Content-Based Research

Researchers carry out content-based research in order to reveal the composition of the object under study. There are many reasons why a researcher may want to carry out this type of research. Nevertheless, this elementary, broad definition of content-based research hides the tremendous diversity to be found within studies of this type. This suggests that content-based research does not form a homogenous body of work.

1. Why Choose Content-based Research?

A cross-section or a video still of an object that we wish to study are two pos­sible metaphors for understanding the purpose of content-based research. The idea is to describe something statically, as it appears at a given point in time. Time is only taken into account implicitly, and the study is not concerned with the object’s evolution. Content-based research reveals the existence or coexis­tence of a certain number of elements rather than the manner in which the object evolves over time. As we will see further on, this does not mean that the temporal dynamics of the object under study are ignored; an object’s temporal dynamics may be used to explain it, or be included as an element of the object’s context. Temporal dynamics do not, however, directly enter the field of content- based research.

Two types of content-based research can be distinguished. They differ in terms both of the methods used and the types of questions they deal with. The first consists in describing the subject of the research in order to understand it better. The goal of research pursued in this way is not to explain the object under study in the sense of causality research, but rather to describe it. The pur­pose of the description is to improve one’s understanding of the object under study. The idea is to go beyond the object’s perceived complexity. The second type of content-based research aims to show and to explain the causal links between the variables the object is composed of. The researcher attempts to delineate the causes or consequences of a given situation.

More precisely, researchers may be confronted with a new problematic for which little or no empirical material or theoretical research exists. In a situation like this, it seems appropriate to devote a certain effort to describing the object under study. This is the case, for example, when new practices appear or when researchers study an aspect that has not yet been empirically studied in any depth. Bailyn et al. (1997) describe an intervention technique aimed at harmon­izing the needs of individuals and the goals of the organization.

Example: Descriptive content research on the ‘dual agenda’ method

Bailyn et al. (1997) describe a method of resolving the opposition between indivi­duals’ goals and the organization’s objectives. From the theoretical and empirical observation of the difficulties many employees have in reconciling their private lives and their commitment to a company, the authors describe several situations in which they were able, within the framework of action research, to identify the con­tents of this method, and thus to define the manner in which to proceed. Their method consists of carrying out group interviews with a double goal. These inter­views should encourage actors to consider how the way in which they organize their work affects their private lives. This increases the group’s awareness of the ways in which conflicts between their professional and their private lives affects their performance at work. The discussion is then oriented towards identifying ‘leverage points’ that may help reduce these conflicts.

These interviews should result in concrete suggestions, which can then be tested by the enterprise. The authors call this process a ‘dual agenda’. By explaining in detail the way they proceeded in several different companies, they contribute to a better understanding of the articulation between professional and personal agen­das, and add to our knowledge of a new technique.

It can also be rewarding to look into methods belonging to another domain or culture. Japanese management techniques and their practical applications in American companies have, for example, inspired a large number of empirical and conceptual studies. The simple act of describing the successful way that Japanese companies have been run and drawing implications this may have on managing western companies was a contribution in and of itself. It then becomes possible to consider the difficulties that arise in adapting techniques from different cultures. This type of descriptive content-based research is illustrated by the following example.

Example: Descriptive content-based research into adapting Japanese-style teamwork to the USA

Cutcher-Gershenfeld et al. (1994) studied the adaptation of different levels of Japanese-style teamwork to American-Japanese joint ventures in the USA. This study was part of a larger research program analyzing the exchange of American and Japanese managerial techniques. Its purpose is to understand teamwork better: its organization, the context in which it is used and the difficulties encountered in its application. The study looked at eight joint ventures, and included 150 indivi­dual and 75 group interviews. Three types of teamwork were detected, and these were described along with the contexts in which they appear.

The ‘lean production’ system optimizes the production process but reduces work­ers’ autonomy. The ‘social-technical’ system is based on a high level of worker autonomy, although this autonomy can turn out to be costly. The third system, ‘off-line teams’, resolves specific problems but cannot be applied to daily operations.

