Potentials and Limits of Secondary Analysis in a Specific Applied Context: The Case of EDF—Verbatim

Abstract: In this paper we present the results of secondary analyses conducted on documents recorded in VERBATIM, the qualitative data archiving system and database developed by the Sociological Surveys Department (GRETS) of Electricité de France (EDF).

The archived studies, unique in France, offer some specific methodological and theoretical features. In order to explain their properties, constraints, and diversity we present four re-analysis studies, which provide an overview of the research aims and topics, the body of textual data, the methods of the secondary analysis and the conditions under which re-analysis can be judged as reliable. We also explore the use of qualitative software in these analyses. Furthermore, we focus on the Sociological Surveys Department researchers' attitudes towards secondary analysis, which shed light on the future development of secondary analysis in France.

Finally, we present a future project concerning scientific collaboration with the Centre d'Informatisation des Données Socio-Politiques (CIDSP) which aims to develop more systematic qualitative data archiving and secondary analysis in France.

Keywords: secondary analysis, qualitative data, archiving, content analysis, applied and industrial context, methodology, qualitative software

3.2 Re-use of several inquiries in order to investigate a precise theme: the use of domestic electricity during night and day

3.3 Re-use of an important set of data on an extended theme: domestic electric comfort; practices and reasoning

3.4 Re-use of multiple qualitative data to explore a new specific area: house renovation

In a previous issue of FQS (Le ROUX & VIDAL, 2000), an overview of the development of the database, VERBATIM was presented. This project has been carried out for the past six years in the Sociological Surveys Department of Electricité de France (GRETS) and it has the task of undertaking various qualitative surveys in order to understand customers' requirements and the problems encountered by the company's employees. VERBATIM is a qualitative archiving database, aiming to keep track of the interviews carried out by the company's sociologists in order to re-use them by recording all that is necessary to perform a secondary analysis, including: the socio-demographic characteristics of the interviewees; the researcher's concerns; the methodology; and the results of the study. Since 2000, VERBATIM has thrived (60 studies, 158 MO) and we have improved our knowledge of methodology about secondary analysis. Here, we discuss the results of this experience. [1]

The principal aim of VERBATIM is neither teaching nor research. Our objective, in an industrial context, is to re-use the archived enquiries in order to capitalise knowledge on some important topics: what kind of representation do customers have of home electric comfort, how do they use electricity by day and by night, and what are their needs about domestic improvements. However, even though the purpose of our analyses is applied-oriented, it is clear that the methodological problems are taken into account, to avoid the risk of obtaining unreliable results. [2]

We sum up the experiences of four years of secondary analysis practice using the VERBATIM database. We will describe the methodological knowledge that we have acquired, the limits of such an exercise and our use of software chosen to analyse the data. [3]

One of the main problems with secondary analysis, especially in an applied context, concerns the aim of the analysis. We have carried out 3 main types of secondary analysis according to different goals:

The purpose of this paper is not to propose a concrete typology but only to discuss our practice. However, we may distinguish several categories according to the purpose of analysis, its depth, the constitution of the corpus (the data of one or several inquiries may be re-used), and the subject under investigation (this one may sometimes be quite far from the hypothesis and themes being investigated throughout the initial inquiry). [5]

These examples we will cite clearly show that there is no unique type of secondary analysis but that practices are quite different. However, we may point out some common key features, especially those concerning the methodological precautions to be taken in order to ensure the reliability of results. [6]

Based upon HEATON's typology (1998) concerning the forms of secondary analysis, those that we have carried out used data from either a single or multiple qualitative data sets and focus either on an additional in-depth, on sub-set analysis, or on a new perspective. [7]

We will present and discuss four different examples of applied-oriented secondary analysis. The table summing up the methodological features of those secondary analyses is presented below.

