Empirical Data and Theory Construction: An Example of Application in Social Science Research
University Roma Tre, Italy
Des Données Empiriques à la Construction de la Théorie : L’approche classique de la sociologie et des autres sciences aussi se rapport à l’usage des hypothèses et des procédures de vérification des hypothèses pour établir les effets, s’il y en a, de certaines variables dans les cas étudiés. Mais d’autres savants suggèrent d’autres solutions pour éliminer la présence des hypothèses et pour privilégier une approche interprétative – opposée à celle hypothétique et déductive –. De toute façon dans cette dernière approche il y a des problèmes pour le chercheur : comment approcher les données et les analyser pour assurer la fiabilité des résultats ? La suggestion de Blumer d’utiliser les “concepts sensibilisants” semble correcte pour arriver à une interprétation fiable mais capable aussi de construire une théorie. Cette procédure peut se réaliser grâce à l’aide des logiciels.
The classical approach in sociology, as much as in other scientific fields, concerns the use of hypotheses and hypothesis testing processes to determine what, if any, effects can be attributed to particular factors in the field being studied. But other scholars suggest another procedure which eliminates the presence of previous hypotheses and privileges an interpretive – as opposed to a hypothetic-deductive – approach to data analysis. However, in the latter approach crucial questions appear for the researcher: how does one approach the data and data analyses to ensure the credibility of findings? Blumer’s suggestion about “sensitizing concepts” seems to be appropriate in order to reach not only a reliable interpretation of data but also the possibility of “building theory”. This process could be supported by computer-assisted research.
Méthodologie, Approche Qualitative, “Concepts Sensibilisants”, Analyse Assistée par l’Ordinateur, Grounded Theory
Methodology, Qualitative Approach, “Sensitizing Concepts”, Computer-Assisted Analysis, “Grounded Theory”
Roberto Cipriani, via del Mascherino 75, 00193 Roma, Italia
A methodological solution becomes more and more reliable: to build a theory starting from data rather than from theory-driven hypothesis testing. The latter opens a new debate between competing and perhaps conflicting positions: the classical approach through previous hypotheses, and an orientation influenced by Grounded Theory (Glaser and Strauss 1967; Charmaz 2006; Corbin and Strauss 2008), which eliminates any kind of hypotheses and offers a range of different scientific views, also thanks to the added presence of “sensitizing concepts” (Blumer 1954) coming out from empirical results, therefore not previously established but inserted after the fieldwork.
The Grounded Theory approach
Actually, in adherence to the Grounded Theory the aim is to build sociological theory starting from the collected data: before the analysis the researchers are involved in the construction of a list of nodes, keywords, items, considered meaningful for both the analysis of material and the purpose of the research. A first draft of this list can be useful to check if chosen “sensitizing concepts” are present or not in the contents of interviews. The purpose is to construct a reliable grid of concepts for the analysis and interpretation of data, a kind of “dictionary”, formed by words selected to support the qualitative approach.
It is not by chance that Paul Boghossian (2006) has suggested an overcoming of constructionist approach to come back to the empirical data in themselves. Of course, the existing computer packets, and probably those that will developed in the future, are helpful tools but cannot substitute for the work of a researcher, engaged in a theory building process (Urquhart 2007: 348-350). The Grounded Theory approach refers exactly to a research process where “the researcher analyzes the data and identifies analytic leads and tentative categories to develop through further data collection. A grounded theory of a studied topic starts with concrete data and ends with rendering them in an explanatory theory” (Charmaz 2007: 2023).
Many programs of qualitative analysis have been developed to respond to theoretical demands of contemporary researchers. But rather than producing results in the way that statistical processing packages analyse data sets and produce statistics, QDA (Qualitative Data Analysis) programs (Lewins and Silver 2007) are powerful supports in treating and controlling data. In the richness of elements offered by qualitative contents researchers can find new concepts, new categories, raising other issues, and generating theories.
