Semantic vs. Latent Thematic Analysis: Which Tells a Better Story?
Qualitative research is never just about collecting words—it’s about uncovering meaning. But meaning isn’t a one-size-fits-all concept. Sometimes, we want to stay close to what participants literally say. Other times, we want to interpret the deeper assumptions beneath their words. This is where two approaches diverge: semantic thematic analysis and latent thematic analysis.
Both are powerful qualitative coding strategies, but they tell very different stories about the same data. So how do you know which one is right for your project? Let’s explore the trade-offs, strengths, and limitations of each.
What Is Semantic Thematic Analysis?
Semantic thematic analysis works on the surface level of the data. Researchers code what participants explicitly say… their words, phrases, and straightforward meanings.
Think of it as taking participants at face value. If a student says, “I find online learning isolating,” a semantic code might be “isolation in online learning.”
When to Use It
Policy-oriented research: Clear, traceable insights are easier to apply in practice.
Customer or student feedback: You want to capture repeated concerns directly.
Time-sensitive projects: Semantic coding is often faster and easier to defend.
Strengths
Transparent and reproducible. Any reader can trace findings back to the raw data.
Minimises researcher interpretation, reducing the risk of overreach.
Limitations
May miss subtle cues, unspoken assumptions, or broader cultural contexts.
Can feel “flat” if your research question demands depth.
What Is Latent Thematic Analysis?
Latent thematic analysis, by contrast, looks beneath the surface. Researchers ask What does this statement reveal about underlying beliefs, values, or social structures?
For example, if a student says, “I find online learning isolating,” a latent code might interpret this as “digital environments undermine community.”
When to Use It
Cultural or social research: Exploring ideology, power, or hidden dynamics.
Identity and experience studies: Where personal meaning goes beyond words.
Complex or exploratory topics: When surface descriptions aren’t enough.
Strengths
Reveals the bigger picture. Social norms, hidden biases, and underlying frameworks.
Generates richer theoretical contributions.
Limitations
More subjective. Findings rely heavily on researcher interpretation.
Requires deep reflexivity to avoid over-interpreting the data.
Can be harder to justify to sceptical reviewers seeking traceability.
Case Study Comparison: Student Experiences of Online Learning
To see the difference, imagine two research teams analysing the same dataset: interviews with university students about online learning.
Team A (Semantic Analysis): Codes focus on what students say- “lack of interaction,” “technical problems,” “flexibility.” Their findings highlight the practical challenges and benefits of online learning.
Team B (Latent Analysis): Codes dig deeper- “erosion of academic community,” “digital inequality,” “shifting norms of self-discipline.” Their findings reveal the structural and cultural implications of digital education.
Both are valid, but they tell different stories. Semantic analysis produces a set of actionable insights for improving platforms and policies. Latent analysis builds a theoretical critique of how technology reshapes education.
Which Should You Choose?
The choice between semantic and latent thematic analysis depends on your research goals.
Ask yourself:
Do I need findings that are clear, transparent, and easy to act on? → Go semantic.
Do I need findings that are interpretative, critical, and theory-building? → Go latent.
Remember, it’s not a binary. Some projects benefit from starting semantically and moving into latent interpretation once patterns are established.
So, which tells a better story: semantic or latent thematic analysis?
The truth is, better depends on your purpose. Semantic analysis tells a story rooted in participants’ words, ideal for clarity and action. Latent analysis tells a story of hidden structures and meanings, ideal for depth and critique.
The real skill lies not in choosing one over the other, but in knowing which lens your project needs. After all, qualitative research isn’t about forcing data into one mould—it’s about finding the story that matters most.
