Animals do havethe ability to make decisions due to the use of neurotransmitters like dopamineand the development of prefrontal cortex. This paper willdiscuss the different experiments that were conducted to determine whether ornot animals are capable of making simple decisions or more complex decisions. Iwill go through articles that talk about the choices that animals make whenfaced with different controlled and not controlled conditions. I will concludemy paper by restating the thesis and whether it was supported by the evidencein the articles I chose as well as state the limitations that were found in thearticles. I will also talk about some future research that is needed in thefield of animal research and decision making. Animals have beencompared to humans for many years.
Some theories say that we were created asanimals first and then evolved into what we are today. With our brains beingvery similar to those of animals such as chimpanzees and bottlenose dolphins,it’s hard to not support the evidence that has been presented about animals andhow they are able to do things like make logical decisions just like us. Bardgett andcolleagues (2009) conducted a study to determine the individual contribution ofdifferent dopamine receptors to effort-based decision making in rats. Just likehumans, different chemicals in the brain can influence decision making. In thiscase, they looked at rats (Rattus norvegicus). There were 9 adult rats whowere grouped in 3 different cages with 3 rats per cage. They all had access tofood and water and had lighting that went along with day and night lighting.
Ratswere trained in a T-maze to choose a large-reward arm that contained 8 pelletsof food over a small-reward arm that contained 2 pellets of food. The rats thenwere trained to climb progressively higher barriers to obtain the food from thelarge-reward arm. They used a discounting procedure on each test day, which wasup to nine trials. The rats had 3 days of testing with the discounted procedurebefore any test. The rats could eat the rewards from both the large and smallrewarded arms.
After this test, it went back to the regular testing which wasthe rats had to pick one side of the arm and they got either a large or smallreward. It was found that rats were more likely to choose the small reward armafter treatment with the D1 antagonist. However, on the first trial day oftraining there was a barrier on the large reward arm and the number of ratsthat went for that arm decreased. On the second and third day, without thebarrier, the number or rats increased for the large reward arm. Dopamine has alot to do with the everyday choices that we as humans make. In animals it’s thesame, they were presented with large amounts of food rewards but the barrierwas in the way so they didn’t want to have to work for it so they just went forthe small reward.
The next article will also look at dopamine and noveltybehaviors. Costa andcolleagues looked at whether dopamine helps with the choice to seek noveltythings versus the familiar options in monkeys (Macaca mulatta). They administeredsystemic injections of saline which is a selective dopamine transporter (DAT)inhibitor, to the monkeys and looked at their novelty seeking behavior during adecision-making task. The task that the monkeys had to do involved pseudorandomintroductions of novel choice options. This allowed monkeys the opportunity toexplore novel options or to explore familiar options.
After testing they foundthat dopamine increased the monkeys’ preference for novel options. Areinforcement learning (RL) model fit to the monkeys’ choice. The data showed increasednovelty seeking after DAT was driven by an increase in the initial value the monkeysassigned to novel options (Costa et al.
2014). The test consisted of 3 malerhesus monkeys ages 5-6. On test days, they monkeys were injected with eitherdopamine or saline and they had to wait for at least a half an hour beforegetting tested.
The monkeys did 6 blocks of tests. Each test consisted of 650trials where they had 3 images on a screen. When introducing a novel item, itwas replaced with one of the three images shown. A total of 32 novel items wereshown in the test. They found that the monkeys would choose the novel iteminstead of the familiar items almost every time. With dopamine having an effectof seeking novel items, we can put this on the list of things that help animalsmake decisions. The next article I read looks at responses in both humans andrats.
Another article I looked at was calledOptimal Response Rates in Humans and Rats. The article had 60 rats and 15humans. The rats had to press a lever to obtain food and the humans had topress the spacebar on a computer to get points that would later be turned intomoney. Both were only rewarded if the time between two consecutive responseswere greater than the target interval. Humans had a white square that was shownon a black screen and they had to press the spacebar when they thought thetarget interval passes. Once they pressed the bar the square would disappear. Awhite vertical line indicated the target interval and a red line was thereproduction interval in that single trial. After the reproduction block theywere tested in 8, 5 minute blocks of differential reinforcement of low rates(DRL) testing.
