Academic writing

Your Action Research Findings Are a Diary, Not a Discovery

Action research findings are not a data recitation exercise. They are a change narrative: what shifted, how it shifted, and what the shift reveals. Description without interpretation is a field log, not a finding.

Your supervisor's feedback probably looked something like this: "The findings section reads as a sequence of observations. Where is the analytical layer?" In plain English: you told me what happened. You didn't tell me what it means. This is the single most common way education action research findings die. Not grammar. Not vocabulary. You mistook description for discovery. You spent three weeks collecting interview transcripts, observation notes, and student work samples, and the findings section you produced is a parade of "Student A said... Participant B reported... The data shows..." The reader finishes and thinks one thing: so what?

Here's what makes action research findings different from experimental results: you don't have a large sample, you don't have a control group, you don't have a p-value. You are a teacher who tried something in your own classroom and wanted to see if it worked. Your findings section's job is not to prove — it's to show change. You need an arc: what things looked like before → what you did → what things looked like after → what that shift tells us. Most international students get stuck here because they default to the experimental template: "The results demonstrated a statistically significant improvement in..." You have 25 students. Don't say significant to me.

The first trap is monotone syntax. Your entire findings section runs on "Participants reported that..." and "The data revealed that..." Count them. If either phrase appears more than five times, your findings read like a recording device, not an analyst. ❌ "Participant 3 reported that she felt more confident in class. Participant 5 stated that the group discussions helped her understand the material better. The data revealed that most students felt less anxious after the intervention." ✅ "Confidence emerged as the most frequent theme across interviews (7 of 12 students). Where pre-intervention responses were hedged — 'I guess I could try,' 'maybe I'll participate' — post-intervention language shifted to claims: 'I raised my hand three times today.' Participant 5's shift was particularly marked: her pre-interview noted silence during group work; by week six, she was facilitating her group." The second version has a theme, layered evidence, a timeline, and a specific person. That's what a finding looks like.

The second trap is fake precision. "The results showed a 73% improvement in participation rates." Where did that 73% come from? You observed for six weeks, jotted down rough counts each session, and divided totals? This pseudo-quantification reads worse than no number at all — it signals you don't know what you're doing. Action research findings can be vague, but the vagueness has to be principled: "Participation increased noticeably from week two onward, particularly during group tasks." That sentence carries more information than the 73% because it specifies the condition (group tasks), the time window (week two onward), and acknowledges that this is observation, not measurement. Honest small-sample language is worth a hundred times more than inflated precision.

The third trap is the deadliest: describing what happened without saying what it means. If your findings section doesn't contain sentences like "This suggests...", "This pattern indicates...", or "The shift from X to Y points to...", then you haven't written findings — you've written field notes. Action research findings require an interpretive layer, even if your sample is one person. ❌ "After implementing the peer-feedback strategy, students submitted higher-quality drafts. Most students expressed satisfaction with the new process." ✅ "The quality shift was most visible in argumentation: pre-intervention drafts typically contained one claim with two examples; by week four, the median was two claims with three examples each. This pattern aligns with what Student 7 described in her interview — 'I started thinking about what my peer would ask me to justify.' The peer-voice functioned less as evaluation and more as a prompt for deeper reasoning." That last clause — translating the phenomenon into a mechanism — is exactly the analysis your supervisor is asking for. One sentence. That's what separates a finding from an observation.

So here's the structure that actually works for action research findings: state the theme → provide evidence (quotes, data, observations) → interpret what the evidence points to → state what this means for your practice. Four moves. None of them are optional. You can mix data types freely — interview quotes, observation notes, work samples can all appear in the same paragraph — but every paragraph must land on interpretation. If you can't summarize a paragraph in one sentence that starts with "So...", that paragraph is filler. Delete it or rewrite it. The most counterintuitive thing about action research findings: the more it sounds like a generic academic paper, the less it sounds like a finding. You don't need to hedge to death. You don't need "It could be argued that..." in every sentence. You don't need to pretend you ran a randomized trial. Your advantage is that you were in that room, watching those students change. Write that presence in. It's more persuasive than any academic cliché.

That said, once the substance is right, the language layer still needs work. Repeated "Participant X stated that she felt" constructions, tense drift between past and present, the Firstly / Secondly / In addition triple — none of these are fatal, but they're fatal to how your writing gets received. I don't fix these by hand. I run the section through easydue and let it pull my findings from "readable but obviously non-native" to "my supervisor won't stop me on language." The time I save goes into sharpening the interpretive layer — which is where your findings section actually earns its place in the paper.