Description can thus be a first approach; an improvement in our under­standing of a phenomenon that until now was relatively unexplored by the scientific community. The purpose is essentially exploratory. But a major chal­lenge for all researchers is to demonstrate the advantages of their work clearly. Insofar as the researcher is simply describing a phenomenon without being able to indicate its causes, it is absolutely essential to insist upon the theoretical and managerial advantages gained by carrying out the description. This empiri­cal work is also indispensable before explicatory content-based research can be done. It is only with an in-depth knowledge of the elements that make up a phenomenon that researchers can attempt to understand the causal links that develop among these elements and which, in the end, determine the form of the phenomenon under study.

2. Troubleshooting Content-based Research

Research questions, as well as the tools and methods employed, will differ depending on whether the researcher’s goal is to describe or to explain the object of the content-based research.

In carrying out content-based research, the researcher aims to reveal the composition of a variable under study, either descriptively or explicatively. Problems may be encountered in the following areas:

  1. Researchers should pay particular attention to carefully defining the object they are studying. For example, a researcher wishing to learn about the structure of an organization should specify if he or she intends to study the firm’s formal structure, as it might appear in an organizational chart, or include its informal structure as well. In the latter case, the research may bring to light intangible elements of this informal structure. The researcher must clearly formulate the research question relative to either the organiza­tion’s formal structure, or both the formal and informal dimensions of the structure.
  2. This general problem raises a second one, relative to the theoretical or empirical models that researchers may employ in order to understand the object of a study. With the exception of exploratory research on an item that has been little studied in management research, a wealth of theoretical models exists with which to describe or explain phenomena. The questions formulated and methodologies used will vary depending on whether the researcher intends to analyze an item’s content on the basis of theory, or from empirical data.
  3. Researchers must pay particular attention to carefully defining their level of analysis. By defining their level of analysis, researchers may decide they need to refine their study, to study their research object in greater or lesser depth. For example, a researcher who wants to use description to compare the structure of different companies (to look at the functions and organization of their structural elements) should specify ahead to what level he or she wants to take this description. Such a study could include the vari­ous departments represented on the organizational chart and the liaisons between them, the services making up each department and the liaisons between them, or go as far as the individuals making up and working in each service. The choice of the level of decomposition and description will depend above all on the purpose of the research, but also on the material available.

2.1. Principal problems in content-descriptive research

We will present here two possible approaches out of the many available for this type of research. The first approach consists in breaking down, or ‘decompos­ing’ the object under study into a certain number of elementary characteristics. The second approach is more global and aims to better apprehend the object as a whole (identification of its form), rather than to break it into parts. In most cases, however, researchers will need to employ both approaches simultane­ously, without making such a sharp distinction between them.

How is descriptive content-based research done through decomposition? One type of descriptive research aims in particular at improving understanding through decomposing the object. The research then follows the classic Cartesian process. In this situation, the corresponding research question is: what is the object under study made up of? What are its elements? Mintzberg’s 1973 study of the activities of company directors illustrates this type of process.

Example: Researching the activities of company directors

One example of descriptive content analysis is Mintzberg’s (1973) study of the activities of company directors. Mintzberg wanted to describe directors’ actual activities, the way they use their time. The methodology he selected can be broken into three stages. Data was first collected on meetings scheduled over a one-month period, on the organization the manager was part of and on the directors them­selves. Then came a period of structured observation. The researcher observed the directors in action. Each event was coded according to various criteria. In order to avoid an overly restrictive coding, the criteria were not determined by analyzing existing literature but were established during and after the observation process. The study followed five experienced company presidents for a period of one week each. This research, proceeding through decomposition only, enabled the identifi­cation of ten key roles around which directors structure their time (for example, negotiator, intermediary, source of information or company spokesperson). It chal­lenged Fayol’s classic vision of directors wielding tight control over all elements of their organization.