Though different, these secondary analyses share some similarities. In every case, the materials (corpus) have been taken from the VERBATIM database with the collaboration of the sociologist who had collected many of the original research interviews. This shows how important it is to rely on the experience and knowledge of the expert of a sociological field. Secondary analysis such as we practice always results from the collaboration between both researchers and disciplines. [9]

In this example, we re-used the results of an inquiry among the employees of the EDF Research Division: During the inquiry which had been commissioned by the head of the Division, the sociologists were eager to understand how those researchers adapt their work to the important changes in the company (loss of the monopoly of electricity production and distribution, adjustment to competition). But they never asked questions about stress-related problems or mental health at work. [10]

About two years after the end of the inquiry, we were asked to re-explore its results in order to understand better the problems of stress and depression with the employees due to the changes in the company (Le ROUX, 2004a). The initial inquiry had allowed us to collate a very important amount of qualitative data, which could not have been completely analysed. Choices always have to be made during the analysis of data, especially in an industrial context when the sponsor of the study is eager to get its results quickly. Thus there are constraints related to efficiency which compel researchers to focus on some aspects only and to neglect others. The principal interest of such a secondary analysis was obviously to be able to explore a thematical field which had not been addressed initially. [11]

Although the initial inquiry did not focus at all on these problems, people had spontaneously spoken about them and we were thus able, by re-using the interviews, to give some indications on stress at work by different populations such as managers, employees, researchers and secretaries. In this case, the secondary analysis was conducted on one set of data by the sociologists who had originally collected the data and carried out the initial study. The software used to assist the analysis was the CAQDAS, Atlas.ti. [12]

However, when giving back our sponsor the results of the secondary analysis, we thought it essential to point out the limits of such an approach. First that stress at work had not been a subject that had been asked about in any depth during the initial inquiry, and second that the data were already two years old and things were changing very quickly. Our conclusion was that there was a genuine problem with stress in the workplace and that we could help elucidate more clearly the different types of interpretation about it as they related to different functions of the company. However, to be able to say more about it, it would be necessary to undertake a new inquiry focusing on this subject. In this case, the secondary analysis had been useful in order to quickly form an opinion about the problem and to define more precisely the nature of a new inquiry. [13]

3.2 Re-use of several inquiries in order to investigate a precise theme: the use of domestic electricity during night and day

The second example that we discuss here is a secondary analysis dealing with the problem of the use of electricity at night by individuals (Le ROUX, 2004b). In France, night units¹⁾ are cheaper than day units for customers who have made a special tariff agreement with EDF. For example, in this case, they turn on their dishwasher and laundry machine at night. [14]

This study had been commissioned in order to define the new energy supply conditions of the company. The marketing division needed precise information about the way people deal with that tariff. Since the statistics could not answer their questions properly it was considered important to obtain qualitative data to understand better the needs of customers. [15]

In order to do that, we extracted a large corpus out of the VERBATIM database by using the natural language query tool TOPIC which is part of every Lotus Notes database. In this case it was quite easy to carry out such research even on a significant amount of data, as it corresponds to a precise terminology, in French "heures creuses", that is the concept of night units for electricity tariffs. As we will see further on, the problem of the corpus definition is much more complex in other cases. We chose to keep all the extracted studies even when they did not precisely focus on tariff problems and to re-use all the interviews collected where people spoke about their experience of "heures creuses" and night-time domestic use of electricity. [16]

The results of this secondary analysis allowed us to point out the main uses of electricity during the "heures creuses" or cheaper periods. It was also extremely useful to define categories of customers in relation to their different patterns of behaviour. For example, their relationship to "heures creuses" depends on several criteria such as: the type of family; the age of people; and the place where they live etc. The results that we obtained from the secondary analysis helped us to define customers' profiles in order to establish special commercial offers about electricity. [17]

