DS2 Case study – Leaf allometry
In the Data Science 2 module, students conduct independent case studies applying the full data analysis workflow — from their own data collection and preparation to statistical modelling and visualisation. They use R Notebooks for the first time, as a stepping stone towards fully reproducible R Markdown and Quarto documents used in advanced courses.
The task
In small groups, students collect leaf measurements (length and width in mm) from various flowering plants in Hamburg and surrounding areas. Along with the measurements, they record metadata such as site coordinates, time of measurement, and taxonomic identification of plant species (using apps like PlantNet or iNaturalist). Raw data is first recorded on paper protocols and then digitised in standardised spreadsheet templates.
The statistical analysis includes:
- Descriptive statistics: Calculating means and 95% confidence intervals for leaf length and width
- Two-sample comparisons: Comparing leaf length between groups (same or different species) with checks for normality and homogeneity of variance
- Multi-sample comparisons: Comparing leaf width across multiple groups
- Linear regression: Modelling the relationship between leaf length and width, including verification of model assumptions
- Power analysis: Assessing whether sample sizes were sufficient for the tests performed
- Evaluation: Critical discussion of the experimental design and analytical methods
Results overview
The following documents summarise the collected data from all groups per semester. They were created using the UHHformats Quarto template.