VIEWS: 48 PAGES: 20 POSTED ON: 2/21/2010
An exploratory test of two solution focused therapy fidelity instruments Peter Lehmann, PhD Randall Basham, PhD Emily Spence, PhD University of Texas at Arlington Purpose • Contribute towards the movement of SFT as an evidence-based model • Test the reliability and validity of two SFT fidelity instruments Solution Focused Therapy • Brief, goal-directed therapeutic model (de Shazar, 1982, 1985, 1988) • Strengths based • Used for individual, couples, family & group treatment • Our application of SFT is delivered mostly by Masters-level student interns in a university- based counseling center primarily serving youth and families referred by a school district Intervention Fidelity • The degree to which the delivery of an intervention adheres to the original model (Mowbray, Holter, Teague & Bybee, 2003) • When combined with outcome evaluation, fidelity assessment becomes an instrumental part of determining whether or not an intervention is effective. Research Design • Time series design with assessments made during the 3rd session and every 3 sessions thereafter. • Instruments: – Solution Focused Fidelity Instrument Therapist (SFT- FIT) – Solution Focused Fidelity Instrument Consumer (SFT- FIC) • Assessed reliability (internal consistency) and factorial validity with 66 clients and 12 student therapists at the UTA Community Services Center Instruments • SFT FIC and FIT include 18 mirror items worded for the respondent • Items are measured on a scale of 1 (Not at all) to 7 (Very clearly and specifically) • The first 17 items assess fidelity and the last item refers to consistency with previous sessions • The FIT is in it’s 8th version and the FIC in it’s 3rd version • Face and content validity was promoted via the review and revision of the instruments collaboratively with ~20 SFT experts on an international SFT listserv Sample • N=169 SFT-FIT’s and FIC’s • Valid Cases: N=160 (excludes missing data) • 70 FIT’s completed by 11 student interns and 1 therapist • 99 FIC’s completed by 66 clients Respondent Demographics • Therapists • Consumers – 11 Masters Social Work – 32 Clients seen individually interns – 34 seen in family sessions – 1 Licensed Therapist • Age – 3 Males, 9 Females – 44.4% 17 years or under – Age – 31.2% 18-35 years • 7 18-35 years old – 24.6% over 35 years • 5 over 35 years – Ethnicity • Ethnicity • 6 Caucasian-Anglo – 51.2% African-decent • 3 African-decent – 27.9% Hispanic • 1 Asian – 14% Caucasian/Anglo – 7% Asian Results: Reliability • Overall internal consistency for combined FIT and FIC instruments was alpha=.89 • Overall scale mean= 96.78 (SD=17.36; possible range is 18-126) • First administration alpha=.88 • Differences observed between the FIT and FIC’s Internal Consistency • FIT • FIC – Alpha=.76 – Alpha=.93 – Scale mean=91.50 – Scale Mean=100.79 – Standard Dev=12.49 – Standard Dev=19.42 – Mean of items: 5.08 – Mean of items:5.60 Factorial Validity • Principal component analysis used to determine the least number of factors that can account for common variance among items • Factors extracted using Varimax rotation with Kaiser Normalization • Eigenvalues set at 1.00 • KMO measure of sampling adequacy: .90 for FIC & .721 for FIT • Bartlett's Test of Sphericity: p=.000 for both; reject null hypothesis of identity matrix Factor Analysis Combined FIT & FIC • Total Explained variance = 52.43% • Items are inter-correlated and represent a one-dimensional scale • Psychometric properties of SFT depend on summative item contributions- thus individual item analysis not recommended SFT-FIC Factor Analysis • 3 components with eigenvalues greater than zero which explain 65.68% of the variance FIC Factor Analysis • Component One: – Explains 52.4% of variance – Loads 15 items (majority of scale) • Component Two – Explains 7.4% of variance – Loads one item (rating scale) • Component Three – Explains 5.89% of variance – Loads two items (miracle question and homework) FIT Factor Analysis • 5 components with eigenvalues greater than 1 explain 57.8% of variance FIT Factor Analysis • Component One: – Explains 24.9% of variance & loads 9 items (mostly solution talk oriented) • Component Two: – Explains 10.5% of variance & loads 3 items (homework, compliments, ID of what’s better) • Component Three: – Explains 9.8% of variance & loads two items (rating scale and miracle questions) • Component Four: – Explains 6.8% of variance & loads one item (goal-related) • Component Five: – Explains 5.8% of variance & loads 1 item (coping questions) Limitations/Implications • Small sample size warrant further testing to assess for similar factorial patterns • Need to consider wording of task-related model components that may not occur every session • Need further examination with larger sample of FIT variability & strengthen research design with: – comparison to professional sample – comparison to video-taped sessions, or two-way mirror observation sessions • Next stage of development is to combine with outcome evaluation and use in different treatment contexts/venues References • de Shazar, S. (1982). Patterns of brief family therapy. New York: Guilford. • de Shazar, S. (1985). Keys to solution in brief therapy. New York: W.W. Norton. • de Shazar, S. (1988). Clues: Investigating solutions in brief therapy. New York: W.W. Norton. • de Shazar, D., Berg, I. K., Lipchik, E., Nunnally, E., Molnar, A., Gingerich, W., & Weiner-Davis, M. (1986). Brief therapy: Focused Solution Development. Family Process, 25, 207-221. • Gingerich, W.J., & Eisengart, S. (2000). Solution-focused brief therapy: A review of the outcome research. Family Process, 29, 477-498. • Mowbray, C. T., Holter, M. C., Teague, G. B., & Bybee, D. (2003). Fidelity criteria: Development, measurement, and validation. American Journal of Evaluation, 24, 315-340. • Trepper, T. S., Dolan, Y., McCollum, E. E., & Nelson, T. (2006). Steve de Shazar and the future of solution-focused therapy. Journal of Marital & Family Therapy, 32, 133-140.
"An exploratory test of two solution focused therapy"