The Perils of Flawed Experimentation

IMG_0030

Cover Letter

The Bobo Doll experiment conducted by Bandura is commonly used as support for the argument that observing aggressive behavior, perhaps on television or in person, results in the spread of such aggressive behavior. However, the Bobo Doll experiment contains flaws in its methods and ethical viability. This does not completely discount the results of the experiment, but does call into question whether it should be so commonly used as evidence for such a serious claim, or if more thorough experiments should receive more attention in regards to the confirmation of arguments.

             Within the field of behavioral psychology, papers tend to center around studies and research more than anything else. However, just because a scientist writes a paper claiming that their study proves or disproves something in particular, that does not make their research valid if there are already underlying problems with the study they conducted. The famous Bobo doll experiment, conducted by Bandura in 1961, is a fitting example of a study that gets some parts right, but is faulty nonetheless. It is vital that studies such as this are critically analyzed so that future experiments can learn from their mistakes and successes to become more accurate.

In any study, a sample is an important part of the experiment process. In order to successfully apply the results of the study to a wide population, the sample must be diverse and representative of that population. In addition, in order to ensure that no sampling bias is present, the sample must be random. The sample that was tested in the Bobo doll experiment is, at first glance, acceptable. An equal number of boys and girls were tested, all of a similar age, and they were also divided equally among three broad groups, which each had two adults, one male and one female, to participate in the study (McLeod). However, all the boys and girls selected for the study came from the same nursery school, which brings into question whether or not the results of Bandura’s study can be applied to larger populations with more diversity (McLeod). In addition, although Bandura’s sample was fairly sizeable, the number of groups and sub-groups he had ended up spreading his sample pretty thin, with only six children in each distinct group, leaving a large amount of room for error based simply on human difference. That being said, Bandura was right to include both male and female models of aggression, in order to eliminate gender as a factor in the likelihood of children to copy aggressive behaviors.

The application of the study was also limited by the very isolated conditions of the experiment, which involved one child at a time with a single adult present, a situation much less likely to occur organically outside of the laboratory. Like the sample, experimental conditions must be applicable to conditions possibly experienced in the course of peoples’ everyday lives, or else the results of the study will only stand up under very specific circumstances. In real settings, children often have many adult and non-adult role models in their lives, not just one. With the conditions of the Bobo doll experiment set up as they were, isolation could be counted as a factor that made the children feel dependent on their one adult role model, and could invalidate the results.

The way Bandura conducted his experiment involved first making the children angry by denying them toys. Then the model of aggression, an adult treating various objects violently, was shown to them (McLeod). This was intended to test the presence of vicarious conditioning, which is the learning of behaviors through the observation of such behaviors (Ungvarsky). However, his experiment brings up the issue of personality, and pre-existing inclinations towards aggression. If some children were already more inclined to show aggression due to their personalities, then their aggressive behaviors may have had little to do with the models they watched, but rather just a predisposition for violent behavior. Another factor that Bandura failed to take into account was the perceived mood of the models as they displayed aggression. The facial expressions and other body language displayed while the models kicked at the Bobo dolls and verbally abused them could have also had an effect on the likelihood of the children to copy aggressive behavior. If the model they observed smiled or generally seemed happy while abusing the Bobo doll, then the child’s impression of the severity of the act could be lessened, while an angry expression could have caused some of the children to feel afraid and deter them from acts of violence. There is also the question of whether the children that repeated the aggressive behavior they had seen did so because they thought it was fine, or because they simply wanted to please the adults that had carried out the behavior. This means that the results of the experiment could indicate the desire for praise felt by children rather than the likelihood of aggression increasing through modeled behavior in general.

One of the most glaring oversights of Bandura’s Bobo doll experiment was the lack of consequences given for aggressive behavior. Fear of consequences, after all, is one of the main factors in keeping people lawful and cooperative with others. In the Bobo doll experiment, not only were there no outside punishments given by society, or a parent for misbehavior, but there were none of the ever present natural deterrents either. In reality, if you were to hit an actual person, they would bleed, or fall down, or hit you back, and if you verbally abused someone, they might cry. The Bobo dolls were, of course, unable to respond to the abuse in such a manner. They were designed to pop back up after being hit, which is highly unrealistic (McLeod). To the children, these dolls would seem to be nothing more than amusing toys that they could take their frustrations out on. Bobo dolls cannot be hurt. Perhaps if the dolls remained knocked over when hit, then the children might have understood the effect of their actions, but as it was it seemed that there was nothing unfavorable about what they were doing, which could definitely skew the likelihood of them acting aggressively. Bandura later conducted studies that took consequences into account, but for this first experiment, they were blatantly lacking. These later experiments did not include the presences of an adult in the same room as the child and involved an example of someone being punished or rewarded for their actions (Ungvarsky). This would fix some of the problems with the lack of consequences in Bandura’s earlier experiment, but would not account for the unrealistic nature of the object of aggression, the bobo doll.

One highly debated flaw within Bandura’s experiment is ethics. In the field of psychology, ethics are of great concern, as experiments often center around people, and the potential for lasting damage cannot be allowed. Ethical restrictions are some of the most discussed aspects of research. Nowadays, ethics committees have to review psychological experiments before they take place in order to ensure that the benefits of the study outweigh any risks. It is very unlikely that an experiment such as Bandura’s would be approved by an ethics committee in the present. Experiments that center around children always face more scrutiny than others, as children are very impressionable at a young age. The Bobo doll experiment focused on the premise that aggression was learned from models of violent behavior. Thus, if aggression was indeed a learned behavior, some the children who participated in the study would face the risk of higher levels of aggression in their later lives. This may seem like a minor or unlikely problem, but the subject of ethics allows for no lenience. Without ethical guidelines people could run experiments on unassuming strangers all the time, and so no allowances can be made for experiments that have the potential to cause issues for participants.

In a society that relies so heavily on media to provide truthful information, one has to be especially careful when conducting an experiment, particularly due to the fact that experiments are designed not only to test correlation, but causation. Bandura’s study led to widespread fear that images of violence in media would turn everyone into aggressive beings, but his study was flawed. Unfortunately, many people take the results of experiments at face value, without actually considering the study itself and whether or not it might contain fundamental problems. Most experiments have some useful elements and do at least seem well-constructed on the surface, but ultimately many have problems that also need to be considered. The results of studies can very rarely be applied to every person or all settings, and s the restrictions or failings of experiments need to be taken into account before they are broadcasted as worldwide truths.

Works Cited

  1. McLeod, S. A. (2014). Bobo Doll Experiment. Retrieved from www.simplypsychology.org/bobo-doll.html
  2. Ungvarsky, Janine. “Vicarious Conditioning.” Salem Press Encyclopedia (2016): Research Starters. Web. 14 Apr. 2016.