Naslov
Can presentation vocabulary influence the perception of research misconduct: a randomized study

Autori
Buljan, Ivan ; Tokalić, Ružica ; Viđak, Marin ; Tomić, Vicko ; Marušić, Ana

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, stručni

Izvornik
6th World Conference On Research Integrity ; Abstract book / - , 2019, 142-142

Skup
6th World Conference On Research Integrity

Mjesto i datum
Hong Kong, Kina, 02.06.2019. - 05.06.2019

Vrsta sudjelovanja
Poster

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
research integrity ; research misconduct ; randomized trial

Sažetak
Objective: We will perform a randomized controlled trial to assess whether the vocabulary used to describe cases of detrimental research practices can influence the perception of research integrity. Method: Randomized controlled trial will be performed on a non-expert population. The task for the participants will be to assess the appropriateness of the behavior described in the scenarios that describe three different types of misconduct: data fabrication, authorship manipulation and non-publication. The intervention in the trial is the language used in scenarios. Scenarios will be described using the vocabulary related to Moral Foundations Theory. In one group, the case will be described using words associated only with Harm or Fairness moral foundations, and in the other using words associated with Authority, Ingroup and Purity moral foundations. All scenarios will be standardized in terms of the number of words and emotional tone. The participants will also be asked to provide the reasons for their decision after reading each scenario and assessment of appropriateness of the behavior. The analysis of their free text responses will be made using natural language processing software LWIC – Linguistic Inquiry and Word Count. Results: The scenarios were piloted on a small sample of medical students (N=46). No issues in clarification of the content or other changes were needed, and based on the pilot results we calculated that we will need 52 participants per group to obtain the 10% difference between groups with α=0.05 and 80% power. The groups in randomized controlled trials will be analyzed using t-test for independent samples (if the distributions of results are normal) or Mann Whitney test (if the distributions are not normal). Qualitative answers about the reasons for their decision will be analyzed and compared between two groups regarding the word characteristics, usage of moral foundations vocabulary and emotional tone.

Izvorni jezik
Engleski

Znanstvena područja
Informacijske i komunikacijske znanosti

POVEZANOST RADA