CC BY
12
Статья / Article
УДК / UDC 2 (450)
Quality of religion in Italy. A multi-methods approach
Roberto Cipriani
Roma Tre University (Università degli Studi Roma Tre), Roma, Italy
Introduction. A research on religiosity in Italy has been carried out in 2017, covering 164 subjects appropriately selected, without the pretension of representativeness. The solution of a completely open interview was tested. For the rest of the 86 subjects, the first part was entirely free and the second part was on some concepts-stimuli: daily and festive life, happiness and pain, life and death, God, prayer, religious institutions and Pope Francis. The results were corroborated with T2K (Text-to-Knowledge), Lexical Correspondences Analysis and VoSpec procedure (Vocabulaire Spécifique des Groupes d'individus). An analysis sheet similar to a semi-structured questionnaire was then prepared.
Materials and methods. The subjects to be interviewed were selected through a stratification of three categories related to the degree of study obtained (level of obligation, upper-middle school diploma, degree), the distinction of gender (man, woman), residence (small towns, medium municipalities, large cities), geographical distribution (north, centre, south and islands) and age (young, adults, elderly).
Results. For almost half of the qualitative sample, i.e. 78 cases, the interviewers tried to obtain narratives, reflections, interpretations and evaluations that were not at all solicited through direct or indirect questions on the theme of religiosity. On the other hand, for the rest of the 86 subjects interviewed, a twofold method was used within the same interview, that is, a first part that could be freely used by the interviewee in terms of content and a second part to be addressed, based on some concepts-stimuli proposed by the interviewer to solicit answers on daily and festive life, happiness and pain, life and death, God, prayer, religious institutions and pope Francis.
Conclusion. As a result of empirical substantiation and theoretical development of the main categories, five fundamental concepts have emerged that can give rise to a theory: (1) "Pastoral care" (believers who help other people by drawing inspiration from the values of Christianity), (2) "Rhetoric of charitable humanitarianism" (people who perceive other people as different and dangerous and, at the same time, tell of the effort to welcome and accept the needy), (3) "Prisoners of despair" (they experience the difficulties of life in a passive and resigned way and takes refuge in God and prayer), (4) "Embracing faith" (practicing believers, guided by Christian precepts and living the faith as dogma) and (5) "Beyond everything" (individuals, who are discordant on the themes of life, death, and the afterlife, but, beyond ideas, actions, and skepticism, they have faith in God).
© Roberto Cipriani, 2020
Key words: religiosity, Italy, qualitative analysis, T2K2 (Text-to-Knowledge, v. 2), lexical correspondences analysis, specific vocabulary procedure, content analysis as investigation, Grounded Theory, analysis of discursive dynamics, bimatrix and multidimensional analysis.
For citation: Cipriani, R. (2020). Quality of religion in Italy. A multi-methods approach. Vestnik Leningradskogo gosudarstvennogo universiteta imeni A. S. Pushkina - Pushkin Leningrad State University Journal. No 3. pp. 193-207. (In English).
Introduction
Twenty-five years after the research on religiosity in Italy [4], the new survey carried out in 2017 on the same theme presents a significant novelty: the approach was no longer only quantitative, by administering a questionnaire to a statistically representative sample of the entire Italian population, but also concerned a group of 164 subjects suitably selected on the national territory, following criteria not far from the overall demographic picture, even if there is no claim to statistical representativeness and therefore to a generalization of the results.
The subjects to be interviewed were selected through a stratification of three categories related to the degree of study obtained (level of obligation, upper-middle school diploma, degree), the distinction of gender (man, woman), residence (small towns, medium municipalities, large cities), geographical distribution (north, centre, south and islands) and age (young, adults, elderly).
For almost half of the qualitative sample, i.e. 78 cases, the interviewers tried to obtain narratives, reflections, interpretations and evaluations that were not at all solicited through direct or indirect questions on the theme of religiosity. On the other hand, for the rest of the 86 subjects interviewed, a twofold method was used within the same interview, that is, a first part that could be freely used by the interviewee in terms of content and a second part to be addressed, based on some concepts-stimuli proposed by the interviewer to solicit answers on daily and festive life, happiness and pain, life and death, God, prayer, religious institutions and pope Francis.
