The Adult Attachment Interview Protocol: THOUGHTS FROM A CODER
Addendum: Maintaining Gentle Control

Helen Deane Dozier

DeaneThese notes were prepared following interviews I listened to where, in one case, the subject rambled collaboratively in an apparent attempt to fill the “45 minutes to an hour-and-a-half” that had been suggested as a time frame, and in another one, where the subject began to digress into troubling and emotional memories when asked for five adjectives and, like the first subject, seemed perfectly willing to talk endlessly to fill whatever space was allowed her to answer each question.

In both cases, the interviewers were going out of their way to allow subjects to digress, but the subjects were especially talkative and the interviews seemed to be getting out of control.  (This topic is addressed by Mary Main’s “AAI Protocol,” which should be studied by interviewers from time to time.)  In my earlier suggestions for new interviewers, I had not properly stressed the importance of keeping gentle but subtle control of the interview, guiding subjects to provide information on some questions, such as the question designed to collect five adjectives, while allowing them to free associate and ramble on others. In my experience, interviewers are more apt to cut a speaker off prematurely than to allow a subject to ramble on questions not designed for rambling.  A response that is cut off when the subject has been asked to give specific support for an adjective or to discuss loss or abuse, for example, can be difficult to score properly.  On the other hand, if a subject wanders unrestrained on all questions, the integrity of the entire interview can be compromised, and can become so lengthy and tiring that the subject and the interviewer run out of energy and interest before all the questions can be asked. 

This “gentle control” is a skill that will improve with practice.   While it’s important to set a relaxed and conversational tone, it’s also important to maintain control of the interview to the point that the questions are asked as “interview questions” so that the subject hears the entire question and understands when you are looking for somewhat precise information.

One key to understanding the importance of asking a question in its entirety is to realize that AAIs are a precise instrument for attachment research, and the wording and suggested follow-up probes have been designed by Mary Main to be asked according to protocol in order to provide information that can be scored and classified correctly.  In determining at what point an especially loquacious subject needs to be urged gently along (indeed there are some who will talk almost endlessly), it may help to know that the coder will be trying to determine such things as whether or not the subject is moving away from the question asked in free association, getting lost in unfruitful dialog, or staying on topic collaboratively but at too much length for an acceptable “turn” in conversation.

Let’s break down the interview and look at some of the questions that are asked to gather information and some that are for more general discussion by the participant.

Question #1:  “ Could you start by helping me get oriented to your early family situation, and where you lived and so on? If you could tell me where you were born, whether you moved around much, what your family did at various times for a living?”

General background, family history:  This question is designed as a warm-up for the subject and as a place for the interviewer to collect demographics and find out who the family members were, who the important caregivers were, and when the caregivers were involved in the subject’s life so that information can then be collected about relationships with these important caregivers in subsequent questions.  It is not a place for the subject to start talking about relationships, and Mary Main’s “AAI Protocol” suggests probes to keep the subject on topic.  She says “no more than two or three minutes at most should be used for this question.”

Questions 3 and 4: 

3. Now I'd like to ask you to choose five adjectives or words that reflect your relationship with your mother starting from as far back as you can remember in early childhood--as early as you can go, but say, age 5 to 12 is fine. I know this may take a bit of time, so go ahead and think for a minute...then I'd like to ask you why you chose them. I'll write each one down as you give them to me.”

The request for five adjectives for each parent and the follow-up probe for examples are at the heart of the AAI, and again, it’s important that you state the questions clearly and completely and according to Mary Main’s protocol. 

Collecting the five adjectives: 

The first part of Question 3 is designed to collect adjectives before moving on to ask for specific examples.  Many subjects pause as they think about this unusual request for words to describe a childhood relationship, and after a comfortable silence, it is fine to repeat the question, possibly rephrasing it a bit, which serves to give the subject more time to consider an answer, and letting them know it’s fine to take some time to think.  The importance of allowing the subject time to think will be better understood by interviewers who have, while practicing, subjected themselves to the interview.  Not all people can immediately come up with words to describe childhood relationships without time to go back in their minds and think about their childhood.  It’s important to help them become comfortable with some moments of quiet time as they consider what words they want to use.

Here is a method employed by many interviewers to keep the subject on topic while coming up with the full set of five adjectives. Say, “Okay, so far I have ‘easy,’ ‘fun,’ and ‘teaching.’ Can you think of two more words or adjectives that describe your relationship with your mother as a young child?”  Note that you are repeating the important parts of the question, which serves to remind the subject of the task and also gives the subject more time to think, plus you are helping the respondent understand that you are counting words as opposed to looking for general discussion.  If the subject starts talking in sentences instead of providing words and phrases, you can says something like, “You mentioned that your mother was controlling.  Does that relate to early childhood and do you want to use that for one of your words?”  If a subject continues to talk conversationally, you could say, “Okay, so what word would you use to sum up what you are describing about this part of your relationship?”  Or, depending on how much information the subject is offering, you might come up with this: “Okay, while you were talking, I wrote down “even-keeled,” “stress-free,” “easy going,” and “no hard feelings” as words you used to describe your relationship with your father as a young child.  That’s four.  Can you think of another word or adjective that describes the relationship from when you were little?  It’s okay to take a little time to think about it – answering this question often takes people a minute or two of time to think about before coming up with words.”

