Skip to content

Changes

Page history
Update BA DS Research Software authored by Barget, Monika (HISTORY)'s avatar Barget, Monika (HISTORY)
Hide whitespace changes
Inline Side-by-side
BA-DS-Research-Software.md
View page @ f6780495
Research software plays an important role both in quantitative and qualitative research, but (thesis) students often struggle with finding software that is easy to use, GDPR-compliant, safe, and ethically sound.
Ideally, students should abstain from buying commercial software of their own. Especially cheap (temporary) licenses for cloud-based tools often imply that ingested data is used for tool improvement, such as AI training, and this can conflict with academic requirements for storing and processing research data.
❗ Ideally, students should abstain from buying commercial software of their own. Especially cheap (temporary) licenses for cloud-based tools often imply that ingested data is used for tool improvement, such as AI training, and this can conflict with academic requirements for storing and processing research data. ❗
Students' first point of information on research software should, therefore, be the UM library, who provide an overview of tools for which we have official UM student subscriptions:
......@@ -15,7 +15,11 @@ In addition, there is a growing pool of recommended open-source research softwar
</summary>
[Atlas.ti](https://atlasti.com/) is a tool for qualitative research, e.g., coding interview data, for which Maastricht University provides a license key to students and staff. The software is available through the university's software store, where students can select a specialised student version. This version of _Atlas.ti_ features a dashboard that facilitates the easy creation of various projects. The top bar offers functions for adding new documents and accommodates multimedia files with a wide range of file extensions. _Atlas.ti_ allows the interactive categorisation and tagging of your data and includes visualisation options (e.g network diagrams and word clouds). However, you should only use the software if you are thoroughly familiar with its functionality and can interpret visualisations correctly. More traditional and equally effective ways of coding your data are pen and paper or spreadsheets (such as EXCEL).
[Atlas.ti](https://atlasti.com/) is a tool for qualitative research, e.g., coding interview data, for which Maastricht University provides a license key to students and staff.
The software is available through the university's software store, where students can select a specialised student version. This version of _Atlas.ti_ features a dashboard that facilitates the easy creation of various projects. The top bar offers functions for adding new documents and accommodates multimedia files with a wide range of file extensions. _Atlas.ti_ allows the interactive categorisation and tagging of your data and includes visualisation options (e.g network diagrams and word clouds).
❗ However, you should only use the software if you are thoroughly familiar with its functionality and can interpret visualisations correctly. More traditional and equally effective ways of coding your data are pen and paper or spreadsheets (such as EXCEL). ❗
In the BA DS programme, _Atlas.ti_ was demonstrated in the _Qualitative Research Methods_ course. In the lecture by Thom Frissen, the sample dataset consisted of nine cooking recipes, which were used to explain the process of [closed coding](https://uh-dcm.github.io/qualitative-research-and-computers/closed-coding/). Closed coding, in qualitative research, means assigning predefined codes to the research material. The term "coding" in this context refers to a deliberate process of data reduction based on a specific theoretical framework. Codes are developed within the _Atlas.ti_ "[code manager](https://doc.atlasti.com/ManualWin.v9/Managers/ManagerForCodes.html)" and then incorporated into a codebook. All codes are presented in a dropdown menu, organised alphabetically by default. Within the _Atlas.ti_ software, users can open a document and manually highlight individual lines of text or even specific words to apply their codes. The software also offers automated "[search and code](https://doc.atlasti.com/ManualWin.v9/SearchAndCode/SearchAndCode.html)" functionalities. It should be noted, however, that cooking recipes, which are heavy in nouns and short, repetitive instructions, are not necessarily representative of the texts that humanities and social sciences scholars normally use for research. Therefore, the analytical process applied to this material needs to be modified when working with interviews or social media data. For instance, _Atlas.ti_ permits the grouping of interview transcripts with field observation notes.
......@@ -32,7 +36,8 @@ Atlas.ti maintains [an official YouTube channel](https://www.youtube.com/channel
### aTrain (automated audio / video transcription)
</summary>
According to its developer Jürgen Fleiss, "[aTrain](https://github.com/JuergenFleiss/aTrain) is a tool for automatically transcribing speech recordings utilizing state-of-the-art machine learning models without uploading any data." It is one of two tools for audio and video transcription that we officially recommend at FASoS because it is GDPR-compliant. You can try and install aTrain on your computer, but your machine may not be powerful enough. If that is the case, you can contact The Plant at FASoS and use one of their high-end PCs with a pre-installed version of aTrain for your transcriptions.
