In June MethodSpace is focusing on Collecting Data with Digital or Virtual Methods. In this guest post Rebecca Whiting and Katrina Pritchard discuss ways to use tracking or trawling to collect data online.
Explore the whole 2021 series on stages of the research process: Finding the Question, Choosing Methodology and Methods, Designing an Ethical Study, Collecting Data from & with Participants, and Finding Data in Documents and Datasets.
Collecting Data Online using Tracking and Trawling
by Rebecca Whiting and Katrina Pritchard
Organizational use of the Internet and digital technology has significantly changed existing working practices and introduced new ones. This has never been truer than during the Covid-19 pandemic. We have seen a dramatic growth of interest in digital research methods, as researchers explore digital work and organization but also look for new ways to supplement traditional face-to-face methods. Relatedly, digital data – and the organizations that collect, use and sell this – are frequently in the news whilst debates rage about the (usually negative) impact of social media and digital devices on contemporary society. This makes engaging with digital data and/or online research both exciting and daunting.
In our book – Collecting qualitative data using digital methods – we focus on what is often called ‘secondary’ data. This can suggest that a data set already exists and awaits analysis. We find this unhelpful; rather we consider ‘online data’ requires a systematic approach. This involves careful research design since although our research might use every day Internet tools, research is much more than browsing the Web for information about a topic.
Tracking and Trawling
We propose two ways of thinking about collecting online qualitative data – tracking and trawling. In collecting Internet material in this way, tracking and trawling follow in the tradition of collecting documents and undertaking documentary analysis in business and management research (Lee, 2012), taking this approach online and accommodating the inherent multimodality of Internet material. This makes it possible to collect a wide range of different types of material from sources including government, newspapers, courts, campaign groups and new web-enabled genres such as social media. These sources are likely to feature texts, image and sometimes sound. This range of material and its multimodality enable the researcher to develop and address a wide range of research questions.
Table 1: Overview of tracking and trawling
|Tracking||Use a variety of digital means to track (or follow) a particular event and/or people or groups of interest and/or a concept due to their engagement with a specific topic of relevance to the research. It is usually prospective in that it involves tracking from the start of the project onwards in time to capture new material as it appears online.|
|Trawling||Uses specific, focused searches to provide access to potentially relevant material from a variety of source types (e.g. websites, blogs, Twitter). It is usually retrospective in that it involves trawling the Internet for existing material that has already been published or posted before the start of the research project.|
We chose the term ‘tracking’ to reflect the idea that a researcher is following digital footprints as they emerge, whereas in contrast when we are ‘trawling’ we are looking back across the Internet to capture material that is already online. In this way, our approaches are temporally sensitive and can be combined, as a researcher may often be interested in moving back and forth. Below we summarise the similarities and differences between these approaches.
Table 2. Similarities and Differences
|Automated search, usually set up to repeat over time||Manual one-off search|
|Prospective in time||Retrospective in time|
|New search criteria cannot be added retrospectively, only prospectively.||Researcher can amend criteria or platforms, and repeat.|
|Require clear search criteria, but can be performed for |
multiple sets of criteria
|Can use proprietary tools, but are influenced by their algorithms|
|Can be combined with each other and other methods of data collection|
|Can capture different forms of Internet data|
|Require researcher to process and organise the material identified|
Having provided a brief introduction, below we highlight some key areas for consideration when planning to collect qualitative data using digital methods.
Understanding multimodal data
Qualitative internet data are often inherently multimodal, combining many different forms, including image, text and sound. You therefore need to decide what forms of data you are interested in and how you will manage these across your research project. It might be tempting to extract one form of data, especially since most qualitative analytic methods are geared towards textual data. However, it is important to consider whether a more in-depth, multimodal study will offer the potential to generate more insight than one that focuses on a particular mode.
Considering flow, transience and circularity of data
While we often talk about ‘online’ and ‘the Internet’ as though it were a single space, of course it is vastly complex. Something we might recognise as data, and want to collect, might often only become ‘fixed’ through our research activity e.g. downloading and storing. A photo, text or video might appear on multiple sites, and equally disappear overnight. On the other hand, sometimes a post or tweet can have a long lasting impact and resurface unexpectedly. Language is another example of this contextual information, and our own linguistic abilities can shape the way in which we can both track and trawl online. For this reason, it is vital that we appropriately contextualise our data collection and keep metadata related to our research activities.
Using online tools
The availability of online tools, particularly of the accessible, intuitive and user-friendly variety, make this type of research particularly appealing. Yet these tools are active in the shaping the research process, not objective, silent partners. Nor are they static: the much-discussed algorithms that lie at the heart of many proprietary products are regularly revised. For those who already have more advanced technical skills, or are afforded the time to develop them, it may be possible to use an increasing array of more specialised tools. However, irrespective of the particular toolset you deploy, it is important to pilot how these will be implemented within your research project, including considering how you will manage the data that you collect.
The ethics of collecting online data has been a developing area in recent years, with related discussions of copyright emerging at the same time. There is not the space to offer a full consideration in this blog, but we highlight the importance of engaging with your own institutional guidance and consulting expert resources such as the Association of Internet Researchers. (Learn more with this open-access chapter, “Digital Ethics.”)
Managing your data
Given the points noted already, it will come as no surprise that we suggest data management as a key challenge for those collecting qualitative data using digital methods. You will need to consider: keeping the data in a stable format (particularly given the inherently temporary and interactive nature of Web material); maintaining data security; ensuring compliance with applicable legislative, regulatory and ethical requirements; enabling data search, retrieval and analysis. We recommend checking what data management systems are available within your own institution, especially where this makes otherwise expensive software free to use.
Qualitative methods have much to offer as we explore experiences shaped through and by the internet. Tracking and trawling offer a great starting point for these explorations.
Katrina and Rebecca recently ran a related webinar for University of Bath Centre for Qualitative Research, and the recording and related materials are available via their Age at Work Blog.
Relevant MethodSpace Posts
- Facebook Groups as Research Method
- The One Netnography Tool You Should Never Be Without: The Immersion Journal—Part 1 of 4
- Practical Tips for Getting Started with Harvesting and Analyzing Online Text
- Online Research: Analyze Talk
- Thinking about collecting qualitative data using digital methods? Introducing Tracking and Trawling