For instance a software like NVivo or Atlas-ti supplies many functions besides retrieval and encoding. NVivo, in particular,supports qualitative research, and it is classifiable, at least in part, as tailored to the construction and representation of theory. According to Bryant and Charmaz (2007: 24), however
ultimately the research process must remain under the control of researcher(s). Glaser and others are correct to be wary of use of software, particularly when researchers come to rely upon it. Yet, cases abound where use of some form of electronic repository, plus sorting and retrieval facilities has proved useful. Researchers must understand both the benefits and the dangers of use (and reliance upon) software support.
It is in the choice of “sensitizing concepts” that researcher plays a key role because “such theoretical categories can sensitize the researcher to identify theoretically relevant phenomena in their field” (Kelle 2007: 207). This helps to distinguish between common sense categories, coming from common sense language and knowledge, and directions along which to look in a methodological and heuristic perspective. As a matter of fact
sensitizing concepts can fulfil an important role in empirical research, since their lack of empirical contents permits researchers to apply them to a wide array of phenomena. Regardless how empirically contentless and vague they are, they may serve as heuristic devices for the construction of empirically grounded categories (Kelle 2007: 208).
In the context of “building theory” process, Grounded Theory has gone through many and significant changes since its first introduction. “Quite apart from the question of whether it is desirable to defer theoretical reflection, the notion that one may conduct research in a theory-neutral way is open to some doubt” says Alan Bryman (1988: 84-85), a quantitative and qualitative methodologist. Besides
there may be considerable practical difficulties associated with field-work conducted within a grounded theory approach. For instance Hammersley (1984), drawing on his experience of conducting a school ethnography, has suggested that when field-work entails tape recording of conversations, interviews, lessons, and the like, the time needed to transcribe such materials may render the grounded theory framework, of a constant interweaving of categories and data, almost impossible to accomplish. One might also question whether what the grounded theory approach provides really is a theory. Much of the discussion of the approach and its associated procedures seems to concentrate on the generation of categories rather than theory as such (Bryman 1988: 85).
Some changes had been initially co-proposed by Glaser and Strauss (1967), and then developed separately by these two, thus generating two different groups following one or the other scientist.
According to Bryant and Charmaz (2007: 33),
through developing this method, Glaser and Strauss aimed to provide a clear basis for systematic qualitative research, although Glaser has always argued that the method applies equally to quantitative inquiry. They intended to show how such research projects could produce outcomes of equal significance to those produced by the predominant statistical-quantitative, primarily mass survey methods of the day. What they also achieved was a redirection of positivist-oriented concern among qualitative researchers seeking reliability and validity in response to criticism from quantitative methodologists. Glaser and Strauss offered a method with a solid core of data analysis and theory construction. Their method contrasted with the strategy of those who sought procedural respectability through collection of vast amounts of unanalysed, and often un-analysable, data.
After Strauss died, Glaser has been holding the stage with his strong committed attitude and fully oriented towards ‘his’ Grounded Theory. Barney Glaser (1992) has pushed through his scientific and methodological option far off the former threshold, which was already quite revolutionary for sociological tradition in the US, and not only there. The most problematic issue of Grounded Theory is the total absence of an orientating perspective in sociological research. In other words, if the idea of giving up detailed hypothesis for the research may be considered, not the same can be said regarding other potential inquiries to introduce in the guidelines of the research. However, once refused the traditional initial research-hypothesis, an incipit can be a good solution to start and, though temporary and open, it can represent a referring point, a focus, an orientating spot, a common basis from where inquiries can take off. Therefore, reconsidering Grounded Theory is even more possible if we take into account what Herbert Blumer maintained, the same Blumer who criticized the supposed absence of methodology of Thomas and Znaniecki (1918-1920) in their famous qualitative fieldwork on The Polish Peasant. On this note, as Blumer suggested (1954: 7), “sensitizing concepts” can be a good solution, so to have at least some conceptually defined contents, which may represent some sort of guiding lines, or a rough draft that, even though of generic references, could be able to guide operative choices during the research.
