Your complete guide to grounded theory research

What is grounded theory in research?

Grounded theory is a systematic qualitative research method that collects empirical data first, and then creates a theory ‘grounded’ in the results.

The constant comparative method was developed by Glaser and Strauss, described in their book, Awareness of Dying (1965). They are seen as the founders of classic grounded theory.

Research teams use grounded theory to analyse social processes and relationships.

Because of the important role of data, there are key stages like data collection and data analysis that need to happen in order for the resulting data to be useful.

The grounded research results are compared to strengthen the validity of the findings to arrive at stronger defined theories. Once the data analysis cannot continue to refine the new theories down, a final theory is confirmed.

Grounded research is different from experimental research or scientific inquiry as it does not need a hypothesis theory at the start to verify. Instead, the evolving theory is based on facts and evidence discovered during each stage.Also, grounded research also doesn’t have a preconceived understanding of events or happenings before the qualitative research commences.

Learn how Qualtrics Text iQ can automatically uncover trends, problems, and opportunities

When should you use grounded theory research?

Grounded theory research is useful for businesses when a researcher wants to look into a topic that has existing theory or no current research available. This means that the qualitative research results will be unique and can open the doors to the social phenomena being investigated.

In addition, businesses can use this qualitative research as the primary evidence needed to understand whether it’s worth placing investment into a new line of product or services, if the research identifies key themes and concepts that point to a solvable commercial problem.

Grounded theory methodology

There are several stages in the grounded theory process:

1. Data planning

The researcher decides what area they’re interested in.

They may create a guide to what they will be collecting during the grounded theory methodology. They will refer to this guide when they want to check the suitability of the qualitative data, as they collect it, to avoid preconceived ideas of what they know impacting the research.

A researcher can set up a grounded theory coding framework to identify the correct data. Coding is associating words, or labels, that are useful to the social phenomena that is being investigated. So, when the researcher sees these words, they assign the data to that category or theme.

In this stage, you’ll also want to create your open-ended initial research questions. Here are the main differences between open and closed-ended questions:

These will need to be adapted as the research goes on and more tangents and areas to explore are discovered. To help you create your questions, ask yourself:

• What are you trying to explain?
• What experiences do you need to ask about?
• Who will you ask and why?

2. Data collection and analysis

Data analysis happens at the same time as data collection. In grounded theory analysis, this is also known as constant comparative analysis, or theoretical sampling.

The researcher collects qualitative data by asking open-ended questions in interviews and surveys, studying historical or archival data, or observing participants and interpreting what is seen. This collected data is transferred into transcripts.

The categories or themes are compared and further refined by data, until there are only a few strong categories or themes remaining. Here is where coding occurs, and there are different levels of coding as the categories or themes are refined down:

• Data collection (Initial coding stage): Read through the data line by line
• Open coding stage: Read through the transcript data several times, breaking down the qualitative research data into excerpts, and make summaries of the concept or theme.
• Axial coding stage: Read through and compare further data collection to summarise concepts or themes to look for similarities and differences. Make defined summaries that help shape an emerging theory.
• Selective coding stage: Use the defined summaries to identify a strong core concept or theme.

Image from StatisticsHowTo.com

During analysis, the researcher will apply theoretical sensitivity to the collected data they uncover, so that the meaning of nuances in what they see can be fully understood.

This coding process repeats until the researcher has reached theoretical saturation. In grounded theory analysis, this is where all data has been researched and there are no more possible categories or themes to explore.

3. Data analysis is turned into a final theory

The researcher takes the core categories and themes that they have gathered and integrates them into one central idea (a new theory) using selective code. This final grounded theory concludes the research.

The new theory should be a few simple sentences that describe the research, indicating what was and was not covered in it.

An example of using grounded theory in business

One example of how grounded theory may be used in business is to support HR teams by analysing data to explore reasons why people leave a company.

For example, a company with a high attrition rate that has not done any research on this area before may choose grounded theory to understand key reasons why people choose to leave.

Researchers may start looking at the quantitative data around departures over the year and look for patterns. Coupled with this, they may conduct qualitative data research through employee engagement surveys, interview panels for current employees, and exit interviews with leaving employees.

From this information, they may start coding transcripts to find similarities and differences (coding) picking up on general themes and concepts. For example, a group of excepts like:

• “The hours I worked were far too long and I hated traveling home in the dark”
• “My manager didn’t appreciate the work I was doing, especially when I worked late”
• There are no good night bus routes home that I could take safely”

Using open coding, a researcher could compare excerpts and suggest the themes of managerial issues, a culture of long hours and lack of traveling routes at night.

With more samples and information, through axial coding, stronger themes of lack of recognition and having too much work (which led people to working late), could be drawn out from the summaries of the concepts and themes.

This could lead to a selective coding conclusion that people left because they were ‘overworked and under-appreciated’.

With this information, a grounded theory can help HR teams look at what teams do day to day, exploring ways to spread workloads or reduce them. Also, there could be training supplied to management and employees to engage professional development conversations better.

 Advantages of grounded theory

• No need for hypothesis – Researchers don’t need to know the details about the topic they want to investigate in advance, as the grounded theory methodology will bring up the information.
• Lots of flexibility – Researchers can take the topic in whichever direction they think is best, based on what the data is telling them. This means that exploration avenues that may be off-limits in traditional experimental research can be included.
• Multiple stages improve conclusion – Having a series of coding stages that refine the data into clear and strong concepts or themes means that the grounded theory will be more useful, relevant and defined.
• Data-first – Grounded theory relies on data analysis in the first instance, so the conclusion is based on information that has strong data behind it. This could be seen as having more validity.

Disadvantages of grounded theory

• Theoretical sensitivity dulled – If a researcher does not know enough about the topic being investigated, then their theoretical sensitivity about what data means may be lower and information may be missed if it is not coded properly.
• Large topics take time – There is a significant time resource required by the researcher to properly conduct research, evaluate the results and compare and analyse each excerpt. If the research process finds more avenues for investigation, for example, when excerpts contradict each other, then the researcher is required to spend more time doing qualitative inquiry.
• Bias in interpreting qualitative data – As the researcher is responsible for interpreting the qualitative data results, and putting their own observations into text, there can be researcher bias that would skew the data and possibly impact the final grounded theory.
• Qualitative research is harder to analyse than quantitative data – unlike numerical factual data from quantitative sources, qualitative data is harder to analyse as researchers will need to look at the words used, the sentiment and what is being said.
• Not repeatable – while the grounded theory can present a fact-based hypothesis, the actual data analysis from the research process cannot be repeated easily as opinions, beliefs and people may change over time. This may impact the validity of the grounded theory result.

What tools will help with grounded theory?

Evaluating qualitative research can be tough when there are several analytics platforms to manage and lots of subjective data sources to compare. Some tools are already part of the office toolset, like video conferencing tools and excel spreadsheets.

However, most tools are not purpose-built for research, so researchers will be manually collecting and managing these files – in the worst case scenario, by pen and paper!

Use a best-in-breed management technology solution to collect all qualitative research and manage it in an organised way without large time resources or additional training required.

Qualtrics provides a number of qualitative research analysis tools, like Text iQ, powered by Qualtrics iQ, provides powerful machine learning and native language processing to help you discover patterns and trends in text.

This also provides you with research process tools:

• Sentiment analysis — a technique to help identify the underlying sentiment (say positive, neutral, and/or negative) in qualitative research text responses
• Topic detection/categorisation — The solution makes it easy to add new qualitative research codes and group by theme. Easily group or bucket of similar themes that can be relevant for the business and the industry (eg. ‘Food quality’, ‘Staff efficiency’ or ‘Product availability’)