The data formats, process and tools needed for qualitative data analysis
Qualitative data comes in a number of different forms, so
there is no ‘one-size fits all’ answer. This article provides an overview of the different research
methods and the data gathered, with options for how to practically go about
your analysis of qualitative data in research.
Qualitative data collection is in-depth and focussed on a
smaller sample size than quantitative data. Undertaken usually in-person with individuals or in small groups, it
investigates perceptions, beliefs, attitudes, reasons why decisions are made,
and aims to understand and observe behaviour. Qualitative research output is based on interpretation and is expressed
in language, images, or even sound bites and video.
Qualitative data can be collected in a broad number of ways,
both via direct questioning, observation, and recording by the subject
themselves. Data can be collected
proactively from individuals or groups, or by gathering data shared via social
media. If interviewing a set of
respondents (either one-to-one interviews or in groups), a sampling method will be
applied to get a cross-section of the target audience (e.g. purchasers,
non-purchasers, age, gender, region).
Qualitative data can be collected via;
From such a range of data collection methods, it’s no
surprise that you can produce huge volumes of data, in a range of different
formats.Data can consist of text, both
verbally delivered or written, or can be a set of visual images recording
activity, behaviour or emotional observations in real, observed situations e.g.
reaction to new product, actions when shopping. Individual interviews and groups are usually
recorded with a transcript provided.
Different interpretation approaches are used to analyse qualitative
data depending on the objective, whether it’s understanding specific content
elements such as mention or recall of brands in content analysis, the
stories people tell using narrative analysis, themes and topics, such as
from those providing feedback on a service within thematic analysis, the
development of grounded theory or understanding customer norms or ideas
within discourse analysis.
The key challenge of qualitative data analysis, whatever the
objective, is the broad range of data collection techniques and multiple
formats of results: how is it possible to bring all this together to create a
cohesive set of conclusions?
Qualitative data analysis needs to enable;
The sheer volume of data can be difficult to sift through to
uncover themes.Various tools are
available to help summarise and filter the data;
The text heavy nature of qualitative data means that by
reading large volumes alone, it’s impossible to get a balanced sense of the
overall themes without some form of quantification. This quantification is done with a process
called coding. Coding enables you to
group open-ended data into broad or narrow themes which can be applied to the
whole dataset in a structured way. The
data can then be quantified into these themes or categories and even analysed
by the different respondent profiles. We cover separately the seven stages of analysing open-ended question data using the concept of coding.
Coding can be done manually either in Excel, within a
qualitative data analysis package or using a dedicated platform such as Codeit. Although Excel is
available widely, it is a poor choice for this kind of large scale qualitative data, so we compare the other platforms below.
There is no one-size-fits all for qualitative analysis, with
the options falling broadly into three categories;
For single-source text answers, it’s possible to use a dedicated
coding platform, such as Codeit. For
multi-source text and non-text data, there are dedicated qualitative analysis
platforms designed to assist.
Dedicated coding software tools such as Codeit are designed
specifically for open-ended survey questions, with multiple functions built in
to make the process efficient, fast and cost effective. Such tools enable accurate coding to be
completed fast, including the generation of topic themes within seconds. Further, for ongoing studies of large
volumes, Machine Learning is harnessed to automate further coding based
on the initial human-curated data.
For studies encompassing images, behaviour observation and
research across multiple sources, dedicated platforms exist to assist with the
organisation of analysis and coding across differently structured data sets and
The table below compares the two options;
If your survey is single source, with open-ended responses,
you can find out more about how to code by reading the seven stages of analysing open-ended questions and the practical steps of how to analyse open ended questions.
If you have multiple qualitative data sources and formats,
you’ll need to use a dedicated qualitative software for your analysis. However, if the majority of your data is text-based
answers from a small number of different sources, you can consider using a
dedicated coding platform such as Codeit which will give you all the power of
coding, with results available fast in a structured format of themes, enabling
easy, actionable, research findings.
you’d like to give Codeit a go, you can take up their 30 day free trial
to help you start coding and analysing your open-ended question data today.
Try it for Free
Anything we can help you with? Ask us