Attachment-Alliance Meta-Analysis

ATTACHMENT-ALLIANCE META-ANALYSIS

Supplementary Table 1

Summary of Articles Included in the Attachment-Alliance Meta-Analysis

Publication / Sample / Findings / Individual Effect Sizes / Included Effect Sizes (Aggregated Results Where Applicable)
Bruck, Winston, Aderholt, & Muran (2006) / 46 patients in a psychiatry outpatient department randomly assigned to 30 sessions of Cognitive-Behavioral Therapy (CBT) or Short-Term Dynamic Psychotherapy (STDP) / (a) Nonsignificant correlation between the Secure subscale of the Relationship Style Questionnaire (RSQ) and scores from therapist ratings of the 12-item version of the Working Alliance Inventory (WAI-12; averaged across the three subscales of the WAI filled out after each of the first six sessions) / r = .00, n = 46 / = .04 (= 46)
(b) Nonsignificant correlation between the Secure subscale of the RSQ and scores from patient ratings of the WAI / r = .00, n = 46
(c) Nonsignificant correlation between the Fearful subscale of the RSQ and scores from therapistratings of the WAI / r = .00, n = 46
(d) Nonsignificant correlation between the Fearful subscale of the RSQ and scores from patient ratings of the WAI / r = .00, n = 46
(e) Nonsignificant correlation between the Preoccupied subscale of the RSQ and scores from therapistratings of the WAI / r = .00, n = 46
(f) Significant negative correlation in the predicted direction between the Preoccupied subscale of the RSQ and scores from patient ratings of the WAI / r = .33, n = 46
(g) Nonsignificant correlation between the Dismissing subscale of the RSQ and scores from therapistratings of the WAI / r = .00, n = 46
(h) Nonsignificant correlation between the Dismissing subscale of the RSQ and scores from patient ratings of the WAI / r = .00, n = 46
Dolan, Arnkoff, & Glass (1993) / 24 patients at a university counseling center / (a) Negative correlation in the predicted direction (with a trend towards statistical significance, p < .10) between Ambivalent subscale of the Attachment History Questionnaire (AHQ) and patient ratings on the WAI completed after the third session / r = .18, n = 13 / = .12 (= 17.67)
(b) Nonsignificant correlation (p > .10) in the nonpredicted direction between Ambivalent subscale of the AHQ and therapist ratings on the WAI / r = -.02, n = 13
(c) Nonsignificant correlation (p > .10) in the nonpredicted direction between Avoidant subscale of the AHQ and patient ratings on the WAI / r = -.04, n = 24
(d) Significant correlation (p < .05) in the predicted direction between Avoidant subscale of the AHQ and therapist ratings on the WAI / r = .26, n = 22
(e) Nonsignificant correlation (p > .05) in the predicted direction between Secure subscale of the AHQ and patient ratings on the WAI / r = .04, n = 18
(f) Significant correlation (p < .05) in the predicted direction between Secure subscale of the AHQ and therapist ratings on the WAI / r = .27, n = 16
Eames & Roth(2000) / 30 patients in five outpatient clinics / (a) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Secure subscale of the Relationship Styles Questionnaire (RSQ) and session 2 patient-rated WAI scores / r = .05, n = 30 / = .13 (= 25.13)
(b) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Fearful subscale of the RSQ and session 2 patient-rated WAI scores / r = .03, n = 30
(c) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Preoccupied subscale of the RSQ and session 2 patient-rated WAI scores
(d) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Dismissing subscale of the RSQ and session 2 patient-rated WAI scores / r = .16, n = 30
r = .03, n = 30
(e) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Secure subscale of the RSQ and session 3 patient-rated WAI scores / r = .27, n = 25
(f) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Fearful subscale of the RSQ and session 3 patient-rated WAI scores / r = .39, n = 25
(g) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Preoccupied subscale of the RSQ and session 3 patient-rated WAI scores / r = .23, n = 25
(h) Nonsignificant correlation (p > .05) in the nonpredicted direction between scores on the Dismissing subscale of the RSQ and session 3 patient-rated WAI scores / r = -.33, n = 25
