward change, the person-centered approach remains underutilized in organizational change research. While recent studies (e.g. Edwards and Clinton, 2023; Kanitz et al., 2024; Straatmann et al., 2018; Tapak et al., 2024) recognize the value of person-centered methods, its application in change research remains limited. To address this gap opportunity, we present a practical guide for organizational change researchers, highlighting the distinct benefits of person-centered approaches over traditional variable-centered methods. Moving away from variable-centered methods prioritizing parsimony and generalizability, person-centered approaches capture the nuanced profiles within groups that may otherwise be overlooked. This debate between parsimony and specificity is particularly relevant in change management, where identifying unique patterns within groups can inform more tailored and effective strategies.

Notably, Howard and Hoffman (2018) discuss the value of the person-specific approach, as a complement to both the person-centered and variable-centered approaches. The person-specific approach focuses on individual differences and unique behavioral patterns, which may be valuable for understanding personal trajectories or idiosyncratic responses. However, organizational change typically requires a broader understanding of how different groups or subpopulations respond to change. Hence, in addition to examining individual variations, it is important to pinpoint how these variations aggregate into patterns that impact group or organizational outcomes (Bouckenooghe et al., 2023). The person-specific approach may overlook broader organizational patterns and subgroup variations that are critical to understanding how change unfolds across different units of an organization. This limits its utility when the focus is on collective behavior and the identification of subgroup responses to change. While the person-specific approach is beneficial for examining individual-level processes, it may not fully capture the complex interactions that occur across various levels within an organization during change processes. By contrast, the person-centered approach is better suited for detecting subgroups with similar responses to change, offering a more comprehensive view of how different parts of an organization respond to and manage change. In short, this article is focused on introducing the person-centered approach as an alternative to the variable-centered approach. The person-specific approach represents yet another perspective, but whether and how it applies to change-related contexts is beyond its scope. The primary goal of the article is to advocate for the inclusion of person-centered methods in organizational change research to provide richer and more nuanced insights.

3. The case for person-centered methods in organizational change research and its complementary role to variable-centered methods

Variable-centered and person-centered research methods each address different types of research questions (see Table 1). Both approaches offer valuable insights, and an over-reliance on just one type of analytical tool could hinder the field’s progress. As the saying goes, if the only tool you have is a hammer, every problem starts to look like a nail. The variable-centered approach is well-suited to answer content questions focused on relationships between variables. For example, it can address questions like “how do organizational culture subdimensions relate to organizational change responses?” or “how do these change responses predict employee support for change?” Conversely, the person-centered approach is tailored to answer questions focused on patterns and profiles. For instance, it can explore “are there distinct subgroups based on change responses?,” examine “how organizational culture subdimensions predict membership in these profiles?,” or “assess whether these profiles differ in levels of employee support for change?” By combining both approaches, researchers can capture a fuller picture of change-related phenomena, enabling more nuanced and effective strategies for organizational change.

Table 1.

Distinction between variable-centered and person-centered approach in change research.

Characteristics	Variable-centered approach	Person-centered approach
Purpose	Examine variables of interest and their relationships in organizational change context.	Identify and study subgroups in population based on a set of variables.
Sample Research Questions	• How do organizational culture subdimensions relate to organizational change responses? • How do organizational change responses predict employee behavioral support for change? • How do organizational change processes (e.g. communication, participation in decision-making) predict the content and nature of the newly proposed change?	• Are there profiles of subgroups of employees according to their change responses (e.g. resistance to change, excitement about change, cynicism about change)? • How do organizational culture subdimensions predict membership of these change response profiles? • Are there differences among these change response profiles in the level of employee behavioral support for change? • Are there ‘change process’ profiles of units or organizations based on the quality of their organizational change communication and level of employees’ participation in decision-making? • Are there differences among these change process profiles in the content and nature of the newly proposed change? • Are there distinct profiles of internal characteristics (i.e. organizational politics and collective change efficacy) of organizational changeability?
Core Assumption	Population homogeneity regarding variable and outcome relationships.	Even if individuals are unique, there are certain patterns that make them classifiable as similar.
Popular Analysis Techniques	• Linear Regression • Multilevel Analysis • Structural Equation Modeling	• Latent Profile Analyses • Latent Class Analyses • Latent Transition Analysis
Implications to Research and Practices	Extending theory and guiding organizational change practices based on relationships between variables and outcomes. For instance, scholars inform change leaders if participation in change-decision-making results in favorable employee responses.	Extend theory and guide practices via exploring different profiles of subgroups during change. For instance, scholars inform change leaders a set of specific actions that belong to the profiles of a positive change process and if this drives favorable employee change responses.

In the context of organizational change, person-centered methods can uncover distinct profiles of organizations or teams that respond differently to change initiatives. For example, some employees may emerge as “cautious change navigators” who feel hesitant about change and doubt their own abilities but display resilience and security in taking action. Another group, the “change skeptics,” might approach change with reluctance and low confidence, feeling unsafe and showing only moderate resilience. Other profiles might include “bystanders,” who have a slightly positive outlook toward change with moderate self-confidence but experience insecurity and below-average resilience. Meanwhile, “ambivalent change experts” may appear reluctant but confident in their skills, feeling safe and displaying strong resilience. Finally, “unbridled change enthusiasts” could eagerly embrace change despite some self-doubt, feeling psychologically safe but exhibiting slightly below-average resilience. These illustrative profiles demonstrate how attitudes, beliefs, and emotional responses can interact to shape individuals’ engagement with change. By using person-centered analysis, researchers and managers gain a nuanced understanding that allows for tailored strategies suited to each subgroup’s unique characteristics, rather than relying on a one-size-fits-all approach.

By integrating person-centered approaches more frequently, research on organizational change can become more nuanced and effective, recognizing the complexity and variety within organizations (Woo et al., 2018). This shift could greatly enhance our understanding and management of change processes, by focusing on the specific needs and characteristics of different groups within a study, rather than generalizing across a homogeneous population. Notwithstanding its growing presence, there are few seminal studies published in the field of organizational change that employ person-centered research, despite having illustrated their value by employing methods like LPA to delve into employee reactions to various change scenarios. For instance, Edwards and Clinton (2023) utilized LPA to identify distinct profiles—such as “flourishers,” “recoverers,” and “ambivalents”—among employees affected by public sector restructuring and downsizing. This approach clarified the range of psychological reactions to organizational change and informed tailored strategies to support employee resilience.

Similarly, Kanitz et al. (2024) applied LPA to assess reactions to diversity initiatives within organizations, identifying profiles like “excited supporters” and “discontent opponents,” highlighting the need to consider various employee perspectives during policy implementations and cultural shifts. Köhler et al. (2024), in turn, leveraged LPA to identify six profiles—“proactives,” “acceptors,” “opens,” “neutrals,” “reluctants” and “deniers”—to assess organizational readiness for change among employees who are highly affected by the change. Expanding on these applications, Straatmann et al. (2018) integrated LPA with classical variable-centered linear analytics to explore the psychological processes influencing employees’ intentions to support organizational change, using data from two measurements of city council employees undergoing continuous complex changes. This research underscored the importance of configurational patterns, showing how different combinations of change-related attitudes, subjective norms, and perceived behavioral control contribute to strong supportive intentions toward change. These findings not only enhance strategies to foster such intentions, but also demonstrate how person-centered methods like LPA can significantly enrich theory-building in change management and the broader study of behavioral intention formation. Finally, Tapak et al. (2024) utilized LPA to classify employees at a medical university in Iran into distinct learning organization (LO) profiles. Using the Senge Learning Organization Questionnaire (SLOQ), they identified three latent classes: low, moderate, and high LO levels. Most participants were categorized as moderate, highlighting areas for improvement, particularly in “common vision” and “systemic thinking.” This research demonstrated the utility of LPA applied to organizational unit-level analysis by focusing on understanding organizational learning orientations and identifying target areas for development.