However, the results were evidently dependent on the data that had been analysed. For example, we had the opportunity to re-use a very important study composed of a lot of interviews about rural and poor populations, which could misrepresent the behaviour of the global French population. Thus we had to be very careful about the sample definition imposed by the database contents and to resist the temptation to give quantitative results. The secondary analysis could only give qualitative information to be confirmed by a quantitative survey if necessary. Fortunately, the number of studies and interviews involved in the secondary analysis was sufficient to ensure a reliability of qualitative results. [18]

3.3 Re-use of an important set of data on an extended theme: domestic electric comfort; practices and reasoning

This research deals with the question of domestic electric comfort. The electric comfort theme had been initially chosen in order to compare various methodologies of secondary analysis and different research goals (LABBE & BRUGIDOU, 2003; ZAPATA, 2002). A subset of the data re-used for those previous analyses was derived and submitted to a final secondary analysis which was mainly exploratory and focused on people's practices and logic when using thermic systems. Results of this research were to contribute to an EDF website, devoted to providing advice on the use of thermic systems. Here, we present results of this precise re-analysis (DARGENTAS, 2003a). Compared to the previous study we described earlier (that related to the price of electricity), regarding the corpus definition, this secondary analysis featured a distinct problem. The topic of practices when using thermic systems doesn't necessarily correspond to a precise terminology; thus, it is very difficult to define the data to be re-analysed. This is the reason for the choice of a larger corpus (that of electric comfort) in which those topics are more likely to occur. This secondary analysis has been agreed upon because of a large amount of archived interviews and studies, arising from diverse research situations and samples and relating to different domestic electric systems. [19]

It is interesting to present here this secondary analysis for three reasons. First, the problems are linked to the initial objectives associated with these research limits; second, the analysis process; and third, the kind of software used. [20]

Regarding the first point, this secondary analysis was difficult to carry out, mainly because of the objectives targeted by the company. These had an overly technical orientation and were uneasy to examine as such, due to the secondary nature of this research and to the verbal status of the available data. In fact, according to the initial aims, "deviated"²⁾ practices when using domestic electric systems should be identified, and this should lead to their typology at the same time taking into account the interviewees' characteristics. Consequently, our first task was to transform these aims relating to the engineers' logic into more sociological aims; that is, we would be more interested in the appropriation of the systems and the study of uses linked to reasoning behind them. Moreover, limits due to the secondary character of this analysis reduced the scope of the initial objectives in that the primary data were not designed to examine the issue of "deviated" practices. These practices depended on a diversity of aims, the samples, and the interview guides. As a result, the theme of practices does appear in a very different manner through primary studies and data. Limits concerning the verbal content of our data also reduced the scope of goals. Thus, a typology of practices was impossible to derive for the following reasons: "deviated" practices would not be representative in this corpus as their recurrence depended on many factors such as the manner in which the interview was conducted; moreover, it would not be possible to draw up an exhaustive list of practices, just the self-reported ones by interviewees. Lastly, this re-analysis illustrates the methodological problems related to the study of any practice available in these data. People talk about their practices in different ways and sometimes practices are deduced from the general context. Thus we can imagine that "deviated" practices are not meaningful for people, but two other possibilities might be proposed: either people recognise that they use improperly thermic systems, due to some specific reason (for example thermic discomfort), or their practices are not recognised as improper because they are part of their appropriation of thermic systems. [21]

In light of the above explanations, the initial objectives were significantly altered, and we decided to work on practices by paying attention to the general context in which they appeared, as well as to the reasoning behind the uses of systems and the subjects' representations. [22]

Moreover, instead of only taking into account the research contexts of the primary studies—as it is commonly recommended for secondary analysis (SANTACROCE, DEATRICK & LEDLIE, 2000), we also opted for the contextualisation of the studied topics (practices and reasoning) in the specific content of the primary data. This way to carry out our analysis suggests a holistic analysis procedure, which aims at preserving the specific content of primary data: thus, themes which are interesting to explore should also be interpreted with regards to their relationship to the content of the whole data set. [23]

