Since early in my career as a research professional, I have been fascinated by the process of writing qualitative research reports out of transcripts. Comparing many responses, distilling information, and ensuring valid conclusions are no easy tasks. This is the reason why I am curious about the evolution of qualitative analysis. For over 10 years or longer, I have been noticing that some companies or groups bypass the transcript production and analysis step. Instead, these groups engage in a dialectic or harmonization analysis, memory, and notes. Though this approach is subject to bias and memory lapses, this approach is defensible when the objectives are very limited and focused results can be tracked manually as the research progresses.
Another approach to qualitative data is to rely more fully on personal notes. The question I set to answer for this piece was whether it is acceptable for a moderator to rely on their own notes captured during interviews given that there is an alternative to work out of transcripts to categorize and summarize information. Let’s face this, transcripts are time consuming and expensive. What does the academic research literature say about this?
Answering this question is important to both clients and research practitioners. Both audiences need to ensure that a thorough analysis is performed and that reports contain the least amount of bias. In the drive toward efficiency, it might be tempting for practitioners to come to rely progressively less on transcripts. But is the method of relying on notes incorporating biases or leaving important information behind?
Before tackling the question, let us define what we understand by transcripts and notes (see Table 1).
Table 1. Definitions of Transcripts and Notes for this Article
| Term | Definition |
| Transcripts | These vary in the amount of detail provided. Verbatim transcripts are word-for-word representations of everything that was said in interviews. Summarized transcripts, on the other hand, leave out some of the non-essential details, like the more prosaic interjections, and verbal mannerisms. The latter are commonly accepted just as thorough as verbatim transcripts. |
| Notes | Notes are information bits compiled through keywords or short sentences. Note takers may use their own shorthand codes for later interpretation during analysis. Notes can be taken by the moderator or by an analyst or scribe, in real time or by listening to audio or video recordings. Notes can include some observations, but in qualitative research they largely coincide with quotes from respondents. |
Now that we have defined transcripts and notes, let’s remind ourselves of our main question. Is it OK to bypass transcripts and rely on notes? A literature search produced a short list of articles, and five of them addressed this question more closely. These articles are summarized in Table 2.
Table 2. Five Academic Papers on the Use of Transcripts and Notes in Qualitative Research.
| Reference | Summary |
| Eaton and Ohan (2019) | These authors analyzed the same interviews with two datasets of qualitative data, each obtained via different capture methods: transcriptions and scribe notes. Scribing produced less data but, on the balance, was superior economically. The number and content themes coincided between the two methods. The scriber role was separate from the interviewer role. |
| Halcolm and Davidson (2006). | In the context of healthcare research, the authors recommend that the choice to use transcripts or other methods of qualitative data capture be consistent with the goals and methodological underpinnings of the research. Thus, if a person needs to conduct a mixed-methods analysis, then detailed transcriptions should be produced. |
| Pope, Ziebland and Mays (2000) | In an article about how to analyze qualitative data, these collaborators implicitly accept the use of notes, fieldnotes, diaries, chronological accounts as well as transcripts. However, their goal was to describe a standard method of qualitative analysis, not to judge the superiority of any data capture method. Notes are admissible because they have some level of value, not necessarily because they are necessarily thorough accounts of interviews. |
| Rutakumwa et al. (2020) | This group compared data produced by processing audio recordings into transcripts and by notes from memory immediately after each interview. The quality of the data was comparable in terms of the detail captured. The topic of the study was sexual partnerships and sexual risk behavior among Ugandan women. |
| Tessier (2012) | In this article, the author argues for admission of transcripts (complete or incomplete), recordings themselves, field notes, and other materials into the construction of a qualitative dataset. Each method brings its own strength. Here notes are not regarded as a substitution for transcripts, but as a complement. |
The review of the five articles above leads us to think that notes are an admissible form of data capture in qualitative research. In fact, there are times where they provide added benefits when other forms of data are also included in the dataset. However, there are two important caveats:
- If notes or summary accounts are produced from memory an interviewer needs to set time aside for their written recount the same day of the interview.
- If written notes are made contemporaneously, then a scribe or notetaker other than the moderator should produce the notes.
It could be argued that relying on memory for notes may not be advisable when the moderator conducts another interview before they are able to produce their notes. There would be cognitive overload and some information might be attributed to the wrong participant. Because it is common to have moderators interview respondents non-stop, back-to-back, we do not recommend a memory approach for practitioners unless there is ample time between interviews.
Moderators can become quite good at taking their own notes, and such notes may help them recall additional information if they happen to be the main analyst in their projects. If a study has many interviews, and the moderator relies on notes, then there is risk of information loss due to the time lag between interview and analysis. A better approach might be for the moderator to take notes after the interviews using recordings if the moderator happens to be their own research analyst.
The implications of this short literature review for the qualitative research profession is that notes are admissible as main source of qualitative data for analysis under one of these three conditions: (1) that they are produced by third-person either during interviews or after interviews using recordings, (2) that they are produced by the moderator after the interviews if they have ample time to work from memory the same day of the interview but before a future interview, or (3) that they are produced by the moderator by listening to interview recordings. In short, we do not advise relying on contemporaneous notes by the moderator as source of data due to the potential of data loss. Regarding this matter, Eaton and Ohan (2019) state:
“Interviewer-produced field notes, however, are fundamentally disadvantaged given that the interviewer needs to simultaneously engage in the process of the interview and in making field notes. Interviewer note-taking potentially disrupts the interview, compromising both the notes and the interview exchange… Moreover, it may result in “thin” datasets, replete with missing data and an underrepresentation of participant voices…” (p587).
Buyers of qualitative research services might want to ask about the specific approach to qualitative analysis that would be conducted for their project. Absent transcripts, notes are acceptable as long as dedicated resources (i.e., a scribe, a notetaker, and time aside by the moderator) are employed. What everyone wants is a report that gets as close as possible to the truth of the matter under study. Reliance on memory and divided cognitive resources, such as contemporaneous moderator notetaking, can result in “thin databases,” as Eaton and Ohan (2019) argue.
References
Eaton, K., Stritzke, W. G., & Ohan, J. L. (2019). Using Scribes in Qualitative Research as an Alternative to Transcription.The Qualitative Report, 24(3), 586-605. Retrieved from https://nsuworks.nova.edu/tqr/vol24/iss3/12
Halcomb, E. J., & Davidson, P. M. (2006). Is verbatim transcription of interview data always necessary?. Applied nursing research, 19(1), 38-42.
Pope, C., Ziebland, S., & Mays, N. (2000). Analysing qualitative data. BMJ, 320(7227), 114-116.
Rutakumwa, R., Mugisha, J. O., Bernays, S., Kabunga, E., Tumwekwase, G., Mbonye, M., & Seeley, J. (2020). Conducting in-depth interviews with and without voice recorders: a comparative analysis. Qualitative Research, 20(5), 565-581.
Tessier, S. (2012). From field notes, to transcripts, to tape recordings: evolution or combination?. International journal of qualitative methods, 11(4), 446-460.
