conclusions of subsequently published work - The Complete Academic Support Hub

Discussion Response 1	Reference: 2 Harvard reference + citation	Andrew Word count: 100 words
The case study chosen (Whitbeck, 2016) is concerned about the ethical issues that occur when one researcher suspects another of fabricating data and invalidating the conclusions of subsequently published work. The brief synopsis of the case study is that a student is trying to improve a previous researcher’s algorithms (who is also the student’s professor). These algorithms were used to increase performance of some attribute in disk caching and is being used in industry. The student asks the researcher for the simulation code and for advice, which the researcher could not provide. After the student rebuilt the simulation to test the previously produced algorithms it was noticed that performance results were significantly worse than those published by the researcher. The student had previously used the results in their previous work, which would cause issues for their earlier research too. The main ethical challenge is that published work has been produced with ‘bad’ data. There is also an ethical issue in that there is a suspicion that the data was falsely fabricated. Even if it is not an act of bad practice, but a real mistake there are ramifications for the original researcher (loss of reputation, maybe even leading to loss of job). There might be a temptation for the student to not confront their professor, however this would mean they would become complicit and sooner or later there is potential for a third party to realise the result data is compromised. The case study does not provide a resolution, instead it asks the question to the reader as to what they would do. A similar real-life case study (Editage.com, n.d.) proposes that the researcher should offer a clear explanation of the data’s credibility. This would be the starting point for the student, asking the researcher that question. This would allow the researcher to offer a retraction for their previous work. If the student continues with their work, then they have a duty to provide the correct results. The project I have chosen aims to automate the validity of financial controls and also producing a number of algorithms I will be using data and producing scripts that run simulations that will produce performance characteristics and testing the validity of the algorithms utilised. I would want to ensure that the results would be repeatable by anyone reading my dissertation. As such there would be a requirement to provide the simulation code and methods of acquiring the data. This would demonstrate that there would be no malicious intent on my part. There is a requirement to keep in mind the best practices of data quality (Horne, n.d.); Don’t cherry pick data- don’t choose data sets that only prove the hypothesis you are undertaking. Understand the margin of error – utilising a lot of data should minimise this. Empirical data should remove any anomalies. Consistency – the data should be cross referenceable with the same results. Completeness – there should be no missing values.
Requirement:
Engage with your colleagues in a follow-up discussion on the implications of ethics within the context of research, suggesting ways in which risk might be minimised and offering insightful comments regarding the legal, social, ethical and professional challenges within the context of a researcher in the computing discipline Dissertation.