Given the principles and assumptions of the person-centered approach, along with insights from related research and the handful of studies that have demonstrated its effectiveness, several new questions can be explored regarding organizational change (see Table 1). For instance, many studies have addressed change implementation (see Armenakis and Bedeian, 1999; Bouckenooghe, 2010), while noting problems with and pessimism about change efforts. In particular, the optimal routes to successful change implementations remain unclear, as are the precise factors that facilitate, hinder, or contribute to far-reaching change that can benefit the organization. Regardless, previous research has identified that such positive organizational change outcomes depend critically on the presence of organizational changeability, which requires a firm’s flexibility and capacity for change (Reinhart and Grunwald, 2001; Ross et al., 2008). Thus, as part of broadening organizational change literature, research on change implementation may benefit from adopting more person-centered research.

Building on this foundation, the remainder of this article provides a nontechnical, step-by-step guide to applying person-centered methods in organizational change research, focusing on uncovering different configurations that either enable or hinder changeability. This approach draws on insights from previous research on organizational responses to change, which highlight various components and processes contributing to positive change outcomes or intentions (e.g. Buono and Kerber, 2010; Judge and Douglas, 2009). Mainstream research has traditionally focused on mapping sample-wide linear relationships, which often oversimplify the complex interactions and interdependencies inherent in organizational responses to change. Recognizing this limitation, there has been a growing call for more heterogeneous, configurational, and holistic descriptions of reactions to change that better capture their multidimensional and complex nature (Bouckenooghe, 2010; Michard and Bouchaud, 2005; Oreg et al., 2018). Addressing this challenge, and acknowledging the promise of person-centered methods in advancing organizational change research, we elaborate on LPA as one of the most relevant statistical techniques for such studies. Through this method we aim to illustrate how person-centered approaches can enrich understanding of organizational change by identifying nuanced profiles in responses to change initiatives.

4. LPA: a flagship tool of person-centered data analysis

Our focus on offering a primer on LPA stems from LPA’s established status as a flagship model within the person-centered approach (see, Howard and Hoffman, 2018; Meyer and Morin, 2016; Morin et al., 2011; Spurk et al., 2020; Vermunt and Magidson, 2002). LPA is central in the field due to its ability to uncover distinct, meaningful profiles within complex data, which aligns well with the configurational nature of organizational change data (Bouckenooghe et al., 2023). In addition, many other analytical techniques, such as Latent Transition Analysis, build on LPA’s foundational principles, making it essential for researchers to have a solid understanding of this method to explore more advanced extensions (see Nylund et al., 2007). Given these observations, focusing on LPA as person-centered data analysis is suitable to address the intricate dynamics in change management research.

LPA is an example of latent variable mixture models with the goal of identifying subpopulations within a population based on a certain set of variables (Collins and Lanza, 2009). With LPA, cases or individuals are assigned varying probabilities to different subpopulations based on configural profiles, this method is characterized by structural parsimony, inferring a limited number of profiles within the data, based on shared attributes. For instance, in examining how employees adapt to organizational changes, as we noted above, LPA might uncover different profiles (e.g. navigators, skeptics) based on variables like positive attitude to change, change self-efficacy, resilience and stress psychological safety. This difference shows how LPA can simplify the complexity of change management by categorizing employees according to their response patterns, thereby aiding targeted intervention strategies. Based on profiles, this grouping reflects the natural features of how cognition represents structures underlying data (Macrae and Bodenhausen, 2000). In this context, classifying based on LPA is both conceptually and methodologically meaningful for developing configurations (Costa et al., 2002).

Through means of an example, we offer a step-by-step approach for conducting LPA, structured into three main phases, each with multiple subphases (see Figure 1). The first phase, “Research and Design,” focuses on determining how a person-centered approach and LPA can effectively address the research question. Selecting high-quality variables and indicators for the study is crucial in this phase. The second phase, “Statistical Consideration and Data Management,” deals with key issues such as ensuring the sample size is adequate to discern the true number of profiles, managing missing data, choosing the appropriate model estimator based on data distribution, and addressing local maxima in the data. The final phase, “Model Selection and Data Interpretation,” involves assessing the robustness of the models, choosing the right fit statistics to identify the best-fitting model, and verifying the validity of the findings. Rigorous attention to these phases and their questions is essential for successful application.

Figure 1.

Flowchart guide for conducting LPA.

5. Application of LPA: illustrative study

5.1. Background

As part of this article’s broader push to develop curated thinking on the person-centered approach to applied organizational change, we illustrate best practice in applying a profile-based approach to organizational changeability as our illustrative study.

Changeability is core to the strategic success of many organizations and refers to the combined forces of organizational flexibility and capacity in generating change (Bennebroek Gravenhorst et al., 2003). Despite the importance of collective dynamics in shaping organizational change outcomes, there is a significant lack of empirical research that examines how different configurations of these dynamics affect an organization’s adaptability and readiness for change. Exploring the interplay between organizational politics and collective change efficacy fills a critical research gap by deepening our understanding of how collective dynamics influence organizational change. Current literature often overlooks these collective dynamics, despite their significant impact on change outcomes (Schwarz et al., 2021). Identifying and understanding the complex interactions between internal organizational politics and collective capabilities, organizations can better tailor their change management strategies. This approach promises both to improve the success rates of organizational transformations and also adds a valuable dimension to theoretical frameworks in organizational change research. Addressing the question “whether there are distinct profiles of internal characteristics (i.e. organizational politics and collective change efficacy) for organizational changeability?,” helps us to illustrate the application of LPA as a person-centered approach in organizational change context.

Organizational politics are the informal, unofficial, and behind-the-scenes efforts to sell ideas, gain power, or achieve other goals within an organization (Brandon and Seldman, 2004; Hochwarter et al., 2000). Rooted in the dynamics of power and control (Clegg et al., 2006; Pfeffer, 1981), these political activities are pivotal in shaping intra-organizational dynamics and are especially relevant in the context of organizational change (Pettigrew, 1975; Tichy, 1983). Organizational politics can either facilitate or hinder change, depending on how they interact with an organization’s collective change efficacy—the shared belief among members about their collective capability to implement change effectively (Chen and Bliese, 2002; Shamir et al., 2000). For our example, we differentiated between four key elements of organizational politics: (1) politicking, (2) distrust in leadership, (3) limited involvement in decision-making, and (4) lack of cohesion.