In this way, on the one hand, we tried to collect the so-called natural data that the most recent sociological literature, in the field of qualitative approach, hopes for, on several occasions, and, on the other hand, we tried to retrace once again the path of focused interviews, which have already produced a long series of reliable scientific results. The intention was to propose a path of mixed methods, and diversified solutions, in the conviction of being able to better understand the sociological phenomena under examination.
The mixed solution does not only concern the interview model, defined from time to time as UNI (i.e. uniform in its total willingness to manage the interview by the interviewee), or MIX (in its twofold possibility: first of all, open speech and then contents guided by some specific questions, suggested by the interviewer to focus the interview on socio-religious issues).
Materials and methods
Quantitative tools and qualitative analysis
In our research on religiosity in Italy we tried to corroborate the results of the qualitative analysis with sophisticated quantitative instruments, in terms of mixed methods [2; 19; 24; 25], that offered significant inputs for the whole study. So, as in the past, on the occasion of another qualitative research in the field [13, pp. 216262], the Institute of Computational Linguistics of the National Research Council (based in Pisa), now named after Antonio Zampolli and directed by Simonetta Mon-temagni, intervened on the subject. In particular, Andrea Cimino, Felice dell'Orletta and Giulia Venturi applied to the texts of the 164 interviews an analysis program, called T2K (Text-to-Knowledge), recently developed [17].
The need to access large amounts of data, mostly textual, has given great impetus to the development of technologies for the acquisition, classification and automatic management of textual content and their increasingly widespread use in a myriad of application contexts.
T2K is a platform aimed at acquiring semantic-lexical information from domain corpora. Through the combined use of statistical techniques and advanced tools of Automatic Language Processing, T2K can analyze the content of documents, extract the most potentially significant terms and entities (names, places, people), identify the relationships that bind terms and entities, index the starting body and reconstruct a multidimensional map of domain knowledge contained in the document collection.
It is a platform for the automatic extraction of linguistic and domain-specific information from document collections. It offers a structured organization of extracted knowledge and indexes the analyzed texts against the extracted information. It depends on a set of tools for Natural Language Processing (NLP), statistical analysis of texts and automatic learning, which are dynamically integrated to provide an accurate representation of language information and specific content.
Graph. 1 - Automatic text analysis: Text-to-Knowledge (T2K)
Contexts-interviews
Lj
Dimensions
Graphs
Graphs -only concepts Graphs -terms concepts
Interviews-all-texts.txt
Binary and Trinary Relationships
Binary and Tiinary Relationships Graphs (binary relationships graphs/trinary relationships graphs)
Matrix Relationships (binary reía tionsliips/ümary relationships)
Sentiment analysis
Extracted terms + sensitizing concepts
В TNT (Broader Temí / Narrower Term) Glossary Formatted index Maüixes-relatioiisliips
Allinterviews
Verb-term
Glossary Formatted index Matrixes-relationships
AUTOMATIC TEXT ANALYSIS
Extraction of linguistic and contextual information
Concepts
Glossary Formatted index Matrixes-relationships
Contexts
Content analysis as investigation
Another important operation was carried out under the direction of Maria Paola Faggiano, providing a content analysis as investigation [18; 20; 21; 26], applied to the texts of the qualitative interviews. The work of the analysts was coordinated by Stefano Delli Poggi [16].
In practice, a survey form has been prepared that can be assimilated to a semi-structured questionnaire (conceived with an "ideal transcription" in mind, in which all the topics and aspects connected with the interview track were touched upon), first tested on a small nucleus of transcribed texts. The instrument, in its tested and definitive version, was subsequently applied to all interviews with classifying intent, to capture recurring patterns, values and social representations, also according to the variables used in the sampling (age, gender, level of education, geographical area of residence).
