Guidelines for using and describing Altmetric data

These guidelines are for anyone who uses Altmetric data in a research study, publication, presentation, LibGuide, video or other material. They are especially important for researchers with access through the Scientometric Researcher Access to Data (SRAD) program.

This is a living document. If the definitions used in the altmetrics and scientometrics communities change, we will update it. We are also open to suggestions from the research community, so please contact us.

Please credit Altmetric in the acknowledgements section of your paper, presentation or other output, for example:

The authors wish to thank Altmetric for providing this study’s data free of charge for research purposes.

If you can cite data in your reference list, you can use:

Altmetric. (Year of dataset). Altmetric data. Unpublished raw data.

Journals and disciplines have different standards for data citation, so please adjust the citation if needed.

Some Altmetric data is proprietary, so we review requests to share data case by case. Please get written approval before you share Altmetric data for any purpose other than private peer review. To make a request, contact [email protected].

Everyone who wants to use the data must apply for their own access through SRAD, including collaborators and students. If you give others access to the data without an approved application, your own access may be withdrawn.

It is not possible to search Altmetric for news stories by keyword or topic and then see which research outputs they mention. Our team can run these searches for you, but they take time and resources, so we charge a fee. Please contact us for a quote.

Company name. Please use “Altmetric” or “Altmetric.com”. Please do not use “Altmetrics” or “Altmetrics.com”.

Logos. If you use a logo, please use the most recent version, available here.

Altmetric Attention Score. This is our weighted indicator of the volume and nature of the attention that research has received online. Please use the full name, “Altmetric Attention Score”, and avoid “the Altmetric score”, “the altmetric score”, “the Altmetrics score” and “Altmetric count(s)”.

General descriptions. Altmetrics data is not the same as data from Altmetric. We are one provider of altmetrics data, alongside others such as PlumX, Impactstory and Crossref Event Data. To avoid confusion:

• Do not capitalise “altmetrics” and do not use “altmetric counts”.

• Do not use “social media metrics” instead of “altmetrics”, unless you are specifically studying data from social media sites.

• To describe our data, use “Altmetric data”, “Altmetric.com data” or “[data type] sourced from Altmetric.com”.

• To describe altmetrics in general, use “altmetrics” in the same way you would use “citations”, for example: “We can understand the attention paid to research by examining altmetrics sourced from Altmetric.com.”

Altmetric data and the Altmetric Attention Score are indicators of attention, and in some cases public engagement. They are not measures of research quality or impact. In rare cases, the underlying data can point to potential downstream impact, for example citations to research in policy documents or patents. Please avoid describing Altmetric data in general as evidence of quality or impact.

Please also make clear that social media is only one of the types of data we collect. Others include news, blogs, policy documents and peer review. For more on how we define attention, see Cassidy Sugimoto’s blog post, “Attention is Not Impact” (2015).

We support the use of Altmetric data to develop journal-level indicators, as long as they follow responsible metrics practices and are used as part of a “basket of metrics”. We recommend indicators that:

• Account for skew, for example by using coverage percentiles (such as “% of articles in a journal with an Attention Score”), geometric means, or medians (allowed, but discouraged).

• Address bias related to subject area, location and other variables, for example by normalising by subject area.

• Look beyond individual journals, towards network analysis, communications theory and topic-based analysis.

• Are transparent and follow the principles of the Leiden Manifesto.

Please choose your denominator carefully. Altmetric data usually includes only the articles that have been mentioned in a source we track, not all articles published by a journal. We recommend using an independent source, such as Dimensions, to find a reliable denominator.

Please avoid simple sums or averages of the Altmetric Attention Score or of source-specific counts (for example, the number of X posts). These do not account for the many outliers in any sample.