A Multi-Dimensional Measure of Reflective Function for Rating Psychotherapy Sessions

Lindsay L. Hill a , Kenneth N. Levy, Ph.D. a,b , & Kevin B. Meehan, M.A. c
a Department of Psychology, Pennsylvania State University; b Department of Psychiatry, Joan and Sanford I. Weill Medical College, Cornell University; c Clinical Psychology Doctoral Program, The Graduate School and University Center, City University of New York

Introduction The nature of how people understand their minds and the minds of others has long been of tremendous interest to philosophers, neuroscientists, psychologists, and theologians. The psychologist Peter Fonagy drawing on the works of philosophers like Dennett and Brentano has recently articulated a theory of mind which he called reflective functioning. According to Fonagy, reflective function is a persons ability to reflect on what is going on in his/her own mind and to accurately understand wishes, intentions, and motivations underlying their behavior and the behavior of others. Fonagy proposed that high reflective functioning would relate to a range of positive outcomes from parenting to relationship functioning to resiliency during stressful situations. The Current Rating System In order to test his theory, he developed a scale to assess various aspects of reflective function (e.g., understanding a developmental perspective, understanding the opaqueness of mental states, and the possible defensive nature of mental states). Consistent with his theory, Fonagy found that caregiver reflective function predicted their babys functioning in a laboratory procedure. In a later study, Fonagy and colleagues found that high reflective function acted as a buffer against the negative affects of traumatic experiences. Subsequent clinical research using this scale has found that reflective function can improve during the course of psychotherapy even for severely disturbed outpatients with personality disorders (Levy et. al., in press). Levy et. al. (2005) also found that reflective function was related to a number of neurocogntive mechanisms including attentional capacities, executive functioning, and impulsivity. These neurocognitive capacities are central to the difficulties experienced by people with personality disorders. Advantages Fonagy created a scale that is both conceptually rich and practical, without which research in the area of mentalization may not have been possible. The scale consists of an 11 point likert scale ranging from -1 (rejection of reflection) to 9 (highly sophisticated reflection). To increase precision and understanding, Fonagys manual offers an in-depth explanation of the theory behind reflective function and explanations what each level of reflection may look like to the coder. Disadvantages As beneficial as the reflective function concept has been for understanding a wide range of outcomes, research in this area has been hampered by problems with the current scale. The scale was developed to be used with the Adult Attachment Interview (AAI; Main, 1993), which is audio taped and currently has to be transcribed in order to use the coding system. Transcription of these interviews typically takes six to eight hours per interview. The scales dependence on this interview limits its applicability in psychotherapy process research and other areas of clinical research. Additionally, despite the richness of this scale, it is cumbersome and only provides a single score. The single score limits our understanding of the complexity of reflective function and our capacity to examine the psychometrics of the measure (e.g., the factor structure). In order to better understand the validity of this promising construct a more applicably versatile, differentiated measure of reflective function needs to be developed.

Scale Development The goal of this project was to develop a multi-item rating scale of reflective function pertinent to psychotherapy process research. The multi-item measure was developed through an iterative construct validity approach beginning with the identification of relevant domains. Using Fonagys manual for scoring reflective function and other relevant research in the area of mental representations, we identified several key domains of reflective function (See Table 1). Similarly, several subcategories were also identified within each specific domain. Once the domains and subcategories were identified, we then generated potential items that could be included in the measure. Care was taken to ensure that each domain and subcategory is represented within the measure. (Table 1 shows a break-down of the
Table 1: Reflective Function Domains
SPECIFIC DOMAINS
I. Awareness of the nature of mental states.
1. Recognizing the opaqueness of mental states. 2. Mental states as susceptible to disguise. 3. Recognition of the limitations on insight. 4. Mental states tied to expressions of appropriate normative judgments. 5. Awareness of the defensive nature of certain mental states.

# Items
10
2 2 2 2 2

II. The explicit effort to tease out mental states underlying behavior.
1. Accurate attributions of mental states of others. 2. Envisioning the possibility that feelings concerning a situation may be unrelated to observable aspects of it. 3. Recognition of diverse perspectives. 4. Taking into account ones own mental state in interpreting others behavior. 5. Evaluating mental states from the point of view of their impact on the behavior of self and/or others. 6. Taking into account how others perceive one. 7. A freshness of recall and thinking about mental states.