(i) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Secure subscale of the RSQ and session 4 patient-rated WAI scores / r = .28, n = 24
(j) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Fearful subscale of the RSQ and session 4 patient-rated WAI scores / r = .26, n = 24
(k) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Preoccupied subscale of the RSQ and session 4 patient-rated WAI scores / r = .28, n = 24
(l) Nonsignificant correlation (p > .05) in the nonpredicted direction between scores on the Dismissing subscale of the RSQ and session 4 patient-rated WAI scores / r = -.28, n = 24
(m) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Secure subscale of the RSQ and session 5 patient-rated WAI scores / r = .29, n = 20
(n) Significant correlation (p < .05) in the predicted direction between scores on the Fearful subscale of the RSQ and session 5 patient-rated WAI scores / r = .52, n = 20
(o) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Preoccupied subscale of the RSQ and session 5 patient-rated WAI scores / r = .17, n = 20
(p) Nonsignificant correlation (p > .05) in the nonpredicted direction between scores on the Dismissing subscale of the RSQ and session 5 patient-rated WAI scores / r = .26, n = 20
(q) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Secure subscale of the RSQ and session 2 therapist-rated WAI scores / r = .23, n = 28
(r) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Fearful subscale of the RSQ and session 2 therapist-rated WAI scores / r = .24, n = 28
(s) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Preoccupied subscale of the RSQ and session 2 therapist-rated WAI scores / r = .19, n = 28
(t) Nonsignificant correlation (p > .05) in the nonpredicted direction between scores on the Dismissing subscale of the RSQ and session 2 therapist-rated WAI scores / r = -.04, n = 28
(u) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Secure subscale of the RSQ and session 3 therapist-rated WAI scores / r = .14, n = 26
(v) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Fearful subscale of the RSQ and session 3 therapist-rated WAI scores / r = .09, n = 26
(w) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Preoccupied subscale of the RSQ and session 3 therapist-rated WAI scores / r = .10, n = 26
(x) Nonsignificant correlation (p > .05) in the nonpredicted direction between scores on the Dismissing subscale of the RSQ and session 3 therapist-rated WAI scores / r = -.29, n = 26
(y) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Secure subscale of the RSQ and session 4 therapist-rated WAI scores / r = .22, n = 24
(z) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Fearful subscale of the RSQ and session 4 therapist-rated WAI scores / r = .33, n = 24
(aa) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Preoccupied subscale of the RSQ and session 4 therapist-rated WAI scores / r = .01, n = 24
(ab) Nonsignificant correlation (p > .05) in the nonpredicted direction between scores on the Dismissing subscale of the RSQ and session 4 therapist-rated WAI scores / r = -.19, n = 24
(ac) Significant correlation (p < .05) in the predicted direction between scores on the Secure subscale of the RSQ and session 5 therapist-rated WAI scores / r = .42, n = 24
(ad) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Fearful subscale of the RSQ and session 5 therapist-rated WAI scores / r = .37, n = 24
(ae) Nonsignificant correlation (p > .05) in the predicted direction between scores on the Preoccupied subscale of the RSQ and session 5 therapist-rated WAI scores / r = .14, n = 24
(af) Nonsignificant correlation (p > .05) in the nonpredicted direction between scores on the Dismissing subscale of the RSQ and session 5 therapist-rated WAI scores / r = -.03, n = 24
Goldman & Anderson (2007) / 30 patients at university-based outpatient treatment centers / (a) Significant correlation (p < .05) in the predicted direction between the Security dimension of the revised Adult Attachment Scale (AAS; Collins, 1996; the Security dimension is calculated by summing the Close and Depend subscale scores and then subtracting the Anxiety subscale scores; Goldman & Anderson, 2007) and session 1 WAI scores / r = .39, n = 30 / = .29 (= 30)
