Show
Grounded theory is a qualitative method that enables you to study a particular phenomenon or process and discover new theories that are based on the collection and analysis of real world data. Unlike traditional hypothesis-deductive approaches of research, where you come up with a hypothesis and then try to prove/disprove it, grounded theory is an inductive approach where new theories are derived from the data. The process of data collection, data analysis, and theory development happen in an iterative process. Iterative data collection and analysis occurs until you reach theoretical saturation, the point at which additional data adds no additional insight into your new theory. You should consider using grounded theory when there is no existing theory that offers an explanation for a phenomenon that you are studying. It can also be used if there is an existing theory, but it is potentially incomplete as the data used to derive that theory wasn’t collected from the group of participants that you plan on researching. The theories you develop using grounded theory are derived directly from real world participants in real world settings using methods like in depth interviews and observation, so your findings will more accurately represent the real world. This is in contrast to other research approaches that occur in less natural settings like research labs or focus group tables. Findings are tightly connected to the dataBecause grounded theory primarily relies on collected data to determine the final outcome, the findings are tightly connected to that data. This is in contrast to other research approaches that rely more heavily on external research frameworks or theories that are further removed from the data. Great for new discoveriesGrounded theory is a strong, inductive research method for discovering new theories. You don’t go in with any preconceived hypothesis about the outcome, and are not concerned with validation or description. Instead, you allow the data you collect to guide your analysis and theory creation, leading to novel discoveries. Offers strategies for analysisThe process of grounded theory describes specific strategies for analysis that can be incredibly helpful. While grounded theory is a very open ended methodology, the analysis strategies enable you to stay structured and analytical in your discovery process. Data collection and analysis are streamlinedData collection and analysis are tightly interwoven. As you collect data, you analyze it, and as you learn from analysis, you continue to collect more data. This helps ensure that the data you collect is sufficient enough to explain the findings that arise from analysis. Buffers against confirmation biasBecause data collection and analysis are tightly interwoven, you are truly following what is emerging from the data itself. This provides a great buffer against confirming preconceived beliefs about your topic. Want to learn how to do Grounded Theory? Submit your email to request our free grounded theory guide with tips on how to get started with your own thematic analysis. Grounded theory relies on an iterative recruiting process called theoretical sampling where you continuously recruit and conduct new rounds of interviews with new participants and previous participants while you analyze data. The recruiting criteria also evolves and changes based on what you learn. Because the recruiting is not predefined, it can be challenging to continuously find the right participants for your study. Time consuming to collect dataThere is no way to know ahead of time how much data you will need to collect, so you need to be flexible with your time. With grounded theory, you continuously collect and analyze data until you reach theoretical saturation, which is the point at which new data does not contribute new insight to your evolving theory. This means that you are likely to conduct many rounds of data collection before your theory is complete. Challenges in analysisData analysis occurs on a rolling basis and involves making constant comparisons between different excerpts of data. It can be challenging to keep track of your comparisons and findings as you go. It can be helpful to use a qualitative data analysis software like Delve to help you stay organized during your analysis.
Grounded Theory (GT) was first developed by Sociologist Barney Glaser and Anselm Strauss. During this period, they criticized the predominant approach to qualitative research, which they found to be very limited. Qualitative studies at this time were following traditional methods which basically involved coming up with a hypothesis and conducting research to validate it. Glaser and Strauss pioneered a new methodology for discovering theory by taking an inductive approach to qualitative research. They formally presented their newly developed research method by publishing Discovery of Grounded Theory: Strategies for qualitative research (1967). Since then, various evolutions of grounded to theory emerged, including Basics of Qualitative Research: Grounded Theory Procedures and Techniques (1990) by Strauss and Corbin. This shifted from the concept of the natural emergence of theory by designing an analytical coding framework for generating theories from data systematically. In 1990s, Kathy Charmaz published a new approach called constructivist grounded theory, and argued that neither data nor theories are discovered but are constructed through the researchers' past and present experiences. Read more about the history of grounded theory here. This is an overview of how you can approach the process of grounded theory. Know that this isn’t the only way to approach grounded theory, but just a collection of tips and processes derived from various grounded theory resources that you can use to inspire your own grounded theory study.