As Blumer maintains, “sensitizing concepts” offer a wide general sense of reference and guidance in approaching empirical instances. In such a way, there is a close relation between concepts and data. Use of “sensitizing concepts” can be very diversified. Concepts can be taken directly from data, that is the case of Grounded Theory purists, who refuse any other solution that does not start from data (the head of the list in this integral and militant approach is Barney Glaser). According to other scientists, “sensitizing concepts” must be confronted with data and they have to be formulated starting from already existing sociological literature on the subject. In this way the relation with empirical data is quite loose, because conceptualization comes before any data collecting. However, there is another solution that cumulates advantages coming from the store of scientific knowledge, previously accumulated, and empirical information collected. It means to keep memory of the previous research made on the same subject and formulate a first draft of some guidance concepts, or “sensitizing concepts”, to be verified on fieldwork of the research. In operative terms, the researcher can propose his own results, at least at a first step, or results acquired by other researchers on the same subject. Therefore, the researcher can prepare a list of temporary concepts, as much as it is possible.
In the second phase, the temporary list of concepts is compared with empirical data collected, in order to verify, through a simple table of presences and absences related to single concepts, whether or not conceptual formulas can be found in the texts, in interviews, in the data-base at disposal. Such verification can be carried out with a certain precision on all data or on just a significant part of them. However, before setting up a definite series of “sensitizing concepts” to be used in subsequent data analysis steps, a good approach could be that of discussing every choice with all researchers and collectors involved in the research in order to decide which concepts should be definitely included, excluded or added. If research is conducted by an individual the choice can be made by the same person in different steps, one after another, in order to reach a final decision.
Actually, in Kelle’s (2007: 209) terms,
a wide array of sensitizing categories from different theoretical traditions can be used to develop empirically grounded categories. Many researchers find it easier to let categories emerge if one stays with one particular theoretical tradition, however Glaser is certainly when with his frequent warnings that the utilization of a single pet theory will almost necessarily lead to the neglect of heuristic concepts better suited to the specific domain under scrutiny. There are heuristic concepts which capture a broad variety of different processes and events and nevertheless may exclude certain phenomena from being analyzed: thus the extended use of concepts from micro-sociological action theory (e.g. actors, goals, strategies) can preclude a system theory and macro-perspective on the research domain. A strategy of coding which uses different and even competing theoretical perspectives may often be superior to a strategy which remains restricted to a limited number of pet concepts. Furthermore, analysts should always ask themselves whether the chosen heuristic categories lead to the exclusion of certain processes and events from being analyzed and coded, since this would be an attribute of a category with high empirical content which refers to a circumscribed set of phenomena.
Abduction and retroduction
At this point the old dilemma between induction and deduction is solved because the modality of “sensitizing concepts” is to be a sort of abduction or retroduction (Fann 1980) that explains phenomena starting from facts but not only in an inductive way. According to Peirce (1868, 1984), the father of pragmatist movement, one abduces or retroduces with the acceptation of a hypothesis in order to explain a given phenomenon. However, the starting point is always the formulation of a hypothesis that has to be verified by facts. This is Peirce’s logic of science. Furthermore, according to Peirce a concept is significant when it produces effects. For “sensitizing concepts” it is quite evident that they are directly related with data. Finally, categories (or concepts) have to be determined both at the beginning and at the end of the analytical procedure.
One might say that knowledge is based on observing facts (Peirce 2001: 289). The following example can be useful: observing an ink-pot, this is a fact; however, before one can say that, one can have sensitive impressions, in which there is no idea of an ink-pot, or of any separated object, or of a ‘self’, or of the act of observing; and the fact of seeing an ink-pot in front of oneself is the outcome of a number of mental operations over such sensitive impressions. Only when the cognition has developed into a proposition or idea over a fact, one can directly control the process. As a matter of fact it is an idle question to discuss ‘legitimacy’ of what cannot be checked. Therefore, all observations should be accepted as they occur. In a certain way, Peirce’s (1931-1958) use of the language is quite old fashioned, and above we have presented a paraphrase of his speech, but the intuitions contained are significant and anticipate times. Particularly explicit is the following statement: “The first thing to do is to propose a questioning hypothesis. Secondly, one should verify feasibility limits with experimental tests”. Hypothesis is an instrument of Peirce’s scientific and philosophic work (2001: 290). He maintains that with hypothesis he does not only mean a supposition on an observed object; in particular, he means that the initial position of a hypothesis and how is it considerate, either as a simple question and at any degree of faith, can be seen as an inferential step that Peirce suggests could be called abduction. This includes the preference for one hypothesis among others to explain the facts, until such preferred hypothesis is neither proved on a previous relevant knowledge, nor verified over other hypothesis already being proved. Finally, Peirce (2001: 304) mentions some of the problems from absolute absence of hypothesis in research. There are scholars who affirm that no hypothesis can be accepted, not even as a hypothesis, until its rightness or wrongness can be directly perceived. Peirce holds the opinion that this is what Auguste Comte (the social philosopher who actually first formulated this) had in mind. Of course, this abduction presupposes that we should believe only what we see, and there are authors who maintain that to make predictions is not a scientific attitude. Therefore, this should also mean that expecting something from research is not a scientific attitude as well. One’s opinion should be limited to what effectively can be perceived, maintains Peirce, as well as he seems to be perfectly aware that such a position cannot be coherently held. In a certain way, this position is auto-denying because it is an opinion based on more than effectively can be perceived.