The predisposition of a data matrix constituted by the intersection of cases and variables in which, where possible, to manage structured and codified elements (e.g. "overall judgment on the Pope", articulated in the modalities "positive", "ambivalent", "negative"), has allowed, for a part of the collected data, to conduct typical quantitative analyses at mono- and bi-varied level. The quantitative analysis in the strict sense, although simple and descriptive, was carried out in cases where the single information concerned all or almost all the cases and where the modalities of the variable were used or transformed into a standardized scheme. Data processing did not go beyond the bi-varied level, given the small size of the sample and the presence of incomplete profiles; moreover, more or less extensive, articulated, heterogeneous and deep interviews were available.
The matrix concerned, in addition to the socio-demographic variables, a whole series of aspects. It can be said that the analysis format has allowed to "sift" the complete transcriptions of the interviews, allowing in the analysis to manage (recode, synthesize, lead back to broader and more abstract categories, etc.) only the parts of the text related to the interests of research. The interviews are in these terms "disassembled" and "reassembled" with the purposes of the analysis.
The overall report has a predominantly discursive character (for which it is possible to quote numerous and interesting "excerpts of the interview"), in consideration of the already mentioned aspect of the partiality of numerous variables considered (in some columns of the matrix there are a few dozens filled cells). The small number of some indications corresponds more often to the depth and relevance of the materials selected and valued in the matrix. The reconstruction of the phenomenology implemented also makes it possible to combine qualitative and quantitative aspects.
Graph. 2 - Content analysis as investigation
CONTENT ANALYSIS AS
INVESTIGATION
Analysis of the texts of the qualitative interviews
Semi-structured survey form
Topics and aspects related to research
Data matrix
Crossing of 164 " cases with 160 variables
Results
Analysis of discursive dynamics
Maria Paola Piccini analyzed discursive dynamics detectable in qualitative interviews, to discover and highlight significant flows of the direct connection between the terms used, in particular between those identified as relevant concepts for the investigation. She worked on the transcriptions of the interviews coding the autonomous units of meaning into thematic and/or semantic categories that, essentially, correspond to the set of sensitizing concepts, collectively identified by the research team. This set has been integrated by some additional categories considered, howev-
er, relevant and based on the analysis of total occurrences in the body of interviews. This way, a first meta-text was prepared to consist of a complex of 320 conceptual categories. Subsequently, a second meta-text was prepared, by recoding the conceptual categories into 24 conceptual macro-dimensions. On both versions of the corpus, the Lexical Correspondences Analysis [15; 21] and the VoSpec procedure (which, in this case, offers not a specific vocabulary but a specific dictionary) were implemented, depending, respectively, on the following variables: type of interview, gender, age classes, educational qualification, territorial location, size of the municipality of residence, country of origin, employment status and marital status of the interviewees. Using the same information, a cluster analysis was carried out, so that VoSpec could be performed according to the partition into clusters, to make the textual dimensions and the typology of the respondents interact. Correspondences analysis and clustering procedures are included in the SPAD (Système Portable pour l 'Analyse des Données) software. In particular, the Lexical Correspondences Analysis applies to textual data. It is a technique developed in the context of the Analyse des Données by the French school of Jean-Paul Benzécri [3], which, with a factorial procedure, identifies dimensions underlying the data, by synthesizing the relationships between the original variables consisting of the words (or categories of words) present in the corpus under examination. These dimensions are called factors, while the original variables are called active variables. The Lexical Correspondences Analysis allows to graphically representing the associations between words and texts, on planes delimited by two factorial axes. SPAD, on the other hand, allows extra-textual variables to be associated with the text, such as, for example, those relating to the socio-demographic characteristics and to relate the characteristics of the texts themselves, found through the lexical forms or thematic categories present in the whole. Finally, the VoSpec procedure (Vocabulaire Spécifique des Groupes d'individus) identifies, always within the texts, the words or categories of words characteristic of certain groups of individuals (clusters).
