Lauren and Heather’s Blog

The Adventures of Dr. Judy and Maisy

Final Week of Testing December 2, 2008

Filed under: Experiment — laurenheather @ 1:56 pm

We are still doing brightness discrimination just as we were doing before Thanksgiving break with one big change.  We are still using two rooms in the visual discrimination box with white paper as the positive stimulus and black paper as the negative stimulus.  We also went back to letting our rats self-correct because Maisy had trouble with the punishment.  The change is that Maisy is now being tested in dim lighting.  We are going to measure the light this week so exact measurements will be available soon.  She was having a lot of trouble with discriminating in the normal lighting, but once we switched to dim lighting, there was an immediate change.

You can see from the table below all of our data so far from the brightness discrimination task.  Session 8 is especially interesting and there are special notes for that day.  That was the day Maisy was switched to dim lighting.  It appears Dr. Hample did very well that day, however, she only completed three trials.  She would not leave the starting room and only did so three times.  We do not know why she behaved that way because she was fed a similar amount of food and run in the same conditions as earlier.

Session 9 is also problematic because our rats had a five day break from the experiment.  This break did seem to harm their performances so we will just have to see if they improve during the rest of the week.

  1 2 3 4 5 6 7 8 9 10 11 Average
Maisy 43% 46% 44% 41% 36% 24% 34% 67% 43% 43% 57% 43%
Dr. Hample 48% 45% 47% 52% 62% 59% 64% 67% 54% 55% 58% 56%

The percentages in italics represent days Maisy was tested in dim lighting.

This figure shows the data from our brightness discrimination task in graph form so it is easier to compare their performances over time.  Remember, for session 8, Maisy switched to dim lighting and Dr. Hample only completed three trials.  For session 9, we had just returned from break so the rats both had five days off from testing.

brightnessdisc_11

Advertisements
 

On to brightness discrimination… November 20, 2008

Filed under: Experiment — laurenheather @ 4:27 am

After getting feedback from the class last week, we decided it was time to modify our experiment again.  After three days of presenting the stripes and the blank wall, the rats were still not making a discrimination above chance.  There was concern that our reward hallways were a problem in our procedure; they left too much time between seeing the stimulus and receiving the reinforcement.  There was also concern that our stimuli were not salient enough.  Changes clearly had to be made to address these worries so we simplified our experiment.

Our procedure now is for brightness discrimination instead of visual acuity.  If the rats were having trouble discriminating between the three stimuli earlier, this may indicate poor vision.  We want to know just how poor their vision is and we will know it is very poor if they are not able to perform a brightness discrimination.  If they can discriminate brightness, this may indicate they just have trouble discriminating between more complex visual stimuli (but this still does not explain their selection of the plain black stimulus during those trials).

To test brightness discrimination, we are using a white room and a black room.  Each room has a slip on the back wall and one on the floor.  Under each slip is either a black or white sheet of construction paper, depending upon the rooms indicated by the random number generator.  The slips are in place to prevent the rats from taking anything besides visual cues from the stimuli.  The white room is the positive stimulus, resulting in reinforcement, and the black room is the negative stimulus, resulting in no reinforcement.

We also removed the reward hallways and now use a hole on the side of each room to deliver reinforcement, or food, upon choosing the positive stimulus.

Originally, we were letting the rats self-correct.  If they chose the negative stimulus first, they were left in the box until they correctly identified the positive stimulus and received reinforcement.  However, the trial was still marked as incorrect.  Correct trials are those in which the rat goes to the positive stimulus first.

Maisy’s peformance appeared to be worsening so instead of continuing with the same procedure, she started receiving time-outs during the seventh session.  Whenever she chose the negative stimulus, she was blocked in that room for thirty seconds and then went back to the starting hallway to begin the next trial.  This procedure will probably also have to be changed because during her time-out, she started biting the stimuli and the slips.  If she is blocked in there enough times, I think she will destroy the stimuli.  I think she should either go back to self-correcting or use the aquarium for time-outs.  I do worry about using time-outs because she is already having trouble discriminating and using time-outs seems to result in her receiving less reinforcement than she was in the earlier sessions.

We work with each rat for thirty minutes and see how many trials they are able to complete.  They begin the trials in the starting hallway.  We still flip a coin to determine which rat goes first and we use a random number generator to determine which room to put each stimulus in for each trial.

Below, you can see a figure depicting how the rats have been doing with the new brightness discrimination task.

brightnessdisc4

This table shows the data we have gathered from the past seven sessions and the average percentage of correct trials for each rat.

1 2 3 4 5 6 7 Average
Maisy 43% 46% 44% 41% 36% 24% 34% 38%
Dr. Hample 48% 45% 47% 52% 62% 59% 64% 54%

Here is a video showing our procedure before we introduced the time-outs for Maisy.

