The aim of this blogpost is to review data analysis software. None of the qualitative software (nVivo, Atlas.ti, Maxqda, dedoose.) actually “analyse” the data for you in the way that quantitative software does. They are more applicable in systematically organising, coding, annotating, visualising, etc. rather than data analysis. In essence, the analysis is still the user’s burden, and the software only assists in managing and exploring the data. In particular, these software overlap quite heavily in terms of the features they offer, and most of their differences are in the “look and feel” of the interface.
The understanding of human behaviour is facilitated through “methodological triangulation”, a term used in the social sciences to describe the combined use of more than two research methods to reach more reliable results.1 Thematic content analysis consists of three stages: 1) pre-analysis, 2) exploration, and 3) treatment and interpretation.2 MAXQDA can cover all three stages and attribute them to codes, accessing text, writings, memos and results in the form of tables and graphs. All of these functions offer users a wide scale of analytical thinking, in addition it obliges them to follow several ethical regulations.
Ethics within the MAXQDA
My research is composed of two main parts: a strictly anonymous survey distributed online, and anonymised semi-structured interviews conducted among members of the Jewish population in Luxembourg and the Czech Republic. At the beginning, it was difficult to navigate the jargon of a wide variety of values, norms and institutional arrangements in both European countries. The new EU General Data Protection Regulation (GDPR) has facilitated my work since I no longer have to follow two sets of regulations. The final consent letter is valid for all segments of the Jewish population concerned by my research in both countries. My research participants include people that are potentially at risk of psychological harm and invasion of privacy. I focus on the older generation who survived the Holocaust, and how this traumatic experience affected their behaviour and reactions. Any research project must take into account the fragility of respondents and ensure that it does not cross the line of personal intimacy. The new GDPR also protects interviews from being intentionally defamed by respondents. It often happens that people want to see transcriptions of their interviews, and sometimes they ask researchers to “enhance” their answers. The signed consent form allows the researcher to refuse this request. This may however cause the speaker to withdraw the interview entirely.
In my opinion, it is much better to forego one interview than to put your entire project at risk. The MAXQDA offers its own online cloud for simple editing. The user can simply withdraw any data at any time without breaking the whole corpus. It also allows the user to store backups from the data. In short, lawful data collection uploaded and stored in a handy platform enables scholars to work on controversial themes without the ongoing fear that authorities might put an end to the project at a later stage.
Document Portraits and Codes Relations Browser
In my research, I have mostly been using exploratory tools. For a comparative analysis, I use a thematic coding tree. This codebook-tree makes easier to find intersections in my work. One of my favourite tools is the Document Portrait displaying any text as a „painting” of either all or specific, selected codes assigned throughout the text. You can either choose colours for your code (e.g. a special colour for Holocaust (black) or Jewish traditions (green) or select some emoticons to stress the positive and negative aspects. At first glance, factors that played an important role are highlighted and therefore easier to locate in the interviews or any other text . You can of course change the assignment of colours at any time throughout the analytical process with a few clicks. A Document Portrait shows the sequence of codes for a selected document. All coded segments are listed in this overview.
This tool, however, takes the size of the segments into account and “weights” the colour according to the segment’s size. The colour attributes of the codes associated with the document are displayed in a matrix with 1,200 little squares (arranged in 30 rows, each one with 40 squares). In my opinion, this proves quite handy and remains uncluttered for viewers. The representation starts with the first square on the left in the first line, and ends with the last square on the right of the last line. All 1,200 squares are divided up according to the share of the coded segments. If a document just has one segment coded with a blue code, then the Document Portrait would be totally blue. If this segment is coded with two codes, red and blue, the Document Portrait would consist of 600 blue and 600 red squares.
Thus, Document Portrait displays the structure of documents in terms of codes as a colourful picture. This helps to immediately identify the basic tone of a document, for instance, if emotions have been coded with meaningful colours (aggressive= red, tolerant= green, and so on).
I also appreciate the Code Relations Browser’s visualisation, which shows the overlapping of codes in a chosen document and allows a quick identification of possible connections between codes. The tool is also a good way to test the quality of your code system. If you have no intersections in your corpus, it may mean that you set up a wrong coding system. The Codes Relations Browsers charts maps each MAXQDA document on the X-axis, whereas the Y-axis contains the whole code-system. If larger or smaller squares are located between the axes, then the codes overlap. You can decide whether you want to analyse a particular segment or the whole corpus. Nevertheless, it makes sense to employ this tool on the whole dataset to see a complex statistic of intersections.
A double-click on the square symbol shows the linked text. You can show the number of overlapping codes by clicking the “show values” button. The document basis view on the other hand simply indicates whether any codes overlap or not. In addition, I appreciate the export function turning the data chart into an Excel or Word file. The Code RelationBrowser nicely visualizes statistics and the export can easily included in presentations.
Regarding MAXQDA’s Ethic rules, users are defined as those who add structure and conduct qualitative explorations. Such an analysis is still fully dependent on humankind and we cannot ignore our responsibility towards the datasets we are using. MAXQDA is both a virtual analytical platform as well as an intriguing environment to generate new ideas and interpretations. The great freedom can also prove to be a disadvantage, especially for projects without clear boundaries.
Critical overview
To conclude my blog post, I want to stress a few differences between two world’s leading digital tools on various types of analysis. NVivo is a qualitative data analysis (QDA) computer software package produced by QSR International. It has been designed for qualitative researchers working with very rich text-based multimedia information. It is eligible for conducting profound analysis on small or large volumes of data. In my opinion, the NVivo was designed to help users organize and analyze non-numerical and mostly unstructured data. Both applications allow users to classify, sort and arrange information; examine relationships in the data; and combine analysis with linking, shaping, searching and modelling. According to my previous experience, the Maxqda is more capable of testing different theories, identify trends, and cross-examine information. The search engine within the coding tree is helpful in finding a code you need to enter. You can easily build a body of evidence to back up your research questions, but the NVivo accommodates a wide range of engines to make literature reviews and an ethnographical social network. Both applications equally support data formats such as audio files, videos, digital photos, Word, PDF, spreadsheets, rich text, plain text as well as web and social media data. Users can interchange data with applications such as Microsoft Excel, Microsoft Word, IBM SPSS Statistics, EndNote, Microsoft OneNote, Survey Monkey and Evernote.
Lastly but importantly, the biggest advantage of any qualitative data analysis software is that you can manage all your resources in a single place including (audio, images, pdf files, multimedia, etc.). You can easily display your data at any point of the process in the form of (graphs, word clouds, matrixes, etc.). The software also allows users to combine some quantitative data with the qualitative data. To share your project with others may be easier if you all have access to the software. The substantial cons of the NVivo is that the Mac version does not have the same functionalities as the Windows version.