Learning patterns/Conducting a semi-structured interview

Conducting a semi-structured interview for academic research

problemDo you want to learn more about someone's perspective and experiences? Have you ever conducted an interview only to find you've very little usable information at the end?

solutionThis learning pattern provides simple and practical guidance for conducting a semi-structured interview for academic research.

creator• Mssemantics

discuss

endorse

created on17:43, 23 September 2014 (UTC)

status:in progress

What problem does this solve?[edit]

Interviews are more difficult to conduct than one might think. This learning pattern provides simple and practical guidance for conducting a respectful yet informative semi-structured interview. Semi-structured interviews allow for more flexibility than structured interviews, which keep to a rigorous script. However, a semi-structured interview also allows for a mindful way to collect qualitative data so that you can compare different participants' responses to similar questions and see how themes emerge. This learning pattern is designed with academic researchers in mind.

What is the solution?[edit]

Things to consider[edit]

Note: If you are enrolled in any kind of institution and intend to publish any of your findings, you will need to apply for Institutional Review Board (IRB) approval as soon as possible. You cannot approach participants until you have approval. It's also highly recommended that you seek support and feedback from the Wikimedia Research Committee (RCom).

Step 1[edit]

Before conducting your interview, identify your key questions. These may be research questions such as: How do women perceive the Wikipedia gender gap? When do new editors start to feel comfortable and competent? Your key questions will inform how you structure your interview and the types of questions you ask your interviewee.

Step 2[edit]

Once you've identified your key questions, begin to draft questions that help to answer your key questions without directly asking them. For example, for "How do women perceive the Wikipedia gender gap?" you may ask: Tell me about whom you work with on Wikipedia. How have your experiences been influenced by gender? In what ways is Wikipedia a safe space for women? In what ways is it not a safe space for women? Your questions should circle and close around your key questions. Be careful not to ask yes or no questions. Also, avoid leading questions. Sometimes, using an approach like Critical Incident Technique (CIT) can help elicit the most authentic responses.

Step 3[edit]

Next, draft a consent form to send to your participants. It should include information about you, your project, and how/if you will anonymize the participants' data and responses. If you have IRB approval, the consent form should also include the reference number for your IRB and contact information for your institution.

Step 4[edit]

Recruit participants via snowball sampling through known contacts, individual messages on their user pages, email, and/or mailing lists. Be mindful of Wikipedia's Research Committee guidelines, and do not spam the community. Let your participants know how long the interview will take and ask:

How they prefer to be contacted (e.g., email, IM, Skype, Google Hangout, phone)
Which dates and times are best
Which time zone they are in.

Step 5[edit]

Once you've scheduled your interview, prepare the following:

A digital recorder (with extra batteries), a phone tap, or an app such as WireTap Studio [Note: Some apps such as WireTap Studio record only your participant's voice.]
A hard copy of your consent form and interview questions
A notepad and pen or pencil.

Step 6[edit]

At the beginning of the interview, introduce yourself and your project. Ask if your participant has any questions. If the questions you are asking are controversial, you should let the participant know that if they want to redact or clarify any of their statements during the interview, they can simply say so at anytime during the interview. Then, notify your participant before you begin to record your conversation. As you're asking questions and waiting for answers, try to be quiet. This is the most difficult part. Become comfortable with pauses and give your participant time to think and share. You do not need to ask all of the questions you've prepared. This is why the interview is "semi-structured"; listen carefully to what your participant says, ask clarifying questions, and tailor your questions as you proceed.

Step 7[edit]

At the end of your interview, thank your participant and let him/her/them know when and if you'll be sharing the transcript of your conversation or the findings of your work.

Step 8[edit]

Immediately after your interview, take notes and transfer audio files to a password protected, secure folder--preferably one that isn't in the cloud.

Connectivity issues[edit]

If you are conducting your interview using online services such as Skype or Google Hangouts, you may sometimes encounter audio quality issues (i.e. audio cuts out or becomes choppy, to the point where responses are ambiguous or not comprehensible). It can be helpful to have alternative outlets for interviewing scheduled ahead of time, such as using the phone or a text-based chat like IRC or Facebook chat, in case the audio quality is too poor for interviewing.

When to use[edit]

This learning pattern can be used by anyone who is interested in conducting interviews with Wikipedians, particularly if the desired outpoint mode is academic research.

Endorsements[edit]

Very good outline of how to get started with interviewing from ensuring the necessary ethical considerations are taken for those interested in publication to preparing and actually conducting the interviews. I have learned much from this learning pattern in doing interviews in my own IEG. I JethroBT (talk) 22:18, 14 October 2014 (UTC)[reply]
Great tips for interviewing. The step-by-step instructions make this particularly easy to follow. KHarold (WMF) (talk) 18:32, 21 October 2014 (UTC)[reply]