The computer-assisted qualitative analysis
Analysis of qualitative data, according to the methodological indications of Grounded Theory and its feasibility offered by the software NVivo, operates at two major levels: nodes, that is to say the key-concepts, and memos, that is to say observations, considerations, theoretical and scientific perceptions coming from data handling. Sometimes one may work a lot, or mainly, on nodes and just a little with memos, which may be neglected and written very seldom. The absence or insufficient number of memos may represent a serious damage for more important operations within the research. As a matter of fact, not only nodes can be in relation to each other, but also memos, and memos can be in relation with nodes, thus enriching with no limits interpretive potentiality of researchers. The group of memos, but also every single memo, is a vital piece of the chain through which one should pass to build theories, attempt interpretations, set significant relations with different parts and results of a research. Memos are the traces of our thoughts over the research problems. Such thoughts in progress may change, but can also consolidate, according to: central and marginal variables; principal and secondary ones; with a hierarchy at any level; with the tree system, which defines priorities and gradual differences of the considered elements. The next passage, which defines one or more theories, is the peculiar job of social scientist, who assumes the role of transforming empirical data into abstract theories.
Quantity and quality in the research on jubilant people
Many of the results of our inquiry with the questionnaire carried out among pilgrims at the Jubilee 2000 (Cipolla, Cipriani 2002) are validated by qualitative research among the 96 jubilant people interviewed. Not only do we have confirmations, but there are also details that strengthen what is said in the conclusion of the quantitative volume Pellegrini del Giubileo (Pilgrims of the Jubilee).
The Jubilee has represented many different things for the participants, both for individual subjects and for the various ‘pilgrims’ compared. In this sense it is possible to understand a more-or-less strong attitude, which is a praxis as well, that can be considered a form of auto-direction (scarce institutional direction), which includes one-third of the sample. On the other hand, we find faithful believers, who are another third of the sample, with a further third who are in an intermediate position or of a more solipsistic nature…. [Moreover,] from our data emerge a critical ability and an autonomy on the part of jubilant people that is dearly greater than what we would have expected, owing to an interplay of orthodoxy and heterodoxy which is not easy to understand and is highly articulated (Cipolla 2002: 14).
The hypothesis formulated by Luigi Berzano and Daniela Teagno, according to which “the pilgrim, arriving in Rome in 2000, perceived and lived the Jubilee according to his sensibility and religious education” is confirmed. “The importance of tradition, spiritual experience, cultural events, and touristic occasion expresses a wide variety of possibilities of approaching religion (belief, experience, belonging, practice, etc.), which are realized in just as many forms of sensibilities and religious subcultures”. Our group of 96 interviewees is likewise “situated within a religious model of Catholic origins, which is why one seems not to notice the process of secularization: such a strong presence of regular practicing believers is particularly significant”. It is not by chance that the pilgrimage to Rome “is, for the faithful, a way to express one’s real religiosity: which is a religious practice more than an intellectual activity, made of materiality and symbolism”. Moreover it is “difficult to talk about a pilgrimage, not only for quantitative reasons, but also for the great number of participants coming from all over the world, because those have different orientations and behaviors, which lead to a plural modality of interpreting and living the Jubilee event” (Berzano and Teagno, 2002: 44-46).
