For a more general discussion of genetic drift, check out the "What is genetic drift?" post.
In terms of simulations, if you want to explore the impact of population size on allele frequencies, I recommend using the population genetics simulation. Most students are good at memorizing that drift is more likely to occur in small populations, but this is a good way to demonstrate that phenomenon and discuss the statistics involved.
The bottleneck event and founder effect simulations both deal with specific examples of genetic drift. These simulations are less involved than the population genetics simulation, but they both still have randomness included. Students get the opportunity to analyze data and practice with probability. These labs are essentially technology versions of classic demonstrations using colored beads or different types of beans (there are many versions online accessible through a search for "genetic drift bead lab").
In both simulations students click on the screen to progress through the scenarios. The "survivors" of the bottleneck event and the "founders" of the founder effect are selected randomly from the population. In the bottleneck event simulation the starting population has an even split of the red and blue alleles. In the founder effect simulation, the blue allele represents a rare allele at a frequency of 0.1 in the starting population.
The founder effect simulation randomly selects the five "founders" that are on the log each time the simulation is run. The chance of each color being selected is relative to the frequency. Since the blue allele frequency is 0.1, the frequency of blue individuals is 0.01. So, for each founder selected there is only a 1% chance that it is blue. The worksheet I use with my 9th graders includes a number of basic statistics questions, including finding combined probabilities.
The bottleneck event simulation starts with a diverse population. Only four individuals survive a drought. The genetic makeup of these survivors determines the genetic makeup of the final population. The four survivors are randomly selected from the population, which is 25% red, 50% purple, and 25% blue.
Both of these simulations are relatively straightforward to run. The graphs allow students to do quick comparisons between original and new populations. They could serve as a demonstration as a part of discussion or a data analysis lab for students. Despite being two of the simpler simulations available on Biology Simulations, they are a useful tool in an evolution unit.