Paper presented at the conference La Relazione che Cura: Due Scuole, Due Paradigmi a Confronto [The Relationship that Heals: Two Schools, Two Paradigms in Dialogue], Florence, Italy, May, 2008.
We shall not cease from exploration,Concept of the Therapeutic Relationship
and the end of all our exploring
will be to arrive where we started
and know the place for the first time.
-T. S. Eliot, Little Gidding
• Evaluated for its beneficial effects on the client
•Process, not thing
•Emergent property of what happens between client and therapist
•Can’t be attributed to either client or therapist separately
•Both inner experience and observable interaction
• Defined by both intention and effect
A Century of Research on the Therapeutic Relationship: Landmarks on the Journey
A. Era of Participant Observation Research:
Freud et al.: Transference, collaboration, observing ego
=> Zetzel: therapeutic alliance
=> Greenson: transference; real relationship; therapeutic alliance (bond); working alliance (tasks)
2. Interpersonal theory (Adler, Sullivan, Leary)
Social/group theory (Parsons, Bales): Task vs. socio-emotional functions
British object relations school (Klein, Bion, Fairbairn, Winicott)
=> Attachment theory (Bowlby, Ainsworth); mentalization theory (Fonagy, Bateman)
B. Era of Quantitative Research on the Therapeutic Relationship
3. Rogers/Person-Centred therapy
Recording sessions (1940’s)
=> Measuring therapeutic process: therapist response modes (e.g., reflection of feelings)
=> Rogers’ bold statement: The Process Equation (warmth, empathy, genuineness) => client experiencing => client symptom change
=> Measuring facilitative conditions (observer measures; client self-report measures)
=> Research on facilitative conditions and outcome
=> Widespread dissemination of skill training courses beyond evidence base (1960’s – 1970’s)
=> Counter-reaction against Person-centred view of relationship (late 1970’s to 1990):
• Criticisms: poor research, mixed results, overlapping concepts,
=> Reframed in terms of…
4. Modern era of Psychodynamic relationship research:
•Luborsky: Penn Helping Alliance measures (Alliance types: Type 1: Confidence in therapist expertise (passive); Type 2: Active collaboration by client
=> Other alliance measures: Horowitz et al: California Psychotherapy Alliance Scale (CALPAS); Toronto Alliance Relationship Scales; Agnew Relationship Measure
=> Current dominant formulation: Bordin: Tripartite model of Working Alliance
Bond: emotional connection
Goal agreement (what we are working toward)
Task agreement (what we are doing to reach the goals)
=> Horvath & Greenberg, Working Alliance Inventory (WAI))
•Other alliance concepts:
•Client predisposition: Working Capacity
• Client motivation: Commitment (from agreement?
•Therapist Understanding and Involvement (cf. Rogers’ facilitative conditions)
Common Therapeutic Alliance Instruments
1. Working Alliance Inventory (WAI) (36 items; 12 item short form) (Horvath & Greenberg, 1989; Hatcher & Gillaspy, 2006 [12-item revised short form])
•Scale Basis, Points & Time Frame: Frequency; 7 points (revised short form: 5 points)
•Subscales: Bond; Task agreement; Goal agreement;
•Informant: Client, Therapist, Observer
•Non-English Translations: Dutch, Danish, French, Italian
2. California Psychotherapy Alliance Scale (CALPAS; 24 items); (Gaston & Marmar, 1994)
•Scale Basis & Points: Agreement- Disagreement; 6 points
•Subscales: Patient Working Capacity; Patient Commitment; Therapist Understanding and Involvement; Working Strategy Consensus
•Informant: Client, Therapist
•Non-English Translations: French, Portuguese, Italian
3. Penn Helping Alliance Questionnaire-II (Haq-II; 19 items) (Luborsky et al., 1996)
•Scale Basis & Points: Agreement; 7 points
•Subscales: Confidence in Therapist; Active Collaboration
•Informant: Client, Therapist
•Non-English Translations: German, French, Norwegian, Dutch
4. Agnew Relationship Measure (ARM; 28 items; 12- and 5-item short forms) (Agnew-Davies et al., 1998)
•Scale Basis & Points: Agreement, 7 points
•Subscales: Bond, Partnership, Confidence, Openness, Initiative
•Informant: Client, Therapist
5. Positivist view of relationship in Randomized Clinical Trials (RCT):
• Non-specific factor/placebo effect/nuisance variable vs. technique/specific factors:
=> Use of supportive/nondirective relationship control groups in Randomized Clinical Trials
=> Used to establish effectiveness of Cognitive-Behavioral Therapy (CBT)
=> Basis of current “Myth of CBT Superiority” in public health
Results of Quantitative Research on Therapeutic Relationship and Outcome
• Therapeutic Relationship (and its components) is the strongest, most consistent predictor of outcome:
• Therapeutic Alliance: Horvath & Bedi, 2002) meta-analysis of 90 studies: mean r = .21 (weighted by sample size)
• Therapist Empathy: Bohart et al. (2002) meta-analysis of 47 studies: mean r = .32 (weighted)
•Thus, relationship factors appear to account for 4 – 10% of the variance in therapy outcome
• Fairly small effects
• But much larger, more consistent than:
•Technique factors (e.g., interpretation, reflection, two-chair work)
•Type of therapy
•At most: 1% of outcome (Wampold)
Largest predictors of outcome appear to be (Wampold):
1. Client factors (pre-therapy functioning): 40 - 50% of variance
2. Whether you get therapy or not: about 13% of variance
3. Therapist differences (5 - 8% of variance; at least half of this “ Relationship effects)
•What aspects of the relationship are important?
