How to conduct interviews online: What I wish I knew before I started

My PhD research has involved developing an online learning tool that can be used to embed discussion into course content. To evaluate this I have completed two design-based research interventions. As part of the process of evaluating these interventions, one of my methods have been interviews. Due to Covid-19 these interviews have been online. After some tweaking, I have managed to develop a process that is working well. In fact, it has allowed me to operate much more effectively that if I were doing face-to-face interviews.

In this article I share my refined process. It may well be useful if you are looking to develop your own online interview protocol. At the end of the article I share some steps that will help you get started on your interviews. At the very least, it will act as a template I can use for my own future research ambitions.

Recruiting Participants

I am following an opportunistic sampling strategy. My participants are my students; given that most of them are young adults, I like to ensure my interactions with them are as informal as possible. What’s key, is to make sure that your participants are relaxed, open and willing to share their experiences.

Rather than send a formal invite email, I normally send an instant message over Microsoft Teams asking them to take part in my research. Should they agree, most of them do, I send a Microsoft Teams meeting request, attaching the below message to the invite:

Hello Participant,

I very much appreciate you volunteering to help with my research.

The interview should take no more than 30 minutes and is nothing more than an informal chat.

Before we meet, you may want to click this link to read the participant information - it contains a little more detail on what will be involved. Should you have any questions or concerns, I will go over this document with you at the start of our interview.

Other than reading the participant information, there is no preparation required.

Thanks for participating,

Joe

Preparing for the Interview

Prepare a Protocol

I plan beforehand what the interview process is going to look like. I pay specific attention to the questions I am going to ask – ensuring they address my research questions. While I don’t rigidly stick to these questions it does ensure that I have some degree of focus and repeatability across my interview process.

Build Ethics into the Protocol

I build into my interview protocol a process of walking the participant through the participant information sheet and ensuring that they give informed consent to take part in the research. I have been using dochub.com to gain signed informed consent entirely online.

Develop a Participant Information Classification Database.

Decide on the attributes that you want to log for each participant (e.g., age, name, gender). Then set up a database to log this information - something as simple as an Excel spreadsheet can work well. Personally, I store all of my interview data in NVivo. I have been capturing the following attributes for each participant:

• Interview Date
• Nationality
• Interview Length (this is useful if you are paying for transcripts)
• Gender
• Email
• Date of Birth
• Course

Conducting Online Interviews

It's now time to conduct your interview!

Set up multiple recording methods

When conducting online interviews, I use the meeting recording feature in Microsoft Teams, and I also use the free audio recording software Quick Time as a secondary recording method.

Start recording discretely

Nothing kills a relaxed mood more than saying, “right, I am now starting to record this interview”. As part of my protocol, when I go over the participant information, I ensure that the participant understands that the interview is recorded. However, I always discretely start recording - this ensures the atmosphere can remain as relaxed as possible.

Keep the interview on track

I’ve found conducting interviews a subtle balance. I need to be outcome focused and ensure I gather enough data to answer my research questions. However, at the same time, I want to let the conversation flow. I have found keeping list of question in front of me and addressing them when they naturally come up in conversation helpful. This is opposed, to sequentially asking them, which can be somewhat forced and kill the flow.

Increase the frequency of your interviews slowly

Schedule your initial interviews several days apart. Treat them as opportunity to test out and refine your interview protocol. Once you feel confident with your process you can then increase the interview frequency.

After each interview

As soon as the interview has finished I liked to email the interviewee a signed copy of their informed consent.

Dear [[participant name]],

Thanks for participating in my research.

I have attached a signed copy of your informed consent.

As I mentioned our conversation will remain secure and anonymous.

Thanks again for participating,

Joe

Starting early data analysis

Once an interview is conducted it’s tempting to rush forward with the next one. However, I have found that this approach does not lend itself to refining the interview process. As such, I have now started to perform some early-stage analysis as I go along. This is by no means a sophisticated and exhaustive analysis – this will come later. Rather, it is a means for me to start to develop early ideas which can then feed into future interviews.

My process, after some refinement, seems to work well:

1. To transcribe the interviews, I have found the best technique for me is using an automated transcription tool. I use otter.ai to do a very quick transcription of the audio data.
2. Next, I re-listen to the interview - but I slow it down to about 50% speed. As it plays, I manually correct any errors in the automated transcription. This has the bonus of re-familiarising me with the data.
3. I then import the transcript into NVivo, re-read the data and start attaching notes and making initial codes. While I use NVivo, you could use the comments feature in Microsoft Word, or manually annotated your work win a pen!

I have found the above process to be a low friction method of beginning my qualitative analysis process. This is key, as with all of this it is very easy to be paralysed by the process and take some small pieces of action.

Conclusion

I hope this article has made you think a little bit more about conducting your own interviews. What’s key is to take some actions. The beauty of qualitative analysis is you can tweak things as you go along.

Tools Used

• NVivo, for coding the transcripts
• Otter.ai, for automating transcriptions
• dochub.com, for sending out informed consent

Take Action Today

Start working on your interview protocol

• Set up an informed consent delivery platform
• Draft a invite email (use my template above)
• Consider what attributes you want to gather from each participant
• Test your recording methods
• Invite your first participant
• Transcribe your first interview, and think about some initial codes.