Politicking refers to the extent to which organizational members engage in political activities as means of actively exercising influence, with the goal of shaping and changing organizational outcomes to achieve personal interests, and securing highly valued resources in the process (March and Olsen, 1983; Milgrom and Roberts, 1988). Second, a lack of trust in leadership (Parker et al., 1995; Poon, 2003) prompts responses that include parochial, self-serving political behavior, because the intersection of trust, leadership, and change largely determines whether organizational members perceive the change positively or negatively, and thus may resist it (as Lines et al., 2005 show). Third, limited involvement in change decisions imposes a lack of voice, leading employees to engage in self-centered and protective behaviors (e.g. remain silent), in a tacit attempt to achieve beneficial outcomes for themselves (Byrne, 2005). By not voicing their concerns, members subject to the change might avoid conflict and therefore appear to engage in nonpolitical behavior, but this response may also reflect the self-interested, protective behavior typical of organizational politics (Drory and Romm, 1990). Fourth, the lack of cohesion and poor quality of interactions with supervisors and colleagues in an organization prevents employees from sharing crucial knowledge, especially if it might be used against them or prevent them from achieving self-interests. That is, self-interested political goals push employees to adopt competitive strategies to claim value, which may undermine the cooperative strategies needed to create value (Parker et al., 1995; Witt, 1998). Together, these four elements depict organizational politics as an interpretive response, reflecting employees’ appraisal of their willingness to embrace change.

Collective change efficacy is a key internal resource that can help organizational members deal with the demands of the political character of change. This future-oriented judgment pertains to capabilities to undertake a course of action to produce an outcome (Chen and Bliese, 2002). As a coordinated sensemaking response, collective change efficacy comprises three pertinent dimensions: (1) ability to lead the change; (2) attitude of management toward change; (3) involvement in the change process. Ability to lead the change in the group entails the perceived managerial abilities to guide, coach, and motivate group members through the change (Anderson and Anderson, 2010). When a shared attitude of management toward change exists, the collective can adopt the stance that management takes (Judge et al., 1999), which should relate positively to change reactions and openness to change. Finally, involvement in the change process provides a sense of common ownership, due to active and dynamic participation in the change (Morgan and Zeffane, 2003), which emphasizes open sharing and support for the change. These elements enhance the collective’s capacity to handle change by fostering a sense of shared ownership and control over the change process, thus increasing the likelihood of successful change implementation.

Overall, the complex combination of the different dimensions that make up organizational politics and collective efficacy shape organizational changeability (see Figure 2).

Figure 2.

Conceptual model illustrative example.

5.2. Research and design

5.2.1. Data

The data for this illustrative study came from 87 companies in both private (51) and public (36) sectors. These organizations were part of a 6-month change management program at a business school in Europe; only companies that had announced some large-scale change within 6 months of enrolling were included; they all undertook an organizational transformation in complex, unstable environments (Mohrman et al., 1989). With this sample, we reduced the potential variability that can result from collecting data from companies at different phases in their change processes. A representative of each company (generally, the person managing the change program) distributed questionnaires to employees of the work units affected most by the change. To minimize variability, we asked each representative to remind participants to think about the specific change initiative as they completed the questionnaire. A cover letter also explained the purpose of this project and confirmed the voluntary, confidential nature of their participation.

5.2.2. Measures

The core concepts in this illustrative application of LPA are based on measures borrowed from the OCQ-CPR framework by Bouckenooghe et al. (2009) allowing for the aggregation of individual perceptions into organizational-level metrics, supported by interrater agreement indices (Rwg(J), ICC(1), and ICC(2)), where the mean Rwg(J) and ICC(2) values exceed .70. Organizational politics is measured through four dimensions: politicking, lack of trust in leadership, lack of voice, and lack of cohesion. Politicking assesses the role of power games (α = .68), lack of trust in leadership uses items such as the fulfillment of promises by the management team (α = .79), lack of voice captures the employees’ ability to raise discussions (α = .79), and lack of cohesion examines openness within the organization (α = .76). A confirmatory factor analysis supports the validity of a four-factor model of organizational politics (CFI = .96, GFI = .97, RMSEA = .05). Collective change efficacy involves three dimensions: involvement in the change process, which examines the consultation of employees about change reasons (α = .86); ability of management to lead change, assessing managerial coaching effectiveness (α = .82); and attitude of management toward change, gauging unconditional support for change processes by the management team (α = .72). These scales demonstrate strong factorial validity and internal consistency. Organizational changeability is measured using a nine-item scale evaluating members’ willingness to invest effort in change processes, showing high internal consistency (α = .84) and excellent model fit (CFI = .99, GFI = .99, RMSEA = .05). These comprehensive measures provide a robust framework for understanding the interplay between internal organizational dynamics and change outcomes.

5.3. Statistical considerations and data management

This article focuses on providing guidance on how to effectively conduct LPA, although despite this focus, we want to emphasize its role as a theory-informed tool. LPA is designed to uncover patterns in the data that align with or extend existing theoretical frameworks, rather than allowing statistical outcomes to dictate interpretation. Its purpose is to generate insights that refine or expand our understanding of complex phenomena. When data reveal novel profiles or configurations, these should be critically evaluated within a theoretical context, fostering abductive theory development—a process that combines inductive observation with deductive reasoning—to advance both theory and practice in organizational change research.

Keeping this in mind, when conducting LPA, there are a number of important steps in terms of statistical considerations and data management; (1) decision on sample size; (2) the selection of estimator, (3) the treatment of missing data, and (4) dealing with local solutions. We will discuss each step in relation to our illustrative example on changeability, and use these steps as a guide for extending this approach for organizational change research. There have been several programs available for estimating LPA models, such as Mplus (Muthén and Muthén, 2017), Latent GOLD (Vermunt and Magidson, 2013), and some R packages (e.g. MCLUST; Fraley et al., 2022).¹

5.3.1. Decision on sample size

Sample size considerations for effective LPA are emphasized in the studies by Sinha et al. (2021) and Tein et al. (2013), which highlight the importance of sufficiently large samples for reliable statistical analysis. Advocating for a minimum sample size typically above 500 to ensure consistent and reliable LPA results, Sinha et al. (2021) note that high-quality indicators may permit smaller sample sizes when classes are larger and more distinct. This is consistent with Tein and colleagues’ findings, which demonstrate that larger sample sizes are crucial when class separations are less distinct, with a recommended sample size of 500 providing robust statistical power across various scenarios. Although the raw data in our example include over 500 individual observations, our analysis focuses on the aggregate level (organization), which means our sample size is actually considered small. However, in our defense, despite this small size, we employed high-quality indicators to determine latent profile solutions, as evidenced by the high entropy values (greater than .90).

Given these recommended sample size thresholds we acknowledge that LPA may seem limited to large-scale research. However, LPA can still be effectively applied in smaller sample contexts when the model is appropriately specified. While larger samples (e.g. 500 or more) are ideal, reliable results can still be achieved with smaller samples (e.g. 200–300), particularly when profiles are well-separated or the number of indicators and latent profiles is limited. Furthermore, strategies such as model simplification, leveraging Bayesian estimation methods, or integrating prior data can help address power issues and broaden the applicability of LPA in organizational change research. This ensures that LPA remains a viable tool even in contexts with moderate sample sizes.

