Qualitative Coding Tutorial: How To Code Qualitative Data For Analysis (4 Steps Examples)

In this video, we are going to dive into the  topic of qualitative coding which you will need to understand if you plan to undertake  qualitative analysis for any dissertation, thesis or research project. We will  explain what exactly qualitative coding is, the different coding approaches and methods and  how to go about coding your data step by step. So, go ahead, grab a cup of coffee, grab a cup of tea  or whatever works for you and let us jump into it.

Hey, welcome to Grad Coach TV where we demystify  and simplify the oftentimes intimidating world of academic research. My name is Emma, and today  we are going to explore qualitative coding, an essential first step in qualitative analysis.  If you would like to learn more about qualitative analysis or research methodology in general we  have also got videos covering those topics so be sure to check them out.

I will include the  links below. if you are new to Grad Coach TV hit that subscribe button for more videos covering  all things research-related. Also, if you are looking for hands-on help with your qualitative  coding check out our one on one coaching services where we hold your hand through the coding  process step by step.

Alternatively, if you are looking to fast track your coding  we also offer a professional coding service where our seasoned qualitative experts code  your data for you ensuring high-quality initial coding. If that sounds interesting  to you, you can learn more and book a free consultation at gradcoach. com.

All right,  with that out of the way let us get into it. To kick things off let us start by understanding  what a code is. At the simplest level, a code is a label that describes a piece of content.

For  example, in the sentence pigeons attacked me and stole my sandwich you could use pigeons as  a code. This code would simply describe that the sentence involves pigeons. Of course, there are  many ways you could code this and this is just one approach.

We will explore the different ways  in which you can code later in this video. So, qualitative coding is simply the process of  creating and assigning codes to categorize data extracts. You will then use these codes  later down the road to derive themes and patterns for your actual qualitative analysis.

For  example, thematic analysis or content analysis. It is worth noting that coding and analysis can  take place simultaneously. In fact, it is pretty much expected that you will notice some themes  emerge while you code.

That said, it is important to note that coding does not necessarily involve  identifying themes instead it refers to the process of labelling and grouping similar types of  data which in turn will make generating themes and analysing the data more manageable. You might be  wondering then why should I bother with coding at all why not just look for themes from the outset?  Well, coding is a way of making sure your data is valid.

In other words, it helps ensure that  your analysis is undertaken systematically and that other researchers can review it. In the world  of research, we call this transparency. In other words, coding is the foundation of high-quality  analysis which makes it an essential first step.

Right, now that we have got a plain language  definition of coding on the table the next step is to understand what types of coding exist.  Let us start with the two main approaches, deductive and inductive coding. With deductive  coding, you as the researcher begin with a set of pre-established codes and apply them to your data  set.

For example, a set of interview transcripts. Inductive coding on the other hand works in  reverse as you start with a blank canvas and create your set of codes based on the data itself.  In other words, the codes emerge from the data.

Let us take a closer look at both of  these approaches. With deductive coding, you will make use of predetermined codes also  called a priori codes which are developed before you interact with the present  data. This usually involves drawing up a set of codes based on a research question or  previous research from your literature review.

You could also use an existing code set from  the codebook of a previous study. For example, if you were studying the eating habits of college  students you might have a research question along the lines of what foods do college students eat  the most? As a result of this research question, you might develop a code set that includes  codes such as sushi, pizza and burgers.

You would then code your data set using only  these codes regardless of what you find in the data. On the upside the deductive approach  allows you to undertake your analysis with a very tightly focused lens and quickly identify  relevant data, avoiding distractions and detours. The downside of course is that you could  miss out on some very valuable insights as a result of this tight predetermined focus. 

Now, let us look at the opposite approach, inductive coding. As I mentioned earlier this  type of coding involves jumping right into the data without predetermined codes and developing  the codes based on what you find within the data. For example, if you were to analyse a set of  open-ended interview question responses you would not necessarily know which direction  the conversation would flow.

If a conversation begins with a discussion of cats it might go on to  include other animals too. And so, you would add these codes as you progress with your analysis.  Simply put with inductive coding you go with the flow of the data.

Inductive coding is great when  you are researching something that is not yet well understood because the coding derived from the  data helps you explore the subject. Therefore this approach to coding is usually adopted when  researchers want to investigate new ideas or concepts or when they want to create new theories.  So, as you can see the inductive and deductive approaches represents two ends of a spectrum  but this does not mean that they are mutually exclusive.