The object under analysis might be one of many things: organizational struc­ture, an individual’s cognitive map, the composition of a group or a decision­making process. In all cases, the goal is to find the elements that compose the studied object. A structure is broken down into subunits, a cognitive map into concepts, a group into individuals, a process into its constitutive elements. The links and relations between the elements also form part of what we are aiming to describe. The methodologies used can be extremely diverse. They can include, for example, network analysis (see Chapter 14), discourse and repre­sentation analysis (see Chapter 16). Analysis of social networks effectively enables us to understand an organization through decomposing it; by breaking it down to the level of individuals and the ties existing between them. We can also attempt to understand a firm’s cohesion by studying the ties between indi­viduals who belong to its constitutive units. By the same token, discourse and representation analyses enable concepts and the links between them to emerge by decomposing the discourses or representations in question. In this manner, one can, for example, try to discover the directors’ principal preoccupations by analyzing interviews on how they run their companies. Analyzing discourse and representations can, among other things, enable us to identify the recurrent themes in these interviews and to reveal key concepts.

How is descriptive content-based research achieved through identifying forms? A second type of descriptive content-based research aims to go beyond decom­position in order to apprehend the research subject as a whole; instead of decomposing it, researchers identify forms. In this case, the researcher’s goal is to underline the interdependence of the elements that make up the research subject. The essential point of these theories is to demonstrate that the proper­ties of a form can be of greater importance than the properties of each element the form is comprised of. For example, no single strategic decision determines a firm’s competitiveness, which is determined rather by the coherence of strat­egy with environmental structure and conditions. What matters most is that these three interdependent elements constitute a balanced form.

The logic of form research is used by a number of different schools of thought in management research (who ascribe to theories similar to that of gestalt). The configurational approach can be found among these schools of thought, and is very popular in management research. It concerns such varied domains as strategic groups, organizational configurations, strategy categories, and leader­ship and management styles. The general principle is to study an object by gathering observations into categories – homogenous groups that facilitate our apprehension of reality. Each category is represented by what can be called a configuration or an ideal type. Elements can then be characterized by their simi­larity to the configuration of the category. Through the use of configurations, researchers introduce a certain amount of order into the complexity of discrete, discontinuous and heterogeneous observations. Each category acts as a sort of benchmark, allowing researchers to work on their contents more precisely. To do this, researchers adopt two distinct approaches: taxonomy uses empirical data to establish the configurations; whereas typology uses a theoretical approach.

Establishing a taxonomy involves an empirical and inductive classification process. It can call upon statistical techniques known as classification and struc­turation (see Chapter 13). Research into strategic groups, for example, usually uses this type of statistical tool to describe the competitive situation within a given industry (Ketchen et al., 1993). Taxonomies improve our understanding of the industry analyzed, by allowing us to generate strategic maps. A taxo­nomy can also be the result of a qualitative approach. An example is Goold and Campbell’s (1987) study, in which the objective was to highlight different management styles.

Example: Goold and Campbell’s (1987) taxonomy of management styles

Goold and Campbell (1987) wanted to examine the role of the head office in large companies, and the way it which it affects peripheral units. From an analysis of existing research, they identified two principal areas in which the head office (the center) was active: determining strategy (or planning influence) and monitoring performance (or central influence). The empirical part of the research was made up of 16 case studies. For each case, 5 to 20 interviews with central management were carried out. These were supplemented by direct observation of certain meetings and by collecting information on formal elements. The data enabled an evaluation of the planning influence and the central influence exercised by central manage­ment in each firm studied. The combination of these two influences allowed the authors to define ‘management styles’. Eight ‘management styles’, that is, eight configurations composed of various combinations of the two types of influence, were established in this way. The authors deduced normative implications from their research. They found that the three of the eight styles stood out, in that they were better able to balance the major tensions within the organizations. These were the styles known as financial control, strategic control and strategic planning.

The second mode of classification is typology. Typology has a conceptual foundation: unlike taxonomies, typologies are not born of empirical research. They can be the result of an analysis of existing research or of the researcher’s accumulated experience. Mintzberg (1980) identifies, for example, five distinct organizational configurations (simple structure, mechanical bureaucracy, pro­fessional bureaucracy, divisional form and ‘adhocracy’), obtained by combining a firm’s different components. Research into organizational structure often calls upon this configurational approach.