5.3.2. Selection of estimator

As with any statistical method, it is essential to ensure that underlying assumptions about the data structure are met before proceeding with analysis to ensure the validity and reliability of the results. Choosing the statistical estimator for LPA should be done based on a closer examination of the data structure within a sample. The default is Maximum Likelihood estimation when data are not showing non-normal distributions. However, if multivariate normality cannot be established, Maximum Likelihood with Robust Standard Errors is a better alternative. Further, extreme outliers can also bias the estimation of final profile solutions (Vermunt and Magidson, 2002). In this context, it has been suggested that an exclusion of extreme outlier cases in the data could offer a solution (Hirschi and Valero, 2017). In our illustrative example, we checked our data and observed normal distributions. Furthermore, running a Mahalanobis distance measure test for multivariate outliers, none of the data seemed problematic.

5.3.3. Treatment of missing data

LPA can be conducted with missing data, but the extent and pattern of missingness complicate the estimation process. Swanson et al. (2012) found that the accuracy of Information Criteria deteriorated when data were not missing at random, and model fit worsened as missing data increased from 10% to 20%, particularly in smaller samples. Models with 20% missing data required roughly a 50% increase in sample size to maintain accuracy (Wolf et al., 2013). Because we have no missing data in our example, we will not offer a detailed overview of missing data treatment, yet we would like to highlight that there are three common methods to handle missing data in LPA include deletion, multiple imputation (MI), and full information maximum likelihood (FIML). Deletion is generally discouraged due to significant data loss. MI generates several datasets with varied imputations of the missing values, useful across different models but complex when handling mixed data types. FIML, preferred for its efficiency, uses all available data to estimate model parameters, assuming data are missing at random. Each method has its advantages and limitations, impacting the choice based on specific conditions of the study’s missing data. For more detailed information on this we refer to the vast literature on the topic (Lanza and Cooper, 2016; Little and Rubin, 2014; Schafer and Graham, 2002).

5.3.4. Local maxima and global solutions

In LPA, finding the global maximum log-likelihood value is crucial as it indicates the best-fitting model. Local maxima can mislead the analysis, suggesting inferior solutions as optimal. To counter this, researchers are advised to use multiple starting values to avoid local maxima and ensure the identification of the global solution. This involves replicating the best log-likelihood value across several solutions to confirm stability (Berlin et al., 2014; Vermunt and Magidson, 2002). In practice, many studies employ extensive random starts (up to 7000) and iterations (up to 300 per start) to refine their solutions, following guidelines suggested by Hipp and Bauer (2006). If the best log-likelihood value cannot be replicated, it’s useful to compare parameter estimates from multiple optimized solutions to assess stability. This involves using a consistent random seed for reproducibility, as enabled by options like Mplus’ OPTSEED. In our example we changed the default settings in Mplus to 100 random starts and a final stage optimization of 20. In doing so, the different profile solutions that were produced did not appear to suffer from local maximization.

5.4. Selecting models and data interpretation

The next set of important steps in the LPA research process are (1) the selection of the best-fitting profile solution, (2) interpretation of data, and (3) validation of profiles.

5.4.1. Selecting the number of profiles

Following procedures described by Nylund et al. (2007) to determine the number of profiles to be retained, an iterative process that starts with an estimation of a two-profile model is recommended, and then adds successive profiles until the most optimal solution is found. A series of statistical tests and indices are used to confirm the statistical adequacy of the extracted classes (Meyer et al., 2012; Muthén, 2002). First, lower values for the distribution-free information criteria including the Akaike Information Criterion (AIC; Akaike, 1987) and Bayesian Information Criteria (BIC; Schwarz, 1978) signal a better fitting solution. Second, to decide on the number of profiles to be retained a second useful index to be considered is the sample-adjusted BIC (SABIC) (Sclove, 1987). SABIC helps to select the model with the best fit and fewest parameters from among a set of non-hierarchical models. Third, the bootstrapped likelihood ratio test (BLRT) (McLachlan and Peel, 2000) is used to assess the extent to which a model with k-profiles provides a better fit than a model with k – 1 profiles. Significant p-values for the BLRT indicate that the k – 1 model should be rejected in favor of the k model. Notably, SABIC and BLRT are particularly effective with relatively small samples (Nylund et al., 2007). Fourth, in addition to evaluating fit indices, it is recommended to assess the posterior probabilities associated with each profile, to determine if strong probabilities exist for participants to belong to the profile to which they were assigned and low probabilities of belonging to other profiles. Fifth, according to Pastor et al. (2007), if a profile solution has one or more profiles representing less than 5% of all observations, it should not be retained. This set of criteria together should help to determine how many profiles should be selected.

For our example, Table 2 presents the fit statistics for the possible latent profile structures; the four-profile solution exhibits lower log-likelihood, AIC, BIC, and SABIC values and a significant BLRT compared with the two- or three-profile solutions. The five- and six-profile solutions achieve even lower values, but one or more classes in these solutions included fewer than 5% of the observations, so we retain the four-profile solution. Even though we decided to retain 4 profiles in our example, it is important to note that the number of selected profiles align with the theoretically most meaningful solution and give priority to the theoretically best-fitting solution (if the fit values allow for such). Put simply, theoretical decisions take priority over fit values on the condition that different fit indices allow for different final solutions.

Table 2.

Model fit statistics.

K	LL	AIC	BIC	SABIC	Classes with fewer than 5% of observations	BLRT
2	−824.33	1692.67	1747.18	1677.75	None	83.15
3	−801.99	1663.97	1738.29	1643.62	None	44.70
4	−783.64	1643.27	1737.41	1617.50	None	36.70
5	−761.95	1615.90	1729.86	1584.70	Class 1 includes fewer than 5% the observations	43.37
6	−748.48	1604.95	1738.73	1568.33	Classes 1 and 2 have fewer than 5% of observations	39.90

LL: log-likelihood, AIC: Akaike information criterion, (SA)BIC: (Sample-adjusted), Bayesian information.

5.4.2. Interpreting profile solution

Once the number of retained profiles has been decided, the final step in this process involves the interpretation and labeling of the profiles. A closer examination of the shape and level of scores on the subdimensions of different clusters helps to capture the essence of the respective profiles and helps with the assignment of labels. There are no clear guidelines in terms of assigning labels, however, the decision of labeling usually strongly depends on the investigated topic, the complexity of the extracted profiles, and the terms and labels used in the respective research fields (Spurk et al., 2020). A frequently used method to help with interpretation of profiles is the plotting and comparison of scores of the indicators/subdimensions that make up the profiles.

In our example of a four-profile solution of organizational changeability, we plot standard scores for the four organizational politics elements and the three collective change efficacy dimensions across these four profiles. Figure 3 reveals the patterns of the means. With an analysis of variance, we identify some significant differences across profile groups: politicking (F(3,84) = 26.83, p < .001), lack of trust in leadership (F(3,84) = 48.40, p < .001), lack of voice (F(3,84) = 26.20, p < .001), lack of cohesion (F(3,84) = 5.50, p < .01), involvement in the change process (F(3,84) = 23.66, p < .001), ability of management to lead change (F(3,84) = 24.42, p < .001), and attitude of management toward change (F(3,84) = 12.37, p < .001).

Figure 3.

Profiles based on politicized work context and collective change efficacy.