According to its developer Jürgen Fleiss, "[aTrain](https://github.com/JuergenFleiss/aTrain) is a tool for automatically transcribing speech recordings utilizing state-of-the-art machine learning models without uploading any data." It is one of two tools for audio and video transcription that we officially recommend at FASoS because it is GDPR-compliant. You can try and install aTrain on your computer, but your machine may not be powerful enough. If that is the case, you can contact 🪴 The Plant at FASoS and use one of their high-end PCs with a pre-installed version of aTrain for your transcriptions.
\[_Summary by Monika Barget_\]
</details>
......@@ -46,7 +51,9 @@ According to its developer Jürgen Fleiss, "[aTrain](https://github.com/JuergenF
The [Data Science Research Infrastructure (DSRI)](https://dsri.maastrichtuniversity.nl/) is available to all students at Maastricht University to collect, process, and analyse data. On the DSRI, you can run Python and R code for web scraping, data visualisation, statistical analysis, etc.
Most BA DS students first use the DSRI as part of the QDA course. Some students reuse the projects they created then for their theses. Other students would like to set up new DSRI projects because they intend to work with different methods or different code. This is definitely possible! In both cases, however, you should contact the DSRI Support Team early on to let them know how long you will need access and why the access is important for your research. The Support Team will then make sure to grant you access for the whole thesis trajectory.
Most BA DS students first use the DSRI as part of the QDA course. Some students reuse the projects they created then for their theses. Other students would like to set up new DSRI projects because they intend to work with different methods or different code. This is definitely possible!
❗ In both cases, however, you should contact the DSRI Support Team early on to let them know how long you will need access and why the access is important for your research. The Support Team will then make sure to grant you access for the whole thesis trajectory. ❗
If you are building a new project based on a Git repository or Docker image that includes a specific course name (e.g. "Machines of Knowledge") although you have never taken the course yourself, this is not an issue. It only indicates the material was originally developed for that course. However, you always have to use the most recent repository version. For example, if you are working in 2024, use the repository labeled with "2024" because earlier versions (e.g., "2023") may lack key updates. If you encounter technical issues, contact the DSRI Support Team directly. They have backend access and can troubleshoot issues much faster than your supervisor or course instructor:
......@@ -61,7 +68,7 @@ If you are building a new project based on a Git repository or Docker image that
### noScribe (automated audio / video transcription)
</summary>
[noScribe](https://github.com/kaixxx/noScribe) is an AI-assisted and at the same time GDPR-compliant transcription tool for different operating systems that simplifies the automatic transcription of interviews, discussions, or videos. It offers speaker recognition, multi-language support, and annotation features. The tool was specifically designed for academic use and does not send any of your data to the cloud for model training as you run the software either locally on your machine or in a Docker image. The downside is, of course, that installing noScribe takes up quite some space on your computer, and you need sufficient computational power, too. If your computer is not powerful enough, you can contact The Plant at FASoS and use one of their high-end computers for your transcriptions.
[noScribe](https://github.com/kaixxx/noScribe) is an AI-assisted and at the same time GDPR-compliant transcription tool for different operating systems that simplifies the automatic transcription of interviews, discussions, or videos. It offers speaker recognition, multi-language support, and annotation features. The tool was specifically designed for academic use and does not send any of your data to the cloud for model training as you run the software either locally on your machine or in a Docker image. The downside is, of course, that installing noScribe takes up quite some space on your computer, and you need sufficient computational power, too. If your computer is not powerful enough, you can contact 🪴 The Plant at FASoS and use one of their high-end computers for your transcriptions.
[Summary by Monika Barget]
</details>
......@@ -122,7 +129,9 @@ The programming language [R](https://www.r-project.org/) and its integrated deve
[Voyant Tools](https://voyant-tools.org/) is a collection of analytical tools for computational text analysis (also known as _distant reading_). Distant reading helps you get high-level insights into large text corpora but can also encourage you to understand a small collection of texts in a new way. The focus is on finding patterns: Which words are most frequent? Which words co-occur throughout the text? How does the vocabulary used change in the course of a narrative?
In the BA DS programme, the students learn to perform basic text analysis with R, but not all students are confident using R independently for their theses. Here, Voyant Tools can offer students with fewer computational skills the opportunity to get high-level insights into a text corpus exclusively via a graphic user interface. Voyant Tools is not officially taught in the BA DS programme and should, therefore, only be used in agreement with your thesis supervisor.
In the BA DS programme, the students learn to perform basic text analysis with R, but not all students are confident using R independently for their theses. Here, Voyant Tools can offer students with fewer computational skills the opportunity to get high-level insights into a text corpus exclusively via a graphic user interface.
❗ Voyant Tools is not officially taught in the BA DS programme and should, therefore, only be used in agreement with your thesis supervisor. ❗
Monika Barget created the following slides for teaching essential text analysis / distant reading with Voyant Tools to MA students in the humanities and social sciences:
......