Collecting and Interpreting Lettuce Seed Bioassay Data
Before reading this page, you should first follow the steps on the "Conducting Reference Toxicity Tests" page.

Taking Measurements
At the end of the 5-day growth period, count and record how many seeds in each dish have germinated.

For each sprout, measure the radicle length to the nearest mm. (The radicle is the embryonic root).

Look carefully at the plants to make sure you are measuring just the radicle, not the shoot as well. For example, in the picture below, you would measure just the part between the two arrows, not the shoot and cotelydons to the left.

How Good are Your Data?
Once you have counted how many seeds germinated, and measured their radicle lengths, then what? How can you interpret these results?

Comparison to the Control
The first thing to check is your control (the dishes that contain deionized or distilled water rather than a sample). The purpose of the control is to identify how well the seeds will grow without any added contaminants. Would you expect all of the seeds in your control dishes to germinate? Probably not, just like a gardener does not expect all the seeds in a garden to sprout.

If fewer than 80% of the seeds in your control dishes sprouted, something may have gone wrong in your experiment. Perhaps the seeds were too old or stored improperly, so they were no longer viable. Or maybe something went wrong with the conditions for growth. Did the dishes get too hot, too dry, or contaminated in some way? Did you use tap water for your control, rather than deionized or distilled water? In many cases this works fine, but since tap water is highly variable from source to source, it gives less predictable results.

A Look at Variability
Within each treatment, how much variability did you find in your results? Did the replicate dishes show similar numbers of seeds sprouting, and similar average radicle lengths? If you think the data are much more variable than you would expect, you might want to explore the potential sources of variability for this type of experiment.