Chunking decision information: a way to make big data actionable

Abstract

Today’s decision management systems focus on supporting selected decisions rather than including and linking pre- and post-decision information, that is, big data. The purpose of this research was to understand how decision information is organised in an international, cross-functional organisation. This paper is an attempt to understand how a five-element, object-oriented framework tailors decision information in the international humanitarian programme. How can chunking and organising decision information object-oriented be a way to make big data actionable? The application of the model helps to improve organising decision information.

Keywords:

• Organising decision information
• chunking information
• Erasmus+ programme
• big data

Introduction

Modern decision management systems mainly focus on tracking selected processes within narrow information fields, and they are designed for applications in static companies with established hierarchies of decision-receivers. Studies of decision- making processes in project teams that are geographically dispersed either permanently or temporarily are centred on applications of means of communication within the field of information technology (IT); they do not give a full overview of decision information itself. Even less is known about decision-making in international social and humanitarian programmes, which often combine the organisational characteristics of traditional companies and project teams. Therefore, a full picture of decision information processing in such organisations remains unclear (Agahi, 2014). Individuals who work in social programmes organise information that they obtain through various formal and informal sources, which is a common practice in many organisations.

Table 1. Respondent demographics.

Number of countries	35 (EU countries + Tunisia, Serbia, Israel, Bosnia and Herzegovina, Albania, Ukraine, Russia, Belarus, Georgia, Moldova and Armenia)
Respondents’ average age	31
Gender	52.8% male, 37.5% female

Despite organisational differences between static organisations and more ‘virtual’ social and/or humanitarian programmes, they all face the same problems related to big data (Fan & Bifet, 2012). Thus, there is no evidence that such organisations deal efficiently with decision information evaluation, storage, or visualisation. Information that is used for working processes often remains unorganised. This paper is an attempt to understand how a five-element, object-oriented framework tailors decision information in the international humanitarian programme. How can chunking and organising decision information object-oriented be a way to make big data actionable?

Theoretical background

Erasmus+is a complex programme that on one hand possess features of typical bureaucratic institutions and policies. Information exchange in such organisations are complex and often goes beyond organisational limits, partially due to human actors (Contandriopoulos, Lemire, Denis, & Tremblay, 2010, p. 464). The mechanisms of Decision Information Management often are unclear. Gornitzka and Sverdrup (2011, p. 51) showed that main sources of information are ‘statistics, scientific journals, media reports, lobbying, parties and other EU institutions, as well as the more informal exchanges of information and gossip.’ The policies in these kinds of organisations tend to be overloaded with information which may cause institutional friction, often associated with the resistance to change (Workman, Szymczak-Workman, Collison, Pillai, & Vignali, 2009).

Erasmus+as organisational teams share features with several types of teams: international cross-functional teams, multidisciplinary project teams, communities of practice, and virtual teams. Further we provide an overview of these types of teams and we define the main types of decisions they face. We also summarise the critical factors for the successful organisation of decision information in each team.

Ghobadi (2013, p. 139) defines international cross-functional teams as ‘temporary work-groups that are charged with the responsibility of completing a development project within a limited time frame, and they consist of member representatives drawn from various functional units (e.g., departments, organisations).’ Ratcheva claims that the three main factors that define multidisciplinary teams are numerical, territorial, and epistemological (2009, p. 207). Territorial factors here refer more to the scope of members’ expertise than to actual geographical differences. The successful management of decision information in multidisciplinary project teams is determined by interpersonal interactions and relational capital among members. The creation of new ways of working through intense interaction is also a key to the successful functioning of such teams (Ratcheva, 2009, p. 208, 2013). Since it is quite challenging to unify decisions in all types of multidisciplinary teams, it seems reasonable to make a generalisation that the decisions which such teams often face are connected to the application of contextual knowledge by team members.

Communities of practice (CoPs) refer to informal groups or networks of individuals who share ‘a domain of interest and knowledge about which they communicate’ or have some work-related activity in common (Hamburg, 2010, p. 24; Hislop, 2013, p. 155). Duguid (2005, p. 11) has argued that within a CoP, knowledge is instantiated dynamically in what Giddens (1984) calls knowledgeability, including all the things which actors know tacitly about how to ‘go on’ in the context of social life without being able to give them direct discursive expression. Decision information that members of CoPs possess and share is highly personal and can be shared within a community up to a certain point (Hislop, 2013, p. 155). Thus, the main decisions in CoPs are related to issues of converting theory to practice and tacit knowledge to explicit knowledge. Another aspect is dealing with interculturality, which is a common feature of CoPs. Hamburg argues that ‘culture and cultural differences can have a strong impact, particularly in the case of tacit knowledge such as leadership skills or management know-how’. (2010, p. 25) Decision information transfer tends to be embedded by such factors as individuals, relationships, processes and the environment (Hamburg, 2010, p. 24). So all decisions are framed with consideration for how the process is embedded in humans and/or technology.