(b) Nonsignificant correlation (p > .05) in the predicted direction between the Security dimension of the revised AAS and session 2 WAI scores / r = .30, n = 30
(c) Nonsignificant correlation (p > .05) in the predicted direction between the Security dimension of the revised AAS and session 3 WAI scores / r = .16, n = 30
Hardy, Barkham, Stiles, & Startup (1998)a / 79 patients receiving treatment at a research clinic / (a) Comparison of patient ratings on the Agnew Relationship Measures (ARM) Openness scale between patients classified as Balanced (M = 5.61, SD = 1.12, n = 26) versus those classified as Overinvolved (M = 5.55, SD = 1.12, n = 36; classifications of interpersonal style were done using patient ratings on Overinvolved and Underinvolved scales constructed using the Inventory of Interpersonal Problems) / r = .03, n = 62 / = .0003 (= 52.50)
(b) Comparison of patient ratings on the ARM Openness scale between patients classified as Balanced (M = 5.61, SD = 1.12, n = 26) versus those classified as Underinvolved (M = 5.57, SD = 1.12, n = 17 / r = .02, n = 43
(c) Comparison of therapist ratings on the ARM Openness scale between patients classified as Balanced (M = 5.02, SD = 1.07, n = 26) versus those classified as Overinvolved (M = 5.04, SD = 1.07, n = 36 / r = -.01, n = 62
(d) Comparison of therapist ratings on the ARM Openness scale between patients classified as Balanced (M = 5.02, SD = 1.07, n = 26) versus those classified as Underinvolved (M = 5.09, SD = 1.07, n = 17 / r = -.03, n = 43
(e) Nonsignificant posthoc comparison of patient ratings on the ARM Partnership scale between patients classified as Balanced versus those classified as Overinvolved / r = 0.00, n = 62
(f) Nonsignificant posthoc comparison of patient ratings on the ARM Partnership scale between patients classified as Balanced versus those classified as Underinvolved / r = 0.00, n = 43
(g) Nonsignificant posthoc comparison of therapist ratings on the ARM Partnership scale between patients classified as Balanced versus those classified as Overinvolved / r = 0.00, n = 62
(h) Nonsignificant posthoc comparison of therapist ratings on the ARM Partnership scale between patients classified as Balanced versus those classified as Underinvolved / r = 0.00, n = 43
(i) Nonsignificant posthoc comparison of therapist ratings on the ARM Initiative scale between patients classified as Balanced versus those classified as Overinvolved / r = 0.00, n = 62
(j) Nonsignificant posthoc comparison of therapist ratings on the ARM Initiative scale between patients classified as Balanced versus those classified as Underinvolved / r = 0.00, n = 43
Kivlighan, Patton, & Foote (1998) / 40 patients in a university counseling center / (a) Nonsignificant correlation (p > .05) in the predicted direction between patient scores on the Anxiety subscale of the Adult Attachment Scale (AAS) and third session WAI scores / r = .19, n = 40 / = .31 (= 40.00)
(b) Significant correlation (p < .05) in the predicted direction between patient scores on the Close subscale of the AAS and third session WAI scores / r = .35, n = 40
(c) Significant correlation (p < .05) in the predicted direction between patient scores on the Depend subscale of the AAS and third session WAI scores / r = .38, n = 40
Mallinckrodt, Coble, & Gantt (1995) / 76 patients receiving between 5 and 12 sessions of therapy (out of a larger sample which had larger variation in therapy duration) at a university counseling center, a community college counseling center, an outpatient hospital-based clinic, and a counseling psychology program training clinic / (a) Significant correlation (p < .05) in the predicted direction between patient scores on the Anxiety subscale of the AAS and WAI scores / r = .32, n = 76 / = .19 (= 76.00)
(b) Nonsignificant correlation (p > .05) in the predicted direction between patient scores on the Close subscale of the AAS and WAI scores / r = .22, n = 76
(c) Nonsignificant correlation (p > .05) in the predicted direction between patient scores on the Depend subscale of the AAS and WAI scores / r = .01, n = 76
Mallinckrodt, Porter, & Kivlighan (2005) / 38 patients receiving treatment in a university counseling center who had completed 3-7 sessions / (a) Significant correlation (p < .05) in the predicted direction between patient scores on the Anxiety subscale of the Experiences in Close Relationships Scale (ECRS) and WAI scores / r = .33, n = 38 / = .29 (= 38.00)