Then, using the customized settings of the T-LAB software, the Probabilistic Concatenation Analysis of Semantic Nodes [22; 23] was carried out on the version of the corpus codified in 24 dimensions, finding roles and semantic activities of the different nodes. The software allows calculating the probability of transition from each category to each of those that, respectively, precede and follow it. In this way, we reconstruct the network of probabilistic concatenations that bind together the categories, considered, in fact, as "semantic nodes", for each of which it is possible to calculate the "degree of reception" or "internal semi grade", i.e. the input represented by the nodes that precede it, and the "degree of emission" or "external semi grade", i.e. the output represented by the nodes that follow it. It has been possible, moreover, to consider the semantic activity of each node, which expresses the intensity with which the respective function of attraction, diffraction or transmission is carried out.
Based on all this information, micro maps have been drawn and centered, respectively, on the source, relay and condenser nodes, characterized by the highest semantic activity (work, family, faith). Finally, a map was prepared which graphically represents the overall discursive dynamics of the interviews.
Graph. 3 - Analysis of Discursive Dynamics
ANALYSIS OF
DISCURSIVE DYNAMICS
Analysis of the texts of the qualitative interviews
I
Coding of autonomous units of meaning
Creation of thematic and/or semantic categoiies (219 sensitizing concepts)
Li
Creation of 2 metatexts
Meta-text 1 (320 conceptual categoiies. of which 219 sensitizing concepts) and meta-text 2 decoding of 320 categoiies in 24 dimensions)
Lexical Correspondences Analysis (SPAD)
Analysis of metatexts 1 e 2
Procedure VoSpec (Vocabulaire spécifique) (SPAD)
* Analysis of meta-textes 1 and 2 ( to identify words and categoiies of words characteristic of groups of individuals)
Cluster analysis
* Text size and tvpe of
respondents
Analysis of probabilistic concatenations of semantic nodes (24 dimensions) T-LAB
Micro maps (source, relay, condenser) and macro maps of discursive dynamics
Content quantitative analysis
Gabriella Punziano coordinated a working group (composed of Amalia Caputo, Augusto Cocorullo, Cristiano Felaco and Barbara Saracino) on the textual material resulting from the 164 qualitative interviews, favoring a quantitative content analysis framework [1] and choosing to work on "minimal units" and units of meaning (lemmas and graphic forms) identified as classification units, to develop a sequential multi-methods design. The texts of the interviews were collected and organized in a custom dictionary undergoing pre-treatment, cleaning and lemmatization, lexicaliza-tion, normalization and segmentation operations. Consequently, the textual data, thus organized, were used as the basis for an integrated path analysis that goes from the evaluation of Peculiar Language to Co-occurrences Analysis, Lexical Correspondences Analysis, Cluster Analysis, and finally a Social Network Text Analysis.
The first phase of the analysis was dedicated to the reconstruction of the peculiar language present in the interviews, to find possible significant differences concerning the lexicons used by the interviewees and, therefore, to identify the concepts emerging within the narratives. In this regard, with the help of the T-LAB software, the frequency lexicons for each category taken into consideration (by gender, age, educational qualification, location, type of municipality of origin, employment, number of children, marital status) were investigated and compared with all the individual texts and taken as the standard of comparison to assess the variations in the different sub-sets generated by the categories. Particular attention was paid to the analysis of the occurrences of the individual words: first of all, the forms most frequently examined in the specific lexicons, individually taken, were examined and, subsequently, it was seen how and according to which positioning the forms of the total lexicon are distributed in the specific lexicons. Possible terms present only in the specific lexicons and therefore typical of that particular category of subjects have been highlighted. This analysis was deepened by observing the trend of the lexicon peculiar to the age groups and in its intersection with the gender of the respondents and with the geographical variable, which allowed to contextualize the individual terms concerning the characteristics of the respondents. Based on these results, the analysis focused on the co-occurrences between the terms.