In virtual teams, groups of people work on interdependent tasks while being geographically distributed. They fulfil their core work mainly through electronic media and ‘share responsibility for team outcomes’ (Horwitz, Bravington, & Silvis, 2006, p. 473). Virtual teams often face the need to make decisions about the ‘articulation of distributed activities’ (Schmidt, 1994), which includes decisions about roles, responsibilities and authorities. This need becomes a challenge in conditions of geographical, cultural separation, especially when using insufficiently powerful communication media. Furthermore, we can suggest that virtual teams need structured DI in scope-of-work logistics, e.g. planning encounters, designing communication schemes, and so on.

The clear understanding of goals and responsibilities is a crucial factor for effectiveness in such teams (Horwitz et al., 2006; Gibson & Cohen, 2003). Other factors include the degree of trust and commitment among managers and the existence of support systems for the team (Horwitz et al., 2006, p. 490). The organisation of decision information in a virtual team depends on communication technologies that mediate the interactions between team members who are geographically separated. Finally, culture also plays an important role in DM processes. Oftentimes members of such teams belong to different cultures, which may influence their perceptions of work, conflict resolution strategies, expressions of leadership and time orientation (Gibson & Cohen, 2003).

Thus, the most common decisions that are made by such teams are related to the application, transformation and transmission of decision information or related knowledge. The complexity of decisions has to do with the wide variety of contexts in which they are applied. Additionally, a large amount of unstructured information can lead to complex decision-making processes among these teams.

An object-oriented approach views decision information as an object that has a ‘defined role in the application organisation’ and possesses a certain state, behaviour and identity (Agahi, 2014, p. 126). Let us examine these attributes more carefully. A decision’s state is shaped by such properties as attribute values and relations to other decisions. A decision’s behaviour shows how a decision will act (e.g. how it will be used). Finally, a decision’s identity points to its uniqueness. Even if two decisions possess similar values for their attributes, they remain two separate decisions. It is therefore crucial to define a decision’s attributes precisely in order to use it for transmitting messages. Another reason why it is important to provide a clear definition of a decision’s identity is that one decision often leads to another. Therefore, decisions require strict identification in order to be able to be linked to each other.

Decision information chunk refers to a fragment of information which is related to a certain decision and which is organised in a way that is most efficient for memorising and application. Therefore, by chunking decision information we mean convenient organising of DI units. In practice, chunking decision information object-oriented may be implemented by using decision cards with five components. The decision card is a decision-information representation model and an object consisting of five major elements:

(1)	The ISSUE on which the decision is focused;
(2)	What the decision is BASED-ON, that is, the background and existing intelligence related to the decision;
(3)	The DECISION itself;
(4)	The AGENTS – the decision-makers and decision-receivers – who are involved in the process; and
(5)	FEEDBACK on the experiences of decisions.

Decision cards capture the key issues, decision contexts, established decisions and people involved in the decision-making. In the meantime, the feedback and experiences of various actions can be added to the card. This is what we like to call business intelligence.

Methodology

For this study a mixed research design was chosen because it enables researcher to study socially constructed human interactions through qualitative methods. At the same time it allows to explore ‘observations that are converted into discrete units that can be compared to other units by using statistical analysis’ using quantitate methods (Bergman, 2010). Moreover proposed study reflects so-called seven-stage mixed data analysing process which, among other, includes data reduction, transformation and integration (Onwuegbuzie & Teddlie, 2003).

In our research we focused on the European Commission’s (EC) programme Erasmus+and its mobility projects for young people and youth workers. These projects ‘provide opportunities for young people to experience learning mobility, to develop their interpersonal skills and improve their employment prospects through training and networking opportunities in Europe and beyond’ (European Commission, 2014). The EC programme Erasmus+brings together seven existing European Union (EU) programmes that support work in the transnational partnership and share innovative practices in the fields of education, training and youth. With financial support of the programme from non-governmental organisations (NGOs), educational establishments and social enterprises located in programme countries and neighbouring countries fulfil a wide range of actions and activities in the fields of education, training, youth and sport.