A white block was shown and they could respond at any time. Ifthe inter-response time (IRL) was greater or equal to the target, then a greensquare was shown. If it was earlier than the target interval, a red square wasshown and a buzzer went off. 48 out of the 60 rats were tested daily for 1 hourand the other 12 were tested at night in 3, 1 hour sessions per night.
The ratswere placed into an operant chamber. They were rewarded for spacing leverpresses by at least the target interval. If they pushed it before, they didn’treceive food.
The results were found that both humans and rats came close tooptimizing reinforcement rate, but they respond faster than they were initiallythought to. His article compared both humans and rats. I think that when wehave a reward and are reinforced when doing something then we will continue todo that. This is the same for animals too. There is a lot of evidence andexperiments that show animals doing a desired task in order to receive somesort of reward whether it be food or a toy or just praise.
The next articlelooks at the choices of pigeons based on their prior investments. The article, the effect of aprior investment on choice: The sunk cost effect, looked at this effect inpigeons (Columba livia). The sunk cost effect is the tendency to continue an endeavoronce a prior investment has been made despite a better option being available.There were 5 identical chambers that the pigeons were tested in. In thechambers was a row of 3 keys that were 2.1 cm wide. They were placed 21 cmabove the ground and 6 cm apart from each other.
The pigeons would peck at thekey that was lite up either red, green, or white. If the peck was hard enough,they were rewarded with wheat. Native birds were trained before to peck usingauto-shaping. After this training, they were trained 7 days a week. The sessionslasted for 60 reinforces, 30 red, and 30 green or until 1 hour passed. Therewere 4 different experiments. The first was to demonstrate the sunk cost effectin pigeons. 5 pigeons were in this experiment.
This experiment had matching ina standard concurrent-chains procedure with schedules as outcomes, but with noprior investment before the choice phase. In Experiment 1b, they introduced a20-peck prior investment before the sunk cost. The red components were the leftkey and green was the right. Both number of responses in the choice phasematched the log inverse effort ratio in the outcome phase.
In Experiment 1a,which the conditions in red and green components were identical, there were nodifferences in preference between left and right keys between red and greencomponents. In Experiment 1b there was a tendency to choose the left key in redcomponents and a tendency to choose the right key in green components. Thisresult follows the sunk cost effect, because the prior investment was on theleft in red components and on the right in green components. Experiment 2 wasto examine whether greater investments will lead to a greater sunk cost effectin the present concurrent-chains procedure with pigeons. This experiment had 3pigeons.
The procedure was the same as for Experiment 1b, except they manipulatedthe size of the prior investment differentially within session between red andgreen components. Each pigeon completed six conditions. The pigeons matchedtheir responses in the choice phase to the log inverse effort ratio in theoutcome phase. Experiment 3 had 4 pigeons that completed 6 conditions.
Theconditions were the same as in experiment 1. This experiment confirmed justlike Experiments 1b and 2, that pigeons commit the sunk cost effect. Experiment4 had 8 pigeons. The procedure was the same as in experiment 1 except that thecolor of the keys were white instead of red and green.
There was also a 1second delay in between prior investment and the choice phase. They did atleast 15 sessions. This experiment confirmed the results of Experiment 1b, 2,and 3 by showing an effect of prior investment on current choice or the sunkcost effect. I think that with these experiments, it shows us that animals aremore likely to do something that they already know rather than doing anothertask with the risk of not knowing. Even if the other task it a better option,sticking with the original investment is safer. The next article will look atpigeons again but will focus on sunk costs.