To grasp the complexity contained in these preliminary operations, the positioning of the individual terms in the general space of meanings was observed, concerning the social and personal information of the interviewees. To this end, a multidimensional analysis was performed using both the Lexical Correspondences Analysis, with the help, this time, of the Spad-T software. So the research question invested the whole corpus to try to understand if the sense of religiosity could be broken down into latent dimensions more complex and complementary to the simple explanation of the religious concept itself. The composition of the areas of meaning has been integrated with the study of the structure of the network formed by the interconnections between the different headwords. In this regard, the application of Network Analysis tools to the study of texts (Text Network Analysis) has allowed us to grasp
how the nodes are connected, thus identifying the areas of the neighborhood and identifying those nodes that occupy a position of relevance in different sets or the entire network. By analyzing, in fact, the position of words within the text network, it is possible to trace those with a greater semantic charge, thus identifying the different paths and contexts of meaning.
Graph. 4 - Content quantitative analysis
Analysis of frequency lexicons by category
Comparison with the texts of the qualitative interviews
Word Occurrence Analysis
и
Frequencies in specific lexicons Positioning and distribution in specific lexicons
Identification of terms present only in specific lexicons
Analysis by age. gender and
geographical
Multidimensional analysis with Lexical Correspondences Analysis, Cluster Analysis and Text Network Analysis • Highlight latent dimensions and particular interconnection spaces
Conclusions
Patrizia Ascione, Lavinia Bianchi and Alberto Quagliata have worked on qualitative data applying the indications of a Grounded Theory of constructivist matrix, based on the model proposed by Kathy Charmaz [5; 6; 7; 8; 9; 10; 11; 12; 14].
The procedure for the construction of the theory started from first processing (line by line coding) of the most common coding summarized in provisional categories and their properties. This first processing was carried out by the team of researchers who used the NVivo 10 software. To this material have been added the memos elaborated by the interviewers who have supported, enriched and oriented the following levels of analysis.
Memos are a type of document that allows recording ideas, insights, interpretations or progressive understanding of the material in the research project. They provide a way to keep the analysis separate from (but linked to) the material being analyzed. Reminders can evolve into an important part of the research: for example, they can lead to book chapters or a presentation scheme. Memos can also be used to learn about problems, comments or insights, as well as to create reflections about a node (or sensitizing concept), describing the meaning of the node itself and the patterns or ideas that emerge from the texts. Finally, reminders not directly linked to a concept or a part of the text can also be used to store other information about the research, useful for the progress of the project.
For the initial (first level) coding, inductive logic was mainly used, to identify the labels or categories related to the selected portions of text.
Focused coding (or second-level coding) was then used. The procedure involved a continuous comparison of the process underway, through a constant comparison between analysts. In this phase the coding was characterized as more abstract conceptually, while remaining adherent to the data: it was a matter of delimiting the analysis of the data exclusively to those elements strongly connected to the central categories identified, having the function of guidance for the further phases.
While focused coding represented the process of synthesis of the interpretative categories that emerged, theoretical coding then identified the relationships between the categories. All this has been done by using abductive reasoning, characterized by an increasing level of conceptualization of data, to elaborate systems of relations between the concepts identified. In other words, the conceptual similarities between phenomena belonging to different fields (i.e. similarities between different similarities) were sought at a higher logical level, signaling what they had in common and accepting all the possible interpretative hypotheses related to the observed data. These passages of an abductive nature, therefore, constituted the conceptual leap for-
ward aimed at producing something new, not derived "simply" from the analysis of the data. In practice, the theoretical codification has elaborated, in the form of hypotheses, the conceptual models of relationship developed to theoretically link the initial coding. The conceptual elaboration was concluded when the relations between the concepts, which emerged through the identification and use of appropriate theoretical constructs, allowed to construct a general theoretical framework capable of describing the grounded theory.
The activity of theoretical coding has developed in parallel with the activity of memoing (construction of memos), characterized by a high degree of abstraction and theoretical sensitivity. The classification and integration of memos were the strategic activities to outline the conceptual framework of the emerging theory and to propose a "conceptually dense" grounded theory: if we had omitted the step relating to the systematization of theoretical annotations, the theory would have been rather linear, conceptually weak and not sufficiently articulated in a meaningful and comprehensive set of assumptions.