 

References November 18, 2008

Filed under: Experiment — laurenheather @ 9:58 pm

Bell, J. (1988). Use of visual and tactual cues in learning of simultaneous shape discriminations by albino and pigmented rats (Rattus norvegicus). International Journal of Comparative Psychology, 2 (2), 119-129.

Bennett, M. (1973). Visual deficit following long-term light exposure. Experimental Neurology, 38 (1), 80-89.

Burn, C. (2008). What is it like to be a rat? Rat sensory Perception and its implications for experimental design and rat welfare. Applied Animal Behaviour Science, 112, 1-32.

Campbell, B. & Messing, R. (1969). Aversion thresholds and aversion difference limens for white light in albino and hooded rats. Journal of Experimental Psychology, (82), 2, 353-359.

Chase, R. & Chorover, S. (1968). Persistence of visual pattern discrimination in binocularly-occluded albino rats. Journal of Comparative and Physiological Psychology, 65 (2), 238-245.

Dashell, J. (1958). The role of vision in spatial orientation by the white rat. 522-526.

Eaker, J. (1967). Behaviorally produced illumination change: visual exploration and reinforcement facilitation. Journal of Comparative and Physiological Psychology, 64 (1), 140-145.

Fields, P. (1953). Multiple discrimination learning by white rats.

Gibson, E., Tighe, T., & Walk, R. (1958). Enhancement and deprivation of visual stimulation during rearing as factors in visual discrimination learning. 74-81.

Jeffery, K. & Minini, L. (2006). Do rats use shape to solve discriminations? Learning and  Memory, 13, 287-297.

Lashley, K.S. (1930). The mechanism of vision: III The comparative visual acuity of pigmented and albino rats. Journal of Genetic Psychology, 481-484.

LaVail, M. (1976). Survival of photoreceptor cells in albino rats. Investigative Opthamology, (15), 1.

Manser, C., Broom, D., Overend, P., & Morris, P. (1997). Investiagtion into the preferences of laboratory rats for nesting materials and nest-boxes. as cited in the University of Cambridge, department of clinical veterinary medicine and the SmithKline Beecham Pharmaceuticals Harlow, Essex, UK (32) 23-35?? no publication…

Noell, W., Walker, V., Kang, B., & Berman, S. (1966). Retinal damage by light in rats. Investigative Opthamology, (5), 5.

Stauffacher, M., Peters, A., Jennings, M., Hubrecht, R., Holgate, B., Francis, R., Elliot, H, & Bauman, V, & Hansen, A. (2000). Future priniciples for housing care of laboratory rodents and rabbits. As presented at the  Council of Europe Convention ETS Appendix A?  (1-62).

Wasowicz, M., Morice, C., Ferrari, P., Callebert, J., & Versaux-Botteri, C. (2002). Long-term effects of light damage on the retina of albino and pigmented rats. Investigative Opthamology and Visual Science, (43), 3.

 

Going Back to Positive Stimulus Training

Filed under: Experiment — laurenheather @ 4:31 am

Last week, we had completed over 200 trials with each rat using our visual discrimination box.  We calculated the percentage of times each rat went to each stimulus and we found out the percentage was close to random chance (33%) for all of the stimuli.  This made us wonder if our rats really can not discriminate between any of the stimuli, which would make us believe their visual acuity is poor.  However, before we can draw that conclusion, we want to be sure that our results are not due to a flaw in our methodology.

To make us feel more confident and double check our results, we have started doing positive stimulus training again.  This time we are doing it a little differently.  We are using the two angled stimulus rooms and the center room is blocked.  For each trial, one room has the positive stimulus and the other does not have a stimulus card.  We determine which room to put the positive stimulus in by using a random number generator.  The trial begins when the starting room opens and ends when the rat goes into the reward hallway of the room with the positive stimulus.  If the rat goes into the reward hallway of the blank wall, nothing is done to the rat so she can correct her mistake by exploring until she identifies the positive stimulus.  After choosing the positive stimulus, the rat returns immediately to the starting room to begin the next trial.  We have been doing the positive stimulus training with our rats for three days.

Below is an image of what our visual discrimination box looks like during our positive stimulus training.

heather-005

You can view this video to see how we conduct trials for the positive stimulus training.

