Thursday, October 10, 2013

Not Quite Love at First Sight

These past few weeks have been interesting, nonetheless, with EP and our talented rat, Fi-B. I mentioned in my first ever blog post that I was hesitant to hold Fi-B. Well that hasn't really changed. I can say that my apprehensions have improved and I am able to hold her for a few, short minutes. The pictures below depict one of our most recent training sessions. The picture on the left shows my rejection of Fi-B's attempt at a kiss. The picture on the right shows her crawling up my shoulder - I wasn't too happy about that one.

Overall, I enjoyed this experience. I felt accomplished to see Fi-B learn the behavior so quickly. I was practically fascinated to see her learn to press the bar almost immediately during FR 1 as I expected this behavior to take at least a week. Additionally, I was impressed with how smart she was. Though EP and I may be a bit biased to our rat, she really is a true star! She probably could have her own tv show on TLC - Fi-B: The Princess Rat.

Our rat seemed to be one of the more well-behaved rats in our class. As the reinforcement schedules increased, she increasingly became more concerned with what was going on outside of the operant box than with actually pressing the bar. Thus, EP and I resorted to placing the front door on the box to hide any distractions.

There were only two aspects of this project that I disliked. One was the food deprivation. I, myself, love food. I constantly snack throughout the day. Fi-B, however, does not have this luxury. She relies on EP and myself for her food. It was almost torturing to see her lunge at the placement of food in her cage. She must have been so hungry! The other aspect that I didn't particularly like was extinction. It was heartbreaking to see her so frustrated. At times she would tap on the front window. I imagined her saying, "get me out of here" or "please reinforce my behavior!" I felt terrible to see her like that. I am glad this part is over. I cannot think of any improvements that can be made to this experience. I can only say that I wish I had gotten closer to my rat.

In conclusion, I do believe I gained quite a bit from this experience. It was great to be able to apply what we learn in class and physically see the results. I look forward to (hopefully) completing another experiment with Fi-B in the future.

Fi-B versus Sniffy: Who's the Better Rat?

I found using the program, the Virtual Rat Sniffy, to be very helpful in training my rat, Fi-B. It was a solid foundation in learning how to handle the training session. Before using the software program, I only understood the basics of operant conditioning. However, the program gave me an interactive experience with a rat. I had a better understanding of our project and what we were expected to accomplish over these past few weeks. I would absolutely recommend using the program for future classes. I believe it helped prepare me for the training periods with Fi-B. Though I made a few mistakes in the Sniffy program, I was able to correct for those with my live rat. 

That being said, training a live rat is much different than training a virtual rat. A virtual rat is not realistic. The virtual rat was instructing me exactly how to train a rat, while the live rat is unpredictable. 

Another difference is the interaction with the rats. Since the rat is virtual, we are unable to show it the unconditional love that we showed our live rat. We are also unable to give the virtual rat the large amounts of verbal encouragement that Fi-B received. Further, we couldn't do other activities such as scroll Facebook or Twitter or complete crossword puzzles with our virtual rats. 


Pictured above is my partner, EP, and Fi-B completing a crossword puzzle from the New York Times. I knew our rat was brilliant - but I didn't think she was this brilliant!

Fi-B's Weight Chart

Target Weight: 207 Grams

It was a challenge for EP and I to keep Fi-B at her target weight, especially in the beginning. You can see between the first training day (9/11/13) and 9/20/13 that her weight fluctuated. Her heaviest weight, besides her starting weight, was 213.3 grams (9/23/13) and her lowest weight was 201 grams (10/4/13). EP and I did our best to adjust her food servings based on the serving chart provided in the rat lab. 


A Graph of Responses for Each Reinforcement Schedule


Shaping --- 47
FR 1 ------- 240
FR 2 ------- 269
FR 3 ------- 149
FR 5 ------- 452
FR 7 ------- 540
FR 10 ----- 587
FR 12 ----- 465
FR 15 ----- 200
FR 20 ----- 783
Ext 1 ------ 300
Ext 2 ------ 300

Final Day of Extinction

Today, we began the last day of extinction at 3:41 pm. To reiterate, extinction is the gradual decline of behavior after ceasing reinforcement. I noticed that it took Fi-B some time to make her way to the bar to continue the previous behavior. She did attempt the behavior and had allotted 57 presses in under 4 minutes. Then after 9 and a half minutes, she had 127 presses. However, none of this behavior was reinforced. Since she was not being reinforced, Fi-B continued to get frustrated in the operant box. She had more time to do activities such as, sniffing the particularly appealing left corner, jumping, standing on her hind legs (this was a new one), and attacking the front window. At one point, she consistently pushed the bar for 20 seconds. If she were human, I would attribute this to her thinking that the bar was broken, but that it would be fixed if she continued to push it. This extinction session went better than the last with fewer extinction bursts.

Below you will find a video taken the last day of extinction. You can see that Fi-B is doing other activities in the operant box. However, at 26 seconds, you see her get frustrated with the bar because she isn't receiving a food pellet for her presses.


(Above is the cumulative record for our final extinction day. You can see one extinction burst around 20 minutes.)


Extinction Day 1

Today was the first day for extinction. The goal was to extinguish Fi-B's learned behavior of pressing the bar for reinforcement. Instead, she would receive no reinforcement for any bar presses. I must say that I felt bad for Fi-B watching her get frustrated. She was frustrated because she was no longer receiving the food pellets. Our session began at 3:52 pm.

Fi-B attempted to perform the same behavior that she had learned. After approximately 10 minutes, she appeared to have give up, though she would try pushing the bar every now and then. Towards the end of the session she became very frustrated. She proceeded to attack the top and bottom of the bar - anything to get reinforced. She also tried sticking her head up the hopper to reach the food pellets. We observed her twisting and turning every which way to get inside. The food hopper was filled with the food pellets so I must wonder if she would continue with this behavior if the food hopper didn't have the pellets. I only assume she can smell them.

Fi-B did experience an extinction burst. Extinction bursts occur when there is a sudden increase in the behavior's frequency followed by a decline in the behavior.



(The cumulative record shows our rat's performance during her first day of extinction. You can see her first extinction burst within 5 minutes.)

FR 20

LIke the previous training sessions, the goal for this session was for Fi-B to learn to press the bar 20 times for a reinforcement. Fi-B was quick to complete the task and immediately went to the bar and began pressing. By 16 minutes and 23 seconds, she had accumulated 363 presses. She appeared to cooperate more today and rarely left the back right corner by the bar. By 25 minutes and 47 seconds, she had pushed the bar 674 presses. Finally at the end of our 30 minute training session, she had an astounding 785 pushes for 39 reinforcements! This training session included lots of verbal reinforcement from my partner, EP, the pageant mom. Nevertheless, I found this training session to be quite stressful as Fi-B would wander around at times when she realized it took a large amount of bar presses.

(This is the cumulative record for FR 20. My apologies for the poor quality.)