By the term data-based decisions, I mean to implement independent variables based on measurable dimensions of an organism's behavior. These skills are important because they aren't limited to what you can teach a rat (or for that matter, a virtual rat) to do in an experimental chamber. Instead, they can be applied to almost every level in a person's life. Whether you're trying to break yourself of a bad habit, teach your children to behave appropriately, or select between different methods of law enforcement to vote on, these skills will be useful.
Time for me to get off my soap-box and describe this assignment. For this final project, you are going to teach Sniffy to respond at 75-90 responses per minute while only delivering a reinforcer about once or twice every five minutes. In other words, I'm asking you to show me how fast you can make Sniffy press the lever while delivering as little reinforcement as possible. By the end, your cumulative record should look something like this:
It take's 75 responses on the lever to cause the cumulative record to reset. The amount of time between a dashed vertical line and the solid vertical line is 5 minutes. So, lets use that to try and figure out Sniffy's rate of lever pressing for the center part of the graph.
By counting the number of resets in the center of the image, 5, we can figure out about how many responses Sniffy emitted in 5 minutes and then calculate responses per minute. You're correct if you counted 6 resets. I omitted the first because the beginning of it didn't fit into our 5 minute window. Seventy five responses times five resets divided by five minutes ( 75 * 5 / 5 ) is easy to calculate because the five's cancel out and you're left with 75 responses per minute. That's not exactly how fast Sniffy is pressing the lever because I left out the beginning of that first reset. If we calculate rate given 6 resets (75 * 6 / 5 ) we get 90 responses per minute. So, we can say that Sniffy is pressing the lever between 75 and 90 responses per minute and during that period of time, he only received two reinforcers!
So, your job is to pretty much reproduce that cumulative record. To get to that point you'll need to start at a continuous rate of reinforcement (i.e., FR-1 or CRF) and then fade in leaner and leaner schedules of reinforcement (just like the FR and FI assignments) By the end, I used a variable-ratio 300 (VR-300) schedule. Feel free to use anything you want but you might be more successful if you use VR schedules (that's what I used).
As always, feel free to contact me if you would like additional assistance.
- Sniffy should be pressing the lever at about 75 to 90 responses per minute.
- In at least one 10 minute period with the above rate of responding, no more than 5 reinforcers can be delivered. You should indicate this area of the cumulative record by drawing on it or describing it (e.g., the last 10 minutes of the session).