Using the plotted graph in Figure 3 and the information in Table 3, we labeled the profiles or configurations that emerged from our analyses, numbered in the order that they were generated. Configuration 1 involves a highly political workplace, with a high score on politicking and a poor climate of trust in leadership, lack of voice, and poor cohesion. In terms of collective change efficacy, perceived control over change is limited, with the lowest score being for involvement in change processes and the second lowest scores for attitude of management toward change and ability of management to lead change. This first configuration of companies thus includes highly politicized organizations with weak collective change efficacy. Configuration 2 is characterized by firm beliefs in the ability to manage change: high scores on involvement in the change process, attitude of management toward change, and ability of management to lead change. Yet politicking also is relatively pronounced (second highest score), with low scores for climate of trust in leadership, voice, and cohesion (but not as low as in configuration 1). Accordingly, we refer to this configuration as politicized organizations with strong collective change efficacy. Configuration 3 is almost the opposite of configuration 1, with weak politicking and the highest scores for cohesion, voice, and trust in leadership. Furthermore, configuration 3 is characterized by collective change efficacy (second highest scores for all three dimensions). From these mean scores, we derive the label apolitical organizations with moderately strong collective change efficacy. Finally, Configuration 4 indicates a moderate level of politicking and moderately high scores for cohesion, voice, and trust in leadership. Even if the work context is relatively apolitical, weak collective change efficacy also exists, with the lowest scores for ability of management to lead change and attitude of management toward change. We thus label Configuration 4 moderately politicized organizations with weak collective change efficacy.

Table 3.

Mean component scores for profiles.

	TRUST	COH	VOI	POL	INV	ABIL	ATC
Configuration 1 (Highly politicized organizations with weak collective change efficacy)	−0.854	−0.474	−0.920	0.741	−0.457	−0.368	−0.155
Configuration 2 (Politicized organizations with strong collective change efficacy)	−0.557	−0.438	−0.270	0.514	1.257	1.238	1.030
Configuration 3 (Apolitical organizations with moderately strong change efficacy)	1.508	0.650	1.139	−0.694	0.694	0.725	0.623
Configuration 4 (Moderate politicized organizations with weak collective change efficacy)	0.410	0.317	0.425	−0.318	−0.336	−0.392	−0.425

The values are standardized mean scores (M = 0 and SD = 1). TRUST: Lack of trust in leadership, COH: lack of cohesion, POL: Politicking, INV: Involvement in change process, ABIL: Ability of management to lead change, and ATC: Attitude of management toward change.

5.4.3. Validation of profiles

When designing studies that use LPA, it’s crucial to plan how to validate the profiles that are derived. There are various strategies for this validation process. One common approach is to establish criterion-related validity by testing for differences in means among the profile groups concerning outcomes that are theoretically relevant (Spurk et al., 2020). Another crucial method for validating the final profile solution in LPA studies is to assess its consistency across various samples, contexts, and time periods (Hirschi and Valero, 2017; Woo et al., 2018). Criterion-related validity in this study was assessed by comparing our profile solution against measures of organizational changeability, confirming the relevance and distinctiveness of the identified configurations. We applied General Linear Models (GLM),² using configuration membership (a combination of organizational politics and collective change efficacy) as the independent variable, while controlling for whether the organizations were in the public or private sector. This distinction is relevant as private organizations typically exhibit more entrepreneurial behavior and openness to change than public ones. Our analysis indicated significant differences in organizational changeability between the private and public sectors, with private organizations showing stronger changeability (M = .48) compared to public organizations (M = -.09). There were also significant variations in changeability across the four profiles, with configurations 2 (M = .95) and 3 (M = .41) showing the highest levels, and configurations 1 (M = -.30) and 4 (M = -.26) showing the lowest. These findings demonstrate that organizations with high collective change efficacy are more adaptable, particularly in politicized environments, whereas those with low collective change efficacy are less changeable (see Table 4).

Table 4.

Means for organizational changeability.

	Configuration 1	Configuration 2	Configuration 3	Configuration 4
Organizational changeability	M = −.30 SE = .15	M = .95 SE = .22	M = .41 SE = .38	M = −.26 SE = .12
Pairwise comparison with configuration 1		−1.13 (p < .001)	−.89 (p < .05)	−.15 (ns)
Pairwise comparison with configuration 2			.24 (ns)	.99 (p < .001)
Pairwise comparison with configuration 3				.74 (p < .05)

Taken together, this illustrative example provides a primer of the crucial steps to be performed and reported when considering the use of LPA on organizational change data.

6. Discussion

This article provides a step-by-step guide for the person-centered approach, specifically, LPA and its add-on value to research on organizational change. With this approach, we highlight its potential application in how it can help to answer seminal questions that remain unaddressed by traditional variable-centered research in organizational change literature (refer to Table 1 for overview). That is, questions related to LPA testing models that involve (1) patterns in response to change, commitment to change, or changeability profiles; (2) movement/shifts in response to change profiles over time (e.g. emotions of change recipients may vary in terms of intensity and valence as organizational change unfolds); (3) complex ways in which psychosocial factors shape behavioral reactions to change and impact wellbeing (e.g. for whom do psychosocial factors shape resistance to change or readiness to change, rather than assuming that for everyone in the population psychosocial factors will predict readiness for change); (4) identifying similarities or differences in response patterns across diverse change initiatives, such as whether groups respond consistently across different types of change or if significant variations emerge based on the nature of the change.

Overall, adopting a profile-based approach in organizational change context, implies that researchers are interested in unraveling complex interactions between a series of variables, which usually would go undetected when using variable-centered approaches. For example, traditional variable-centered methods allow studying the interplay of different commitment to change variables (i.e. affective, continuance and normative) through the inclusion of interaction terms, yet this approach becomes impractical when the number of interacting variables increases (Gellatly et al., 2014). Because of the awareness that a single set of parameter estimates cannot be assumed to always hold for an entire population, and because of the limitations of variable-centered methods to capture complex patterns of interactions, person-centered methods (LPA, latent transition analysis, regression mixture models) present a viable supplement to variable-centered research.

Besides offering insights into why person-centered research can add value to organizational change literature, by proposing a simple nontechnical guide on how to perform LPA, we provide a guide on essential steps to consider when performing LPA. These steps include: (1) Research and Design Question Phase (2) Statistical Consideration and Data Management, (3) Selecting Models and Data Interpretation. In offering this review, and its suggestions for approaching change, we hope to encourage researchers to consider person-centered applications of organizational change and development in their future research. With the results reported in this article, we specifically recommend applying a person-centered approach in continued research on collective change responses. The distinct profiles we identify help clarify resource (i.e. collective change efficacy) and demand (i.e. organizational politics) factors that may determine change responses (i.e. changeability), as well as the actual changes that may result, both burgeoning areas of research on change. Furthermore, a person-centered approach may delineate organizational change readiness options, due to its basis in equifinality (see Katz and Kahn, 1978), which assumes the existence of multiple profiles, rather than a single, optimal, measurable outcome (Fiss, 2007).

6.1. Person-centered approach and theory development

From a more general perspective, person-centered methods are particularly interesting if theoretical development is limited or non-existent regarding a certain topic, because the methods offer a way to discover new aspects of phenomena through theoretical induction (Ketokivi and Mantere, 2010). Relying on theory-based induction, subpopulations are not assumed just to be the result of a particular sample setting. Instead, a theoretical understanding of the subject matter is the focus of person-centered research. More specifically, the nature of the person-centered approach, and its ability to detect subpopulations within a larger population, allows for theory-building in change. Thus, the person-centered approach has already shown itself to be useful in two ways.

