By Viljar Storvik
BIO299 – Research Practice in Biology

When I started my BIO299 project, I had a clear goal: examine the pollen of Hedlundia lancifolia, a known hybrid species, and determine whether its pollen is morphologically distinct enough to be recognized in sediment cores. The logic was straightforward: by studying pollen from living individuals, we could learn what it looks like – so that future paleoecological studies might be able to identify it in ancient layers of mud and time.

I imagined microscope work, coding, statistics, and pattern recognition. Instead, I spent hours – then days – staring into a microscope and finding almost nothing. And I thought: I must be doing this wrong.

Introducing the Confusian Matrix

To make sense of this anxiety, I started thinking in terms of a concept from data science: the confusion matrix. It’s a tool for understanding prediction errors – how often our expectations match reality, and how often they don’t.

In science, we label errors like this:

• True Positive: You expect something, and it’s there.
• True Negative: You don’t expect something, and it’s not there.
• False Positive (Type I Error): You think something is there, but it’s not.
• False Negative (Type II Error): You think something isn’t there – but it actually is.

I didn’t know it at the time, but I was living inside this matrix. The confusian matrix.

My Prediction: Pollen Is There

Based on everything I knew, pollen should have been in my samples. That assumption was baked into the project itself. These were relatively fresh anthers, from flowering individuals. My entire methodology – focused on analysing pollen grain morphology – depended on it.

But the microscope didn’t agree. Over and over, I found… nothing.

The Fear of the False Negative

That’s when the questions started creeping in.

“What if the pollen is there and I just didn’t find it?”
“What if I prepared the slides wrong?”
“What if this whole thing is a mistake?”

This is the panic of the false negative – what we often call a Type II Error – when you conclude something isn’t there, but in reality, it is. It’s the most unsettling category in the matrix because it turns your attention inward: Did I fail as a researcher?

Was this a true negative – pollen absent, because it simply wasn’t produced?
Or a false negative – pollen present, but missed due to a human error?

I don’t know. I may never know. And strangely, I’ve come to accept that.

Shifting the Research Question

Eventually, I realized that the absence of pollen might be the real story. If this hybrid produces little to no viable pollen, that tells us something significant about its biology. Could Hedlundia lancifolia be partially sterile? Could its reproduction be more complicated than we assumed? This wasn’t the result I expected – but it was still a result.

What I Learned

I want to be honest: this semester was difficult. I overcommitted with too many courses, underestimated the time BIO299 required, and faced unexpected limitations both in my lab work and personal schedule. I had ambitions of doing statistical analyses and writing code. Instead, I made more and more slides – which one after another taunted me with their blank faces. One positive in this proses was that I became overly attentive on the slide preparation – carefully making sure I did tea lab work correctly. The result stayed the same, and I slowly became more and more assured of  my result – there was no pollen to be found.

But this experience taught me something that no clean dataset ever could. Science isn’t always about answers. It’s about learning to sit with the unknown, and to keep asking questions anyway. Sometimes, the most honest scientific result is: I don’t know. Sometimes, the bravest thing you can do is write: Here’s what I looked for – I didn’t find it – and here’s what I think this might mean.

Final Thoughts

Maybe I saw a true negative.
Maybe I missed a False Negative (Type II Error).
Maybe it doesn’t matter – because what matters more is the process, the integrity, and the curiosity I brought to the project.  I walked into this course wanting clear results. I leave it understanding that research is often messy, inconclusive, and uncomfortable.