•How does the therapeutic relationship develop?
•How does relationship affect outcome?
•How does relationship interact with other factors?
C. Current Era: Systematic Qualitative Research on Therapeutic Relationship:
6. Emergence of qualitative research on the relationship:
a. Helpful factors research (reviewed by Elliott & James, 1989): Interpersonal, relational factors most frequent helpful factors reported by clients
b. Bachelor (1995): phenomenology of clients’ experience of therapeutic relationship: importance of therapist technique in building relationship
7. Research on Significant Relationship Events
a. Elliott (1985) => Taxonomy of relationship events: Understood, Supported, Close to Therapist, Involved in Therapy => Revised Therapeutic Reactions Scale (RSRS)
b. Timulak (2006) qualitative meta-analysis: Reassurance/Support/Safety (all studies); Feeling Understood (more than half of studies)
c. Bedi, Davis & Williams (2005): Critical Incident Method used to identify factors experienced by clients as contributing to the development of the therapeutic alliance (n=40). The four most common (reported by >30% of clients):
1: Technical activity: Outcome-focused techniques and strategies commonly used by psychotherapists (not including active listening): 72.5%.Implication: These recent studies suggest that clients see the therapeutic relationship somewhat differently from current theory:
• e.g., The therapist taught me a grounding technique.
2: Nonverbal communication: The psychotherapist’s body language (e.g., physical attending skills) and other nonverbal means of communication: 47.5%
• The therapist sat leaning forward in his/her chair.
3: Active listening: The psychotherapist’s active listening skills, including empathic and content-based reflections, summaries, connecting current client material with past client
material, verbal prompts (e.g., “uh-huh”), and silence: 37.5%
• The therapist recognized my feelings and put a name to them when I couldn’t.
4: Choices: The psychotherapist’s presenting and accentuating client choices and conveying that the psychotherapy process is client driven: 32.5%
• The therapist verbally told me I had the choice of whether I would answer anything she/he asked me.
• Downplay collaboration
• See technical and environmental factors (e.g., therapy room) as important contributors to the therapeutic relationship.
•Distinction between technique and relationship may be a false dichotomy
•Need to look closely at therapy interaction as relationship develops.
Two Views Forward:
1. Quantitative: Advanced Psychometric Methods: Rasch Analysis
Process Experiential Therapeutic Alliance Formation task (Elliott et al., 2004) hypothesized the following sequence of alliance formation with Bordin model:
0. Client drops out before session 1Elliott, Fox et al. (2004) used Rasch analyses of clients’ Working Alliance Inventory ratings to measure the “difficulty” of items
1. Client physically present
2. Work on trust/bond
3. Work on therapeutic focus
4. Work on goal agreement
5. Work on task agreement
6. Productive working environment
Suggests alternative alliance formation sequence (with some overlap between successive stages):
(1) Absence of misunderstanding2. Intensive look at client-therapist interaction during initial alliance formation:
(2) Trust/confidence in therapist
(3) Agreement on general goals/tasks
(4) Experiencing therapist presence
(5) Agreement on practical specifics
(6) Experienced benefits (=early outcome)
Example: PE-111-1, Initial Empathy Segment-1 (2:15+)
“George”: 63-year old client with panic and bridge phobia
2:15 into Session 1: follows two previous short segments also devoted to alliance formation
This segment involves establishment of empathic stance with client
C1: … No, it’s this, it’s this, fear, of, of,… getting across bridges of all thingsSummary
T Settles into inquisitive, intent body position, looking at C, slouched slightly forward. Tilts head slightly to left during silence, as if trying enter client’s experience
T: I see, OK. [soft slightly thin voice]
C1.1 And height, oh my, I have terrible fear of height and it…
T: OK Deep head nods, bobbing upper body.
T1: And it- that interferes with your life? Intent, soft voice; gestures with right hand
C2.1: Oh, it does. We used to take great long road trips, and don’t do any of that any more.
C2.2: Funny thing is, I’m not afraid to fly, as long as it’s a jet plane and it’s multi-engined.
Right hand to mouth, then strong head nods I see.
Shifts head forward, slight nods.
T2: I see, OK, that doesn’t bother you Nodding, intent; soft voice
• Therapeutic relationship generally relates to outcome
• Both therapist and client contribute
• Other factors are also important
• Deepening our understanding of the development, maintenance and repair of therapeutic relationship complex, elusive… and important
• Will require:
• Participant observation by therapists
• Quantitative research relating of therapy process, client experience, and outcome
• Systematic qualitative research on client and therapist experience, significant alliance events, and moment-to-moment process