In December 2022, ChatGPT had just begun to gain popularity, but tech gurus already hailed it as the wave of the future. But academicians are hesitant to embrace this AI technology.
In fact, the two most esteemed academic journals, Science and Springer Nature, have enacted new editorial guidelines that forbid or severely restrict the use of programmes like ChatGPT in the creation of academic papers.
A recent study, however, might play devil’s advocate for these worries. It asserts that, rather than being a threat, it can help academics with their studies.
“Good Finance Journal”
Researchers in finance, Brian Lucey of Trinity College Dublin and Michael Dowling of Dublin University, recently described the results of their investigation in an article that appeared in The Conversation.
They assert that using ChatGPT, one could produce a finance paper that passed for an academic publication.
They asked the AI tool to produce the four traditional components of a research study—the research problem, associated literature, dataset, and recommendations for testing and examination.
According to the researchers, the results should be deserving of publication in a “good finance journal.”
In a second test, they pasted at least 200 abstracts of pertinent research studies into the ChatGPT window and instructed the software to use these in building the four research stages.
They examined the responses generated by the tool and offered corrections to ensure that their knowledge was still being utilised in the project.
The team then asked a group of 32 reviewers to rate ChatGPT’s results. They were tasked with determining whether it is thorough, accurate, and capable of advancing the subject of finance studies.
Considered Acceptable
According to Lucey and Dowling, the panel largely regarded all of this research as acceptable. This may suggest that, despite being in its early stages, the AI tool may be able to generate good suggestions for academic study.
They found that the rankings for various study fields varied. Positive reviews of the dataset and research plan were common. The assessments of the literature and the testing recommendations received a lower but still respectable rating.
The researchers hypothesise that while ChatGPT excels at connecting a collection of external texts, it struggles with more difficult steps, particularly when there are numerous conceptual steps, such as literature reviews and testing.
“For the time being, we believe researchers should view ChatGPT as a tool rather than a danger. It may be especially helpful for study teams that frequently lack the funding for conventional (human) research assistance. In the article, the researchers wrote:
However, they also point out that while writing journal papers, academics must be mindful of the prohibition on its use. There are different opinions on the consequences of this instrument in the academic world, so it must be used with caution.
Stay connected with postvines for more information!!