The more you do the AAIs, the more you will learn little tactful tricks to help the subject understand the task at hand.  It’s important to get a full set of five adjectives, if possible, before moving on, even if you have to help the subject extract them from the information.

Of course, some subjects might have trouble understanding “adjectives for relationships,” but Mary Main stresses the nuances of how to ask this question in her “AAI Protocol” in this way:

“Not all participants will be able to think of five adjectives right away. Be sure to make the word relationship clear enough to be heard in this sentence. Some participants do use "relationship" adjectives to describe the par­ent, but some just describe the parent herself --e.g., "pretty"... "efficient manager"--as though they had only been asked to "pick adjectives to describe your mother". These individual differences are of interest only if the participant has heard the phrase, "that reflect your childhood relationship" with your mother. The word should be spoken clearly, but with only slight stress or emphasis.”

Follow-up Probe for Examples

The follow-up question to probe for examples after collecting the full set of adjectives is this:
“Okay, now let me go through some more questions about your description of your childhood relation­ship with your mother. You say your relationships with her was (you used the phrase) “______”.  Are there any memories or incidents that come to mind with respect to (the subject’s word)?”

Followed by this, for each adjective…

“You described your childhood relationship with your mother as (or, “your second adjective was", or "the second word you used was") “_____”.  Can you think of a memory or an incident that would illustrate why you chose “_____” to describe the relationship?

When asked to support each of their five adjectives with specific examples, some subjects will digress as they respond.  Keep in mind that the digression will take on more significance to the coder if the coder can be sure the subject heard the interviewer request a specific example or memory, and that the interviewer reminded the subject their adjective was chosen to describe the early childhood relationship.  But when a question is not asked according to protocol, it’s much harder for the coder to ascertain how much of a subject’s digression might be caused because the subject didn’t hear or understand the question.

The reason behind the precise way the AAI questions are phrased takes on more meaning when you witness the problems caused by not using the precise phrasing, such as when a subject begins talking about an adjective related to the present relationship because it was not made clear that the interviewer was looking for a discussion about the adjectives as they relate to childhood.   

Be sure you go through each adjective and try to get one or two specific examples for each one in a fairly structured way.  If subjects digress off-topic or reverts to more recent memories, that’s fine and there is no need to stop them or correct them as long as the coder can be sure the question was asked properly.  For example, “Can you think of a specific time in early childhood when something happened that made you feel you had a ‘loving’ relationship with your mother?

I cannot overstate the importance of probing for specific examples for each adjective according to Mary Main’s AAI Protocol.  An AAI with no probe for examples of the adjectives would need to be conducted a second time.

 Question 7Separations.  The question regarding the first memory of being separated from parents often induces a question for clarification (subjects want to know what you’re looking for), and it can be enlightening to see what subjects come up with on their own with some general guidance such as, “Just whatever stands out in your mind when you think of any early separation from your parents.”  Some subjects may mention a first day of school, and others may mention their parents’ divorce – very different “separations” and the question is open-ended so that the subjects can discuss what the word “separation” evokes in their minds.
If a subject can’t seem to come up with anything, the interviewer can suggest it could be just anything, big or small – first day of school, going to a baby sitter’s, getting separated in a store, going off to camp, a sleepover at a friend’s, a parent leaving on a trip—whatever they remember when they think of separations. 

Question 9:  Threatening behavior, abuse.  Whenever a subject brings up strict discipline, beatings or spankings, it’s important to the coder to find out such things as what is meant by “strict” and whether any beatings ever left marks and how frightening they were to the subject.  If a casual dialog can be established regarding a specific incident (providing the subject is comfortable discussing it), it can offer important evidence for identifying and scoring abuse.  Read the excellent suggestions on how to probe for Question 9 in Mary Main’s “AAI Protocol.”

Question 13: Loss.  After finding out what losses a subject has experienced, the interviewer must decide which losses are important enough to probe according to protocol.  For instance, a subject who says her grandparents who died lived overseas and she had only seen them twice, that loss would not be as important as an aunt who had helped raise the subject, and it’s okay to move on without letting them digress too far about the overseas grandmother. 

Regarding important losses, the subject should be allowed to answer questions fully without interrupting (within reason). Important:  In probing loss and abuse, the interviewer should allow plenty of time after an unfinished sentence to see if the subject is going to finish it.   

This Addendum to my "Thoughts froma Coder" are designed to help interviewers produce transcripts that can be scored correctly. This paper is a work in progress and will be updated and expanded on from time to time.

Back to Clients Page

Back to AAI Coding Home Page