You can also take a hybrid approach  where you utilize a mix of both. For example, if you have got a set of codes you have derived  from a literature review or a previous study, in other words, a deductive approach but you still  do not have a rich enough code set to capture the depth of your qualitative data you can combine  deductive and inductive approaches which we call a hybrid approach. To adopt a hybrid  approach you will begin your analysis with a set of a priori codes, in other words, a deductive  approach and then add new codes, in other words, an inductive approach as you work your way through  the data.

Essentially, the hybrid coding approach provides the best of both worlds which is why  it is pretty common to see this in research. All right, now that we have covered what  qualitative coding is and the overarching approaches let us dive into the actual coding  process and look at how to undertake the coding. So, let us take a look at the actual coding  process step by step.

Whether you adopt an inductive or deductive approach your coding will  consist of two stages, initial coding and line by line coding. In the initial coding stage the  objective is to get a general overview of the data by reading through and understanding  it. If you are using an inductive approach this is also where you will develop an initial set  of codes.

Then in the second stage line-by-line coding you will delve deeper into the data and  organize it into a formalized set of codes. Let us take a look at these stages  of qualitative coding in more detail. Stage one, initial coding.

The first step of the  coding process is to identify the essence of the text and code it accordingly. While there are many  qualitative analysis software options available you can just as easily code text-based data using  Microsoft Word's comments feature. In fact, if it is your first time coding it is oftentimes best  to just stick with Word as this eliminates the additional need to learn new software.

Importantly  you should avoid the temptation of any sort of automated coding software or service. No matter  what promises they make automated software simply cannot compare to human-based coding as it cannot  understand the subtleties of language and context. Do not waste your time with this.

In all  likelihood, you will just end up having to recode everything yourself anyway. Okay, so let us take a  look at a practical example of the coding process. Assume you had the following interview data  from two interviewees, in the initial stage of coding you could assign the code of pets or  animals.

These are just initial fairly broad codes that you can and will develop and refine  later. In the initial stage broad rough codes are fine they are just a starting point which  you will build onto later when you undertake line by line coding. So, at this stage, you are  probably wondering how to decide what codes to use especially when there are so many ways to  read and interpret any given sentence.

Well, there are a few different coding methods you can  adopt and the right method will depend on your research aims and research questions. In other  words, the way you code will depend on what you are trying to achieve with your research. Five  common methods utilized in the initial coding stage include in vivo coding, process coding,  descriptive coding, structural coding and value coding.

These are not the only methods  available but they are a useful starting point. Let us take a look at each of them to understand  how and when each method could be useful. Method number one in vivo coding.

When you use  in vivo coding you make use of a participant's own words rather than your interpretation of the  data. In other words, you use direct quotes from participants as your codes. By doing this you will  avoid trying to infer meaning by staying as close to the original phrases and words as possible.

In  vivo coding is particularly useful when your data are derived from participants who speak different  languages or come from different cultures. In cases like these, it is often difficult to  accurately infer meaning thanks to linguistic and or cultural differences. For example,  English speakers typically view the future as in front of them and the past as behind them  however this is not the same in all cultures.

Speakers of Aymara view the past as in front  of them and the future as behind them. Why, because the future is unknown it must be out of  sight or behind them. They know what happened in the past so their perspective is that it  is positioned in front of them where they can see it.

In a scenario like this one, it is not  possible to derive the reason for viewing the past as in front and the future as behind without  knowing the Aymara culture's perception of time. Therefore, in vivo coding is particularly  useful as it avoids interpretation errors. While this case is a unique one it illustrates  the point that different languages and cultures can view the same things very differently  which would have major impacts on your data.

Method number two, process coding. Next  up there is process coding which makes use of action-based codes. Action-based  codes are codes that indicate a movement or procedure.

These actions are often indicated by  gerunds that are words ending in ing. For example, running, jumping or singing. Process coding  is useful as it allows you to code parts of data that are not necessarily spoken but that  are still important to understand the meaning of the texts.

For example, you may have action  codes such as describing a panda, singing a song or arguing with a relative. Another example would  be if a participant were to say something like I have no idea where she is. A sentence like  this could be interpreted in many different ways depending on the context and movements of the  participant.

