Measurement
How a newspaper questionnaire became a research instrument
Attachment theory began as an account of infant behaviour. John Bowlby set it out across three volumes published between 1969 and 1980, describing a behavioural system that organises an infant's proximity-seeking toward a caregiver under conditions of threat. Mary Ainsworth and colleagues gave the theory an observational method: the Strange Situation, described in Ainsworth, Blehar, Waters and Wall (1978), a laboratory procedure in which infant behaviour during separation from and reunion with a caregiver was classified as secure, avoidant, or anxious-ambivalent.
The extension to adults arrived nearly a decade later. Cindy Hazan and Phillip Shaver proposed in 1987 that romantic relationships could be described in attachment terms, and measured the proposal with a single forced-choice item: three paragraphs, each describing a way of relating to a partner, from which respondents selected the one that fit best. The item was distributed through a newspaper survey and subsequently administered to a student sample.
What a single item can and cannot support
As an exploratory instrument the item did its job. Its limits, though, were structural rather than incidental. A measure consisting of one item yields no internal-consistency estimate, so the reliability of the score cannot be examined from the data it produces. Forced choice among three descriptions records category membership rather than degree, which means two respondents who chose the same paragraph for very different reasons become indistinguishable. And a survey circulated through a newspaper returns a self-selected readership, from which population-level inference is not available.
The revisions that followed addressed these points in sequence. Kim Bartholomew and Leonard Horowitz (1991) proposed a four-category model generated by crossing two underlying dimensions — a working model of the self and a working model of others — which added a fourth pattern to the three inherited from the infant literature. The move mattered less for the extra category than for the reasoning: categories were now presented as regions of a two-dimensional space rather than as natural kinds.
Kelly Brennan, Catherine Clark and Phillip Shaver (1998) took the step that reorganised the field. They pooled the items from the self-report attachment scales then in circulation, administered the combined pool, and factor-analysed the result. Two dimensions recurred: one indexing anxiety about abandonment and one indexing avoidance of closeness. The Experiences in Close Relationships scale was constructed from that analysis. R. Chris Fraley, Niels Waller and Kelly Brennan (2000) later rebuilt the item set using item response theory, producing a revised version with better measurement precision at the secure end of both dimensions.
Categories are a convention, not a finding
Whether adult attachment is organised into types or into continua is an empirical question, and it has been examined directly. Fraley and Waller (1998) applied taxometric methods and reported structure consistent with dimensions rather than discrete classes. On that evidence, describing a person as belonging to one of four types is a reporting convention imposed on continuous scores, not a result the measures deliver. Popular summaries that present the four labels as categories a person occupies are describing the convention.
Two limits carry through the whole literature. Much of it rests on undergraduate convenience samples recruited in North America and Western Europe, so the generalisability of parameter estimates across ages and cultural settings remains open. And most self-report scales ask about romantic relationships in general rather than about a specific relationship, which produces a summary score whose relationship to any single partnership is itself a research question.
The instrument determines what can be claimed. A finding reported from the 1987 item, the four-category model, and the revised dimensional scales are three different kinds of statement, and the differences do not disappear when the results are summarised under a shared vocabulary.
Sources
- Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of Attachment.
- Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults. Journal of Personality and Social Psychology.
- Bowlby, J. (1969–1980). Attachment and Loss, vols. I–III.
- Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment. In Attachment Theory and Close Relationships.
- Fraley, R. C., & Waller, N. G. (1998). Adult attachment patterns: a test of the typological model.
- Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis of self-report measures of adult attachment. Journal of Personality and Social Psychology.
- Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology.
Corrections to any citation on this page can be sent to support@haldenquarterly.site and are listed on the standards page once made.