DOI: 10.3897/ese.2024.e123288

Authors: Esteban Giraldo-González, Edgar García-Valencia, Juan Felipe Córdoba-Restrepo, Elea Giménez-Toledo

Abstract: Background: The project Cartografía de la Edición Académica Iberoamericana aims to analyze the production of academic books in Spanish- and Portuguese-speaking countries in the Americas. Following the path opened by similar studies in Colombia and Brazil, we present the results for Mexico.Objectives: To analyze academic books published in Mexico between 2013 and 2019 to examine the entities that published the books and their respective shares in the total output.Methods: A mix of quantitative and qualitative approaches was used to characterize the Mexican publishers of academic books based on data on ISBNs, the International Standard Book Numbers. The data comprised the information provided to the agency that assigns a unique ISBN to each book. We also used the Delphi method and formed discussion groups of experts. The groups were set up on the basis of responses to semi-structured questionnaires that sought to determine the criteria an entity must satisfy to be considered an academic publisher.Conclusions: Of the 196 533 ISBNs issued in Mexico between 2013 and 2019, 117 929 (60%) were issued for books dealing with academic subjects. Commercial publishers accounted for the largest share of those books (63 044 ISBNs, or 53.4% of all the academic books), followed by university presses (29 628 ISBNs, or 25.1%). The group of experts suggested that among the 1289 publishers that requested ISBNs for academic books, only 151 (11.7%) can be considered truly academic publishers; 678 (52.6%) cannot; and 460 (35.7%) were borderline cases, as they meet some but not all the criteria for them to be considered truly academic.

HTML

PDF

DOI: 10.3897/ese.2024.e114734

Authors: Gina Joubert

Abstract: Background: Statistical analyses are a key component of quantitative research in health sciences. Objectives: To review the instructions for authors on reporting and presentation of statistical methods by all health sciences journals based in South Africa. Methods: Health sciences journals based in South Africa that publish original quantitative research articles were identified using three sources, namely the list of accredited South African journals compiled by the South African Department of Higher Education and Training in 2022, relevant journals covered in Scopus, and web pages of major health sciences publishers in South Africa. The list was cross-checked against the listing of journals in Sabinet, an online database covering South Africa, under the category ‘Collection: Medicine and Health’. The instructions for authors given by the journals were accessed through their websites. The form for recording data was based on items listed in the ‘Statistical Analyses and Methods in the Published Literature’ (SAMPL) guidelines. Results: All except one of the 52 journals could be located online. Of the 51, 13 (25%) made no mention of statistics in their instructions, and 11 (22%) made only a general statement regarding statistical content with no further guidance. The statistical item most frequently mentioned was the P value (45% of journals), whereas the rest of the items appeared in the instructions of 20% or fewer journals. Nine journals (18%) referred to the EQUATOR guidelines, mainly CONSORT (10%). Conclusion: Nearly half of the health sciences journals based in South Africa either did not mention statistics at all in their instructions for authors or made only a cursory reference to statistics. The study thus emphasizes that these journals, in their instructions for authors, need to cover in greater detail the reporting and presentation of statistical methods in articles reporting quantitative research.

HTML

PDF

DOI: 10.3897/ese.2021.e63780

Authors: Ksenija Bazdaric, Dina Sverko, Ivan Salaric, Anna Martinovic, Marko Lucijanic

Abstract: Regression analysis is a widely used statistical technique to build a model from a set of data on two or more variables. Linear regression is based on linear correlation, and assumes that change in one variable is accompanied by a proportional change in another variable. Simple linear regression, or bivariate regression, is used for predicting the value of one variable from another variable (predictor); however, multiple linear regression, which enables us to analyse more than one predictor or variable, is more commonly used. This paper explains both simple and multiple linear regressions illustrated with an example of analysis and also discusses some common errors in presenting the results of regression, including inappropriate titles, causal language, inappropriate conclusions, and misinterpretation.

HTML

XML

PDF