The fundamental problem researchers are confronted with when they attempt to define forms is linked to the difficulties involved in establishing boundaries between the forms identified. As we indicated above, each configu­ration corresponds to an ideal type. Even when these ideal types are precisely defined, organizational reality never matches them perfectly. The question is then what status researchers should give the configurations defined by manage­ment research. Another question concerns taxonomies in particular, for which the classification method used has a determinant effect on the configurations obtained. Researchers must be able to justify with precision the choices they are inevitably called upon to make.

2.2. Principal problems in content-explanatory research

This type of research often uses the hypothetical-deductive approach combined with a quantitative method.

Using a hypothetical-deductive method to explain content This is the most frequently used approach in management research. A certain number of hypo­theses are formulated in relation to causal links between variables known as explicative and explained. These links between variables are then tested and interpreted to establish the existence of causality (see Chapter 12). The aim is to explain the variance of the dependent variable, and to understand why it is in a given state. The contingency school has inspired a great amount of explicative content research. The idea of contingency corresponds to a firm’s need to con­stantly adapt to its environment if it is to survive. Govindarajan’s (1988) research follows this line of reasoning.

Example: A contingent approach at the level of the operational unit

Govindarajan (1988) studied the administrative mechanisms governing the rela­tions between general management and the operational units of diverse groups. He observed that different strategies are adopted in operational units depending on their local context. The question of what type of supervision mechanisms should be adopted is then posed. Starting from a review of existing research, the author formu­lated a series of hypotheses on the supervision mechanisms best suited to a given strategy. The hypotheses were of the following type: ‘for an operational unit employ­ing a strategy of cost-domination, reinforcing the importance of achieving goals is associated with high performance’. Data on supervision mechanisms and opera­tional unit performance were collected by questionnaire from 121 unit managers in 24 companies. The hypotheses were then tested by performing a multiple-regression analysis. The research demonstrated the necessity of adapting supervision mecha­nisms to the strategy adopted by each individual operational unit.

Explicative content research often calls upon the results of descriptive con­tent studies. These studies provide the concepts or configurations required to formulate hypotheses as well as the operationalization of research variables.

Other possibilities for explicative content research Quantitative hypothetical- deductive research has dominated research management for a long time. It aims for high external validity and encourages an accumulation of knowledge. Still, it presents two limitations. First, the use of numeric data often requires proxy variables to be defined, which transforms reality. Second, the rigid framework of the hypothetical-deductive process can impede the emergence of new ideas.

Other possibilities do exist. Researchers can use a qualitative process, and rediscover all the subtlety of the reality studied. This does not exclude the formulation of propositions, which are then confronted with reality via case studies. Researchers may also use an inductive process; in this case causal links emerge directly through fieldwork. Tellis and Golder’s (1996) study illustrates the possibility of revealing causal links through an inductive and qualitative approach.

Example: An inductive and qualitative approach to finding causality

Tellis and Golder (1996) observed that the pioneer advantage innovators have only rarely translates into a dominant situation in the middle term. They then set out to establish what caused certain pioneers to maintain their dominant position. In order to achieve this, Tellis and Golder used a historical reconstitution method, which allowed them to study the positions of firms in some 50 product categories. They triangulated their study by analyzing both documents from the period studied (1,500 periodicals and 275 books) and data gathered directly from experts. They were therefore able to reveal determinants which had not been perceptible to actors in the sector itself. The authors identified five factors that conditioned high perfor­mance: a vision of the existing mass market based on an innovative idea, manage­rial perseverance, allocation of sufficient financial resources, continual innovation and synergy.

Source: Thietart Raymond-Alain et al. (2001), Doing Management Research: A Comprehensive Guide, SAGE Publications Ltd; 1 edition.

Leave a Reply

Your email address will not be published. Required fields are marked *