So, in order to examine these objectives and respect the need for a secondary analysis to emphasize the specificities of primary studies, we adopted a procedure aimed at characterizing the whole of the data set and at subsequently undertaking a more in-depth analysis of practices and patterns. [24]

This research proposes an analysis through several types of software, which should, by their methodological perspectives, examine different empirical questions. The data have been explored through two types of software devoted to text analysis. The first stage of the analysis related to the content trends of the whole corpus and has been carried out using the French Automatic Textual Data Analysis software (Tropes). The latter helped to highlight the themes appearing in our data and their importance in terms of frequency. It also offered results related to the linguistic specificities of the data. Through this first analysis we have been able to make a general description of the whole corpus in a somewhat "quantitative" approach. On the other hand, the second stage was devoted to a more in-depth analysis, which aimed at exploring the practices and people's representations, by focusing on thermic systems; the Tropes software was used once again. Finally, there was a need for a more complete description of the use of thermic systems, and a decision was taken to study a unique interview using Atlas.ti software. [25]

3.4 Re-use of multiple qualitative data to explore a new specific area: house renovation

This last piece of research focuses on the theme of house renovation (DARGENTAS, 2004). This secondary analysis was an exploratory study focusing on many aspects linked to renovation, for example, what is the object of renovation? What causes it? What are the problems related to it? How are decisions to renovate taken? Who are the different actors involved in renovation? [26]

The primary studies were chosen from those archived in the database VERBATIM, through a theme search according to their goals and their links to the theme of renovation. In cases where surveys had not been purposely designed to study renovation, this topic had sometimes been raised spontaneously or had sometimes been a marginal topic undertaken as part of the primary study. The secondary analysis was chosen because of the diversity of research situations (corpus, samples, objectives, guides of interviews, etc.). As a result, the context in which the theme of renovation was raised varied and this might have lead to the possibility of a more multidimensional study of the situation regarding renovation. [27]

This study is very similar to the previous one described in 3.3 because of its methodology. It is interesting not only because of the holistic method used, but also because of the active role of the analyst during the interpretation of results. The researcher manages to identify the biases due to the primary data, and also the results are not seen as representative of the situation surrounding renovation, but rather as informing some issues related to this topic. [28]

In order to deal with the limitations of our analysis and to give meaning to results obtained in this secondary analysis, we had to study the primary studies thoroughly, with regards to their theoretical and empirical properties. We also had to adopt a holistic kind of analysis centred on the thematic context in which the renovation theme appeared in the whole data set. We used the software Alceste, which allowed us to gain a general idea of the thematic network dominating the interviewees' language and the main variables. Afterwards, using the same type of software we had to make a more in-depth analysis of the renovation theme, by viewing the data more precisely, as well as the variables that were linked to its different aspects. [29]

The limitations of this research were linked to the non-equivalent and non-homogeneous presence of the theme of renovation through the various primary studies. Indeed, the theme of renovation is referred to in a different manner depending on the research context (e.g. sample, thermic systems, guides of interview) and on objectives. It may be a spontaneous theme or one that is induced, or it may be dominant or marginal throughout interviews. Lastly, its meaning varies among actors (tenants, landlords, caretakers, etc.). The researcher is then obliged to interpret results on renovation as being non-representative. [30]

When interpreting the results, we were able to discover some biases related to the secondary nature of our analysis. That is, some findings showed that in 75% of cases the main object of renovation was heating systems, whenever this theme was raised. Also, this kind of renovation was significantly linked among other variables to people aged 40 and below, whose houses had been built after 1970. In spite of these results, obtained through the Alceste software, we did not take into account these findings, which were considered as consequences of gathering our primary corpus, that is the secondary data analysed mainly represented: houses built after 1970; people aged up to 40; and studies that focused primarily on heating systems. Thus, we were able to interpret these results as the consequence of our corpus structure, regarding its contents and its variables. Additionally, another finding seemed to oppose the existing statistics on renovation. The complete renovation of housing does not appear through our findings, which focus mainly on renovation of thermic systems, or on the partial renovation of housing. In this investigation, we collaborated closely with the sociologist who helped us to select the primary studies and who had undertaken some of the original studies. [31]

