Agnew Relationship Measure – 5 (ARM-5)

Abstract

The Agnew Relationship Measure – 5 (ARM-5) is the 5-item, self-report measure derived from the original 28-item Agnew Relationship Measure (ARM; Agnew et al., 1998). Developed and validated by Cahill et al. (2012), the ARM-5 is designed to efficiently assess the quality of the client-therapist alliance, which is consistently identified as one of the strongest predictors of positive treatment outcomes.

The scale is intended for frequent, routine administration—typically at the end of every therapy session, or every second or third session—to provide timely feedback to clinicians. This tool is crucial for identifying potential risks of client dropout and monitoring any ruptures or repairs occurring within the alliance. The ARM-5 assesses three core dimensions of the alliance: Bond, Partnership, and Confidence in therapy, and has been demonstrated to be psychometrically equivalent to other established scales, such as the Session Rating Scale (SRS).

Keywords

Therapeutic Alliance, Client-Therapist Relationship, Psychotherapy Outcome, Agnew Relationship Measure, ARM-5, Self-Report, Feedback Informed Treatment, Alliance Rupture, Treatment Efficacy.

Authors

Jane Cahill, William B. Stiles, Michael Barkham, Gillian E. Hardy, Gregory Stone, Roxane Agnew-Davies, Gisela Unsworth (Primary developers of the ARM-5 short form); Roxane Agnew-Davies, W.B. Stiles, G.E. Hardy, M. Barkham, & D.A. Shapiro (Developers of the original ARM).

[quads id=5]

Purpose

The primary purpose of the ARM-5 is to provide clinicians with a rapid, reliable, and valid instrument for continuously monitoring the therapeutic alliance from the client’s perspective. Its highly efficient 5-item structure minimizes session time devoted to measurement, allowing for integration into high-frequency monitoring protocols.

Clinically, the ARM-5 is widely adopted by practitioners engaging in Feedback Informed Treatment (FIT). By using the scale at regular intervals, therapists can track the trajectory of the alliance, identify moments of strain or disengagement (ruptures), and intervene proactively to strengthen the working relationship, thereby enhancing the likelihood of achieving successful client outcomes and reducing premature termination.

Construct

The ARM-5 measures the core construct of the Client-Therapist Alliance, which encompasses the quality and strength of the working relationship between the client and the therapist. This construct is recognized across diverse theoretical orientations as a fundamental common factor influencing therapeutic change.

The scale is structured to assess three specific, inter-related dimensions of the alliance that are considered essential for treatment efficacy:

1. Bond: This dimension captures the classic emotional connection, trust, and feelings of therapist supportiveness shared between the client and the therapist.
2. Partnership: This dimension assesses the extent of mutual agreement regarding the specific tasks and overall goals of the therapy, reflecting collaborative effort.
3. Confidence in Therapy: This dimension measures the client’s belief in the treatment approach and the perceived competence and self-confidence demonstrated by the therapist in their techniques. Research suggests confidence is the strongest predictor of positive outcome.

Validity

The ARM-5 has demonstrated strong validity, particularly in relation to its longer predecessor. Cahill et al. (2011) evaluated the ARM-5’s psychometric properties against the full 28-item Agnew Relationship Measure and concluded that the short form was psychometrically approximate to the Core Alliance indexes of the longer measure, supporting its construct validity as a measure of the therapeutic alliance.

Predictive validity is high, as evidenced by studies showing that consistently higher ARM-5 scores are predictive of improved client outcomes, including greater treatment gains and symptom reduction, particularly when used within modalities like Cognitive Behavioral Therapy (CBT) for depression. However, a significant limitation noted in the literature is the presence of ceiling effects. Data from the Sheffield Psychotherapy Project (n=1073) showed a highly skewed distribution (mean rating of 5.76 on a 7-point scale), indicating that clients frequently rate therapists very highly, which may limit the scale’s ability to differentiate between high levels of alliance.

Reliability

The ARM-5 exhibits acceptable levels of reliability for clinical use. Cahill et al. (2011) reported acceptable levels of internal consistency when evaluating the 5-item measure against the full ARM scale. Furthermore, the scale demonstrated acceptable alternative forms reliability, confirming that the short version reliably captures the core dynamics of the alliance measured by the original, longer instrument.

For clinical interpretation, research suggests that the most reliable and useful score is the total alliance score. This recommendation stems from findings that show minimal psychometrically significant distinction between the three proposed subscales (Bond, Partnership, and Confidence), implying that the measure functions most robustly as a single-factor assessment of overall alliance quality.

Factor Analysis

Although the ARM-5 is conceptually structured around three subscales (Bond, Partnership, and Confidence), analyses suggest that these dimensions are highly intercorrelated. The evidence indicates that there is minimal psychometrically significant distinction among the three subscales, effectively supporting the use of the measure as a unidimensional scale.

Consequently, factor analysis supports the clinical guidance that practitioners should primarily utilize the total raw score (ranging from 5 to 35) or the derived mean score (1 to 7) as the most reliable indicator of therapeutic alliance, rather than attempting detailed interpretation based on the individual subscale scores.

Instrument

Test Type: Self-Report Measure (Client-completed)

Format: 5-item Likert scale utilizing a 7-point response range (Strongly Disagree to Strongly Agree).

Language Available: English (Original).

Population Group: Clients or patients actively participating in psychotherapy or counseling.

Age Group: Primarily adults and adolescents engaged in clinical treatment.

Population Details: Validated across diverse clinical settings, including use in Cognitive Behavioral Therapy (CBT) for depression, based on the Sheffield Psychotherapy Project sample (n=1073).

Test Methodology: Brief, routine assessment administered at regular intervals (e.g., every session, every second session, or every third session) often integrated into Feedback Informed Treatment (FIT) protocols.

Keywords

Psychotherapy, Alliance Measurement, Outcome Prediction, Client Dropout, Agnew Relationship Measure, Bond, Partnership, Confidence.

[quads id=5]

Authors

Author ORCID Identifier: Not provided in source.

Affiliation Email addresses: Not provided in source.

Correspondence Address: Not provided in source.

Permissions & Fee and Test Year

The original Agnew Relationship Measure (ARM) was published in 1998 (Agnew et al.). The ARM-5 short form was psychometrically validated and published in 2012 by Jane Cahill et al. in the journal Psychotherapy Research, with the DOI: 10.1080/10503307.2011.643253. Specific details regarding licensing fees or formal permissions for clinical use are not included in the source material, but the scale is widely used in research and clinical practice.

Reference’s

Agnew‐Davies, R., Stiles, W. B., Hardy, G. E., Barkham, M., & Shapiro, D. A. (1998). Alliance structure assessed by the Agnew Relationship Measure (ARM). British Journal of Clinical Psychology, 37(2), 155-172.

Bouchard, D. (2018). Assessment of the Therapeutic Alliance Scales: A Reliability and Validity Evaluation. School of Psychology, University of Ottawa.

Cahill, J., Stiles, W. B., Barkham, M., Hardy, G. E., Stone, G., Agnew-Davies, R., & Unsworth, G. (2012). Two short forms of the Agnew Relationship Measure: The ARM-5 and ARM-12. Psychotherapy Research, 22(3), 241-255. DOI: 10.1080/10503307.2011.643253.

[quads id=5]

Items of the Agnew Relationship Measure – 5 (ARM-5)

Instructions: Thinking about today’s meeting, please indicate how strongly you agree or disagree with each statement.