First, in the context of developmental systems theory (DST) which states that development entails simultaneous influences, reflecting a constructive process with nonlinear dynamics (Molenaar and Nesselroade, 2015). As DST seeks to explain development, it also accounts for the mechanisms (i.e. emotional and cognitive) that engender change (Ford and Lerner, 1992). Doing so, this theory responds to calls to develop more process-oriented theories in the context of organizational change (Langley et al., 2013). Advocates of DST claim that person-centered analysis may be ideal to test the complex interactions between cognitive and emotional mechanisms shaping development and change.

Second, in the application of theory of planned behavior (TPB) as an integrative framework for explaining change reactions. As Straatmann et al. (2018) indicate, TPB determinants (i.e. change-related attitude, change-related perceived behavioral control, change-related subjective norms) generally influence each other’s effects or must even co-exist for high change-supportive intentions to occur—an issue that gets lost in traditional variable-centered methods. With this gap in mind, and through our illustrative example, we emphasize the importance of considering the interdependencies of change-related variables. Doing so, we caution change researchers and practitioners against focusing only on one factor for interventions aimed at change support (e.g. usually choose the factor with the strongest beta weight in regression analysis), as this neglects important interdependencies and may not result in the expected impact.

6.2. LPA and future research directions

LPA helps to offer a more detailed understanding of employee reactions to change initiatives by categorizing them into profiles that reflect varying levels of engagement, resistance, or acceptance of change. The recent development of a circumplex model for measuring responses to change (Oreg et al., 2024) provides an opportunity for LPA to delve into the relationships between these response types and their impact on change outcomes, identifying four response types: change acceptance (positive and passive), change proactivity (positive and active), change disengagement (negative and passive), and change resistance (negative and active). This framework aligns with the person-centered approach, indicating that LPA could be valuable in identifying distinct patterns within these response types. By applying LPA, future research could uncover clusters of employees with varying response patterns, leading to deeper insights into the complexity of change reactions.

Person-centered methods can also shed light on the context, process, and content of change (Pettigrew, 1987). By applying LPA to explore these elements, researchers can refine theoretical models and improve the practical application of change management strategies (Bouckenooghe et al., 2021). Understanding these organizational change configurations allows practitioners to design targeted interventions to meet the specific needs of each subgroup or configuration, enhancing the effectiveness of change strategies (Kanitz et al., 2024). Ultimately, the person-centered approach and LPA offer researchers and practitioners a robust framework for understanding and managing organizational change. By uncovering distinct response profiles and exploring their implications, organizations can develop more effective and adaptable strategies for managing transformations and fostering a positive workplace culture.

While several of the references and examples used in this article focus on individual-level profiles (e.g. employee’s change adoption profiles), LPA can also be applied at the organizational level to explore broader outcomes, such as profitability, innovation, or other indicators of organizational change effects. For instance, whether distinct organizational readiness profiles emerge across different industries or change initiatives. Considering the multilevel nature of change-related data, multilevel LPA would allow researchers to explore how micro-level dynamics (e.g. individual attitudes or behaviors) interact with macro-level outcomes, offering a richer understanding of the complexities of organizational change (Mäkikangas et al., 2018). Thus, adopting a multilevel perspective would broaden the scope of LPA, making it even a more versatile tool for studying change across individual, team, and organizational levels.

6.3. Potential pitfalls with person-centered research

With a focus on the benefits of using person-centered research, and given our interest in encouraging the use of this approach in organizational change, we have discussed in detail LPA. While noting the advantages that come with using LPA, we also acknowledge it involves limitations. First, LPA is sensitive to model specification, relying on researchers to select the number of profiles to extract, a decision that can be subjective and impact the interpretation of results (Nylund et al., 2007). Moreover, the optimization process in LPA may converge to local rather than global optima, leading to suboptimal solutions (Celeux and Soromenho, 1996). Sample size requirements present another challenge, as LPA typically necessitates a large sample size for accurate parameter estimation and profile identification (Nylund et al., 2007). Identification problems, such as non-identification and local identifiability, further complicate model estimation (Collins and Lanza, 2009). Interpretation of profiles can be challenging due to their complexity or overlap, requiring careful examination and validation (Nagin, 2005). Finally, generalizability of LPA findings to other populations or contexts may be limited, necessitating robustness checks across diverse samples and settings (Vermunt and Magidson, 2002). As outlined in this guide, addressing these limitations requires careful consideration of model assumptions, rigorous methodological approaches, and thorough interpretation of results in the context of the research questions and data characteristics.

Key Research and Practical Implications

This article familiarizes organizational change scholars with person-centered research approach, particularly by highlighting latent profile analysis as a key technique. It offers a detailed guide outlining the steps necessary for successfully applying this method to change-related data. Person-centered research methods allow scholars to address key questions that are difficult to capture using variable-centered approaches alone. In essence, the article is designed to be a low-entry primer, making latent profile analysis more accessible to organizational change researchers.

This article can help organizational practitioners to develop targeted change interventions that meet the particular requirements of distinct subgroups or configurations of employees, thereby increasing the performance of these interventions. Through the person-centered approach, change practitioners are provided with a robust tool through which they can more effectively understand and implement organizational change initiatives. By identifying different response profiles and understanding their implications, they can design more successful and flexible strategies for managing organization-wide transformations and contributing to better work climates.

Footnotes

Acknowledgements

We would like to express our sincere gratitude to the editorial team and the anonymous reviewers for their constructive feedback and thoughtful suggestions throughout the review process. Their insights significantly improved the clarity, rigor, and overall quality of the manuscript. We are especially thankful for their time and effort in helping us make this work publishable.

Final transcript accepted 10 February 2025 by Muhammad Ali (Deputy Editor).

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Dave Bouckenooghe

Gavin M Schwarz

Phong Thanh Nguyen

Dirk De Clercq

Notes

References

Aguinis

Edwards

(2014) Methodological wishes for the next decade and how to make wishes come true. Journal of Management Studies 51: 143–174.

Akaike

(1987) Factor analysis and AIC. Psychometrika 52: 317–332.

Anderson

(2010) Beyond Change Management: How to Achieve Breakthrough Results Through Conscious Change Leadership. San Francisco, CA: Wiley.

Anderson

Lenz

(2001) Modeling the impact of organizational change: A Bayesian network approach. Organizational Research Methods 4: 112–130.

Armenakis

Bedeian

(1999) Organizational change: A review of theory and research in the 1990s. Journal of Management 25: 293–315.

Bartunek

Moch

(1987) First-order, second-order, and third-order change and organization development interventions: A cognitive approach. Journal of Applied Behavioral Science 23: 483–500.

Bennebroek Gravenhorst

Werkman

Boonstra

(2003) The change capacity of organisations: General assessment and five configurations. Applied Psychology 52: 83–105.

Bergman

Magnusson

(1997) A person-oriented approach in research on developmental psychopathology. Development and Psychopathology 9: 291–319.