Through the studies described above, we can see that the researcher adopts a reflexive position during the secondary analysis, thereby enhancing the quality of re-analysis. This position varies across the different stages of the analysis, depending on the object of the study and its theoretical and empirical properties. In these studies, the following situations were relevant: the choice of the primary studies and interviews; the limits of the secondary analysis; the depth of knowledge of the primary research contexts; the consequences of the objectives; the type of analysis that should be undertaken (i.e. general or in-depth analysis); the interpretation of findings (e.g. influence of variables, studying representations rather than practices only, etc.); and the collaboration of the original investigating sociologist and producer of the primary studies. [32]

We also would like to emphasize two further points: first the "holistic" procedure adopted in two of these secondary analyses and second, the use of software—CAQDAS (Atlas.ti) or the French Automatic Textual Data Analysis (Tropes, Alceste). Concerning the "holistic" procedure, in some of the secondary analyses we undertook a contextualisation of the topic that was the focus of the secondary study taking into account all of the topics featured in our data. This procedure was sometimes made necessary by the goals and specificities of the research. Moreover, we suggested this way of secondary analysis as an additional means to elucidate links with primary data and their thematic specificities. Secondly, the use of the French software helped us to carry out this method of analysis. There are programmes, that originated in the French sociolinguistic tradition, that contribute to the analysis of huge amounts of data; relying on linguistics and statistics, and that permit a quasi-automatic textual data analysis without the analyst imposing his or her own units of analysis³⁾. This is a complementary way to proceed compared to the CAQDAS types of software; the latter enabling a more in-depth investigation with regard to people's reasoning. We tried to use these softwares in a complementary way and to answer different empirical questions through the use of them. [33]

The studies we described belong mainly to an applied empirical perspective. These secondary analyses are unique in France, where this methodological trend is absent from everyday practice. The approaches have certain peculiarities, advantages and limits. Let us examine some peculiarities and advantages first:

Thus, in the light of the above specificities, and in spite of the limited topical focus of the analyses, the past or future secondary analysis in the GRETS seems to be conducted in a somewhat ideal context as opposed to most of the secondary analyses undertaken by other researchers and institutions. [35]

Nevertheless, there are some key problems that reduce the quality of secondary analysis and thus prevent the development of the archiving and of secondary analysis in this sociological EDF Department⁵⁾:

These problems are not new. Scholars have already identified most of them in scientific contexts outside France (BROWN, 2002; CORTI & THOMPSON, 2004; CORTI, 2000; CORTI, DAY & BACKHOUSE, 2000; FIELDING, 2000; GOODWIN & O'CONNOR, 2002; HEATON, 1998, 2004; HINDS, VOGEL & CLARKE-STEFFEN, 1997; SANTACROCE et al., 2000; THORNE, 1994, 1998). Thus, they add to knowledge about and experiences of the development of archiving and secondary analysis of qualitative data. [37]

Consequently, the expansion of secondary analysis in the future depends on the systematic use of archiving, on the production of methodologically rigorous studies that should demonstrate the usefulness of applying secondary analysis, and on knowledge of appropriate qualitative research software. [38]

From our experience, we are better able to foresee the future of archiving and secondary analysis, to appreciate the problems raised, and to understand better their development inside French academic circles. We should point out here that these topics have not been discussed in the French research context. Finally, a form of scientific cooperation, based on the experience of GRETS, is planned in order to think and prepare for the development of these areas in French academic circles. This cooperation is taking place between GRETS and CIDSP (Centre d'Informatisation des Données Socio-Politiques, Grenoble). The latter is already devoted to the archiving and secondary analysis of quantitative data and would like to extend these activities to qualitative data. [39]

This article could not have been written without the collaboration of a research team especially Véronique BEILLAN, Mathieu BRUGIDOU and Luz ZAPATA to whom we are very thankful.

