Simscript

We decided to use Simscript which allows us to create entities (students) with different attributes, which is the tool we realized we needed after our attempt using anyLogic. This function is necessary to represent the different types of students moving through the system. The base attributes are defined as school, year, and gender. Each student is also assigned the following attributes: willingness to utilize mental health resources, knowledge of Let’s Talk, self-stigma restriction, likelihood of crisis, likelihood of A3N, and accessibility to each of various locations. The attributes are determined a randomly generated variable that is distributed (0, 1).

As described in the previous section, we defined six barriers an entity would have to overcome in order to attend Let’s Talk. The barriers that are independent of school, year and gender required only one statement to determine if they passed the barrier: if the randomly generated variable is less than the barrier criteria, then the entity will move forward in the system; otherwise, it will exit the system. The other dependent barriers were more difficult to set up. They required cases based on the student’s school, year, and gender, such that depending on which of the categories the student fit into, different probabilities would apply. For example, when determining whether or not a student is in crisis, each characteristic needs to be checked and classified in order to determine their crisis status.

In order to simulate each location Let’s Talk occurs, separate Simscript models were written for each location. Each simulation is set to run for 1.5 hours per day for 13 days, representing 1.5 hour sessions that occur once a week for one 13-week semester. While it initially would make more sense and seem more straight forward to simply run a simulation 1.5 hour simulation 13 separate times, the output of each run would be identical. This is a result of Simscript’s random number generator which uses a multiplicative congruence algorithm and can only generate pseudo-random variables. Each time the simulation is run, the random number generator starts with the same default seed and therefore gives you the same stream of random numbers. This is why it is necessary to simulate the 13 Let’s Talk sessions continuously instead of simulating 13 separate sessions.