An experiment on the Jubilee of the year 2000
Our results of a qualitative research on jubilant people, to say pilgrims participating in the Jubilee of the Year 2000 in Rome, give a contribution to the sociological analysis and interpretation of a collective phenomenon of mobility (be it religious or not), and to the process of “building theory”.
The qualitative research carried out in 2000 and the years after, starting from data collected over 96 pilgrims in Rome (from 18 countries, and speaking 8 different languages) in occasion of the Great Jubilee of 2000 (Cipriani 2003; 2006), has been, as far as we know, the first research in Italy based on Grounded Theory and realized with NVivo. Actually not all resources of the software have been used. However, the results are to be considered quite promising, even if further interpretations of collected data are still possible. This research shows the outcome of a triangulation between the software DiscAn, invented by Pierre Maranda (a Canadian anthropologist), the Analysis of Lexical Correspondences, and the results of operations with NVivo.
The interviewees, both men and women, presented diverse backgrounds, resulting from their freedom of expression, without any kind of restriction and with no pre-defined questions and pre-coded responses, giving way to a high level of spontaneity in the answers and as consequence a deeper knowledge of some issues concerning the Jubilee.
When the qualitative and quantitative approach are used simultaneously the outcome is very rich and permits to reach more evidences. To give an example, at the end the most important concepts are situated in the following manner:
Religiosity (religiosità) appears as the core category, together with jubilee (giubileo) and faith (fede), and also emotion (emozione). Church is marginal, and the family too is peripheral. Another similar graphic construction is suggested through the software NVivo, which presents a clear relationship between the same primary concepts (religiosity, faith, jubilee, emotion):
To complete the illustrations of results a semantic map, created through DiscAn, confirms the key role of Jubilee (as a central relay=R), religiosity (as a source=F), and faith (as a relay=R). In this case, however, emotion doesn’t seem so relevant: it is a relay but isolated.
The documents gathered could be used for other interpretations too, but our outcome has been useful to build a provisional theory of events based on collective religious mobility. The final theoretical framework is the following, in short:
PRIMITIVE GENERAL TERMS
PRIMITIVE PARTICULAR TERMS
SECONDARY PARTICULAR TERMS
If there is a background religiosity a participation in the events like Jubilee is foreseeable.
Religiosity is the motivation for pilgrims mobility.
Church as institution
It is rare that a non believer could be a pilgrim.
If an individual is Catholic he will participate in Jubilee.
Possibility of enlargement
Higher religiosity corresponds to higher participation in the events of collective religious mobility.
Faith is a good support to participate in pilgrimages.
Possibility of modifications
Jubilee is much more a relay than a source of collective. behavior.
Faith comes from official Church teaching.
Use of a general condition
Church is marginal in the phenomenon of collective mobility.
Motivations to participate in pilgrimages do not come from official Church teaching.
No details of specific conditions
Jubilee motivations are complex.
Temporary, local, and undefined applicability
Quantitative approach and qualitative approach can be very fertile if treated and balanced together with methodological rigor and scientific attention, which means to aim at keeping the best of both different approaches. These can unveil useful elements over social action motivations, the preeminent values orienting experience, more recurring sociological categories in the perception of social reality, and finally, the connections that motivate most significant choices.
As a general rule, according to the analysis of content of the 96 biographies gathered, some conclusions can be drawn. Without pushing through a certain limit the level of analysis, according to the usual criteria of Grounded Theory (Glaser and Strauss 1967; Strauss and Corbin 1990), we suggest some interpretation, the validity of which remains partial and temporary, but can be considered rather innovative in light of previous attempts. The plurality of scientific modalities and techniques applied, from the software NVivo to psycho-sociological perspectives, offers sufficient guarantees of reliability not always acquired with other solutions. Hence, with these cautions, it is possible to explain some significant or exemplary trends.
Berzano L and Teagno D (2002) Pratiche giubilari tra fede e cultura. In: Cipolla C and Cipriani R (eds) Pellegrini del Giubileo. Milan: Franco Angeli, pp. 42-61.