The participant could for example shrug their shoulders which would indicate  that they genuinely do not know where the girl is. Alternatively, they could wink suggesting  that they do actually know where the girl is. Simply put, process coding is useful as it allows  you to in a concise manner identify occurrences in a set of data that are not necessarily spoken  and to provide a dynamic account of events.

Method number three, descriptive coding.  Descriptive coding is a popular coding method that aims to summarize extracts by using a single word  that encapsulates the general idea of the data. These words will typically describe the data in  a highly condensed manner which allows you as the researcher to quickly refer to the content. 

For example, a descriptive code could be food when coding a video clip that involves a group  of people discussing what they ate throughout the day, or cooking, when coding an image showing  the steps of a recipe. Descriptive coding is very useful when dealing with data that appear in forms  other than text. For example, video clips, sound recordings or images.

It is also particularly  useful when you want to organize a large data set by topic area. This makes descriptive coding  a popular choice for many research projects. Method number four, structural coding.

True to  its name structural coding involves labelling and describing specific structural attributes of the  data. Generally, it includes coding according to answers of the questions of who, what, where and  how rather than the actual topics expressed in the data. For example, if you were coding a collection  of dissertations which would be quite a large data set, structural coding might be useful as you  could code according to different sections within each of these documents.

Coding what centric  labels such as hypotheses, literature review and methodology would help you to efficiently  refer to sections and navigate without having to work through sections of data all over again.  So, structural coding is useful when you want to access segments of data quickly and it can help  tremendously when you are dealing with large data sets. Structural coding can also be useful for  data from open-ended survey questions.

This data may initially be difficult to code as they lack  the set structure of other forms of data such as an interview with a strict closed set of questions  to be answered. In this case, it would be useful to code sections of data that answer certain  questions such as who, what, where and how. Method number five, values coding.

Last  but not least values-based coding involves coding excerpts that relate to the participants'  worldviews. Typically this type of coding focuses on excerpts that provide insight regarding the  values, attitudes and beliefs of the participants. In practical terms this means you would be looking  for instances where your participants say things like I feel, I think that I need and it is  important that as these sorts of statements often provide insight into their values,  attitudes and beliefs.

Values coding is therefore very useful when your research aims and research  questions seek to explore cultural values and interpersonal experiences and actions or when you  are looking to learn about the human experience. All right, so we have looked at  five popular methods that can be used in the initial coding stage. As I  mentioned this is not a comprehensive list so if none of these sound relevant to your project  be sure to look up alternative coding methods to find the right fit for your research aims. 

The five methods we have discussed allow you to arrange your data so that it is easier to navigate  during the next stage, line-by-line coding. While these methods can all be used individually it is  important to know that it is possible and quite often beneficial to combine them. For example,  when conducting initial coding with interview data you could begin by using structural coding to  indicate who speaks when.

Then as a next step, you could apply descriptive coding so that you  can navigate to and between conversation topics easily. As with all design choices, the right  method or combination of methods depends on your research aims and research questions. So, think  carefully about what you are trying to achieve with your research then select the method  or methods that make sense in light of that.

So to recap, the aim of initial coding is to  understand and familiarize yourself with your data to develop an initial code set if  you are taking an inductive approach and to take the first shot at  coding your data. Once that is done you can move on to the next stage,  line by line coding. Let us do it.

Stage two, line by line coding. Line by line  coding is pretty much exactly what it sounds like reviewing your data line by line, digging deeper,  refining your codes and assigning additional codes to each line. With line-by-line coding, the  objective is to pay close attention to your data to refine and expand upon your coding especially  when adopting an inductive approach.

For example, if you have a discussion of beverages and you  previously just coded this as beverages you could now go deeper and code more specifically, such  as coffee, tea and orange juice. The aim here is to scratch below the surface. This is the time to  get detailed and specific so that you can capture as much richness from the data as possible.

In the  line-by-line coding process, it is useful to code as much data as possible even if you do not think  you are going to use it. As you go through this process your coding will become more thorough  and detailed and you will have a much better understanding of your data as a result of this.  This will be incredibly valuable in the analysis phase so do not cut corners here.

Take your time  to work through your data line by line and apply your mind to see how you refine your coding as  much as possible. Keep in mind that coding is an iterative process which means that you will move  back and forth between interviews or documents to apply the codes consistently throughout your  data set. Be careful to clearly define each code and update previously coded excerpts if you  adjust or update the definition of any code or if you split any code into narrower codes. 

