Which of the Following Are Not Research Data? A Clear Guide
You've just finished a massive research project. That said, you've got notebooks full of observations, survey responses, lab measurements, interview transcripts, and — oh yeah — that 47-page literature review you wrote to frame your study. Here's a quick question: does that literature review count as research data?
If you're unsure, you're not alone. Which means this is one of the most common points of confusion in research methods, and it matters more than you might think. Funding agencies, academic journals, and institutional review boards all have specific expectations about what counts as research data — and getting it wrong can cause real problems down the line Simple, but easy to overlook. Nothing fancy..
Not obvious, but once you see it — you'll see it everywhere.
So let's clear this up Small thing, real impact..
What Is Research Data, Exactly?
Research data refers to the raw materials you collect or generate during a research project that serve as the evidence for your findings. It's the stuff you analyze to draw conclusions. Think of it as the building blocks — the concrete observations, measurements, and responses that exist independently of your interpretation Easy to understand, harder to ignore..
This is where a lot of people lose the thread Simple, but easy to overlook..
Research data can take many forms:
- Survey responses or interview transcripts
- Experimental measurements and sensor readings
- Statistical datasets or spreadsheets
- Audio or video recordings
- Genomic sequences or chemical formulas
- Field observations and notes
- Images or photographs collected for analysis
The key characteristic? Consider this: it exists before you start interpreting it. Research data is something you collected or generated as part of your systematic inquiry. It's the evidence, not the argument The details matter here..
The Gray Areas Worth Understanding
Here's where things get interesting. Some items sit in a fuzzy zone:
- Code or scripts you wrote to analyze data — these are often considered research data in computational fields, but less so in traditional qualitative research
- Metadata — information about your data (when it was collected, from whom, under what conditions) is usually considered separate from the data itself, though it's critically important
- Preliminary analyses — if you ran some tests just to see what patterns might exist, those outputs typically aren't the core research data
But there's a cleaner way to think about this: if you could hand someone the raw evidence and they could draw their own conclusions from it, that's data. If they'd need your interpretation to make sense of it, you're probably looking at something else.
Why This Distinction Actually Matters
You might be wondering why any of this matters. Here's why: research data management has become a huge deal in academia.
Funding agencies like the NSF and NIH now require data management plans. Journals increasingly ask authors to share their data or explain why they can't. Universities have policies about data retention and ownership. And if you're a graduate student, your committee is going to have opinions about what belongs in your appendices versus your analysis chapters Not complicated — just consistent..
Getting this wrong means:
- Your data management plan might not satisfy grant requirements
- You could waste time "preserving" things that don't need preserving
- You might lose access to materials you actually needed to keep
- Peer reviewers might question your methodology
Real talk: I've seen researchers lose points on dissertations because they couldn't clearly articulate what their data was. Don't let that be you.
What Is NOT Research Data
Now let's get to the heart of your question. Which of the following are generally not considered research data:
Literature Reviews and Background Materials
That comprehensive review of existing scholarship you wrote? And that's not research data — it's part of your analysis and argument. But you're synthesizing others' work, not generating new evidence. The articles, books, and papers you read are sources, not data (unless, of course, you're doing a meta-analysis and systematically coding their findings).
Published Articles and Scholarly Publications
You might have collected a mountain of journal articles to inform your study, but these are reference materials. The exception? They're what you used to design your research, not what you generated through it. If you're analyzing those publications as data — say, studying trends in a field over time — then they become your dataset But it adds up..
Your Interpretations and Conclusions
This should be obvious, but it's worth stating clearly. The meaning you derive from your data, the arguments you build, the conclusions you draw — these are not data. They're what you do with the data. Your analysis, your discussion section, your theoretical framework: all products of working with data, not the data itself.
Administrative and Procedural Documents
IRB approval letters, consent forms (the templates, not the completed forms with participant info), grant applications, meeting notes with your advisor — these are part of the research process, but they're not research data. They're administrative overhead Worth keeping that in mind..
Analysis Code and Software (Usually)
This one depends on context. In others, they're treated as tools. In some fields, the scripts you write to process data are considered part of the data package. If you're in computer science or computational biology, your code might absolutely be part of your research data. If you're in sociology conducting interviews, your Stata do-files are more like analytical tools Nothing fancy..
Short version: it depends. Long version — keep reading.
Drafts and Working Papers
Early versions of your analysis, rough drafts of your findings section, notes to yourself about what you think is going on — these are working documents, not data. They document your thought process, but they're not the evidence itself That's the part that actually makes a difference..
Common Mistakes People Make
Here's what I see most often:
Treating everything as data. Some researchers get so paranoid about data loss that they try to archive their entire project folder — including every draft, every random note, every email thread. That's not data management; that's hoarding. Focus on the actual evidence.
Confusing sources with data. If you downloaded a dataset from someone else and then added your own observations to it, you have two different things: their data (which might be your source) and your data (what you contributed). Be clear about the distinction.
Forgetting about data derivatives. If you cleaned a messy dataset and created a "clean" version, you've created a new file — but is it data? Usually, yes. Processed data is still data. Just make sure you can document what processing you did Simple as that..
Ignoring field-specific norms. What counts as data in anthropology differs from what counts in physics. Talk to people in your discipline. Check what journals in your field expect. Norms vary more than you'd think Most people skip this — try not to..
Practical Tips for Getting This Right
Start by asking a simple question about each component of your project: Could someone else take this and independently answer my research question with it?
If yes, it's likely data. If no — if they'd need my framework, my interpretation, my synthesis — it's probably not.
Here are a few other things that help:
- Create a data dictionary early in your project. Write down what each dataset contains, how it was collected, and what format it's in. This forces you to be explicit about what your data actually is.
- Separate your files as you go. Keep raw data separate from processed data, and keep both separate from your analysis scripts and your writing. You'll thank yourself later.
- Check requirements before you need to comply with them. Look up your funding agency's data management requirements. Check your target journal's data sharing policy. Know what's expected.
- Ask your advisor if you're unsure. This is not a question you want to guess wrong on.
FAQ
Are interview transcripts research data? Yes. The transcripts are the record of what participants said — that's your evidence. The audio recordings themselves are also data Most people skip this — try not to..
What about notes I took during observations? Those are data, assuming they record what you observed. If they're just your interpretations of what you saw, they might be closer to analysis notes. Be precise about the difference That alone is useful..
Is a survey questionnaire research data? The blank questionnaire template? That's more like an instrument or tool — similar to analysis software. But the completed surveys with actual responses? Those are definitely data That's the whole idea..
Does my thesis or dissertation count as research data? No. Your thesis is your written argument. It's the product of your research, not the data itself.
What if I'm using someone else's data? Then their data is your data, in a sense. You're using existing evidence to answer your question. Just be sure to document where it came from and how you accessed it No workaround needed..
The Bottom Line
Research data is the evidence — the observations, measurements, responses, and recordings you collect to answer your question. Everything else: your literature review, your analysis, your conclusions, your drafts — that's the work you do with the data, not the data itself.
Most guides skip this. Don't.
Get clear on this distinction early, and you'll save yourself a lot of confusion. Your committee will appreciate it. Your future self, digging through files three years later, will definitely appreciate it.
Now go make sure your data management plan actually covers the right things.
