The Session Reactions Scale-5 (SRS-5) is a five-item client-report measure of the helpful impact of a therapy session (Buchanan, Hegarty, & Řiháček, 2026), for use with adults aged 18 and over. It is for psychologists, counsellors, and other mental health clinicians who practise feedback-informed treatment or routine outcome monitoring.
The Session Reactions Scale-5 (SRS-5) is a five-item client-report measure of the helpful impact of a therapy session (Buchanan, Hegarty, & Řiháček, 2026), for use with adults aged 18 and over. It asks clients how much they took from their most recent session across five domains of helpful reaction. The SRS-5 was developed as an ultra-brief measure for feedback-informed treatment, drawn from the Helpful Reactions items of the Session Reactions Scale-3-Brief (SRS-3-B; Řiháček et al., 2024), and is designed for administration after every session.
Feedback-informed treatment (FIT) is a structured approach in which clinicians routinely collect brief client-report data and use it to adjust care while therapy is underway (Miller et al., 2016). Two streams are distinguished: outcome feedback, tracking symptom change between sessions, and process feedback, capturing the client’s experience of the session. Routine outcome feedback is associated with improved outcomes and lower drop-out (de Jong et al., 2021), and in a multisite trial it reduced deterioration among clients who were not progressing as expected (Delgadillo et al., 2018).
Process feedback has centred on the therapeutic alliance (Duncan et al., 2003; Cahill et al., 2012). The alliance is a moderate but robust predictor of outcome across therapies (Flückiger et al., 2018), and the SRS-5 occupies the same slot in routine practice, but asks how productive the client found the session instead of how they regard the therapist. Sessions differ in what they produce for a client, and client and therapist ratings of the alliance correlate only moderately (Tryon et al., 2007), so a client-report measure can capture variation the clinician does not observe.
Client ratings of the alliance and of session quality are typically high, which limits how far such measures can tell sessions apart. Asking what a client took from a session, rather than how they regard the therapist, is intended to produce a wider spread of responses. A low or falling score is an early prompt to ask what was missing rather than a verdict on the session.
The SRS-5 covers five distinct domains of helpful in-session impact, which together produce a single total score:
The SRS-5 is suited to routine outpatient psychotherapy and any setting practising feedback-informed treatment or routine outcome monitoring. Its brevity is a deliberate design feature, since session-by-session measures are only sustained in real-world practice when they impose minimal burden.
The primary use is progress monitoring. Administered after each session, the SRS-5 tracks how helpful sessions are experienced over a course of therapy, and the average item score is plotted across administrations. A single low-scoring session is itself clinically useful: it provides an early, concrete prompt to explore what a client did or did not get from a session, and to adjust the approach before disengagement or drop-out follows. The measure also supports client feedback and reflection, since clients complete it themselves and can review the result with their clinician as part of the session.
The SRS-5 total score is the sum of items 1 to 5 and ranges from 5 to 25, and is also reported as an average item score from 1 to 5. A higher score indicates a more helpful session as experienced by the client. The interpretive categories are defined on the average, which is what the report charts; change between sessions is reported in raw points.
Example results from a client’s fifth administration:
| Score | Raw Score (5-25) | Average (1-5) | Percentile | Descriptor |
|---|---|---|---|---|
| Total Score | 19 | 3.8 | 66 | Neutral Reaction |
Each of the five items covers one domain of helpful in-session impact, and the total is the primary output. The individual responses still carry clinical detail the total cannot, since they identify which kind of help was missing:
The following categories are applied to the average item score, with the equivalent raw totals given alongside. They are derived from the response scale anchors, which run from “not at all” (1) to “very much” (5), so the bands describe the client’s average level of helpful reaction rather than a position in the reference distribution.
A total of 23 is the lowest score whose average reaches the halfway point between “quite a bit” and “very much”, which is where the top boundary sits.
A percentile places the total score against the distribution of sessions in the NovoPsych routine-care sample (Buchanan et al., 2026). Because the reference group is other therapy sessions rather than the general population, it answers how helpful this session was compared with sessions in routine care, where the 50th percentile marks a typical session. A score of 18 sits at about the 57th percentile, while 22 sits near the 88th, more helpful than most (Buchanan et al., 2026).
Session and alliance measures attract a high proportion of maximum ratings, which limits their ability to distinguish between sessions. The SRS-5 addresses this in its design, by asking what the client took from the session rather than how they regard the therapist. A maximum total therefore indicates a session the client experienced as highly helpful in every domain.
Change between sessions is reported in raw points. Session impact varies legitimately from one session to the next, so a single change in either direction says more about that session than about the direction of therapy. The pattern carries the interpretive weight: a sustained decline, or a run of sessions in the lower bands, indicates the work is repeatedly landing with little impact for the client.
