Keywords

1 Introduction

There are few companies these days that can remain competitive without Information Technology. Be it at the core of the business or limited to opportunistic explorations, IT projects can determine the success or failure of an organization. These projects are managed in many different ways, from ad hoc processes to the waterfall model using Gantt charts.

At the turn of century, noticing the chaos of the absence of a proper process and the frequent inefficiency of the waterfall model [1], a group of software developers proposed “agile” as a more appropriate way to manage IT projects [2]. Highsmith [3], one of the signatories of the Agile Manifesto [4] states that the growth and the criticism of agile methods have to do with values and culture, and not taking cultural factors into account is the main cause of failure in the adoption of software development methodologies. Research by Begel and Nagappan [5] identified the incompatibility with culture as one of the barriers to adoption of agile methods.

In contrast, Robinson and Sharp [6] indicate that agile practices can be adapted to produce a methodology that is appropriate for a particular culture, and cultural analysis can help prepare an organization to introduce changes [7].

Often, there is a conflicting relationship between culture and practices that one seeks to adopt in IT organizations, which can harm change initiatives, particularly in the context of agile methodologies adoption. Thus, this study aims to:

  1. (a)

    Identify in literature how the various characteristics of an organizational culture are described and how they can be measured;

  2. (b)

    Identify the practices that best represent, for the purposes of this work, the principles behind agile methodologies;

  3. (c)

    Identify which basic assumptions of organizational culture are more or less related to the adoption of agile practices by mapping the basic assumptions prevailing in organizational culture of different companies and their correlation with the adoption of agile practices in the implementation of information systems in said companies.

2 Main Elements of Study Execution

This study was based on cultural elements that can be classified as national, regional, organizational, group or individual, focusing on organizational elements. Besides these, elements of agile methods and its main practices were analyzed and mapped.

2.1 Culture Elements

Ali and Brooks [8] define culture as “shared patterns of behavior.” Within the organization, the definition by Ed Schein, also used by Fleury et al. [9] was used:

The culture of a group can now be defined as a pattern of shared basic assumptions learned by a group as it solved its problems of external adaptation and internal integration, which has worked well enough to be considered valid and, therefore, to be taught to new members as the correct way to perceive, think, and feel in relation to those problems [10].

Both Hofstede and Hofstede [11] and Schein [10] suggest the use of different dimensions to describe an organizational culture in aspects relevant to empirical studies. For Schein basic assumptions are organized in the following dimensions [12]:

Nature of Human Activity:

Between Environment dominant (the organization acts as if subordinate to the environment it operates) and Organization dominant (the organization believes in its ability to determine the environment and the market). Between Proactive (the organization accepts that individuals can cause instability in the search for improvement) and Reactive/fatalistic (the organization expects the protocol to be followed and when unforeseen results happen, it accepts the consequences).

Nature of Reality and Truth:

Between Moralistic authority (the organization trusts the boss/expert/manual to determine the proper way of doing things) and Pragmatism (the organization tries to obtain objective information and believes truth emerges from the exchange of ideas among all).

Nature of Space:

Between Favors privacy (the organization creates private spaces and discloses information as needed) and Favors communication (the organization adopts an environment that facilitates the rapid exchange of ideas).

Nature of Time:

Between Near future oriented (the organization uses its planning as basis for decisions) and Past oriented (the organization revisits previous experiences to look for solutions to problems). Between Long time units (organization favors a long-term view, disregards minor delays) and Short time units (the organization plans its activities with a short horizon of time and sees small delays as significant).

Nature of Human Nature:

Between Humans are basically good (the organization believes that properly motivated people will exceed expectations) and Humans are basically evil (the organization suspects that people will act inappropriately if given the chance). Between Human nature is fixed (the organization believes that people are what they are and cannot change) and Human nature is mutable (the organization believes that people adapt to different situations and can improve depending on the context).

