A note on using the inductive and the deductive approaches in analyzing qualitative research findings: for MBA students

A note on using the inductive and the deductive approaches in analyzing qualitative research findings for writing an MBA dissertation report chapter 5 (discussion of findings)

 

In the context of doing an MBA dissertation project, please offer a few straightforward advices on how to employ the inductive and the deductive approaches to analyze qualitative research findings, notably for writing of the dissertation report chapter 5 (discussion of findings).

For an MBA qualitative dissertation, the most practical option is usually a hybrid analysis: use deduction to organise findings around your research questions and literature-based concepts, then use induction to capture participants’ unexpected experiences and explanations. In Chapter 5, do not merely restate themes; explain what they mean, how they relate to prior research, and what they imply for managers.

Distinguish the two approaches

Approach	Starting point	What you look for	Best use in Chapter 5
Inductive (“bottom-up”)	Interview transcripts or other raw data	Patterns, meanings, and themes that emerge from participant accounts	Explain new, contextual, or surprising findings
Deductive (“top-down”)	Research questions, conceptual framework, or literature	Evidence that supports, refines, contradicts, or fails to show expected concepts	Evaluate your framework and connect findings with the literature review

In inductive thematic analysis, codes and themes are developed from the data without forcing them into an existing coding frame. Deductive analysis instead uses concepts or topics brought to the data by the researcher—for example, constructs identified in Chapter 2.

A straightforward workflow

1.    Set up deductive parent codes.
Derive a small number of initial coding categories from your research questions and literature review. For a study of online retail customer loyalty, these might be trust, perceived value, service responsiveness, and repurchase intention.

2.    Code openly within and beyond those categories.
As you read transcripts, retain codes that do not fit the framework. Do not discard an idea simply because Chapter 2 did not anticipate it.

3.    Develop inductive subthemes.
Group repeated, meaningful observations into themes. For example, under “trust,” participants may repeatedly describe “fear of counterfeit products,” which becomes a data-led subtheme.

4.    Create a theme-to-literature matrix.
For every final theme, record:

o   The relevant research question

o   A concise interpretation of the finding

o   Supporting participant quotations or evidence

o   The related theory or prior study from Chapter 2

o   Whether your finding confirms, extends, qualifies, or contradicts that literature

o   Its managerial implication

5.    Write Chapter 5 by research question or major theme—not participant by participant.
This makes the discussion analytical rather than descriptive.

The familiar thematic-analysis sequence—data familiarisation, initial coding, theme development, theme review, definition/ naming, and report writing—supports either an inductive, deductive, or combined design.

Write Chapter 5 in four moves

Use the following structure for each research question or major theme:

1.    State the finding.
“Participants generally viewed rapid response to enquiries as a key basis for trusting an online seller.”

2.    Interpret its meaning.
Explain the underlying mechanism: “Here, responsiveness appears to signal accountability and reduce the perceived risk of buying without physically inspecting a product.”

3.    Compare it with literature.
“This supports the service-quality literature’s emphasis on responsiveness, but the findings further suggest that the effect is especially strong where authenticity and delivery reliability are uncertain.”

4.    Draw the implication.
“Online retailers should treat pre-sale response time as a trust-building process rather than simply a customer-service operational metric.”

That final comparison is where deduction is most visible. The added insight about authenticity concerns is where induction contributes.

Example wording

Consistent with the proposed conceptual framework, perceived value was important to participants’ intention to repurchase. However, interview accounts suggest that value was not understood solely as a lower price. Participants frequently referred to transparent delivery charges, product authenticity, and the effort saved by reliable post-sale support. This inductively derived interpretation extends the conventional price–quality view of perceived value by showing its operational and risk-reduction dimensions in the Hong Kong online retail context. For managers, the implication is that communicating price alone may be insufficient; transparent fulfilment and after-sales policies may be equally important in creating perceived value.

Avoid common mistakes

• Do not claim induction if all themes were fixed before coding. If you began with a theoretical framework, describe the work as primarily deductive or hybrid.
• Do not treat every quote as a finding. Quotes illustrate an interpretation; they do not replace one.
• Do not force all evidence to support your theory. Contradictory or absent evidence can refine the model and often makes the dissertation more credible.
• Do not repeat Chapter 4. Chapter 4 should present what was found; Chapter 5 should explain why it matters in relation to the research questions, theory, context, and practice.
• Be transparent in Chapter 3. State how you began with literature-informed codes, how open codes were retained, how themes were reviewed, and how you checked that interpretations were grounded in the dataset.

A useful one-sentence methodological description is:

“The study adopted a predominantly deductive thematic analysis guided by the research questions and conceptual framework, while retaining an inductive element to identify unanticipated themes emerging from participants’ accounts.”

This is normally defensible for an MBA project because it links your empirical material directly to established management knowledge while preserving room for context-specific practical insight.

A collection of blog notes on using chatgpt for research purpose.