Steps for grounded theory
Note: Approach your research iterativelyGrounded theory is not a linear process where you collect data, analyze it, and then you’re done. It is an iterative research methodology that involves cycling through the steps iteratively. Part of what made Grounded Theory revolutionary was that it mixed data collection with analysis. It emphasized going back to the field even after conducting some analysis. You will recruit some participants, gather data and analyse it, and go back into the field again with a different recruiting strategy and focus of inquiry. Then you’ll incorporate those findings into further rounds of analysis. Grounded theory is deliberately cyclical in nature. 1. Determine initial research questionsStart off with your initial research questions. Have an idea for what phenomenon you are trying to explain. These initial questions will help guide your first steps in recruiting and data analysis but know that the questions may evolve as you observe and learn more from the data you collect. 2. Recruit and collect data using theoretical samplingWith grounded theory, recruiting participants is iterative. Instead of pre-determining a specified recruiting criteria ahead of time, you will practice what is called theoretical sampling. With theoretical sampling, you start with recruiting a small group of participants loosely based on your initial research questions. Once you have some data, such as recordings from in depth interviews, prepare that data for analysis by turning them into transcripts. After you do some initial analysis of that data, which we detail in the following steps, you use what you learned from that analysis to determine who to recruit next. Read more about how to do theoretical sampling here. After you have collected some data, such as transcriptions from interviews*, you can begin open coding. Open coding is when you take your transcripts, and break it into individual excerpts. Then, take the excerpts and continuously compare and contrast them with other excerpts This act of comparison is part of a core grounded theory method called constant comparative method, which you will use throughout various phases of your analysis. Notice similarities and differences between excerpts.
For example, in a study about the COVID-19 lockdown in New York City, you may read an excerpt that describes a person having trouble sleeping. You should take that excerpt and compare that to other people who also experienced trouble sleeping. Take notice of any similarities or differences between those experiences. *For the purpose of this article we will refer to collected data as ‘interview transcripts’ and ‘transcript excerpts’, but you can use any type of qualitative data such as observations, notes, etc. Read more about open coding here. Reflect on thoughts and contradictions by writing memos during analysisReflect on your analytical thoughts and write them down in the form of memos. Think of memos as your “notes to self” to record your train of thought, and to keep a record of your reflections. The act of writing memos can be a great way to reflect on any contradictions you find in the data. Your memos may eventually turn into the building blocks for your theory. Learn more about analytical memos here. 4. Group excerpts together into codes using open codingAs you make comparisons between excerpts of data, look for sets of excerpts that represent the same central idea or concept, and group them together. You can use a “code” to encapsulate these groups of excerpts. Codes are like tags or labels that are assigned to excerpts of text. For example, suppose you were comparing these excerpts:
All of these represent the concept of “trouble sleeping”. So if you are using qualitative data analysis software, you can create a code called “trouble sleeping” and bring all of these excerpts under the code “trouble sleeping”. Once you have a code called “trouble sleeping”, all future excerpts that you analyze should not only be compared to other excerpts, but they should also be compared to the code “trouble sleeping”, and any other code that you have.
As you gradually develop a list of codes that bring together sets of excerpts, you should also begin to also compare codes with other codes. When you find connections between multiple codes, you can group them together into a ‘category’. This step of grounded theory is called ‘axial coding’, where you find the axes that connect various codes together. If you are using qualitative data analysis software, these categories are represented by a series of ‘nested codes’ which are stacked in a hierarchy. For example, in the previous step, we had a code called “trouble sleeping”. Suppose you also had another code, “experiencing panic attacks”. You may find that there is a relationship between these 2 codes and they can be grouped under a category called “Reacting negatively to the pandemic with anxiety”.