1) Cheaper units are also extended to some other periods such as lunch hours or weekends (that is periods of less global electricity consumption). <back>

2) According to engineers of EDF, "deviated" practices are those which prevent thermic systems from functioning properly. They are responsible for thermic deterioration. For example: keeping windows open when electric systems are in use, blocking air vents. <back>

4) For example, HEATON (2001) is based on 58 interviews from one primary study. Another study, that of PICKENS (1999) analyses 19 interviews extracted from two prior studies. <back>

5) Most of these points have been raised by ZAPATA (2001) and DARGENTAS (2003b). <back>

Brown, David (2002). Going Digital and Staying Qualitative: Some Alternative Strategies for Digitalizing the Qualitative Research Process. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 3(2), Art. 12. Available at: http://www.qualitative-research.net/fqs-texte/2-02/2-02brown-e.htm [Date of Access: 01.09.2004].

Corti, Louise (2000). Progress and Problems of Preserving and Providing Access to Qualitative Data for Social Research—The International Picture of an Emerging Culture. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 1(3), Art. 2. Available at: http://www.qualitative-research.net/fqs-texte/3-00/3-00corti-e.htm [Date of Access: 01.09.2004].

Corti, Louise & Thompson, Paul (2004). Secondary Analysis of Archived Data. In Clive Seale, Giampetro Gobo, Jaber Gubrium & David Silverman (Eds.), Qualitative Research Practice (pp.327-343). London: Sage.

Corti, Louise; Day, Annette & Backhouse, Gill (2000). Confidentiality and Informed Consent: Issues for Consideration in the Preservation of and Provision of Access to Qualitative Data Archives. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 1(3), Art. 7. Available at: http://www.qualitative-research.net/fqs-texte/3-00/3-00cortietal-e.htm [Date of Access: 01.09.2004].

Dargentas, Magda (2003a). Pratiques, "déviance" et logique des sujets à l'utilisation des systèmes thermiques: Contribution à une étude exploratoire d'offres de conseil autour de pratiques de gestion des systèmes thermiques. Analyse secondaire d'entretiens qualitatifs de clients résidentiels. Note interne Edf: HE-71/03/004/A.

Dargentas, Magda (2003b). Evaluation par les chercheurs du GRETS de la base Verbatim. Rapport interne EDF, unpublished.

Dargentas, Magda (2004). La Rénovation du logement: Analyse secondaire d'entretiens. Note Interne EDF, to be published.

Fielding, Nigel (2000). The Shared Fate of Two Innovations in Qualitative Methodology: The Relationship of Qualitative Software and Secondary Analysis of Archived Qualitative Data. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 1(3), Art. 22. Available at: http://www.qualitative-research.net/fqs-texte/3-00/3-00fielding-e.htm [Date of Access: 01.09.2004].

Goodwin, John & O'Connor, Henrietta (2002). "They Had Horrible Wallpaper": Representations of Respondents and the Interview Process in Interviewer Notes. CLMS Working Paper No 39. Available at: http://www.clms.le.ac.uk/publications/workingpapers/working_paper39.pdf [Date of Access: 01.09.2004].

Heaton, Janet (1998). Secondary Analysis of Qualitative Data. UniS—Social Research Update, 22. Available at: http://www.soc.surrey.ac.uk/sru/SRU22.html [Date of Access: 01.09.2004].

Heaton, Janet (2001). Hospital Discharge and the Temporal Regulation of Bodies. Time & Society, 10(1), 93-111.

Hinds, Pamela S.; Vogel, Ralph J. & Clarke-Steffen, Laura (1997). The Possibilities and Pitfalls of Doing a Secondary Analysis of a Qualitative Data Set. Qualitative Health Research, 7(3), 408-424.