Berlin

Williams

Parra

(2014) An introduction to latent variable mixture modeling (part 1): Overview and cross-sectional latent class and latent profile analyses. Journal of Pediatric Psychology 39: 174–187.

10.

Bouckenooghe

(2010) Positioning change recipients’ attitudes toward change in the organizational change literature. Journal of Applied Behavioral Science 46: 500–531.

11.

Bouckenooghe

De Clercq

Raja

(2019) A person-centered, latent profile analysis of psychological capital. Australian Journal of Management 44: 91–108.

12.

Bouckenooghe

Devos

Van den Broeck

(2009) Organizational change questionnaire–climate of change, processes, and readiness: Development of a new instrument. The Journal of Psychology 143: 559–599.

13.

Bouckenooghe

Schwarz

Kanar

, et al (2021) Revisiting research on attitudes toward organizational change: Bibliometric analysis and content facet analysis. Journal of Business Research 135: 137–148.

14.

Bouckenooghe

Schwarz

Sanders

, et al (2023) The multiple faces of collective responses to organizational change: Taking stock and moving forward. Journal of Organizational Behavior 44: 997–1014.

15.

Brandon

Seldman

(2004) Survival of the Savvy: High-Integrity Political Tactics for Career and Company Success. New York: Simon & Schuster.

16.

Buono

Kerber

(2010) Creating a sustainable approach to change: Building organizational change capacity. SAM Advanced Management Journal 75: 4–14.

17.

Byrne

(2005) Social Exclusion. Maidenhead: McGraw Hill Education.

18.

Celeux

Soromenho

(1996) An entropy criterion for assessing the number of clusters in a mixture model. Journal of Classification 13: 195–212.

19.

Chen

Bliese

(2002) The role of different levels of leadership in predicting self-and collective efficacy: Evidence for discontinuity. Journal of Applied Psychology 87: 549–556.

20.

Chou

Sibley

Liu

, et al (2015) Paternalistic leadership profiles: A person-centered approach. Group & Organization Management 40: 685–710.

21.

Clegg

Courpasson

Phillips

(2006) Power and Organizations. Thousand Oaks, CA: Sage.

22.

Collins

Lanza

(2009) Latent Class and Latent Transition Analysis: With Applications in the Social, Behavioral, and Health Sciences. Hoboken, NJ: Wiley.

23.

Costa

Jr Herbst

McCrae

, et al (2002) The replicability and utility of three personality types. European Journal of Personality 16: S73–S87.

24.

Drory

Romm

(1990) The definition of organizational politics: A review. Human Relations 43: 1133–1154.

25.

Edwards

Clinton

(2023) Profiling employee psychological responses during restructuring and downsizing in the public sector: Flourishers, recovers and ambivalents. Personnel Review 52: 1916–1935.

26.

Fiss

(2007) A set-theoretic approach to organizational configurations. Academy of Management Review 32: 1180–1198.

27.

Ford

Lerner

(1992) Developmental Systems Theory: An Integrative Approach. Thousand Oaks, CA: Sage.

28.

Fraley

Raftery

Scrucca

(2022) mclust: Gaussian mixture modelling for model-based clustering, classification, and density estimation. Available at: https://CRAN.R-project.org/package=mclust

29.

Gabriel

Daniels

Diefendorff

, et al (2015) Emotional labor actors: A latent profile analysis of emotional labor strategies. Journal of Applied Psychology 100: 863–879.

30.

Gellatly

Cowden

Cummings

(2014) Staff nurse commitment, work relationships, and turnover intentions: A latent profile analysis. Nursing Research 63: 170–181.

31.

Hagl

Kanitz

Gonzalez

, et al (2024) Change management interventions: Taking stock and moving forward. Human Resource Management Review 34: 101000.

32.

Harley

Cornelissen

(2022) Rigor with or without templates? The pursuit of methodological rigor in qualitative research. Organizational Research Methods 25: 239–261.

33.

Hipp

Bauer

(2006) Local solutions in the estimation of growth mixture models. Psychological Methods 11: 36–53.

34.

Hirschi

Valero

(2017) Chance events and career decidedness: Latent profiles in relation to work motivation. The Career Development Quarterly 65: 2–15.

35.

Hochwarter

Witt

Kacmar

(2000) Perceptions of organizational politics as a moderator of the relationship between consciousness and job performance. Journal of Applied Psychology 85: 472.

36.

Hofmans

Wille

Schreurs

(2020) Person-centered methods in vocational research. Journal of Vocational Behavior 118: 103398.

37.

Howard

Hoffman

(2018) Variable-centered, person-centered, and person-specific approaches: Where theory meets the method. Organizational Research Methods 21: 846–876.

38.

Judge

Thoresen

Pucik

, et al (1999) Managerial coping with organizational change: A dispositional perspective. Journal of Applied Psychology 84: 107–122.

39.

Judge

Douglas

(2009) Organizational change capacity: The systematic development of a scale. Journal of Organizational Change Management 22: 635–649.

40.

Kanitz

Reinwald

Gonzalez

, et al (2024) Supportive, resistant, or both? A person-centric view on employee responses to diversity initiatives. Journal of Applied Psychology 109: 1635–1658.

41.

Katz

Kahn

(1978) The Social Psychology of Organizations. New York: Wiley.

42.

Ketokivi

Mantere

(2010) Two strategies for inductive reasoning in organizational research. Academy of Management Review 35: 315–333.

43.

Köhler

Ritter

Kauffeld

(2024) A typology of organizational readiness for change based on a latent profile analysis. Frontiers in Psychology 15: 1453836.

44.

Langley

Smallman

Tsoukas

, et al (2013) Process studies of change in organization and management: Unveiling temporality, activity, and flow. Academy of Management Journal 56: 1–13.

45.

Lanza

Cooper

(2016) Latent class analysis for developmental research. Child Development Perspectives 10: 59–64.

46.

Lanza

Tan

Bray

(2013) Latent class analysis with distal outcomes: A flexible model-based approach. Structural Equation Modeling 20: 1–26.

47.

Laursen

Hoff

(2006) Person-centered and variable-centered approaches to longitudinal data. Merrill-palmer Quarterly 52: 377–389.

48.

Lines

Selart

Espedal

, et al (2005) The production of trust during organizational change. Journal of Change Management 5: 221–245.

49.

Little

Rubin

(2014) Statistical Analysis with Missing Data. Hoboken, NJ: John Wiley & Sons.

50.

Macrae

Bodenhausen

(2000) Social cognition: Thinking categorically about others. Annual Review of Psychology 51: 93–120.

51.

Magnusson

(1988) Individual Development from and Interactional Perspective: A Longitudinal study. Lawrence Erlbaum Associates.

52.

Magnusson

Törestad

(1993) A holistic view of personality: A model revisited. Annual Review of Psychology 44: 427–452.

53.

Mäkikangas

Tolvanen

Aunola

, et al (2018) Multilevel latent profile analysis with covariates: Identifying job characteristics profiles in hierarchical data as an example. Organizational Research Methods 21: 931–954.

54.

March

Olsen

(1983) The new institutionalism: Organizational factors in political life. American Political Science Review 78: 734–749.

55.

Marsh

Lüdtke

Trautwein

, et al (2009) Classical latent profile analysis of academic self-concept dimensions: Synergy of person- and variable-centered approaches to theoretical models of self-concept. Structural Equation Modeling 16: 191–225.

