29.10.12 12:43 1 Assessing the impact of online academic videos Mike Thelwall, Kayvan Kousha Statistical Cybermetrics Research Group, School of Technology, University of Wolverhampton, Wulfruna Street, Wolverhampton WV1 1LY, UK. E-mail: m.thelwall@wlv.ac.uk, k.kousha@wlv.ac.uk Tel: +44 1902 321470, +44 1902 321000 Fax: +44 1902 321478 Katrin Weller Dept. of Information Science, Heinrich-Heine-University, Düsseldorf, Germany E-mail: weller@uni-duesseldorf.de Tel: +49 (0) 211 8110803 Cornelius Puschmann Dept. for English Language and Linguistics, Heinrich-Heine-University, Düsseldorf, Germany E-mail: cornelius.puschmann@uni-duesseldorf.de Tel: +49 (0) 211 8115927 The web provides scholars with mechanisms to publish new types of outputs, including videos. Little is known about which scholarly videos are successful, however, and whether their impact can be measured to give appropriate credit to their creators. This article examines online academic videos to discover which types are popular and whether view counts could be used to judge their value. The study uses a content analysis of YouTube videos tweeted by academics: one random sample and one popular sample. The results show that the most popular videos produced by identifiable academics are those aimed at a general audience and which are edited rather than a simple format. It seems that the audience for typical academic videos is so small that video production in most cases cannot be justified in terms of viewer numbers alone. For the typical scholar, videos should be produced for niche audiences to support other activities rather than as an end in themselves. For dissemination videos, in contrast, view counts can be used as a good indicator of failure or popularity, although translating popularity into impact is not straightforward. Introduction Some academics are taking advantage of the web to publish pictures, videos, presentations, and other things that previous generations may not have produced or may have shared with a limited circle of students or colleagues. In theory, these extra resources should be valuable to the scientific enterprise and attempts have been taken to assess the impact of non-standard academic objects such as presentations, blog mentions, syllabuses (Kousha, Thelwall, & Rezaie, 2010b) or scientific images (Kousha, Thelwall, & Rezaie, 2010a). Moreover, innovative outputs may help to communicate and progress science in new ways. Nevertheless, researchers devoting time to non-traditional content production may not be appropriately recognised or rewarded because their products may be excluded from standard evaluation information (e.g., citations, prizes and grants). It is therefore important to assess the value of new types of publishing and develop mechanisms to evaluate individual contributions. This article focuses on online academic videos in YouTube, about which there is surprisingly little research. Video production is increasingly simple with modern digital video cameras or mobile phones and editing suites. The results can be easily shared on popular sites like YouTube. One example of an innovative video-based scholarly initiative is the Journal of