Jenny, Jacques (1997). Méthodes et pratiques formalisées d'analyse de contenu et de discours dans la recherche sociologique française contemporaine: Etat des lieux et essai de classification. Bulletin de Méthodologie Sociologique (B.M.S.), 54, 64-112.

Labbé, Dominique & Brugidou, Mathieu (2003). Analyse des représentations du confort électrique à partir d'un corpus d'entretiens. Note interne EDF: HE-75/03/003/A.

Le Roux, Dominique (2004a). Le stress chez les agents techniques à la R&D: analyse secondaire. Note interne Edf, to be published.

Le Roux, Dominique (2004b). Populations, tarifs et comportements: zoom sur les ménages, les heures creuses, le week-end. Note interne EDF: HE-71/01/026/A.

Le Roux, Dominique & Vidal, Jean (2000). Verbatim: Qualitative Data Archiving and Secondary Analysis in a French Company. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 1(3), Art. 15. Available at: http://www.qualitative-research.net/fqs-texte/3-00/3-00rouxvidal-e.htm [Date of Access: 01.09.2004].

Pickens, Judith M. (1999). Living with Serious Mental Illness: The Desire for Normalcy. Nursing Science Quarterly, 12(3), 233-239.

Santacroce, Sheila J.; Deatrick, Janet A. & Ledlie, Susan W. (2000). Secondary Analysis of Qualitative Data: A Means of Collaboration in HIV-Related Research. Journal of the Association of Nurses in AIDS Care, 11(3), 99-104.

Thorne, Sally (1994). Secondary Analysis in Qualitative Research: Issues and Implications. In Jan Morse (Ed.), Critical Issues in Qualitative Research Methods (p.263-279). Newbury Park, CA: Sage.

Thorne, Sally (1998). Pearls, Pith, and Provocation: Ethical and Representational Issues in Qualitative Secondary Analysis. Qualitative Health Research, 8(4), 547-555.

Zapata, Luz (2001). Compte-Rendu des Entretiens Informels sur VERBATIM. Rapport interne EDF, unpublished.

Zapata, Luz (2002). Exploration de la question du 'Confort' à travers l'analyse avec la méthode ALCESTE d'un corpus d'entretiens. Note interne Edf: HE-71/02/022/A.

Magda DARGENTAS: Ph.D. candidate in Social Psychology at the EHESS, Researcher under contract with the Social Science Research Group in the Research and Development Division of Electricité de France and with the CIDSP-CNRS.

EDF / DRD / ICAME / GRETS
1 avenue du General de Gaulle
92141 CLAMART Cedex
France

Dominique Le ROUX, Ph.D. Linguistics and Information Science, Sociologist, Research Engineer in the Social Science Research Group in the Research and Development Division of Electricité de France. Manager of the VERBATIM project (Qualitative data archiving and secondary analysis).

EDF / DRD / ICAME / GRETS
1 avenue du General de Gaulle
92141 CLAMART Cedex
France

Dargentas, Magda & Le Roux, Dominique (2005). Potentials and Limits of Secondary Analysis in a Specific Applied Context: The Case of EDF—Verbatim [39 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 6(1), Art. 40, http://nbn-resolving.de/urn:nbn:de:0114-fqs0501404.

Forum Qualitative Sozialforschung / Forum: Qualitative Social Research (FQS)

ISSN 1438-5627

Creative Commons Attribution 4.0 International License

Theme of Secondary Analysis	Number of Primary Studies	Number of Interviews of Primary Studies	Software Used	HEATON's perspective	Aims
Use of electricity at night	14	93	Atlas-ti	New focus	Applied
Stress at work	1	146	Atlas-ti	New focus	Applied
Thermic comfort	15	138	Tropes, Atlas-ti	New focus, Subset	Methodological and Applied
Renovation	8	76	Alceste	New focus, Additional in-depth analysis	Applied and Methodological