56.

McLachlan

Peel

(2000) Finite Mixture Models. New York: Wiley.

57.

Meyer

Morin

AJS

(2016) A person-centered approach to commitment research: Theory, research, and methodology. Journal of Organizational Behavior 37: 584–612.

58.

Meyer

Stanley

Parfyonova

(2012) Employee commitment in context: The nature and implication of commitment profiles. Journal of Vocational Behavior 80: 1–16.

59.

Michard

Bouchaud

J-P

(2005) Theory of collective opinion shifts: From smooth trends to abrupt swings. The European Physical Journal B-Condensed Matter and Complex Systems 47: 151–159.

60.

Milgrom

Roberts

(1988) An economic approach to influence activities in organizations. American Journal of Sociology 94: S154–S179.

61.

Mohrman

Jr Mohrman

Ledford

, et al (1989) Large-Scale Organizational Change. Hoboken, NJ: Jossey-Bass.

62.

Molenaar

PCM

Nesselroade

(2015) Systems methods for developmental research. In: Overton

Molenaar

PCM

Lerner

(eds) Handbook of Child Psychology and Developmental Science: Theory and Method. Hoboken, NJ: John Wiley & Sons, pp. 652–682.

63.

Morgan

Zeffane

(2003) Employee involvement, organizational change and trust in management. International Journal of Human Resource Management 14: 55–75.

64.

Morin

AJS

Morizot

Boudrias

J-S

, et al (2011) A multifoci person-centered perspective on workplace affective commitment: A latent profile/factor mixture analysis. Organizational Research Methods 14: 58–90.

65.

Muthén

(2017) Mplus. In: Van der Linden

(ed.) Handbook of Item Response Theory. Chapman and Hall/CRC, pp. 507–518.

66.

Muthén

(2002) Beyond SEM: General latent variable modeling. Behaviormetrika 2: 981–117.

67.

Nagin

(2005) Group-Based Modeling of Development. Cambridge, MA: Harvard University Press.

68.

Nguyen

Stinglhamber

(2020) Workplace mistreatment and emotional labor: A latent profile analysis. Motivation & Emotion 44: 474–490.

69.

Nylund

Asparouhov

Muthén

(2007) Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Structural Equation Modeling 14: 535–569.

70.

Oreg

Bartunek

Lee

, et al (2018) An affect-based model of recipients’ responses to organizational change events. Academy of Management Review 43: 65–86.

71.

Oreg

Sverdlik

Paine

, et al (2024) Activation and valence in responses to organizational change: Development and validation of the change response circumplex scale. Journal of Applied Psychology 109: 135–155.

72.

Oreg

Vakola

Armenakis

(2011) Change recipients’ reactions to organizational change: A 60-year review of quantitative studies. Journal of Applied Behavioral Science 47: 461–524.

73.

Parker

Dipboye

Jackson

(1995) Perceptions of organizational politics: An investigation of antecedents and consequences. Journal of Management 21: 891–912.

74.

Pastor

Barron

Miller

, et al (2007) A latent profile analysis of college students' achievement goal orientation. Contemporary Educational Psychology 32: 8–47.

75.

Pettigrew

(1975) Towards a political theory of organizational intervention. Human Relations 28: 191–208.

76.

Pettigrew

(1987) Context and action in the transformation of the firm. Journal. Of Management Studies 24: 649–670.

77.

Pfeffer

(1981) Power in Organizations. Marshfield, MA: Pitman Publishing.

78.

Poon

(2003) Situational antecedents and outcomes of organizational politics perceptions. Journal of Managerial Psychology 18: 138–155.

79.

Reinhart

Grunwald

(2001) Changeability through flexible and integrated product design and assembly planning. In: Proceedings of the 2001 IEEE international symposium on assembly and task planning (ISATP2001) assembly and disassembly in the twenty-first century, Fukuoka, Japan, 29 May, pp. 318–323. New York: IEEE.

80.

Ross

Rhodes

Hastings

(2008) Defining changeability: Reconciling flexibility, adaptability, scalability, modifiability, and robustness for maintaining system lifecycle value. Systems Engineering 11: 246–262.

81.

Schafer

Graham

(2002) Missing data: Our view of the state of the art. Psychological Methods 7: 147–177.

82.

Schwarz

(1978) Estimating the dimension of a model. The Annals of Statistics 6: 461–464.

83.

Schwarz

Bouckenooghe

Vakola

(2021) Organizational change failure: Framing the process of failing. Human Relations 74: 159–179.

84.

Sclove

(1987) Application of model-selection criteria to some problems in multivariate analysis. Psychometrika 52: 333–343.

85.

Shamir

Brainin

Zakay

, et al (2000) Perceived combat readiness as collective efficacy: Individual- and group-level analysis. Military Psychology 12: 105–119.

86.

Sinha

Calfee

Delucchi

(2021) Practitioner’s guide to latent class analysis: Methodological considerations and common pitfalls. Critical Mare Medicine 49(1): e63–e79.

87.

Spurk

Hirschi

Wang

, et al (2020) Latent profile analysis: A review and how to guide of its application within vocational behavior research. Journal of Vocational Behavior 120: 103445.

88.

Stouten

Rousseau

De Cremer

(2018) Successful organizational change: Integrating the management practice and scholarly literatures. Academy of Management Annals 12: 752–788.

89.

Straatmann

Rothenhofer

Meier

, et al (2018) A configurational perspective on the theory of planned behaviour to understand employees’ change supportive intentions. Applied Psychology: An International Review 67: 91–135.

90.

Swanson

Lindenberg

Bauer

, et al (2012) A Monte Carlo investigation of factors influencing latent class analysis: An application to eating disorder research. International Journal of Eating Disorders 45: 677–684.

91.

Tapak

Hamidi

Toosi

(2024) Using the learning organization questionnaire in combination with latent profile analysis: An example on how to identify latent classes in an organization to improve learning potential. The Learning Organization 31: 884–898.

92.

Tein

J-Y

Coxe

Cham

(2013) Statistical power to detect the correct number of classes in latent profile analysis. Structural Equation Modeling: A Multidisciplinary Journal 20: 640–657.

93.

Tichy

(1983) Managing Strategic Change: Technical, Political, and Cultural Dynamics. New York: Wiley.

94.

Vermunt

Magidson

(2002) Latent class cluster analysis. In: Hagenaars

McCutcheon

(eds) Applied Latent Class Analysis. Cambridge: Cambridge University Press, pp. 89–106.

95.

Vermunt

Magidson

(2013) Technical Guide for Latent GOLD 5.0: Basic, Advanced, and Syntax. Belmont, MA: Statistical Innovations Inc.

96.

Witt

(1998) Enhancing organizational goal congruence: A solution to organizational politics. Journal of Applied Psychology 83: 666–674.

97.

Wolf

Harrington

Clark

, et al (2013) Sample size requirements for structural equation models: An evaluation of power, bias and solution propriety. Educational and Psychological Measurement 73: 913–934.

98.

Woo

Jebb

Tay

, et al (2018) Putting the person in the center: Review and synthesis of person-centered approaches and methods in organizational science. Organizational Research Methods 21: 814–845.