Comparative analysis between artificial and human intelligence in the teaching of higher education journalism studies
Objectives: In order to carry out this research, two objectives have been set: O1) To compare the skills of journalism students with AI in writing report headlines. O2) To recognise the capabilities of AI as a producer of journalistic content. Methods: This research compares the journalistic writing...
| Autores: | , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2025 |
| País: | España |
| Institución: | Universidad Complutense de Madrid (UCM) |
| Repositorio: | Docta Complutense |
| Idioma: | español |
| OAI Identifier: | oai:docta.ucm.es:20.500.14352/126506 |
| Acceso en línea: | https://hdl.handle.net/20.500.14352/126506 |
| Access Level: | acceso abierto |
| Palabra clave: | 070 316.77 004.8 Inteligencia artificial generativa Periodismo Experiencia docente Estudiantes de periodismo ChatGPT Inteligência artificial generativa Jornalismo Estudantes de jornalismo Generative artificial intelligence Journalism Teaching experience Journalism students Comunicación social Inteligencia artificial (Informática) 5910.02 Medios de Comunicación de Masas 1203.04 Inteligencia Artificial |
| Sumario: | Objectives: In order to carry out this research, two objectives have been set: O1) To compare the skills of journalism students with AI in writing report headlines. O2) To recognise the capabilities of AI as a producer of journalistic content. Methods: This research compares the journalistic writing of journalism students and the IAG. Fourth-year students were asked to produce two entries for a news report: one written by them and one using the IAG through Editmaker, a software developed by Cibeles Group with OpenAI's GPT-3.5 TURBO technology. The analysis included 72 entries, half from students and half from the IAG. Variables such as number of words, use of the 5Ws of journalism, coherence, sentence types, syntactic complexity and voice were evaluated. Results: The analysis shows differences in the entries between students and Generative Artificial Intelligence (GAI). GAI uses, on average, 7.25 words more than students, with averages of 91.94 and 84.69 words, respectively. The dispersion is greater among the students, with the entries varying between 27 and 171 words, while the GIIs range between 44 and 159. Regarding the use of the 5Ws of journalism, the GIIs outperform the students, with an average of 3.67 vs. 3.08. The biggest difference is found in the question "where?", with 22.22% in favour of the IAG. Who?" and "When?" also stand out, with a difference of 11.11%. The IAG answers all 5Ws in at least 50% of the cases, except "when". In contrast, students only exceed 50% for "what?" (97.22%) and "how?" (55.56%). Conclusions: The study shows that, although the differences between students and Generative Artificial Intelligence (GAI) in journalistic headline writing are not large, GAI performs better in all the variables analysed. GAI uses at least 50% of the 5Ws in most of its headlines, while students only reach that frequency in two of the 5Ws. In addition, the IAG is more effective in answering more questions and generating more entertaining texts, with greater variety in tone. KEYWORDS: Generative artificial intelligence. Journalism. Teaching experience. Journalism students. ChatGPT. |
|---|