PARTAGER L’ARTICLE SUR
As the name suggests, Thematic Analysis means analysing the patterns of themes in data. It is a method of qualitative data analysis. It means, this method can be used to analyse non-numerical data such as audio, video, text, etc.
Example:
An interview transcript. A researcher will have to go through the entire transcript and look for meaningful patterns in themes across the data.
The patterns can be analysed by repetitive data reading, data coding, and theme creation. The below picture depicts a flow of the process.
The patterns can be analysed by repetitive data reading, data coding, and theme creation. The below picture depicts a flow of the process. Let’s take an example to understand how thematic analysis helps in research questions.
Example: How has social media changed over the years?
The above research will need you to gather data from sources, blogs, news, interviews posted online. Interview a few new generation users of the platforms and the old users to gather intel about how they use the social platforms and what their experience is.
Some of the established players have started implementing Thematic Analysis to improve their Manual Rules processes but tend to produce a list of terms that are difficult to review. This approach works well for text analytics platforms that are focused on improving the customer experience. However, it avoids generating generic solutions that are usually not designed to solve the problem. Only a small portion of feedback is linked to the top 10 themes. Uncategorized feedback means that you can’t slice the data to get deeper insights. Thematic Analysis is a method that can easily analyze text-based feedback from multiple sources, such as email, social media, and real estate brokers.
Thematic analysis is an unsupervised approach that enables you to create categories and perform statistical tests without having to set up any rules or procedures in advance.
Thematic analysis is an unsupervised approach that enables you to create categories and perform statistical tests without having to set up any rules or procedures in advance.
Thematic analysis is typically phrase-based. Sometimes, it can’t capture the meaning of a phrase correctly. For instance, in a complex narrative, it can’t capture the customer’s intent to stop using the service.
With an adequate understanding of Thematic Analysis, let us dive into the various approaches we can choose to work with.
Focuses on developing a theory. This approach is used when there is not much information available on a topic and you have to build a theory straight from scratch. You can always validate this approach but it is hard to prove that observation made from this approach is correct. The inductive approach consists of three stages:
Focuses on testing an existing theory. It totally depends on the Inductive approach as you start from working on an already existing theory. You go on formulating the theory and derive a conclusion out of it. The genuineness of deductive theory depends on how much true inductive theory is. The deductive theory has four stages:
Analyse the result (does collected data reject or validate the hypothesis)- since all the roads are busy during working hours -> support a hypothesis.
Focuses on the details of the data. We research the data on the grounds that it has some secondary meaning and purpose to it. This will help to construct insights and information regarding how the data was being used.
Focuses beyond the semantics of the data and works more on the underlying meanings, concepts, and assumptions that we made earlier with the semantic approach.
In order to choose the best-fit approach for your study, go through its requirements and which approach or combination of approaches will best align with your data.
Once you have gathered adequate data and chosen your suitable approach, it is time to follow the following steps to build your thematic analysis for your problem statement.
It is important to be familiar with the data before we begin to dig deep into the individual topics. This can include re-reading the whole data, having an overview of its context, and taking out personal notes if necessary. This is will help you to know your data.
This includes highlighting or labelling certain words or group of words or even phrases in the data that all together indicates something. This something will come in handy when you are trying to grab the essence of the data. Let’s take an example to understand this:
Example: How has social media changed over the years?
Let’s say we are interviewing an old social media user here and her opinion on the problem statement. She says, « I think the social media platforms are not for us oldies anymore. The trends are rapidly changing and there is always something new on the wall every day. It becomes difficult for people like me to keep up with those. Hence we often feel disconnected. » Now we can derive codes for the highlighted phrases like; Fast change | Uninterested | Discomfort
Now that we have our codes, we can derive themes from them. Themes can have several codes indicating the same expressions. As for our above example, we can have a theme called « not satisfied » for all the codes we derived from the interview. This will give an idea about how many codes are being used again and again and which ones of them serve no purpose so we can just discard them.
Here we compare the themes with our original data and look for any missing points or irrelevant results. We can modify our themes depending on how they satisfy and justify the data after tracing them back to it.
Further ahead, we can name the themes depending on what they indicate and what we get to understand from it about the data.
For the last step, we will the results that we have come to and the conclusion that our thematic analysis has helped us to understand. As per our example, we can conclude that social media has changed so much that the older generations find it hard to interact with and result in their dissatisfaction on the matter.
So that is how to regulate your perfect Thematic Analysis for the next time you decide to research a problem statement.
Nous utilisons des cookies sur notre site web pour vous offrir la meilleure expérience de navigation et pour adapter la publicité. En continuant à utiliser notre site web, vous nous donnez votre accord pour l'utilisation des cookies. Lire la suite
English
French