The bibliographical references (BR) in scientific articles exert several functions (Stapleton 1987). They report on the history of the topic and its status quo, and show authors, institutions and countries that have examined it. They also report on the date and place where the reports on that subject have been published. It is ethical to refer all related studies or at least contemporaries. The BR does not appear in title, summary and Materials and Methods, but they are in introduction, Materials and Methods and discussion (Lindsay 1984, Stapleton 1987 and Correa 2003).
Currently, the quality of scientific articles is usually evaluated by the site of their publication (Bauman 2009, Rickert 2009 and Soto 2016), which is usually within lists made by institutions, which require these qualifications in order to periodically evaluate the performance of scientists, and also to give or not financial support to carry out new researches. In general, it is the so-called impact factor of the publication, the index most commonly used for these evaluations. In turn, high-impact publication sites tend to be preferred by possible authors to spread and disseminate their experimental results.
Possibly one of the features that could contribute to the quality of scientific publications could be the use and manipulation of the bibliography consulted and referred to in a publication. This topic, which is not easy to evaluate, has been little examined, in places such as instructions for writing papers in periodically published scientific journals, or in manuals for writing documents that report on experimental works (Lindsay 1983 and Stapleton 1987). The objective of this study is to inform about a research of the status quo of BR in high impact periodic publications on the topic of fibrous resources in pigs, during the period 2006 and 2017.
Materials and Methods
An examination of the referred documents or bibliographical citations in scientific articles was random carried out. The articles were published between 2006 and 2017 on the topic of using fibrous resources in pigs rearing. The common characteristics to all the evaluated articles were that they reported on unpublished experimental work and that they appeared in journals that published systematically and periodically, materials of all types of livestock and in different topics. Journals published out of Cuba and Mexico also had the characteristics of publishing only or preferably in English. The list of journals appears in table 1.
The number of documents referred per article, their distribution in the article (introduction, Materials and Methods, results and discussion), the average year of publication of the documents referred to in each article and the age or temporal distribution of those documents from the closest date to the furthest publication, by groups that comply with a geometric progression: 1, 2, 4, 8, 16 and 32 years of publication were evaluated. All these measures were compared between six years, following a simple classification (Steel et al. 1997) to research the effect of the two groups of six years, 2006/2011 and 2012/2017. The manipulation of the data was performed with the help of an appropriate statistical package (Minitab 2014).
Results
It was found that the average BR per article was 33.5. There was no significant effect (P> 0.10) when the six-year periods 2006/2011 and 2012/2017 were compared. In addition, its distribution among introduction, Materials and Methods and discussion was 33.5. 20.5 and 46.0 %, respectively, without significant effect (P> 0.10) of six-year period in any case (table 2).
The percent of BR contained in the introduction was inversely proportional (r, - 0.873, P <0.001) to those of the Discussion (table 3). There was no significant correspondence between the percent in Materials and Methods with respect to the other two sites where BR appeared. The correlation matrix found is shown in table 3.
The average publication date as well as the temporal distribution of publications that were referred in the 12 articles studied, appears in table 4. There was no significant effect of six year period (P <0.192) for the average publication date, within the cycle of 12 years that were reviewed. In the chronogram of referred publications, neither the six year periods exerted any influence (P> 0.10), although between 2 and 8 years of previous publications, there could be some influence, with fewer documents cited in the second six year period. In the chronology of publication of BR, there was not treatment effect (P> 0.10), with an accumulated of 1.7, 5.6, 17.6, 35.0, 66.1 and 93.9 % of the total mentioned with an age of 1, 2, 4, 8, 16 and 32 years, perhaps forming a temporary cycle.
The polynomial link between age and accumulated percent was highly significant (R2, 0.999, P <0.001).
This interdependence is shown in figure 1.
1Express the average of the year in which all BR of each article were published
2Express the cumulative number of documents referred to in the period
Discussion
This evaluation was based on the assumption that all BR reported in each of the evaluated documents were adequate and there was no lack of any. This is due to the ethical principle of the authors, in the sense that it is not possible to ignore any document because its reflects contradictory data with those reported in the reference article (Apel 2003, Gardner 2010, Dorta 2013 and Soto 2016). Even so, subjective and contingent omissions could occur for reasons not inherent to the will of authors, and equally, there could be objective impediments to making the citation, such as the inaccessibility of the document. These aspects were not taken into account in the exam published here.
In this study it was evidenced that the number of BR per article, as well as the average date of references per article were not subject to much variation, despite the diversity of publication sites in the 12 selected years, in the high-impact journals that were considered. More tendencies to variation could be found in the temporal distribution of the referred publications. It was not possible to make comparisons with other studies performed in this regard, which are not usually mentioned in the scientific writing manuals (Correa 2003, Marín 2006, Luby and Southern 2015 and Smith and Fudge 2017) or dismissed as a routine office task (Ferreira and Abreu 2007), or by subjecting the quality of a scientific publication exclusively to its content (Muñoz 2007). Perhaps an age of eight years in these referred publications could be a possible index that could qualify the scientific articles, at least those similar to the bias selected in this research. Undoubtedly, more research is required to support this suggestion. In this way, it would be possible to contribute to the evaluation of each scientific article per se, and not because of the impact factor of the publication site.
The BR are essential to know and interpret experimental results, as well as to project new researches. Consequently, BRs can have a considerable influence on the evaluation of the quality of scientific publications. It is recommended to continue the study of the function performed by the previous references cited in the scientific articles. At least pigs raised with fibrous foods.