Hi Varuna,

I think it would really depend on your question(s), your respondents age and the burden on the respondents that you thought was appropriate. There are many self-report measures about and I have developed a brief measure of attachment that is fairly efficient and has good validity and reliability. I have attached it I hope it give you some ideas and is useful. Of course other methods provide the option of depth and use of the respondents immediate reaction. All the best, Terry

Hi Varuna, 

I am not sure we can claim that one measure is more valid than the other, since all three ones you mention have been widely validated in previous research. In accordance with Terry, I would say that your choice depends on the specific intererests and resources available for your research. 

Best, 

Georgios. 

This chapter coauthored with Dr. Judith Solomon provides use and validity information for child attachment assessments. 

The Adult Attachment Interview utilizes a semi-structured interview format, focusing on an individual’s descriptions and evaluation of salient early attachment experiences and the effects of these experiences on current personality and functioning (George et al., 1985/1996). Based on continuous nine-point ratings of the speaker’s inferred childhood experiences and current state of mind, interview transcripts are classified into one of five general categories: secure-autonomous, dismissing, preoccupied, unresolved, and cannot classify (Main & Goldwyn, 1984/1998; Main et al., 2003; Main, Hesse, & Goldwyn, 2008). Classification focus is on the quality of discourse (rather than content), the extent to which communication is collaborative and provides a coherent, free flowing picture of the participant’s experiences and related feelings (Main & Goldwin, 1984/ 1998). Both validity and reliability among coders have been documented (see also Hesse, 2008, and Main et al., 2008, for overviews).

The Strange Situation is a standardized laboratory assessment that consists of eight brief episodes designed to activate infant attachment behavior through an increasingly stressful series of infant–mother separations and reunions (Ainsworth et al., 1978). Individual differences in attachment relationships are coded with respect to the infant’s gaining comfort in the mother’s presence when stressed, and using the mother as a secure base from which to explore. Based on the patterning of the infant’s behavior across all episodes, infant–mother dyads are assigned to one of four major classifications: secure, anxious avoidant, anxious resistant, or disorganized/ disoriented (Ainsworth et al., 1978; Main & Solomon, 1990).

The Attachment Q-set covers a broad range of secure base and exploratory be- havior, affective response, social referencing, and other aspects of social cognition. Accordingly, the Q-set can be construed as an overview of the entire domain of attachment relevant behavior, as currently understood within an ethological/control systems perspective. Each item in the attachment behavior Q-set consists of an item title and more specific descriptive statements printed individually on cards. These items constitute a standard vocabulary for describing 

individual differences within a particular domain of personality, attitudes, or behavior. Q-set items can easily be written to refer to specific behaviors or to behavior in specific contexts. In addition to describing individuals, Q-sets can also be used to operationalize constructs in terms of an array of scores on a specific set of relevant items.Use of a standard multiple-item vocabulary for assessment has many advantages. The development of the Q-set itself demands close examination of theory and reference to extensive clinical or observational data. What is the range of behavior relevant to a particular construct or set of constructs that might be assessed with the Q-set? What contexts are salient in evaluating construct relevant behavior? Construction of a Q-set forces us to consider process- oriented models of behavior and behavioral organization in detail. We are also forced to clarify distinctions and ambiguities that are more easily glossed over in designing rating scales. A well-designed Q- set is thus a powerful tool for transferring theoretical and behavior sophistication to new observers (Waters & Deane, 1995). 

As first author of the Adult Attachment Interview, I can say definitely that the trend in the field of attachment to use the AAI as a measure of child attachment is not scientifically sound.  There is a significant statistical relation between AAI classification and child classification, yes that is true.  The statistic significance depends on how many secure children in the sample.  Secure children have integrated parents, and they are integrated secure on the AAI.  The correspondence weakens for insecure children.  If you have a high risk sample, for example, you are likely to see the U AAI group predominantly associated with D children's attachment.  But parents who are Ds and E on the AAI have U children too.  Solomon & George (2011) reported on this pattern.  Further, Susan Spieker (2011) reported on lack of continuity of the AAI classification over time in high risk groups.  Mothers appeared to monitor what kinds of trauma they described on the AAI.  Trauma rating is the core of the U classification. Especially, therefore, if one is integrating attachment into child clinical intervention, one should be careful of making the assumption of congruence.  The congruence is not 100% and we can develop our intervention plans under false assumptions. The AAI can be used for adult treatment, but assessing parenting is also needed for best practice.

The only adult measures that have a solid conceptual and not just a statistical correspondence to child attachment are caregiving measures.  Many use the Parent Development Interview and we use a form of that interview in our work called the Caregiving Interview.  See the attached Handbook of Attachment chapter for more information.