1 Lesson 1: Basic concept of ATLAS/ti

A few words about the history of the application: ATLAS/ti in its original form was developed as a DOS-application in a university research project. A commercial version (in German and English) was released thereafter. In the past two years, developers have constructed an extended version based on the initial template, introducing major changes for the Windows 3.1, Windows 95 and Windows NT operating systems.

The previous version, which imposed some inconveniences when used in a Windows-environment, was distributed globally and was being used mainly in the USA, Canada, Finland and the UK

In 1995, Eben Weizman and Matthew Miles published a book with test-results and comparative evaluations in which they describe ATLAS/ti, even the previous version, to be among the two most comprehensive and powerful qualsoft applications in the world. (The other one is NUDIST from Australia).

A number of new, additional features have been implemented in the latest release and ATLAS/ti is now likely to be the most powerful application of its kind available on the market.

Notes on marketing qualitative software

Software for qualitative data-analysis serves a very small but in particular highly specialized market mainly consisting of public institutions.

This further limits the sales-potential dramatically. (Sales-quantities similar to standard-software packages for offices are completely illusory) and therefore has an impact on the development-efforts for such applications. Large development-teams, helpdesk-centers cannot be financed. The continued development often depends on a particular person (Thomas Muhr for ATLAS/ti, Mr. and Mrs. Richards for NUDIST, John Seidel for ETHNOGRAPH, Alan Cartwright for CODE-A-TEXT, Udo Kuckartz for WINMAX

The interest of commercial developers of such applications (who have almost always emerged of university research-activities) must therefore be drawn at broadening the spectrum of applications.

This has the following impact on the qualitative social research: An application of that kind cannot afford to simply focus on one or two particular processes or methodologies within the field of qualitative sociology but must endeavor to offer tools for as many processes as possible. ATLAS/ti for example is oriented at the Grounded Theory and the qualitative content analysis but has become a universal tool for the entire field of the qualitative social research (including interfaces to the qualitative social research!).

Furthermore such applications are increasingly trying to open markets beyond social research. ATLAS/ti for example now offers the ability to analyze and code image- and audio-documents, which will certainly open new fields of application (e.g.: for Radiology and Sonography as well as for history of art and criminology.)

The basic purpose of ATLAS/ti

If your purpose is to qualitatively analyze, interpret, sort and administer textual, graphical as well as audio and video data and the goal is the acquisition of analytic ideas and the foundation of full theories based on that data, then the application provides you with the right tools. There are multiple functions to administer, extract, analyze, compare or aggregate meaningful data from the stack of all collected data.

An important approach is also to leave a space for creativity and flexibility to the researcher, which at the same time allows to work systematically.

ATLAS/ti does virtually apply no limits to the scope of data to be analyzed or neither to the number of created objects within the analysis (Concepts etc.) nor to the complexity of the entire analytic structure.

A typical work-procedure with ATLAS/ti might look like the following:

A typical work-procedure

  1. A project (a so-called hermeneutic unit "HU") is created as the first step. HU's serve the purpose of organizing the total number of findings, codes, memos, structures and data within a research-task as a name and register a main file around it.
  2. Secondly, all data-files, which may reside anywhere on the computer or on the network, are linked to the HU. This newly created data-skeleton for a given topic is already a useful tool. All data contained therein may be opened through this single HU-file.
  3. The next step is typically reading of data-texts, marking or quoting analytically interesting sections as well as assigning codes and memos to those. This is the main task in most of the projects at least with regard to the timely effort.
  4. Now, usually a comparative analysis of coded texts and if necessary, linking further data-texts (for example through ‚theoretical sampling‘), follows.
  5. A further step (does not necessarily have to be carried out at this time!) is to organize the various object-types (PDs, codes, memos) into groups or „families" (for example to organize the entire observation-protocols, all interview-transcripts and all codes into labor-division processes), which may then also be analyzed selectively or in contrast.
  6. Acquiring conceptual, semantic or logically expressive networks from the previously generated codes. Those networks lay the foundation to form the subject-related theory.
  7. Finally, a report that will incorporate many parts of the HU needs to be created.

    Variability of the working-steps

    The above steps and their order are neither fixed nor obligatory but they do outline the typical process of research-actions (a "Script"). ATLAS/ti allows you to vary these steps (but you won't be able to form a subject-related theory without having previously analyzed or coded data). It is especially possible to unroll working-steps integrated into one another within an iterative cyclical process, which is up to the maxims of the qualitative social-research. Another important conceptual background of ATLAS/ti is the effort to facilitate a superior support of teamwork-processes within the field of research (for example merging separately acquired HU's, administering multiple editors for a single HU, exchanging HU's through the Internet etc.).