Multivariate analysis of credit risk and bankruptcy research data: a bibliometric study involving different knowledge fields (1968–2014)

Bibliometrics are important evaluation tools to the scientific production. A bibliometric study was used to evaluate researches about credit risk and bankruptcy. Credit support technique studies on economic and social orders are relevant in this knowledge field. Therefore, the aim of the current stu...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Scientometrics Ročník 106; číslo 3; s. 1007 - 1029
Hlavní autoři: do Prado, José Willer, de Castro Alcântara, Valderí, de Melo Carvalho, Francisval, Vieira, Kelly Carvalho, Machado, Luiz Kennedy Cruz, Tonelli, Dany Flávio
Médium: Journal Article
Jazyk:angličtina
Vydáno: Dordrecht Springer Netherlands 01.03.2016
Témata:
ISSN:0138-9130, 1588-2861
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Bibliometrics are important evaluation tools to the scientific production. A bibliometric study was used to evaluate researches about credit risk and bankruptcy. Credit support technique studies on economic and social orders are relevant in this knowledge field. Therefore, the aim of the current study is to identify and describe the application of multivariate data analysis techniques to credit risk and bankruptcy scenarios. The herein presented data were collected in publications indexed to Thomson Reuters’ Web of Science database between 1968 and 2014. The results corroborate information in the literature and in previous bibliometric reviews, as well as highlight other indications regarding the construction and development of research fields. Since the 1990’s, the neural networks became relevant due to their increased use as study object in publications. However, both the discriminant analysis (J Finance 23(4):589–609, 1968 . doi:10.2307/2978933) and the logistic regression (J Account Res 18(1):109–131, 1980 . doi:10.2307/2490395) are still often used in researches, fact that shows the tendency to find articles using more than one technique or hybrid models, artificial intelligence techniques and complex computer systems. This field appears to be multidisciplinary in journals and Web Science databases involving the business and economy, operational research, management, mathematics, data processing, engineering and statistics fields. Another relevant discovery was the increased number of publications about this subject launched right after the 2008 crisis.
ISSN:0138-9130
1588-2861
DOI:10.1007/s11192-015-1829-6