(b) Nonsignificant correlation (p > .05) in the predicted direction between patient scores on the Avoidance subscale of the ECRS and WAI scores / r = .24, n = 38
Marmarosh, Gelso, Markin, Majors, Mallery, & Choi (2009) / 48 patients who remained in therapy until at least the third session at a university counseling center / (a) Nonsignificant correlation (p > .05) in the predicted direction between patient ratings on the Anxiety subscale of the ECRS and third session patient WAI scores / r = .11, n = 48 / = .09 (= 48.00)
(b) Nonsignificant correlation (p > .05) in the nonpredicted direction between patient ratings on the Anxiety subscale of the ECRS and third session therapist WAI scores / r = -.08, n = 48
(c) Significant correlation (p < .05) in the predicted direction between patient ratings on the Avoidance subscale of the ECRS and third session patient WAI scores / r = .32, n = 48
(d) Nonsignificant correlation (p > .05) in the predicted direction between patient ratings on the Avoidance subscale of the ECRS and third session therapist WAI scores / r = .01, n = 48
Parish & Eagle (2003; personal communication, M. Parish, September 5, 2007) / 107 participants who were current or recent doctoral-level students at several training programs in psychodynamic therapy, individuals in psychoanalytic treatment who received the surveys from the aforementioned students, and current patients in a university counseling center conducting psychodynamic psychotherapy / (a) Significant correlation (p = .02) in the predicted direction between patient ratings on the Dismissive subscale of the Relationship Questionnaire (RQ) and WAI scores / r = .22, n = 105 / = .18 (= 104.75)
(b) Significant correlation (p = .04) in the predicted direction between patient ratings on the Fearful subscale of the RQ and WAI scores / r = .20, n = 105
(c) Nonsignificant correlation (p = .66) in the predicted direction between patient ratings on the Preoccupied subscale of the RQ and WAI scores / r = .04, n = 105
(d) Significant correlation (p = .01) in the predicted direction between patient ratings on the Secure subscale of the RQ and WAI scores / r = .25, n = 104
Reis & Grenyer (2004) / 58 patients receiving treatment at an outpatient university clinic / (a) Significant correlation (p < .05) in the predicted direction between patient ratings on the Dismissive subscale of the RQ and third session WAI scores / r = .27, n = 58 / = .14 (= 58.00)
(b) Nonsignificant correlation (p > .05) in the predicted direction between patient ratings on the Fearful subscale of the RQ and third session WAI scores / r = .09, n = 58
(c) Nonsignificant correlation (p > .05) in the nonpredicted direction between patient ratings on the Preoccupied subscale of the RQ and third session WAI scores / r = -.03, n = 58
(d) Nonsignificant correlation (p > .05) in the predicted direction between patient ratings on the Secure subscale of the RQ and third session WAI scores / r = .24, n = 58
Romano, Fitzpatrick, & Janzen (2008) / 59 volunteer clients in enrolled in a counseling course who chose a counseling experience as an optional, experiential component of the class / (a) Nonsignificant correlation (p > .05) in the predicted direction between client ratings on the Anxiety subscale of the ECRS and WAI scores from session 5-9. / r = .20, n = 59 / = .22 (= 59.00)
(b) Nonsignificant correlation (p > .05) in the predicted direction between client ratings on the Avoidance subscale of the ECRS and WAI scores from session 5-9 / r = .24, n = 59
Saatsi, Hardy, & Cahill (2007) / 85 patients out of a larger sample of patients for whom interpersonal style classification was possible and who received treatment at a jointly-managed National Health Service and university research clinic / (a) Comparison between patient-rated alliance scores (using either the ARM completed after every session or the California Psychotherapy Alliance Scales [CALPAS] completed after sessions 1,2, 3, 7, and 12) of patients classified as secure (M = 0.52, SD = 0.68, n = 16) versus ambivalent (M = -0.04, SD = 1.01, n = 28) / r = .29, n = 44 / = .30 (= 50.50)