The use of theoretical memos and theoretical coding were based on the relevant conceptual categories in terms of integrity, accuracy and conceptual fullness. To elaborate them, analysts have deepened the ideas "deposited" and "submerged at a conceptual level" in the memos, seeking the balance between logic, creativity and openness to the unexpected.
In this phase of elaboration of the core categories five founding nuclei have emerged, able to generate the theory: "Pastoral care" (believers who help other people by drawing inspiration from the values of Christianity), "Rhetoric of charitable humanitarianism" (people who perceive other people as different and dangerous and, at the same time, tell of the effort to welcome and accept the needy), "Prisoners of despair" (they experience the difficulties of life in a passive and resigned way and takes refuge in God and prayer), "Embracing faith" (practicing believers, guided by Christian precepts and living the faith as dogma) and "Beyond everything" (individuals, who are discordant on the themes of life, death, and the afterlife, but, beyond ideas, actions, and skepticism, they have faith in God).
According to Charmaz [11], the focus is on the co-construction of meanings, on the process itself of shared co-construction and, therefore, on the fact that the emerging theory is invented by analysts and by the subjects of the study themselves. With the emergence of categories, the level of abstraction increases and the relationships between them are highlighted, which, accumulating, give rise to an integrated central theoretical structure, that constitutes the core of the final emerging theory.
Graph. 5 - Grounded Theory approach
1. Line by line coding (NVivo 10)
1.1. Coding of the texts of qualitative interviews and memos
1.2. Summary of the most common coding
1.2.1. Provisional categories
1.2.2. Properties of provisional categories
2. Initial coding by induction (first level)
2.1. Identification of core categories
2.2. Comparison between analysts, for the co-construction of the theory
3. Focused encoding by abduction (second level)
3.1. Identification of core categories
3.2. Delimitation of the analysis to the core categories only
4. Theoretical coding by abduction (third level)
4.1. Identification of relations between core categories
4.2. Elaboration of systems of relationships between the selected concepts
4.3. Search for similarities between different phenomena
4.4. Formulation of interpretative hypotheses
5. Construction of the general theoretical framework
Graph. 6 - Research general synopsis
References
1. Amaturo, E. and Punziano, G. (2013). Content analysis. Tra comunicazione e politica. Milano: Ledizioni. 328 p.
2. Amaturo, E. and Punziano, G. (2016). I Mixed Methods nella ricerca sociale. Roma: Carocci. 198 p.
3. Benzécri, J.-P. (1973). L 'analyse des données. Vol 1. La taxonomie. Paris: Dunod. 619 p.
4. Cesareo V., Cipriani R., Garelli F., Lanzetti C., Rovati G. (1995). La religiosità in Italia. Milano: Mondadori. 369 p.
5. Charmaz, K. (1995). Between positivism and postmodernism: Implication for methods. Studies in Symbolic Interaction. Vol. 17 / ed. by N. K. Denzin. Greenwich: JAI Press, 1995, pp. 43-72.
6. Charmaz, K. (2000). Grounded Theory: objectivist and constructivists methods. In: Handbook of qualitative research / ed. by N. K. Denzin and Y. S. Lincoln. London: SAGE, pp. 509-535.
7. Charmaz, K. (2003). Grounded Theory. In: Qualitative psychology: A practical guide to research methods / ed. by J. A. Smith. London: SAGE, pp. 81-110.
8. Charmaz, K. (2005). Grounded Theory in the 21st century: A qualitative method for advancing social justice research. In: Handbook of qualitative research / ed. by N. K. Denzin and Y. S. Lincoln. Thousand Oaks (CA): SAGE, pp. 507-535.
9. Charmaz, K. (2008). Constructionism and the Grounded Theory method. In: Handbook of Constructionist Research / James A. Holstein and Jaber F. Gubrium (eds). New York: Guilford Publications, pp. 397-412.
