StoryCorps uses Google Cloud Speech-to-Text and Natural Language API to p rovide machine-generated transcripts. Transcripts have not been checked for accuracy and may contain inaccuracies, misspellings, or omissions. These transcripts should not be considered a completely accurate expression of the recorded audio content. If you find some errors in your transcript, please know that all transcripts appear with an explanation that they are machine-generated and some small errors should be expected in any transcript in our archive.
Because of the exceptionally large number of interviews newly transcribed in the Online Archive, StoryCorps does not have the capacity to offer edits and corrections at this time. We understand that some transcription errors may be particularly concerning for our participants, particularly if:
- the error involves misspelled names of people, places, or other proper nouns;
- the error results in factual inaccuracies on key information in the interview;
- or (in very rare cases) if the overall quality of the transcript renders it unreadable.
StoryCorps will begin offering to make corrections to transcripts on a limited basis in the coming months. If you consider the error in your transcript to be particularly troubling, like the scenarios described above, reach out to StoryCorps Public Support by submitting a request. A member of StoryCorps’ support team will gather your information and details about your concerns so that we can begin the corrections process in the future. If necessary, our staff can also hide the transcript from public view while you are awaiting these corrections.
If you have concerns about errors to your transcript related to language or accessibility, please reach out to StoryCorps Public Support team to request a transcript.