6. Analyze more excerpts using constant comparative methodRemember, grounded theory is a cyclical process! Even after you have created lists of codes, and grouped codes into categories, you should continue to analyze additional interview transcripts, and compare the new excerpts to your existing codes categories. Read more about constant comparative method here. As you make comparisons between your new excerpts to your codes and categories, your excerpts will generally do one of three things: contradict, expand upon, or support your existing codes and categories. Here’s what you should consider in each scenario:
With grounded theory, your goal is not to code or keep track of everything that occurs in every excerpt. For example, once you establish the category that people under COVID-19 lockdown were [Reacting negatively to the pandemic with anxiety], you don’t need to go back and code every single excerpt that refers to that category. However, if you come across an excerpt where a person did not [react negatively to the pandemic with anxiety], this may open the doors to expanding upon or changing your category. 7. Continue collecting data and analyzing until you reach theoretical saturationWith these iterative steps, when do you know that you have analyzed enough? How do you know when you should stop recruiting or analyzing additional data? With grounded theory, you want to continue until you reach the point where additional transcript excerpts do not expand upon your codes and categories. In other words, if you are learning the same thing over and over again even with additional excerpts, that means that your codes and categories have become ‘theoretically saturated’. The excerpts you have collected so far address all relevant aspects of your codes and categories and there is no need to pursue further data collection or analysis for your particular codes and categories. 8. Define the core category using selective codingOnce you feel you have reached theoretical saturation in your codes and categories so far, it is time to pull your findings together with selective coding. With selective coding, you connect all your codes and categories together under one core category. This core category represents the central thesis of your research, and is the core idea behind your theory. This core category can be an existing category that you derived earlier, or it can be a new category that you derive from all your existing findings so far. This core category will be the basis for your new grounded theory. For example, if you have a list of categories like
You may use selective coding to define the central, core category as [Access to lockdown suitable housing, mitigated COVID-19 Anxiety during lockdown], to link all your existing categories together. Learn more about selective coding here. 9. Write your grounded theoryOnce you have determined your core category through selective coding, and are confident that you have reached theoretical saturation, it is time to construct your new theory.
You can do your grounded theory coding by hand, using word processors and spreadsheets such as Microsoft Word and Microsoft Excel, or use Computer Assisted Qualitative Data Analysis Software such as Delve. There are pros and cons to each approach, and you should choose one based off what is most appropriate for your research. Read more about how to code qualitative data.
You can use simple tools like pen, paper, scissors, and highlighters to code by hand. Just print out your transcripts, and do open coding by cutting up the transcripts into individual excerpts. The next steps are done by organizing those papers into piles, as you create your codes and categories. This is a great way to organize data with your hands, but can be come very time consuming, especially with large data sets. And it can be challenging to keep track of your comparisons since you’d have to keep track of all your sheets of paper. Grounded theory coding using a word processorIf you decide to use a word processor such as Microsoft Word or Google Docs, do your open coding round by highlighting excerpts. You can then code by adding comments to those excerpts. To create category, copy and paste excerpts into different documents labeled by the category name. This is a good way to keep your analysis in a digital format, but can feel cumbersome to continually copy and paste your Grounded theory coding using qualitative data analysis softwareSoftware such as the Delve qualitative data analysis software are designed to support processes like grounded theory. You can use Delve to help keep track of your excerpts and codes, and organize your thoughts as you do constant comparisons. The digital interface will help you manage large data sets and keep track of the many comparisons you will do. Additional features such as demographic filters and the ability to search across transcripts can also help streamline your grounded theory process. Make your life easier while doing grounded theory by using grounded theory software like Delve. Codes and categories are constantly evolving during the grounded theory analysis process. Using pen and paper or spreadsheets to analyze qualitative data can get unwieldy and chaotic. In contrast, qualitative data analysis software like Delve helps you make sense of the mess and focus on finding your insights. And don’t just take our word for it. Here’s what researchers say about using Delve:
Online software such as Delve can help streamline how you’re coding your qualitative coding. Try a free trial or watch a demo of Delve. References
Cite this blog post:
|