(b) Comparison between patient-rated alliance scores (using either the ARM or the CALPAS) of patients classified as secure (M = 0.52, SD = 0.68, n = 16) versus avoidant (M = -0.15, SD = 1.00, n = 41) / r = .31, n = 57
Satterfield & Lyddon (1995) / 60 first-time clients seeking treatment at a university-based counseling clinic who completed all measures, out of a larger sample of clients / (a) Nonsignificant correlation (p > .05) in the predicted direction between client ratings on the Anxiety subscale of the AAS and third session WAI scores / r = .19, n = 60 / = .21 (= 60.00)
(b) Nonsignificant correlation (p > .05) in the predicted direction between client ratings on the Close subscale of the AAS and third session WAI scores / r = .12, n = 60
(c) Significant correlation (p < .01) in the predicted direction between client ratings on the Depend subscale of the AAS and third session WAI scores / r = .31, n = 60
Satterfield & Lyddon (1998) / 63 first-time clients seeking treatment at a university counseling center / (a) Nonsignificant correlation (p > .05) in the nonpredicted direction between client ratings on the Dismissing subscale of the RQ and WAI scores (rated at some point after the third session and before the sixth session) / r = .10, n = 63 / = .12 (= 63.00)
(b) Significant correlation (p < .05) in the predicted direction between client ratings on the Fearful subscale of the RQ and WAI scores / r = .20, n = 63
(c) Nonsignificant correlation (p > .05) in the predicted direction between client ratings on the Preoccupied subscale of the RQ and WAI scores / r = .18, n = 63
(d) Significant correlation (p < .05) in the predicted direction between client ratings on the Secure subscale of the RQ and WAI scores / r = .21, n = 63
Sauer, Lopez, & Gormley (2003; personal communication, E. Sauer, October 24, 2007; November 19, 2007; Sauer, 1999) / 28 patients receiving treatment from therapists who planned to see their clients for at least seven sessions and who were recruited from graduate level training courses, university counseling centers, and community counseling centers / (a) Nonsignificant correlation (p > .05) in the predicted direction between patient ratings on the Anxiety subscale of the Adult Attachment Inventory (AAI; Simpson, 1990) self-report measure and first session WAI scores / r = .01, n = 28 / = .03 (= 22.33)
(b) Nonsignificant correlation (p > .05) in the predicted direction between patient ratings on the Avoidance subscale of the AAI and first session WAI scores / r = .14, n = 28
(c) Nonsignificant correlation (p > .05) in the predicted direction between patient ratings on the Anxiety subscale of the AAI and fourth session WAI scores / r = .22, n = 22
(d) Nonsignificant correlation (p > .05) in the predicted direction between patient ratings on the Avoidance subscale of the AAI and fourth session WAI scores / r = .04, n = 22
(e) Nonsignificant correlation (p > .05) in the nonpredicted direction between patient ratings on the Anxiety subscale of the AAI and seventh session WAI scores / r = -.32, n = 17
(f) Nonsignificant correlation (p > .05) in the predicted direction between patient ratings on the Avoidance subscale of the AAI and seventh session WAI scores / r = .07, n = 17
Schiff & Levit (2010) / 95 female patients receiving treatment at four methadone clinics / (a) Nonsignificant correlation (p > .05) in the predicted direction between mean of items on the Anxious dimension of the revised ECRS translated into Hebrew and cube transformations of scores on the 12-item version of the WAI / r = .11, n = 95 / = .14 (= 95.00)
(b) Nonsignificant correlation (p > .05) in the predicted direction between mean of items on the Avoidance dimension of the revised ECRS translated into Hebrew and cube transformations of scores on the 12-item version of the WAI / r = .17, n = 95

ATTACHMENT-ALLIANCE META-ANALYSIS

aSee Method section for details about calculation of effect sizes for this study.

References

Collins, N. L. (1996). Working models of attachment: Implications for explanation, emotion, and behavior. Journal of Personality and

Social Psychology, 71, 810-832. doi:10.1037/0022-3514.71.4.810

Sauer, E. M. (2000). Therapist and client adult attachment: A longitudinal examination of the therapeutic working alliance.