Who is it for?
For transcribing audio and video files, researchers at Vrije Universiteit Amsterdam can use the services of Transcript Online and Amberscript.
What is it and why do we do this via these services?
Conversion of speech to text (transcription) of audio or video files always involves the processing of personal data. For this reason, one cannot hire third parties to transcribe research data without contractual agreements. Both parties, Transcript Online and Amberscript, have signed a blanket processing agreement with Vrije Universiteit Amsterdam, which means researchers do not have to draw up individual processing agreements to share personal data. In other words, Transcript Online and Amberscript can start transcribing immediately since all agreements around privacy legislation have already been arranged with VU in advance.
Which ways are there in terms of transcription?
Transcription is possible as verbatim, literal and automated transcription. In verbatim transcription, hesitations, faltering and silences are not noted and the transcriber delivers a running text. This is mainly used for informal conversations and is not always useful for scholarly research.
Literal transcription also notes all hesitations, faltering, slips and silences. This gives a better understanding of how the conversation took place. This is the type of transcription used for legal and research purposes.
Automated transcription uses software to convert speech to text within a few minutes, after which researchers can check the transcription for themselves, and if necessary, adjust it. Automated transcription falls in the middle between verbatim and literal transcription. The software will not transcribe short hesitations, but an emphatic "eeuhm", sigh or silence will be noted. Although the software distinguishes between different speakers, automated transcription has difficulty transcribing larger groups or voices that speak at the same time. The transcribed text will be delivered in the speakers' language.
Are one of the options better than the other?
The choice between different the types of transcription depends on the type of research and budget. When transcripts are not used as data but convey an idea or event, verbatim transcription and automated transcription is sufficient. When the transcription becomes the basis for exact data, such as legal, linguistic or psychological research, literal transcription is required. Automated transcription can also be used for this (in some cases), as long as it is closely controlled by the researcher.
What is Amberscript about?
Amberscript offers both automated transcription and transcription by transcribers. For automated transcription, credit per hour can be purchased in advance. For verbatim and literal transcription, a quote needs to be requested.