The Erasmus+programme has National Agencies (NAs) in each of the 33 countries that participate in the programme (e.g. Norway, Turkey, Macedonia, Iceland, Liechtenstein, and so on). NAs carry out decentralised actions, they may possess different names and they may be combined with national bodies responsible for youth issues. For example, the NA in Sweden is called the Swedish Agency for Youth and Civil Society. The NAs’ work is coordinated by the Education, Audiovisual and Culture Executive Agency (EACEA) located in Brussels (European Commission, 2013a). NAs and the EACEA are the main decision makers in the Erasmus+programme; they decide whether a project will be funded, and, if so, on what conditions. Decision-receivers are represented by organisations of civil society (NGOs), social enterprises and public bodies from programme countries. They initiate projects that aimed to resolve certain social issues. After an application for project funding is submitted, it goes through review by the NAs or the EACEA. The review process takes 3–6 months, and, after a positive decision has been made and the project has received funding, a whole series of actions and sub-decisions take place. The overall scale of Erasmus+’s area of operation is quite significant. Altogether, since 2013 Erasmus+has approved 22,258 social and educational projects of different scales (European Commission, 2014b).

The review of project applications is based on defined criteria and legislation that frames the whole Erasmus+programme. The main legal act that is used for managing the Erasmus+programme is Regulation (EU) No. 1288/2013 (European Commission, 2013a). This document explains the scope of the programme, provides the main definitions, objectives, and actions of the programme and provides other necessary information for further project evaluation. The Erasmus+Programme Guide contains more specific information about award criteria and application evaluation (European Commission, 2014a). Figure 1 presents examples of chunking of decision information object-oriented in the Decision Card framework.

Figure 1. Examples of chunking of decision information object-oriented in the Decision Card framework.

A sequential mixed method data collection strategy was used in this study. This approach involves data collection ‘in an iterative process whereby the data collected in one phase contribute to the data collected in the next’ (Driscoll et al., 2007, p. 21). Data collection was organised in three stages. First, an online survey was conducted among Erasmus+practitioners selected by a convenience sampling method. This sample included 76 individuals who were involved as leaders, trainers, facilitators or managers in the DM processes for more than 60 Erasmus+projects. The cultural diversity of respondents covered 35 countries. Among them, Macedonia and Turkey each had 5 representatives; Romania, Italy and Germany had 4 each; Spain, Austria and Norway had 3 each; Slovakia, Poland, Iceland, Denmark and Bulgaria had 2 each; and Malta, Lithuania, Ireland, Hungary, Greece, Estonia, Cyprus, Croatia, France, Sweden, Bosnia and Belgium had 1 each. Respondents from Erasmus+partner counties such as Tunisia, Serbia, Israel, Bosnia and Herzegovina, Albania, Ukraine, Russia, Belarus, Georgia, Moldova and Armenia also took part in the survey. Although legal bodies from these states are not entitled to be direct decision-makers, activists and organisations take an active part in actions and sub-decisions derived from the main decisions. Therefore, the participation of these respondents was considered relevant. The age of respondents varied between 21 and 60. The average age of the survey respondents was 31, while the most frequent age in the study was 27 years. Of the respondents, 52.8% were male and 37.5% female.

Respondents demographic information

The questionnaire consisted of 27 closed – i.e. multiple-choice – questions. There were three group of questions in the survey: ones related to respondents’ demographic data, ones describing characteristics of their experience within Erasmus+and once directly related to DI management within their teams. The last question, however, was open-ended in order to obtain additional comments on uncovered topics. Here we provide an example of two survey questions together with answer options: What, in your opinion, from mentioned below was/would have been the most helpful for successful DI management in your team? (Clear appointment of leader-follower roles / Avoiding of conflicts / Multi-ways communication and provision of frequent feedback within the team / Strong control of task performance/Mistake tolerance / Monitoring and evaluation of DI management processes /Leadership and /or management commitment). What, in your opinion, was the biggest challenge that your team has faced working in this project? (Unclear goal setting, responsibilities and priorities / Lack of technology that ensures communication quality / Cultural differences / Poor management / leadership / Lack of competences in the given topic / Lack of team members commitment / Not having enough of face-to-face meetings)

Answer options were constructed on the base of findings from studied literature with consideration of success factors for DI management in different types of teams. Data derived from the survey became the basis for a semi-structured interview guide. The purpose of the interviews was to explore certain tendencies that were discovered in the survey more deeply. Thus, survey answers showed that respondents consider clarification of objectives, roles and responsibilities within leaders’ team more as the key success factor for DI management. Therefore, interview respondents were asked about How to make clarification of objectives, roles and responsibilities within leaders team more clear/successful/efficient? Similarly, survey respondents marked that most of D-related process in the team could have been named as creation of DI. Consequently, interview respondents were asked to provide concrete examples of DI creation. Apart of DI-related questions, interview respondents were asked to tell how they saw differences between DI management in Erasmus+and traditional organisations.