Sunk cost: Pigeons (Columba Livia), too, show bias tocomplete a task rather than shift to another, pigeons were tested to see ifthey would complete a task or switch to another one. There were 4 experimentsin the study. Experiment 1 had 8 pigeons and they were kept in cages. All thepigeons had previous experiences in unrelated studies involving simplesimultaneous discriminations and matching to sample discriminations (Pattinson& Zentall, 2012). The procedure was held in an operant chamber. The pigeonsreceived pre-training in which they had to peck 30 times for reinforcement on acolored key. In each 90-session trial each color, either green or red, waspresented 15 times at each of the three response key locations. For the regulartraining, after 5, 10, 15, 20, or 25 pecks to the center key, the key wasturned off and one of the side keys was presented with the 30-peck color, thisone had a reinforcement.
On the test trials, the pigeons were first presentedwith the high response color on the center key. After the pecking up to 25times, it was turned off and this time keys on both sides were lite up. Theexperiment fond that the pigeons tended to complete the 30-peck requirement,even when that choice was not optimal, and it required investing more work andtime than would have been required had they chosen to switch to the fixed,15-peck alternative. Also, this suggests that pigeons, like humans, show a biasto stay with an initial investment. Inexperiment 2 the subjects and apparatus were the same as the first experiment.
The procedure in the second experiment had 30 trials each of two trial typeswhich were forced and choice. In the choice trial the pigeons were presentedwith the 30-peck color on one of the two side keys and the 15-peck color on theother. There were no initial investments on the center key like experiment 1.Reinforcement occurred on the same number of pecks like the first experiment.
After a single peck to either the required key or the 15-peck key they were nolonger lite up. On the forced trial, only one side key was lite up. This wenton for 7 sessions. The results were found that when no initial investment wasrequired, the pigeons initially demonstrated a preference for the 15-peckalternative. In experiment 3 they had 4 pigeons and they were tested in achamber with 3 circular keys separated by a grain feeder. The left and rightkeys were red and green and the center one was white only. Each pigeon wastrained to peck the left (green) and right (red) response key.
The number ofpecks required for reinforcement was gradually increased to 30 pecks. Therewere six kinds of training trails in this experiment, 30-peck trials, 10-pecktrials, choice trials with no investment, choice trails with a fixed 10-peckinvestment, choice trials with a fixed 15-peck investment, and a choice trailwith a fixed 20-peck investment. The results for this experiment were that withno initial investment, all the pigeons preferred the fixed 10-peck alternative.When there was an investment they preferred the 30-peck requirement (Pattinson& Zentall, 2012).
For the last experiment, the number of pigeons and theprocedure were the same as in experiment 3. The results were also the same inthis experiment as they were in 3. This article also states that pigeons wouldrather stick to the easier investment rather than shifting to anther task. Thisdecision tells us that some animals do better when staying a task instead ofswitching it up.
Cache decision making is the topic of my next article. Preston and Jacobs (2015), examined the effect of adominant competitor species on the caching and behavior of Merriam’s kangaroorat (Preston & Jacobs, 2015). In the first experiment, there were 8 malerats.
The test was conducted in an arena with 4 white walls. A food dish withseeds was placed at the space between the two sides of the arena. One side ofthe arena was bare and the one side had decorations. There were three cachingtrials. Premanipulation, manipulation, and postmanipulation. For the first one,each subject was given 100 shelled sunflower seeds to cache.
The experimenterreleased each subject into the arena on a randomly determined side and left theroom. After caching, the subject was removed and returned to the home cage witha new supply of oats and lettuce. The position of all seeds was recorded, andall seeds were replaced for the next trial. For the second, subjects weren’tgiven any new seeds, but all seeds from the premanipulation were available intheir prior locations. Each subject was released into the arena on the sideopposite that they were released on the previous trial and given time tocontinue the cache session from the premanipulation. For the third, theprocedure was identical to the premanipulation, that is, all previous cacheswere removed, 100 new seeds were given to each subject and the arena wascleaned to eliminate any cues from the competitor’s prior presence. The resultswere that on average, subjects cached 14% of their seeds on the bare side inthe premanipulation, 33% in the manipulation, and 56% in the postmanipulation(Preston & Jacobs, 2015). For experiment 2 the methods and the procedure werethe same as in experiment 1.
