Original Article
Health care articles with simple and declarative titles were more likely to be in the Altmetric Top 100

https://doi.org/10.1016/j.jclinepi.2016.11.018 Get rights and content

Abstract

Objectives

The aim of this study was to assess whether specific title characteristics could influence the likelihood of being included in the “Altmetric Top 100.”

Methods

We conducted a 1:2 matched case–control study with the cases being the health care articles included in the “Altmetric Top 100” lists (2013–2015) matched through a random computerized procedure with two health care articles published in the same journal and year. For each title, we extracted the number of characters in the title, the number of uncommon words, and whether the title was declarative. Conditional logistic regression was used to estimate odds ratio (OR) with 95% confidence intervals adjusted for a prespecified baseline confounder (open access).

Results

One hundred eight “Medical and health sciences” articles were retrieved in the 2013–2015 “Altmetric Top 100” and matched to 216 control articles. Titles of the “Altmetric Top 100” articles were 102.6 characters (±42) long, included 3.4 (±2.0) uncommon words, and 29.6% (32/108) were “declarative.” Titles of the matched articles were 109.3 characters (±37.1) long, included 4.7 (±2.4) uncommon words, and 21.8% (47/216) were “declarative.” After multivariate adjustment, declarative titles with a lower number of uncommon words were significantly more represented in the Altmetric list, with declarative titles having 2.8 times the odds of being in the top list (OR: 2.8; 95% confidence interval: 1.2, 6.4). For each additional uncommon word in the title, there was a 1.4 increase in the odds of being a non-Altmetric Top 100 article (1.4; 1.2–1.6).

Conclusion

An easy-to-understand, informative title may help bridge the gap between scholar and social media dissemination.

Introduction

The title is a fundamental part of an article. The objective of the title is to summarize the content of the article, trigger reader's curiosity, and ideally facilitating its indexing. However, there is no consensus on whether a title should always cover the conducted research methods (“indicative title”) or key conclusions (i.e., “declarative title”). Defenders of declarative (syn. informative) titles suggest that they could “attract” busy clinicians, assisting them in their efforts to keep up with the literature [1]. Opponents claim that declarative titles tend to support biased, goal-based research [2]. However, all stakeholders agree that a balanced dissemination of research findings is essential [3], [4], [5].

In the last few years, new article metric tools were developed, that is, altmetrics. The term altmetrics, a generalization of “article-level metrics,” is based on the homonymous hashtag and has been implemented in 2010 [6]. Altmetrics include novel, nontraditional metrics that measure several aspects of dissemination of scientific findings and are proposed as alternative to the traditional citation metrics (e.g., impact factor, h-index). Several types of metrics that attempt to capture various aspects of the impact of a research article are considered altmetrics. As an example, the altmetrics used by the Public Library of Science include the number of times an article has been “Viewed” (i.e., number of HTML views and PDF/XML downloads), “Discussed” (i.e., dissemination of the article in blogs, Wikipedia, and social media, such as Twitter, Facebook, and Reddit) and “Saved” (i.e., number of times that a research article has been bookmarked in Mendeley and CiteULike) [7]. One particular article-level metric, the Altmetric score, tracks how often articles are mentioned online, such as in blogs or social media platforms [8]. Besides measuring social impact, Altmetric scores are also associated with citations [9], [10].

Since 2013, Altmetric publishes a list of the 100 most shared (or discussed) articles of the year. This list therefore shows which research studies in a pertinent year had a significant impact on social media. The objective of our study was to assess whether specific title characteristics could influence the likelihood of being included in the “Altmetric Top 100.”

Section snippets

Study design and outcomes

A matched case–control study was planned. Health care articles (i.e., articles categorized as “Medical & Health Sciences”) included in the top 100 Altmetric lists (2013–2015) were the cases, and randomly selected health care articles matched for journal and year of publication were the controls. The primary outcome was the association between title characteristics (type of title, presence of uncommon words, and title length) and inclusion in the top 100 Altmetric list.

Selection of control articles

Control articles were

Results

Of the 300 articles listed in the 2013–2015 “Altmetric Top 100,” we included the 108 articles of the “Medical and health sciences” category. To these cases, we matched 216 randomly selected articles as controls. The operators mostly agreed in their categorization of titles as “declarative” and “indicative,” with only 11 disagreements on 324 articles. All disagreements were resolved through discussions.

Titles of the 108 “Altmetric Top 100” articles were 102.6 characters (±42) long, included 3.4

Discussion

Based on the results of the present study, there is an association between certain title characteristics of health care articles and inclusion in the “Altmetric Top 100.” This suggests that an easy-to-understand, informative title may help bridge the gap between academia and social media.

The value of alternative metrics, and especially of the “Altmetric Top 100,” as a measure of scientific impact is controversial. Some of the articles included in these lists represent extraordinary scientific

References (23)

  • J.K. Hoang et al.

    Using social media to share your radiology research: how effective is a blog post?

    J Am Coll Radiol

    (2015)
  • N.S. Trueger et al.

    The altmetric score: a new measure for article-level dissemination and impact

    Ann Emerg Med

    (2015)
  • R.B. Haynes

    More informative titles

    ACP J Club

    (1994)
  • J. Aronson

    When I use a word … Declarative titles

    QJM

    (2010)
  • G. Orizaola et al.

    Free the tweet at scientific conferences

    Science

    (2015)
  • F. Bosch et al.

    From scientific advances to public health action: the crucial role of science dissemination

    J Epidemiol Community Health

    (2015)
  • J. Priem et al.

    Altmetrics: A manifesto, 26 October 2010

    (2010)
  • Public Library of Science (PLOS). Available at http://www.lagotto.io/plos/#static-content-wrap. Accessed September...
  • M. Thelwall et al.

    Do altmetrics work? Twitter and ten other social web services

    PLoS One

    (2013)
  • G. Eysenbach

    Can tweets predict citations? Metrics of social impact based on Twitter and correlation with traditional metrics of scientific impact

    J Med Internet Res

    (2011)
  • J. McGowan et al.

    Informative titles described article content

    J Can Health Libr Assoc

    (2005)
  • Cited by (0)

    Conflict of interest: None.

    Funding: None.

    View full text