Line by line coding takes time so do not rush it, be patient and work through your data meticulously  to ensure you develop a high-quality code set. Stage three, moving from coding to analysis.  Once you have completed your initial and line by line coding the next step is to start  your actual qualitative analysis.

Of course, the coding process itself will get you in  analysis mode and you will probably already have some insights and ideas as a result of it so you  should always keep notes of your thoughts as you work through the coding process. When it comes to  qualitative data analysis there are many different methods you can use including content analysis,  thematic analysis and discourse analysis. The analysis method you adopt will depend heavily  on your research aims and research questions.

We cover qualitative analysis methods on the  Grad Coach blog so we are not going to go down that rabbit hole here but we will discuss the  important first steps that build the bridge from qualitative coding to qualitative analysis.  So, how do you get started with your analysis? Well, each analysis will be different but it is  useful to ask yourself the following more general questions to get the wheels turning.

What actions  and interactions are shown in the data, what are the aims of these interactions and excerpts,  how do participants interpret what is happening and how do they speak about it, what does their  language reveal, what are the assumptions made by the participants, what are the participants  doing, why do I want to learn about this, what am I trying to find out? As with initial  coding and line by line coding your qualitative analysis can follow certain steps. The first  two steps will typically be code categorization and theme identification.

Let us look at these two  steps. Code categorization which is the first step is simply the process of reviewing everything you  have coded and then creating categories that can be used to guide your future analysis. In other  words, it is about bundling similar or related codes into categories to help organize your data  effectively.

Let us look at a practical example. If you were discussing different types of animals  your codes may include dogs, llamas and lions. In the process of code categorization, you could  label in other words, categorize these three animals as mammals whereas you could categorize  flies, crickets and beetles as insects.

By creating these code categories you will  be making your data more organized as well as enriching it so that you can see new connections  between different groups of codes. Once you have categorized your codes you can move on to the next  step which is to identify the themes in your data. So, let us look at the theme identification step. 

From the coding and categorization processes, you will naturally start noticing themes  therefore the next logical step is to identify and clearly articulate the themes in your  data set. When you determine themes you will take what you have learned from the coding  and categorization stages and synthesize it to develop themes. This is the part of the  analysis process where you will begin to draw meaning from your data and produce a narrative. 

The nature of this narrative will, of course, depend on your research aims, your research  questions and the analysis method you have chosen, for example, content analysis or thematic  analysis. So, keep these factors front of mind as you scan for themes as they will  help you stay aligned with the big picture. All right, now that we have covered both  the what and the how of qualitative coding I want to quickly share some general tips and  suggestions to help you optimize your coding process.

Let us rapid-fire. One, before you  begin coding plan out the steps you will take and the coding approach and method or methods you  will follow to avoid inconsistencies. Two, when adopting a deductive approach it is best to use a  codebook with detailed descriptions of each code right from the start of the coding  process.

This will ensure that you apply codes consistently based on their descriptions  and will help you keep your work organized. Three, whether you adopt an inductive or deductive  approach keep track of the meanings of your codes and remember to revisit these as you go  along. Four, while coding keep your research aims, research questions, coding methods and analysis  method front of mind.

This will help you to avoid directional drift which happens when coding is  not kept consistent. Five, if you are working in a research team with multiple coders make sure that  everyone has been trained and clearly understands how codes need to be assigned. If multiple coders  are pulling in even slightly different directions you will end up with a mess that needs  to be redone, you do not want that.

So, keep these five tips in mind and you will  be on the fast track to coding success. And there you have it qualitative coding in a  nutshell. Remember, as with every design choice in your dissertation, thesis or research project  your research aims and research questions will have a major influence on how you approach the  coding.

So, keep these two elements front of mind every step of the way and make sure your  coding approach and methods align well. If you enjoyed the video hit the like button  and leave a comment if you have any questions. Also, be sure to subscribe to the channel for  more research-related content.

If you need a helping hand with your qualitative coding or  any part of your research project remember to check out our private coaching service  where we work with you on a one-on-one basis, chapter by chapter to help you craft a winning  piece of research. If that sounds interesting to you book a free consultation with a friendly  coach at gradcoach. com.

As always I will include a link below. That is all for this episode  of Grad Coach TV. Until next time, good luck.