
Join our community of researchers who have used Landmark for over 100,000 transcripts
Get a 30 min free sample. One of our Account Executives will be pleased to guide you through our Platform.
You can cancel your account at any time.
by Diana Carrillo / Landmark Associates Inc. / Work / 06/09/19
Analyzing collected data and being able to see the patterns in order to reach conclusions is an essential part of the research process. But this analysis can be compromised without the proper pooling and filtering of important data from the material that is sourced during the investigation. This source material can come in a variety of formats, such as audio, video,imagery, polling data, and written documents from participants including questionnaires. With text-based versions of all materials, investigators can better identify and prioritize their sources to improve the accuracy of theirfindings.
But, the amount of time, energy, and expertise required to review the audio material for analysis makes this activity practically prohibitive. The industry standard for transcribing one hour’s worth of recording is more than four hours, and it may take even longer if the audio quality is poor and the transcriber is lacking the necessary skills and equipment. If a member of the research team is assigned the task of transcribing, the process of capturing every detail often involves the continuous replaying ofrecorded sections to double and even triple check statements. While this person is busy transcribing content, it also means his or her skills cannot be used elsewhere.
In order to overcome this hurdle, many research teams partner up with professional transcription services to turn their media files into searchablewritten insights that can then be organized and analyzed.
While those on a budget may decide to work with an automatic transcription platform, which relies on speech recognition technology to quickly convert audio to text, at this point there is still a long way to go when it comes to the quality and reliability of the results. Consider that audio quality can vary significantly based on a number of variables, such as the quality of the mic or the recording device being used, it’s distance from the speaker, the presence of accents and industry jargon, and the number of people speaking. While progress has been made, automated systems are still struggling to make sense of it all.
Even where human based transcription is used, the job is often given over to inexperienced freelancers, interns or low-quality outsourcing firms. In this case, the resulting quality and accuracy can actually be much worse than an automated solution.
• Seeing the patterns
• Knowing who said what
• Capturing verbatim versus non-verbatim data
Get a 30 min free sample. One of our Account Executives will be pleased to guide you through our Platform.
You can cancel your account at any time.
Join our community of researchers who have used Landmark for over 100,000 transcripts
Get a 30 min free sample. One of our Account Executives will be pleased to guide you through our Platform.
No credit card required. You can cancel your account at any time.
Contact Us