In order to avoid misunderstandings and duplications of responses, the interviewees did not take the online questionnaire. Since all of the interviewees lived in different countries, interviews were conducted by Skype and recorded on the phone. The average length of the interview was 27 min. Interviews were conducted with decision-makers only.

A further case-study method has been applied. During the case-studies, DI in 4 Erasmus+projects have been studied from the perspective of the five-element, object-oriented model. The typical decision-making process in Erasmus+projects consists of a set of actions (pre-decision phase), the decision itself and a combination of actions and sub-decisions (post-decision phase). For instance, the decision to provide funding for the Erasmus+project’s international training course ‘Communication for NGOs’ was made by a Swedish NA (decision-maker). All actions related to this decision were implemented by a decision-receiver – a non-profit organisation, ‘Newality,’ which applied for this funding. The decision to fund this project was based on the application that was submitted by Newality and the evaluation criteria outlined in the programme documentation. The consequences of this decision included a set of actions primarily connected to the content and logistics of the project. Such actions could be sub-decisions on project dates and venue, the distribution of funds, the selection and involvement of speakers, the design of the daily timetable and so on.

A similar procedure is applicable for all Erasmus+projects. In the pre-decision phase, applicants formulate a project idea and develop a partnership consortium. The projects are different in details such as the number of partner organisations, the duration and the delegation of responsibilities.

Results

Survey results

Application of the mixed data analysis approach allowed us to see presence of different DI-processes in the team work. The majority (76.4%) of respondents stated that the prevailing DI-related process during their cooperation was information sharing. Other DM-related processes such as information development, creation and utilisation were mentioned by 55.6, 41.7 and 30.6% of respondents, respectively. About 69.4%, decision information was related to the theme of the project.

Based on the literature review, we defined main challenges that may impact managing DI in such organisations and suggested respondents to choose most relevant for them. The survey results showed that, the absence of face-to-face meetings, cultural differences and unclear goal-setting make a strong negative influence on quality of DM. Improvement of organisational and technical communication is believed to be the best way to deal with those challenges. One of the respondents named a lack of data-structuring and business analytics as a challenge for successful cooperation: ‘over prepared but under structured is something we work with in my groups.’

Obtained quantitative data shows that more than 43% of respondents used IT communication for the organisation of DI. Among the most usable technologies that were named included those that enable instant communication and basic information storage. At the same time, technologies that allow more advanced information storage, such as Google Docs or online project management tools, were used by less than 18% of respondents (Figure 2).

Figure 2. Respondent use of IT.

The clarification of objectives, roles and responsibilities was named as the main factor for effective DI organisation. Other factors of successful DM support that were named included monitoring/evaluation of the knowledge management process as well as frequent feedback from within the team (Figure 3).

Figure 3. Respondent objectives, roles, and responsibilities.

Interview results

Analysing qualitative data we saw a strong interviewee’s emphasis on the unity of information sharing, modification and creation in DI management processes. Uniting or chunking, and transforming already-existing data creates new information and provides the basis for future decisions. As one respondent remarked: ‘in a way we produced our international knowledge … we incorporated different perspectives’ (Director, Germany). The involvement of all team members in the DM process, the clear presence of a new intellectual product, the possibility for practical applications of this new product, and its attractiveness for future users were named as the main criteria of effectiveness of DI management.

Face-to-face communication that was highly appreciated by survey respondents, by interviewees was associated with participants’ stronger involvement in the issue. According to interviews, it allows team members to be part of the DI organisation process and to make a bigger impact on it. Furthermore, face-to-face meetings and the use of effective communication channels foster the equal involvement of all DM team members and enhances the quality of this decision-making process. Equal participation in the decision-making process has a positive impact on team dynamics and, therefore, on work performance, as one of the interviewees mentioned: ‘it’s more motivating when you are a part of decision-making’ (TCA officer, Norway).

Previously we mentioned that virtuality is an important feature of team interaction within the policy or programme. It is worth mentioning that the team itself is not virtual, though their communication often is. The interview results show that DI management depends, to a large extent, on issues of data storage, data evaluation and transmission. This finding corresponds to what Fan and Biffet name as future challenges in big data management. In particular, efficient data management requires clear visualisation, objective evaluation and user-friendly information compression/storage. Without effective organisation of these processes, the construction of a unified DI system is problematic. In order to let all team members have equal access to information and be updated on all decisions and changes in a timely manner, such a system is a need as a coordinating tool that quickly connects users to relevant information.