10. Charmaz, K. (2011). Grounded Theory methods in social justice research. In: The SAGE Handbook of qualitative research / ed. by N. K. Denzin and Y. S. Lincoln. New York: SAGE, pp. 359-380.
11. Charmaz, K. (2014). Constructing Grounded Theory. London, Thousand Oaks, New Delhi: SAGE. 416 p.
12. Charmaz, K. and Bryant, A. (eds) (2007). The SAGE Book of Grounded Theory. London: SAGE. 657 p.
13. Cipriani, R. (1993). De la religion diffuse à la religion des valeurs. Social Compas. Vol. 40. No. 1. pp. 91-100.
14. Cipriani, R. (ed.) (2008). L'analisi qualitativa. Teorie, metodi, applicazioni. Roma: Armando. 224 p.
15. Della Ratta-Rinaldi, F. (2007). L'analisi multidimensionale dei testi. Ricerca sociale. Dall'analisi esplorativa aldata mining. Roma: Carocci, 2007. Vol. III. pp. 133-150.
16. Delli Poggi S., Bolla F., Fiorentini S., Valletta G. L. (2020). Italiani: scomunicati o liberi pensatori? Eresia, apostasia, scisma o libertà di credere. Profili di cattolici in Italia. - Indipendent publication. URL: https://www.amazon.com/Italiani-scomunicati-pensatori-Apostasia-cattolici/dp/1693714280
17. Dell'Orletta F., Venturi G., Cimino A., Montemagni S. (2014). T2K2: a system for automatically extracting and organizing knowledge from texts. In: Proceedings of 9th International conference on language resources and evaluation (LREC 2014). Reykjavik: European Language Resources Association (ELRA), pp. 2062-2070.
18. Faggiano, M. P. (2016). L 'analisi del contenuto di oggi e di ieri. Testi e contesti on e offline. Milano: FrancoAngeli. 116 p.
19. Johnson, R. B., Onwuegbuzie, A. J. and Turner L. A. (2007). Toward a definition of mixed methods research. Journal of Mixed Methods Research. Vol. 1. No. 2. pp. 112-133.
20. Krippendorff, K. (1980). Content Analysis. An introduction to its methodology. New York: SAGE. 188 p. [Analisi del contenuto. Introduzione metodologica. Torino: ERI, 1983. (In Italian)].
21. Losito, (2002). G. L'analisi del contenuto nella ricerca sociale. 4th ed. Milano: Franco Angeli. 176 p.
22. Maranda, P. (1990). DiscAn: Users Manual. Québec: Nadeau Caron Informatique.
23. Maranda, P. (1995). DiscAn: un programma di analisi reticolare per la costruzione di mappe semantiche. In: Ricerca qualitativa e computer. Teorie, metodi e applicazioni / ed. by R. Cipriani and S. Bolasco. Milano: Franco Angeli, pp. 171-183.
24. Mauceri, S. (2017). L'avvento dell'era dei mixed methods. Nuovo paradigma o deadline di un dibattito? Sociologia e Ricerca Sociale. No. 113. pp. 39-61.
25. Mauceri, S. (2018). Mixed survey research. Le funzioni delle tecniche qualitative nella ricerca standard. Sociologia e ricerca sociale. No. 116. pp. 90-102.
26. Statera, G. (1997). La ricerca sociale. Logica, strategie, tecniche. Roma: Seam. 397 p.
About the author
Roberto Cipriani, Emeritus Professor of Sociology at Rome 3 University, Rome, Italy; ORCID ID: 0000-0001-6722-2183587, e-mail: rciprian@uniroma3.it
Чиприани (Cipriani) Роберто, почетный профессор факультета социологии, Университет Рим-3 (Universita degli Studi Roma Tre), Рим, Италия; ORCID ID: 0000-0001-6722-2183587, e-mail: rciprian@uniroma3.it
Об авторе
Поступила в редакцию: 20.03.2020
Received: 20 March 2020
Принята к публикации: 24.08.2020
Accepted: 24 August 2020
Опубликована: 14.09.2020
Published: 14 September 2020
