RESEARCHING THE REAL WORLD



MAIN MENU

Basics

Orientation Observation In-depth interviews Document analysis and semiology Conversation and discourse analysis Secondary Data Surveys Experiments Ethics Research outcomes
Conclusion

References

Activities

Social Research Glossary

About Researching the Real World

Search

Contact

© Lee Harvey 2012–2019

Page updated 25 January, 2019

Citation reference: Harvey, L., 2012–2019, Researching the Real World, available at qualityresearchinternational.com/methodology
All rights belong to author.


 

A Guide to Methodology

4. In-depth interviews

4.1 Introduction to in-depth interviewing
4.2 Types of in-depth interview
4.3 Methodological approaches to in-depth interviews
4.4 Doing in-depth interviews

4.4.1 Constructing an interview guide
4.4.2 Setting up the interviews
4.4.3 Interviewing
4.4.4 Recording interview data

4.4.4.1 Written recording
4.4.4.2 Mechanical recording
4.4.4.3 Transcription
4.4.4.4 Verbatim recording as naturalistic or objective

4.5 Analysing in-depth interview data
4.6 Summary and conclusion

 

4.4 Doing in-depth interviews

4.4.4.3 Transcription
Transcription is a major bottleneck in qualitative research. Transcription is time consuming and costly. It takes about fours hours to transcribe one hour of spoken material if you are a competent transcriber and have the appropriate tools, such as transscription machine, computer program. Professionals may be faster. In any event, transcription usually slows down your research and increases the cost.

While it is desirable to have full transcripts for analytic purposes, in reality one has to decide whether to expend time and money on making full transcriptions. Would notes made of the content (perhaps with selected quotes) rather than a verbatim transcript be just as valuable? Woud edited transcripts be adequate? Although this raises the question of who does the editing and how?

Increasingly there are technologies that enable the storage of sound bites and written transcriptions in computer-based analytic tool, thus permitting the storage of relevent chunks of conversation without the need to transcribe everything. This, though means working in a different way; listening sequentially, rather than reading, and only transcribing the parts the researcher wants to quote.

When listening, one can also hear tone of voice, which may be relevant and is mainly lost in normal conversation transcription. It does mean that conversational analysis, for example, could be undertaken with without the need for complex transcription and processes (see Section 6.2.5.2). In fields such as discourse analysis and conversation analysis audio recording enables repeated listening, which provides the opportunity to identify subtle and unnoticed aspects of the interaction.

As with any selection of parts form the whole, the researcher must be careful in making selections. Selections of 'relevant' text at an early stage may be based more on preconceptions than a close consideration of the research setting and initial outcomes. Selection may be the result of a premature hypothesising and thus predetermine what is considered as 'useful' evidence.

If selection of snippets, or summary notes are compiled, the researcher is advised to retain the original recordings until the end of the project as there may be a point when the research outcomes suggest a change in direction and it may be therefore necessary to revisit original material to explore alternative interpretations.

The possibility of using voice recognition software to automatically transcribe interviews has been explored but progress on this is somewhat slow and automated transcription has remained elusive.

Transcripts of recorded conversations all provide the opportunity for the researcher to feedback the transcript to the person or persons taking part to check that they can attest to the accuracy of the material. This provides a means to convey to the respondents the raw material of the research that involves them. Some ethical procedures require this (see Section 10).

An issue that arises when the transcript is fed back is that it gives the respondent an opportunity not only to confirm the accuracy but also to request deletion or amendment of sections. When amendment is simply clarification this raises little concern and is usually beneficial. When the respondent want to make changes or deletions that involve substantive changes to what was said, then the researcher faces a dilemma. In some cases the changes may be made for legal or moral reasons, in which case the researcher has little choice but to accommodate them, even if it undermines some of the evidence for the evolving thesis. In other cases, changes may be made because, for whatever other reason, the respondent regrets saying things. In some cases this may be a request to have the whole thing deleted. Does the researcher have to accommodate such requests? Some ethical procedures would say that the researcher just complies with the respondent view. An alternative though, is that given the transcript is not necessarily going to be quoted in any reporting of the research, it could still be retained for analytic purposes.

Another issue that arises from storing transcripts is how much input the respondent has into subsequent interpretation of the transcript material. Some researchers share the intended section of the final draft report, which refers to the transcription, with the respondent, especially if it contains verbatim quotes (albeit anonymised). The respondent sometimes rejects the interpretation and the researcher has to consider carefully whether to amend the draft of the final publication.

Storing transcripts and their subsequent utilisation has to be done with care to ensure whatever level of confidentiality and anonymity the researcher has guaranteed the respondent (See Section 10)

Top

Next 4.4.4.4 Verbatim recording as naturalistic or objective