For the results, most subjects preferred to cacheon the rich side of the arena in the premanipulation. Only one control subjectcached only on the bare side. With the majority of the caching happening in thepostmanipulation trial, this tell us that after learning something the animalswere more comfortable with the task. With the articles I’ve talked about thenext also has to do with dopamine and the effects it has on decision making.According to Salamone’s groupof researchers, forebrain dopamine (DA) systems are thought to be a criticalcomponent of the brain circuitry regulating behavioral activation, work outputduring instrumental behavior, and effort-related decision making (Salamone eaal., 2009). The article describes a novel effort-discounting task that involvesthe modification of a previously developed T-maze choice procedure (Salamone etal.
, 2009). Each arm of the maze contained different amounts of food toreinforce. The goal was to have the rats climb a barrier to get a largerreward. With the choice to climb the higher barrier to get more food or have nobarrier but a small amount of food they could assess the effects of dopaminedrugs. This articles experiments were similar to my first one.
Dopamine beingan important neurotransmitter has a great effect on animal’s decisions. Therats had a choice and with the food being the end goal they had to decide ifthey wanted to go for the easy food or work for the larger reward. Decisionslike these can tell us how much dopamine can play a part in the end decisions.Another article was called, AgeDifferences in Strategy Selection and Risk Preference During Risk-basedDecision Making. This article consisted of 22 rats that were housedindividually in guinea pig containers. The experiment lasted about 4 months.The rats were tested in a chamber that had white noise playing.
There was afood cup and two levers on either side of it. A light was above each of thelevers. The reward was a vanilla ensure liquid and was delivered into the foodcup. The rats were trained on hippocampus-dependent and hippocampus-independentversions of the Morris swim task. They did this training for 4 consecutive dayswith 6 trials a day. These 6 trails were separated by 2 blocks and had a20-minute break in between. There were 7 different releasing locations that therats were released. There was a little platform in the middle of the waterwhich was colored in a way so the rat couldn’t see through the water.
The ratswent through 6 trials where the platform was clearly visible and moved aroundfor each trial. All that rats could find the platform within 20 seconds ofbeing released. The rats were then tested in the chambers. They were given thevanilla ensure in their food to help reduce the effect of neophobic. They had 3days of magazine training where both levers were available and food rewardswere given when the rats would poke their noses on the food cup.
The rats wereable to press the lever within 5 days of training. After the lever pressing taskthere was a reward task. In this task, the rats would press the lever and getthe vanilla ensure and pressing the other lever would give them a reward 4 timethe regular size. Each session contained 36 trials and they were given once perday. The trails were broken up into 2, 12 forced trials where there was onlyone lever and 10 free choice trials where there were two levers. The lightwould appear above the desired lever and it would go off when the rat wouldrespond to that lever and food was given right after.
It was found that the agedrats learned or modified their choices toward them large reward associatedlever more slowly than did young rats. Also, that both young and aged ratslearned to discriminate between the lever associated with the small reward, andthe one associated with the large reward (Samson et al, 2015). Age has a lot todo with making decisions. Animals that are older are less likely to work hardfor things like rewards.
Although pressing a lever wasn’t a strenuous task itwas still found that the old rats were slower than the younger ones. My nextarticle is about discounting functions using concurrent-chain procedures. Researchers used a concurrent-chainsprocedure within sessions combined with an adjusting-amount procedure acrosssessions to determine the present, subjective values of food reinforcers to beobtained after a delay (Vanderveldt et al., 2016). There were 10 male pigeonsin this study that had had previous experience with discounted procedures. Therewere two chambers that the experiments were held in. There were two responsekeys in each that were in front of a panel that was above the ground and hadwhite, red and green lights that would light up. A clicker was used as auditoryfeedback.