The overall DI organisation in Erasmus+projects tends to follow the same pattern where certain actions of decision-receivers lead to the decision itself (e.g. NGOs or public bodies come up with an idea for holding an international project and subsequently apply for funding). These actions are based on existent documents such as calls for proposals and application guidelines, among others. Decisions about funding provisions are made by decision-makers (Erasmus+authorities). Decisions are based on the submitted funding applications as well as evaluation and assessment guidelines.

Once a decision has been made, this leads to a number of sequential actions by decision-receivers. Among such actions are sub-decisions on project logistics (time period, venue, budget allocation) and content (themes, timetable of activities, leaders). These actions are interdependent and often mutually influence each other.

Discussion

A decision in an international humanitarian programme is seen as combination of the decision itself and the action required for the implementation of this decision. In other words, the DI organisation process in an international humanitarian programme is always a complex, systematic process consisting of many levels of interdependent decisions. This study has shown several of the main problems of the decision support methodology for the Erasmus+programme. These problems include the lack of face-to-face communication, the unclear division of responsibilities, the lack of feedback on decision implementation, and the general inability to track the post-decision phase. Each of these drawbacks points to the absence of a single unified system of decision information, which would make it possible to know not only the decision itself, but also that decision’s background, the consequences of the decision and the parties involved in making the decision.

The application of the Decision Card (DC) system as a decision support methodology increases the actionability of DI by making it clearer in all three stages of the DM process (pre-decision, decision and post-decision). By tracking a decision’s feedback, the DC system supports future pre-decisions that appear as consequences of previous decisions. In particular, DC contains multidirectional information about a decision and shows its connection to other decisions. Thus, without imposing any DM model, the DC system offers a way to make DI actionable and accessible for users.

Furthermore, a deeper analysis of interviews from behavioural and decision model perspectives has shown that it is not the absence of face-to-face meetings itself that prevents the efficient management of DI. It is, rather, the absence of the possibility for all team members to be informed about all updates regarding DI in a timely manner. Since such updates often take place during face-to-face meetings or in individual e-mail conversations, some team members risk missing selected decisions as well as the reasons that led to such decisions. Therefore, the inability to follow this information is a challenge for the DI organisation process. The same conclusion relates to goal-setting and the division of responsibilities. Decision-makers and decision-receivers sometimes remain unknown unless they occupy major positions in the process. For this reason, it might be difficult to find the person responsible for a particular action or sub-decision. Limited access to information resources that contain data which is crucial for DM is another challenge. Often only the main documents are easily accessible, while a large amount of supporting documentation gets lost in semi-private communication. The DC system objectifies a variety of decision-making behaviours without giving priority to any of them. In other words, the DC system does not shape the DM process. Considering all of the available data and the endless number of possible combinations of DI, we do not strive to tailor the DM process to several limited patterns. Instead, the DC system organises all DI in an actionable way that can be applied under any DM strategy.

An object-oriented approach offers the DC system as a solution for better analytics and knowledge discovery. Here we suggest an explanation for how named challenges could be solved with the application of the DC system. The ISSUE provides a short summary of the problem/idea to which the decision relates. BASED_ON covers the full range of factors that influence the decision. BASED_ON can be something that Hamburg calls ‘individuals, relationships, processes and environment’ (2010, p. 24). The Based_ON section also includes the documentation of previous decisions, the outline of existing guidelines, and the use of IT media for official conversations. This section aims to be a kind of ‘frequently asked questions’ page where team member can find answers to numerous questions regarding the decision. The Section Agent lets the user know who is responsible for the decision and who is involved in its implementation. Finally, the feedback sections enlighten the post-decision phase of the process. It allows users to see the consequences of decisions that have already been made, to run formal assessments and evaluations, and to store intellectual outcomes. Newly created product that was named among main criteria of successful DM.

The DC system also brings together other key points for promoting efficiency in the DI organisation processes of humanitarian programmes. First, it becomes clear that communication media allow teams to keep all decision information together, which prevents team members from having to look for pieces of it in separate documents or e-mails. Finally, the DC system gives routine procedures more structure, shows how different decisions are connected to each other, lets users build a cause-consequence chain based on facts, and sheds some light on the post-decision phase.

Conclusions

Chunking DI is an efficient way to organise knowledge related to a certain project, policy or programme. The application of an object-oriented model within a humanitarian programme helps to improve DI organisation and create a clear system that allows users to track decisions during the life-cycle of a project. Specifically, an object-oriented approach aims to clarify the post-decision phase. By collecting and systematizing all of the consequences of a decision such as actions, new knowledge or further decisions, the object-oriented decision card system bring order to the routines of such humanitarian programmes as Erasmus+. Moreover, the application of decision card in such projects fosters equality among team members because it enables access to all decision information.

Disclosure statement

No potential conflict of interest was reported by the authors.