Use fresh harvested and graded seeds of maize, apply the ageing test following the recommendation of temperature, use different age times (as many as possible), calculate the coefficient of variation; use replication of 100 seeds for every treatment, no less than six replications, apply EDA - Exploratory Data Analysis; use a control. Best regards, Walter.
As Romain suggests, use top paper in plastic box with adequate quantity of water. Put the seeds (50 to 100 per replicate pending the size of the seeds and of the box) and read germination (counted on radicule protrusion) of the seeds. Calculate the average over the 2, 3 or 4 boxes of each sample...
If each group of students use a different lots of seeds, it will allow you to show them that germination is a lot specific specification.
One could include under "Seed germination efficiency" both germination percentage and germination rate to be assessed for instance through time to 50% germination. In addition to Romain and Benoit suggestions, MaYy I also take in to account the time for 50% germination (Average time to have half of sown seed germinated in replicates with a minimum of three replicates and 50 seeds per replicate ).