There was also a triple cue light that was in the middle of the panelthat would light up green, yellow, and red. There was a pellet dispenser behindthe panel where the food would come out. The procedure had two controlconditions and two experimental conditions. A concurrent-chains procedure wasused in all control and experimental conditions. During the initial link of thechain, both keys were illuminated with white light. Red and green keys wereassociated with either the smaller, immediate reinforcer or the larger, delayedreinforce. In the first control condition, the pigeons chose between 32 pelletsto be delivered immediately and 32 pellets to be delivered after a 10-s delay.In the second control condition, the pigeons chose between 16 pellets and 32pellets, both of which were delivered immediately.
This condition was done for14 sessions. In the experimental condition there were two phases, eachconsisting of five conditions. In the first phase, the delayed reinforcer was32 food pellets, and in the second phase the delayed reinforcer was 16 foodpellets. Each of the two delayed amounts was studied at five delays, and inboth phases, each pigeon experienced the five delays in a different order.
thatas the delay to a reinforcer increased, its present subjective value decreased,and the data were well-fitted by a hyperbolic discounting function (Mazur, 1987). This result supports the use of the combination ofconcurrent-chains with an adjusting-amount procedure as appropriate forstudying delay discounting. This article concluded that even though there is adelay, the value of the food reward is still high. Animals like pigeons wouldn’twork for something like praise so food is the best way to make sure they willdo the task.
Effects of environmental enrichment ondecision-making behavior in pigs (porcus)was another article that looked at about animal’s emotional state, potentiallymodulated by environmental conditions, and its cognitive effects on processessuch as interpretation, judgement and decision making behavior (van der Staay,2017). The pig got to choose between two alternatives. The pigs can makeadvantageous or disadvantageous choices, where advantageous, low risk choicesdeliver smaller, but more frequent rewards, whereas disadvantageous, high risk choicesyield larger, but less frequent rewards (van der Staay, 2017). The study had 20 male piglets and they weretested in a pig gambling test apparatus. The pigs were either in a barren orenriched environment.
There were two goal boxes, each with a food bowl coveredby a plastic ball. The pigs had to be habituated to the two experiments andM&M’s which was their reward. There were 3 sessions per day.
The M’swere scattered in the corners and the pigs were encouraged to explore to getused to the space. A plastic ball was hung above the food which had theM&M’s in it. The pigs were trained to push the ball out of the way to getthe food. The ball was slowly lowered in each session to eventually cover thegoal bowl completely. They pigs were rewarded then in the central food bowl.The pigs were then only rewarded in the central food bowl.
Next both the goalboxes were open and the pigs could choose any box. 10-trial sessions were heldfor 6 days. It was found that housing conditions affected performances.
Barren-housed pig performed better and made more advantageous choices than theenriched-housed pigs, whereas no differences between the two groups of pigswere seen during the retention phase. The barren-housed pigs had higher haircortisol levels which suggested that the barren environment was more stressfulthan the enriched environment. Environmental enrichment is a huge part ofanimals and their overall well-being. Evidence of this has come about I manyarticles.
This particular article shows that the pigs in the barren houses andthose with higher levels of cortisol took more risks. This could be because ofthe stressed environment that they were in. Not having an enriched environmentcan make it so that you have to make more risky decisions and have to fight foryourself. When in an enriched environment, the animals most likely haveeverything they need and are feed, bathed, and treated with care. This willmake it harder for them when coming up against a decision like the ones in thearticle because they aren’t used to doing things like this one their own. With animalshaving the ability to make decisions due to certain drugs and neurotransmitterslike dopamine and the development of prefrontal cortex, my initial hypothesiswas supported with the evidence of these articles. Animals are tested to see ifthey can make their own choices and for the most part they can.
More studiesshould be done to see if there are other things like certain chemicals ordevelopmental processed go into making decisions.