 

Phase Two: Including Punishment November 10, 2008

Filed under: Experiment — laurenheather @ 6:42 pm

After presenting our experiment on Thursday and finding no significant results after running 150 trials, we decided to try a mild form of punishment in the form of a timeout. After every incorrect trial, we are going to punish the rat by placing them in the aquarium for thirty seconds, rather than allowing them to simply begin the next trial. We chose thirty seconds because it is a longer period of time than it takes to switch the backgrounds and is therefore longer than the rat would have to wait in the starting hallway after a correct trial. We are also going to limit every trial to one minute. If the rat just walks around without going into a reward hallway, then we will end that trial after a minute and put the rat in the aquarium for the thirty second timeout.

In addition to the changes we have made to our procedure, we have also started looking at other articles to give us ideas of how to modify and improve our experiment. One study by Fields found that the albino rats had trouble discriminating between a circle and a square even though they were able to discriminate between other stimuli. The researcher believes this could have been due to the visual acuity of the rats. This may be similar to the trouble we are having in our experiment. Perhaps the triangles and stripes do not look similar to us, but they look much more similar to the rats. They also did not control for brightness in their study by using stimuli with similar areas of white, which could have contributed to why the rats could discriminate between some stimuli but not the circle and the square.

One other study we read was by Gibson, Walk, and Tighe. They used painted figures and cutout figures in two separate experiments and felt the cutout stimuli were more effective. We are also using cutouts (the stripes and triangles on the black backgrounds are cutouts from white construction paper), but maybe it is important for us to keep in mind that our results could be different if they are presented differently. These researchers also found that the rats may perform better on discrimination tasks if they are exposed to the stimuli more frequently (including in their cages) before and during discrimination training. This is because the rats have more time to adjust to the stimuli so they can perceive them. However, this facilitation is challenged in other studies, including the second experiment of their own.

 

This video shows an example of an incorrect trial that is punished and a correct trial that is reinforced.

The video below shows an example of a mistrial with Dr. Hample.

 

The first week of testing November 5, 2008

Filed under: Experiment — laurenheather @ 7:15 pm

We started testing this week and so far our procedures have worked fairly well and there are fewer problems with our homemade discrimination box. We have had some issues with Maisy ripping up the foam door that holds her into the starting hallway, but that has not been a huge problem.  The “kinks” in our box have been ironed out fairly well, so this part of the experiment is much better.

During the first three days however, the rats have not been choosing the positive stimulus anymore than chance would expect.  The highest percent correct was only 36 % (you can view this on our results page) , and the rats were also not consistently choosing the two brighter backgrounds over the one blank black background.  We trained them to go to the room with the positive stimulus for a whole week, so we are hoping to eventually start seeing some results.  We are planning to do at least 300 trials throughout this week and next week.

We have started to worry about the possibility that the rats’ training to go into the reward hallway to get food might soon become extinct.  For example, on one of the testing days, Maisy only picked the positive stimulus four times out of thirty, and so we are a little bit anxious that because they are not being reinforced when they go into the reward hallways, this behavior might extinguish. This would be a problem if we hope to get in at least thirty trials per day. Overall, however, the experiment is running smoothly, and hopefully after more testing, we might be able to see some discrimination in action!

You can see some footage of our testing by viewing the following video:

 

Training our rats and our amazing apparatus October 29, 2008

Filed under: Experiment — laurenheather @ 6:16 pm

We have been training Dr. Hample and Maisy this week to go through the reward hallways of the stimuli rooms and receive reinforcement.  We have randomly assigned which stimuli room to use during each training session so as to avoid having them associate their reinforcement with a particular room. We used a random number generator we found online to randomize everything. 

We have encountered some interesting behavioral issues that we didn’t think of before we built our box.  Firstly, it has been hard to keep the rats inside the box.  We have used wire over the top of the box to keep the rats inside.  Another problem was our doors.  The rats were able to sometimes push open our doors, and so Heather has devised a clever rubber band strategy to keep the doors secured in place.  We have also had to reglue our pictures of stimuli because the rats sometimes chew or pick at them. Another problem we are still encountering is that the rats will run very quickly into the reward hallway and grab the food and then run back into the entry way.  We have inferred that they may “be hesitant to be grabbed out of the reward hallway because it may be a bit scary for them.”  They seem to want to avoid staying too long in the reward hallway.  We have tried putting a door down when they come through the reward room, but then we have a hole in the mesh of the ceiling and  the rats try to escape through that.

Dr. Hample also does not always eat the food given to her in the reward hallway.  Her behavior seems to show that she is more interested in escaping because she waits by the exit door of the reward hallway, waiting for it to open rather than eating the food.  We have reduced her food intake considerably down to 10-15 grams in hopes that she will eat the food and be more on task.  Maisy however has not had a problem with this.  She was the first one to learn and did it fairly quickly.

We are excited however to begin testing very soon, and the rats have been doing much better after we ironed out the various kinks in the construction of our box!

Here you can watch a video describing some of our problems and the solutions we have used: