Our qualitative analysis process has involved us working with a shared codebook but on a large number of interviews in different languages from each of the four countries in our study. Before starting, we had a short training session with a MAXQDA expert, in addition to watching numerous tutorials. This gave a good starting point for using a limited number of the available functions, the most essential being creating code systems, and coding interview transcripts.
Overall, it has been enjoyable to learn how to use a new software. Having used statistical software packages before, it felt somewhat intuitive to imagine how some functions, such as pulling out certain parts of the dataset, might work in MAXQDA. However, it was rewarding to attend some basic training in how the programme is specifically suited for different types of qualitative analysis. Neither of us had used any qualitative software before, and felt a similar kind of relief when familiarising ourselves with the coding functions as when moving from referencing by hand to using referencing software, such as Zotero or Mendeley. For a small dataset it might still be realistic to sit on the floor with sticky notes, highlighter pens, and printed transcripts, or whatever the preferred manual method may be, but seeing as we are working with eight field sites and about 200 interview transcripts in total, it is inconceivable to manage and code it all by hand.
Beyond relief, it has been exciting to learn about functions we had not been able to imagine before trying MAXQDA. While we have not yet had a chance to master things like photo or video coding and analysis, or visual tools like MAXMaps, it is great to know that there is much more to this software than what we have done so far.
In addition, we have benefitted from the transcription functions, as again, neither of us had used any tailored software for that before. Our days of jumping between MS Word and a media player are now over! Transcription becomes much faster when the recording can play at an adjusted speed, and be paused with the click of one button in the keyboard (F4, in case you haven’t tried it yet). This may not be revolutionary to more experienced transcribers but it still feels great. However, we do wish MAXQDA could check our spelling while we type away frantically… And, for the good of the project, in more than one language – this is a strong hint to MAXQDA software developers!
As for challenges, we have accepted that we as a large team, or combination of several smaller teams, are in a tricky (but not unusual) situation when it comes to sharing work, and doing joint analysis. We have not been able to connect or sync our work, and so everyone has been working independently and then sending and merging their coded data later on. It was difficult to see how all of this would work at the beginning, and thus we spent a lot of time fine-tuning our shared code system. Since we are many people working on many interviews, it makes sense that the analysis cannot be a fully intuitive, data-driven process anyway. In order to make some sense of it all, we needed to agree on most of the codes to use before starting the coding process. In hindsight, we probably did many of these stages in less efficient ways than might have been possible had we been familiar with the software in the first place.
Another challenge is that sometimes the files disappear and are not found. The backup files are however available but some of the audio tends to disappear. This happened more than once, and particularly when using MAXQDA on a USB stick as opposed to the version installed directly onto a PC. Another problem is that the more data we have in the USB, the slower MAXDA becomes and in some cases, it freezes and the recent changes are lost. Therefore, to attract long term users of MAXQDA, some of these aspects that make the everyday experience of the software more practical and user-friendly need to be addressed.
In sum, there is certainly more to MAXQDA than we have had a chance to experience, but the functions we have experienced could also benefit from some improvements. Becoming proficient in a new software requires a commitment to using it frequently, and so it is important that basic functionality does not put one off early in the process.