---
title: "Questions, hypotheses & predictions"
subtitle: "Analysis of Ecological and Environmental Data<br>QERM 514"
author: "Sarah Converse"
date: "8 May 2026"
output:
  ioslides_presentation:
    wide: true
    css: ../lecture_slides.css
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE)
```


# FOR DISCUSSION


## Models vs Hypotheses

When do candidate models represent distinct scientific hypotheses?


## Addressing uncertainty

How should uncertainty be interpreted when multiple hypotheses remain plausible?


## (Un)conscious bias

What is the statistical cost of favoring one hypothesis too early?


## Competing explanations

Which design features most strongly determine whether competing explanations can actually be distinguished?


## A scenario for discussion

_Suppose two mechanisms could explain observed variation in species abundance: habitat quality and dispersal ability/limitation. You have observational data from 50 sites sampled over 5 years._

What would a design look like that could discriminate between these mechanisms?

How would you translate competing hypotheses into candidate models?

Would model comparison alone be persuasive? Why or why not?


## Some reflection

What might you do differently in your own research after reading these papers?