10
2 1 1 2 1 2 1

number of items in our rating scale that are relevant to each category.) Items were also alternated in positive or negative wording to prevent bias. Once all potential items were established, items were reviewed by the authors for wording, clarity, relevance, readability, etc. We were concerned about the clarity of the items. Reflective function is mostly a psychodynamic/object relations construct. One of our intentions for this rating questionnaire is for it to be . easily used by psychotherapy researchers and trained graduate students. As a result, examples and/or mini-explanations were added to the more confusing items to enhance clarity. After many iterations of this process, a 53 item scale was developed. Each item on the scale is rated from 1 to 5. Each rater is instructed as follows: the statements below describe a number of ways in which patients may perceive, understand, and reflect upon the mental states of the self and others. Please rate the following items on the extent to which they are true of you in your work with your patient, where 1 = not true at all, 3 = somewhat true, and 5 = very true. (See Table 2 for example items). Rater Training Currently two raters are being trained in coding reflective function using Fonagys manual for scoring Adult Attachment Interviews. Once these coders are familiar with this system of scoring reflection function, the new scale items will be applied to already existing AAIs (from Levy et.al. in press; 2005) in order to establish inter-rater reliability and criterion validity. Conclusion The development and validation this measure promises to expand our understanding of this complex construct. A multi-item measure allows for more psychometrically sophisticated, and timely, assessments of reflective functioning. Researchers and therapists will have the opportunity to examine the dimensional components of RF as well as apply this measure to rating changes in psychotherapy patients.
Table 2: Examples of Questions
Domain Sub-Category Question
The patient acknowledges that one can be internally experiencing emotions different from what is being overtly displayed. (i.e. I felt sad but didnt want to show that to her.). The patient often acknowledges how his/her perception of an event might have been distorted by what he/she was feeling or thinking at the time. Awareness of the nature Mental states are of mental states susceptible to disguise. Explicit effort to tease out mental states underlying behavior. Recognizing developmental aspects of mental states Taking into account ones own mental state in interpreting others behavior.

III. Recognizing developmental aspects of mental states.

1. Taking an intergenerational perspective / making links across generations. 2. Taking a developmental perspective. 3. Revising thoughts and feelings about childhood in light of understanding gained since childhood. 4. Envisioning changes of mental states between past and present, and present and future. 5. Envisioning transactional processes between parent and child. 6. Understanding factors which developmentally determine affect regulation. 7. Awareness of family dynamics.

9
1 1 1 2 1 2 1

IV. Mental states in relation to the interviewer / therapist.

1. Acknowledging the separateness of minds. 2. Not assuming knowledge. 3. Emotional attunement.

4
1 2 1

V. Rejection of RF
1. Becomes hostile/defensives when asked to reflect. 2. Perceives questions about motivations as an attack. 3. Avoids opportunities to reflect.

3
1 1 1

VI. Unintegrated, bizarre, or inappropriate RF

1. Applies bizarre explanations to the cause of his/her behavior that are beyond the bounds of common-sense. 2. Employs literal, concrete thinking

3
2 1

VII. Disavowal of RF
1. Normalizes negative experiences in terms of reference to global/cultural trends. 2. Makes generalizations and avoid specificity regarding behaviors of self and/or others. 3. Often side-steps questions about what may have caused a particular feeling/thought.

3
1 1 1

VIII. Distorting or self-serving RF

1. Often expresses certainty about what other people are thinking/feeling. 2. Over-estimates the extent to which he/she may have been the cause of the behaviors of others. 3. Often makes assumptions that are self-aggrandizing. 4. Often assumes the therapist/interviewer is unaffected by what he/she hears.

4
1 1 1 1

Revising thoughts and The patient seems to have revised thoughts about his/her history feelings about childhood based on insights gained through development. (i.e. "Looking back in light of understanding at it now as an adult I can see that") gained since childhood. Becomes hostile/defensive when asked to reflect. Often assumes therapist/interviewer is unaffected by what he/she hears. Not assuming knowledge. Sophisticated/complex understanding of the thoughts and feelings of others. The patient tends to become overtly defensive when asked to reflect on the feelings underlying his/her actions (ex. How the hell should I know why I did that?!). The patient, becoming upset while recounting a painful event, often assumes that the therapist is unaffected by what he/she hears. The patient makes spontaneous efforts to clarify confusing aspects of his/her narrative in session. The patient maintains complex/sophisticated understanding of the thoughts and feelings of others even when dealing with emotionally charged subject matter (such as the death of a loved one).

Rejection of RF

IX. Nave or simplistic RF

1. Attribution of others' behavior are often marked by "splitting". 2. Attributions are superficial, banal, or uses excess clichs.

2
1 1

Distorting or selfserving RF Mental states in relation to the therapist Marked/Sophisticated RF

X. Overly analytical or hyperactive RF

1. Comes across as psychologically minded but his/her language is irrelevant to the topic at hand. 2. Use of jargon or diagnostic language as the sole explanation offered for behavioral causes. 3. Excessively long explanations.

3
1 1 1

XI. Marked/Sophisticated RF
1. Attributions are placed in a causal sequence. 2. Sophisticated/complex understanding of the thoughts and feelings of others. Total:

2
1 1 53

Acknowledgements: This research is supported by a grant from the Presidents Fund for Research (PI: Lindsay L. Hill). Address correspondence to Lindsay L. Hill at llh167@psu.edu and Kenneth N. Levy, Ph.D. at klevy@psu.edu