Nature of human relationships:

Between Groupism (the organization believes that all good things come from the group and strive to create consensus) and Individualism (the organization believes that individual talent is key to problem solving). Between Collegial/participative (authority is determined by the context and the leader defers to the group) and Authoritarian/paternalistic (the organization believes in a strong hierarchy).

2.2 Elements of Agile Methods

The Agile Manifesto [4] and the methodologies associated with it caused a significant change in the way teams develop software. According to Taylor [13]:

Agile methodologies generally promote a project management process that encourages frequent inspection and adaptation, a leadership philosophy that encourages teamwork, self-organization and accountability, a set of engineering best practices that allow for rapid delivery of high-quality software and a business approach that aligns development with customer needs and company goals.

Fowler [29] points out that “one of the hardest parts of introducing agile methods into an organization is the cultural change it causes.” Because of the flexible nature of agile methodologies, it is often better to adapt practices when these prove inadequate to a given context, provided the agile values are respected [2, 6, 14, 15].

In order to compare practices and adoption, a number of agile practices were selected based on a comparison of several authors [2, 5, 14, 16,17,18,19,20] (Table 1). They were selected by measuring how often they were mentioned in the reference literature and mapping them according to the 12 principles of the Agile Manifesto [4] (the principle of “Our highest priority is to satisfy the customer…” was considered as a meta principle, resulting of the conjunction of the others and as a guiding principle for the application of all practices, not mapping exclusively to a single practice).

Table 1. List of selected Agile practices and its corresponding principles. Number of papers considered that reference the practice in parenthesis.
Full size table

3 Study Method

This work is understood as exploratory study [21]. At the time the survey was performed, no other studies were found relating cultural dimensions and agile practices adoption.

This study used a quantitative research method [22], using as instrument an online survey form. It was considered, according to literature [10, 11], to be good enough for an initial search of insights into the role of culture in inhibiting or facilitating organizational change.

The case study was carried out between March and April of 2011 in the company ThoughtWorks Inc., global leader in consulting and the use of Agile methodologies in implementation of software development projects [23]. The source data was collected from ThoughtWorks Inc. consultants delivering software or providing advice on software delivery based on their experience during the delivery or after it has finished. These consultants were distributed globally in many different organizations and cultures, which have different levels of knowledge of Agile.

For cultural dimensions, the model used was adapted from the one proposed by Schein [12] described in Sect. 2.1. A five point scale between the cultural assumptions at each end of a given dimension was used to measure an organization’s manifested behavior with regards to that dimension.

A comparative analysis of previous works was used for the selection of agile practices surveyed (Table 1). The adoption of a given practice was measured using the scale proposed by Boehm and Turner [24] for “Levels of Software Method Understanding and Use” (Table 2).

Table 2. Levels of software method understanding and use.
Full size table

The responses were based on the subjective perception of consultants from ThoughtWorks Inc. on the client’s experience and satisfaction with agile methods; the culture of the client organization; and the understanding and use of agile methods put forward by employees of the organization at the end of the project.

For data analysis, descriptive and multivariate statistical procedures were employed using statistical functions as Pearson correlation and hypothesis test (t test error probability Type I (α) accepted of 0.01) [25,26,27]. No data cleansing process was performed. Each dimension of culture was separated in the two basic assumptions it corresponds to, with assumptions on the left of the scale having the negative correlation value, and assumptions on the right of the scale having the positive correlation value. Thus, negative correlations between a practice and a dimension indicate, in fact, a direct correlation with the assumption on the left in a certain cultural dimension. To avoid drawing conclusions on extreme situations with a low number of responses, only the top 20% of the assumption and practice pairs with the highest correlation were selected for analysis [13].

4 Analysis of Results

The survey consisted of questionnaires distributed to a population of 1400 consultants, with a return rate of 8.14% (114 responses), which allowed calculation of a correlation between the cultural dimensions researched and selected agile practices. The questionnaires were validated by two specialists from ThoughtWorks and the results verified and commented by five others. Both questionnaire and results were reviewed by three members on an examination board. The results can be seen in Table 3. The darker the cell the higher the correlation, either positive or negative. It is important to notice that a high value for modular correlation does not imply that this correlation is significant. This requires a hypothesis test, which was done in selected cases.

