Alcohol's impact on Violent Crimes

January 9, 2019

 Image Sourced from


Does alcohol impact one’s behaviour while under the influence? In “The Impact of Alcohol on Different Types of Violent Incidents” by Felson, Burchfield, and Teasdale (2007) examine conflicts involving alcohol between both strangers and various types of acquaintances. This shows interactions between people during intoxication and is an indication of which violent crimes are most likely committed. One might say their article provided a useful source of information because they collected a representative sample from the National Violence Against Women and Men Survey (NVAW). However, this critical review will set out to prove that the article was not sufficient in its research or findings because of evidence found in the introduction, methods, results, discussion section, the overall outcome of the paper, and the possibilities of future research.


Out of Date Research?

The introduction contains many primary resources that were not current during the period the article was written. This is problematic because it is out-dated research. There could have been more current, updated research that absolved past research assertions. “People are less likely to assault a stranger unless they are drinking” (Felson, Burchfield, & Teasdale, 2007, p. 1058). The authors do not use research to provide evidence that supports this statement. It is possible that the statement is a thesis statement the researchers are looking to validate from their study, but it is not clearly indicated. The introduction of a research article should be factual background research that pertains to the current study. I believe drinking is a factor that may play a role in whether someone assaults a stranger, but I believe it mainly falls on their aggressive nature and level of confidence. Some people look for fights and others may try to reason with people but resort to violence out of frustration. Not everyone can verbally reason with people calmly. I think if someone is frustrated with a stranger they are less likely to walk away because they know it will never be resolved. Whereas if the conflict is between people that know each other they are likely to distance themselves to calm down before solving their dispute.


Qualitative and Quantitative Data

The researchers’ use a qualitative and quantitative approach in their article which is appropriate because two methods increase the validity of the study. Used in combination with each other they provide support for each other. Used solely they provide information but have limitations on the interpretations and conclusions they can provide. The researchers also positively defined definitions that had room for interpretation. This is good because it narrows the focus and does not leave the audience guessing. The article’s methods section outlined how the NVAW collected the data. The authors did not provide a reason as to why they decided to use data from the NVAW. The problem with using this data is the researchers cannot be sure if the researchers collecting data for NVAW did it in an appropriate manner that had no effect on the results (ex. Gender of the researcher could have affected how the participants responded). Using this data would have been quicker than generating their own data and it likely gave them a more diverse/large sample size. If they conducted the study themselves, they could have set up more controls to generate more accurate information such as having an anonymous paper self-report. This would minimize the impact that the researcher could have on the data.



The results showed people were likely to be intoxicated when involved in a violent situation with a stranger. This is likely because it takes confidence to talk to a stranger let alone fight them. Alcohol impairs judgement, giving a feeling of invincibility, which is where the confidence stems from. Many of the other results were considered to not be statistically significant. This is likely because of the degradation of internal and external validity. There were many variables to control for in this study causing some variables to be forgotten which decreases validity. Validity was also decreased from the interview conducting procedure and from the people self reporting because they could have errors in their information making the data inaccurate. For instance, the victim could have been intoxicated as well but did not want to admit to it out of embarrassment. They could have also not admitted ingesting alcohol because they wanted to make the offender look bad and portray themselves as an innocent victim. People care what their peers think of them even if the researchers are strangers. It would have been better if a written self-survey was done because it increases anonymity. The results will also be skewed from the various levels of alcohol in offender’s systems during the time of attack. If victims were asked to recall the amount there would be an error of measures (overestimate of the number of drinks). Many victims also did not know whether their perpetrator was under the influence or not.


“Offenders may be at greater risk of retaliation or criminal sanctions if they assault strangers” (Felson et al, 2007, p. 1066). The authors do not give reasoning or evidence behind this statement. I believe strangers are more likely to retaliate or call the police because they do not know the person on a personal level to care about how criminal sanctions affect their life. Whereas if the offender and victim know each other then the victim may be more sympathetic or thinking about how it personally will affect them if they retaliate or proceed with criminal sanctions (ex. The victim could worry about being shunned by a friend group for retaliating against another friend). I think an important factor of violence induced by alcohol is the location of the attack. Strangers are more likely to share a dispute involving alcohol in a public place such as a bar. Many people go to bars as a way of meeting someone from the opposite sex making most of the people in the bar strangers. If alcohol is consumed in a home, then intimate partner violence is more likely to occur because there is no one else to take their anger out on. Intimate partner violence is more likely to occur behind closed doors especially if the man is the abuser because they do not want strangers judging them hitting a woman (socially unacceptable).


What went Wrong?

There were many factors that were not controlled in this experiment such as the amount of alcohol or drugs that were consumed by the offenders. Some of the victims were unsure whether their attacker was intoxicated at the time of the incident because they were not in a bar setting to know for sure. The only way to know for sure is if the incidents were reported to the police because the police would be able to do a breathalyzer test. Self-reports are not the best method of data collection because people become unreliable when they are asked to report events that happened in the past. The victims would be more unreliable if they were also intoxicated at the time of the incident. It is also unsettling that the data is from 1995 and 1996 because it does not reflect recent conditions. In today’s society there is more security in reporting an incident in comparison to the past where many people did not want to talk about sexual abuse or attacks much less report them. Society has changed what is considered socially acceptable. Another point is the title and abstract indicate that alcohol is being looked at as a factor of violent behaviour. The data collected from NVAW used other drugs as a factor of violent behaviour. Therefore, it is tough to know whether the violence was from drugs, alcohol, or both.



In the future the researchers should use up-to-date research. This way it will be reflective of the current population. Self-reported information should not be used as well so more accurate results are yielded. Instead crime statistics should be used from police recording information about the incidents occurring. This gives more accurate information because the information is recorded facts by a policeman rather than opinions or speculations by an unsettled victim. The problem with this method would be less data and potentially less variety because not all violent disputes are reported. Having factual information will return better results. It would also be better if only certain factors were focused on such as race and gender that way all the results are interpreted. Having fewer variables can allow room for greater elaboration on the results. The authors used results to confirm/refute their hypothesis and gave examples of data interpretation, but they did not interpret all of the data. If the data was not useful they should have excluded it. Data should not be included without an explanation. It is assumed the audience of the article was intended for scholars who could interpret the information from the data tables themselves. However, scholars also are interested in the author’s interpretations of the data since they felt it was necessary to include that specific data.


In conclusion, the authors did not create a successful research article. There are aspects of the article that can be seen as positive elements of a good research article. This includes the use of qualitative and quantitative methods and defining definitions. However, the negative elements outweigh the positives which can be seen as outlined by evidence throughout the article.



Felson, R. B., Burchfield, K. B., & Teasdale, B. (2007). The impact of alcohol on different types of violent incidents. Criminal Justice and Behavior, 34(8), 1057-1068. doi:10.1177/0093854807299651

Please reload

Recent Posts

February 4, 2019

Please reload


Please reload


Please reload

City, University of London
Clerkenwell, London

©2017 City, University of London Student's Union Law Society 
Image copyrights Rajas R. Chitnis, unless otherwise stated.