Blumer H (1954) What is Wrong with Social Theory? American Sociological Rewiew 19: 3-10.
Boghossian P (2006) Fear of Knowledge. Oxford: Oxford University Press.
Bryant A and, Charmaz C (2007) Introduction. In: Bryant A and Charmaz C (eds) The SAGE Handbook of Grounded Theory. London-New Delhi-Thousand Oaks-Singapore: Sage, pp.1-28.
Bryman A (1988) Quantity and Quality in Social Research. London: Unwin Hyman.
Charmaz C (2006) Constructing Grounded Theory. A Practical Guide Through Qualitative Analysis. Thousand Oaks (Ca): Sage.
Charmaz C (2007) Grounded Theory. In: Ritzer G (ed) The Blackwell Encyclopedia of Sociology. Oxford: Blackwell, pp. 2023-2027.
Cipolla C (2002) Introduzione. In: Cipolla C and Cipriani R (eds) Pellegrini del Giubileo. Milan: Franco Angeli, pp. 9-22.
Cipolla C and Cipriani R (eds) (2002) Pellegrini del Giubileo. Milan: FrancoAngeli.
Cipriani R (2002) Conclusioni In: Cipolla C and Cipriani R (eds) Pellegrini del Giubileo. Milan: Franco Angeli, pp. 164-169.
Cipriani R (ed) (2003) Giubilanti del 2000. Percorsi di vita. Milano: FrancoAngeli.
Cipriani R (2006) Jubilee 2000. A Computer-Assisted Analysis of the Religion of Jubilant People. In: Swatos WH (ed) On the Road to Being There. Studies in Pilgrimage and Tourism in Late Modernity. Leiden-Boston: Brill, pp. 219-253.
Corbin J and, Strauss AL (2008) Basics of Qualitative Research. Thousand Oaks (Ca.): Sage.
Fann KT (1980) Peirce’s Theory of Abduction. The Hague: Martinus Nijhoff.
Glaser BG (1992) Emergence vs Forcing: Basics of Grounded Theory Analysis. Mill Valley: Sociology Press.
Glaser BG and Strauss AL (1967) The Discovery of Grounded Theory. Chicago: Aldine.
Hammersley M (1984) The Researcher Exposed: A Natural History. In: Burgess RG (ed) The Research Process in Educational Settings: Ten Case Studies. London: Falmer Press, pp. 39-67.
Kelle U (2007) The Development of Categories: Different Approaches in Grounded Theory. In: Bryant A and Charmaz C (eds) The SAGE Handbook of Grounded Theory. London-New Delhi-Thousand Oaks-Singapore: Sage, pp. 191-213.
Lewins A and Silver C (2007) Using Software in Qualitative Research. A Step by Step Guide. Thousand Oaks (Ca): Sage.
Peirce CS (1868) Some Consequences of Four Incapacities. In: Hartshorne C, Weiss P, Burks A (eds) Collected Papers of Charles Sanders Peirce, 1931-1958, 8 volumes; 1965-1967, 4 double volumes. Cambridge (Mass.): Harvard University Press.
Peirce CS (1931-1958) Collected Papers of Charles Sanders Peirce. Cambridge (Mass.): Harvard University Press.
Peirce CS (1984) Le leggi dell’ipotesi. Milano: Bompiani.
Peirce CS (2001) La logica dell’abduzione. In: Scritti di filosofia. Milano: Fratelli Fabbri, pp. 289-305.
Strauss A and Corbin J (1990) Basics of Qualitatìve Research. Newbury Park, CA: Sage.
Thomas WI and Znaniecki F (1918-1920) The Polish Peasant in Europe and America, 5 volumes. Chicago: University of Chicago Press (transl. (1968) Il contadino polacco in Europa e in America, 2 volumes. Milano: Comunità).
Urquhart C (2007) The Evolving Nature of Grounded Theory Method: The Case of the Information Systems Discipline. In: Bryant A and Charmaz C (eds) The SAGE Handbook of Grounded Theory. London-New Delhi-Thousand Oaks-Singapore: Sage.