Table 3. Map of the correlation between practices and cultural dimensions.
Full size table

Positive or negative correlations represent the influence of different assumptions on the adoption of a practice. In Table 3 cultural dimensions are in columns, so negative values indicate a strong correlation with assumptions to the left in the column heading while positive values indicate a strong correlation with assumptions to the right in the column heading. Only the highest 20% of the correlation results (in bold in Table 3) were selected for a detailed analysis. All correlations highlighted proved significant.

Based on the results, it is worth noting that the basic assumptions of “Pragmatism”, “Favors communication” and “Collegial/participative” are the ones that correlate more with agile practices, with 8 practices with high correlation for each assumption. This can be seen as aligned with the four values of the Agile Manifesto [4] as seen in Table 4.

Table 4. Values of the agile manifesto and basic cultural assumptions.
Full size table

On the other hand, four dimensions showed no correlation among the 20% analyzed: Between “Environment dominant” and “Organization dominant”; Between “Near future oriented” and “Past oriented”; Between “Long time units” and “Short time units”; and between “Human nature is fixed” and “Human nature is mutable”. It is surprising that the last one showed no correlation, since the concept of “Agile Mindset” [28] speaks explicitly about the belief that people can change.

Another interesting observation is that the practices of “co-location”, “Test Driven Development”, “Continuous Integration”, “Refactoring code”, “Pair programming” and “Stand-up meeting” showed no representative correlation with any basic assumption. This suggests that these practices are more culture “agnostic” and possibly can be adopted more easily by any organization. Many of these practices (“Test Driven Development”, “Continuous Integration”, “Refactoring code” and “Pair programming”) are highly technical, which may explain why many adoptions of agile methodologies in organizations are led by software development teams.

Therefore, one could assume that, particularly in organizations with predominantly “Favors privacy”, “Moralistic authority” and “Authoritarian/paternalistic” cultural assumptions, an Agile adoption will be more successful if it starts by implementing the practices of “co-location”, “Test Driven Development”, “Continuous Integration”, “Refactoring code”, “Pair programming” and “Stand-up meeting”.

5 Conclusions and Contributions

Based on the highlighted correlations, one can argue that it was possible to identify cultural assumptions that are related to the adoption of agile practices. This was achieved by seeking in literature instruments which allow the classification of the culture of an organization, listing a set of practices representative of agile principles and looking to measure the correlations between these in real situations.

Although the results are encouraging, one must be aware that their validity is limited due to the inherent limitations and biases of surveys and the small number of responses. It is also a subjective matter within a fairly homogeneous population, limiting extrapolation of results.

Therefore it is considered that this work may contribute to a deeper exploration of the theme by suggesting significant correlations between the adoption of agile practices and cultural assumptions of different organizations.

To the academia, this work serves as an initial model to look into agile methods and practices in the organizational culture domain. We hope that these results encourage researchers to consider organizational culture as an important aspect in studies of adoption of these methodologies. For practitioners, we hope this inspire companies looking to adopt agile methodologies to consider the culture of its organization, and plan strategy accordingly, aiming, for example, to start the journey by the practices that seem less sensitive to culture, and eventually seeking to adapt their culture (or the expected results) according to these limitations.

Several possibilities for future studies arise from this work. To confirm its results, and mitigate any bias that this study could have been subject of, it is necessary to expand the set of sources analyzed, seeking a greater volume and diversity of data. In addition, a longitudinal study with a limited set of customers observing the impact of agile practices in organizational culture can help to establish a causal relationship and deepen the understanding of the nexus between agile practices and organizational culture. We believe it is also important that the practitioners of agile methodologies - from beginners to the more experienced - pay attention to day-to-day moments where culture was an impediment to improvement, or, to the contrary, a boost to change. Conveniently, the Agile Manifesto itself suggests moments of reflection where this debate can be rewarding.