On a first administration, two bar charts follow the results table. SRS-5 Session Reaction Average Score shows the average of the five item responses as a single bar on an axis numbered 1 to 5. The four interpretive category bands are shaded behind the bar and named at the right, because the categories are defined on the average item score. SRS-5 Reaction Domain Responses then shows the response to each of the five items, Insight, Empowerment, Clarity, Skills, and Relief, as one bar per item on an axis that labels each point with its response anchor, from “1 (Not at all)” to “5 (Very much)”, each item in its own colour.
On a repeat administration these are replaced by a matching pair. SRS-5 Session Reaction Average Score Over Time plots the average item score at each administration date on the same numbered axis, with the four bands shaded across the plot and named to the right of the latest date. The points are joined by a curved line. The chart covers the 60 most recent administrations, split evenly across two plots, oldest first, when there are more than 30. The average is printed beside each point when a plot has two to four. Below it, SRS-5 Reaction Domains Over Time gives each item its own panel, tracking the response to that item across the same dates on the axis labelled with the response anchors, in the colour it takes on the first-administration chart.
The SRS-5 is developed from the Helpful Reactions subscale of the Session Reactions Scale-3-Brief (SRS-3-B; Řiháček et al., 2024), a post-session measure whose 14 rated items cover helpful and hindering reactions to in-session events. The SRS-3-B originates from the work of Elliott (1985) on helpful events in therapy, which identified clusters of client-identified impacts spanning the tasks of therapy and the interpersonal relationship.
For the SRS-5, five of the ten SRS-3-B Helpful Reactions items were retained, selected to span distinct domains of helpful impact:
Items were chosen for content breadth across these domains together with favourable classical properties, namely acceptable item-total discrimination and wide response dispersion without strong ceiling effects. Two items were excluded because their content overlaps with the therapeutic alliance (feeling understood by the therapist, and personal investment in therapy), and they also had the largest ceiling effects of the ten Helpful Reactions items. Two more were excluded because they apply only to some sessions (distancing from difficult feelings, and feeling more hopeful about another person), and one because its content overlapped with the retained items (acceptance). In the parent measure’s ten-item model, the retained items were stronger indicators of the helpful-reactions dimension than the five omitted (average loading .74 against .58; NovoPsych routine-care sample, N = 975; Buchanan et al., 2026).
Two estimates describe how much of the parent subscale the short form recovers, and they answer different questions. Against the five omitted items, which share no items with the SRS-5, the total correlates r = .81. Against factor scores from the full ten-item subscale, of which the SRS-5 items are half, it correlates r = .98 (Buchanan et al., 2026); that figure is raised by the shared items and is not an independent estimate. The r = .81 is the conservative reading.
The SRS-5 is scored as a single total of helpful in-session impact, drawn from the Helpful Reactions dimension of the parent SRS-3-B, which resolves into two dimensions, helpful and hindering reactions, in its own validation (Řiháček et al., 2024). Internal consistency for the total is strong (Cronbach’s α = .86; McDonald’s ω = .86), and corrected item-total correlations range from .59 to .72, so each item contributes coherently to the composite. These were computed on the five SRS-5 items across 975 SRS-3-B administrations in the NovoPsych routine-care sample, one per client (Buchanan et al., 2026).
Convergent validity was supported by a moderate correlation with the Agnew Relationship Measure-5 (ARM-5; Cahill et al., 2012), a validated measure of the therapeutic alliance. In a matched subsample the two correlated r = .41, 95% CI [.18, .59], rising to r = .63 among pairs completed on the same day (n = 68 clients, 39 same-day; Buchanan et al., 2026). The measures share about 16% of their variance while remaining clearly distinct, supporting the interpretation that the SRS-5 indexes a helpful-impact construct related to, but not redundant with, the working alliance. Discriminant evidence points the same way: the five items correlate r = -.29 with the parent scale’s four Hindering Reactions items (N = 975; Buchanan et al., 2026), consistent with helpful impact specifically rather than overall session reactivity.
The SRS-5 had not been administered as a standalone scale when these data were assembled, so its reference distribution is derived from the parent measure: SRS-3-B administrations scored on the five items that make up the SRS-5. Because those five items are a subset of the SRS-3-B, each administration provides a complete set of SRS-5 item responses. Each client contributed one administration. The resulting 975 administrations were completed between October 2023 and July 2026 by 975 clients under the care of 50 clinicians (84.0% psychologists, 8.0% counsellors, and 8.0% social workers). Most were recorded by Australia-based clinicians (73.3%), with the remainder from the United States (11.5%), New Zealand (8.1%), the Philippines (4.7%), and other locations (2.4%; Buchanan et al., 2026). Clients were 50.8% female, 27.6% male, and 3.7% gender diverse or other, with 17.9% unspecified. Age was recorded for 89.1% of clients; the median age band was 25 to 34 years, and 9.2% of clients with a recorded age were under 18, so the norms include a small proportion of adolescents.
In this sample the average total score was 16.91 (SD = 4.39; median = 17; Buchanan et al., 2026). Scores were approximately normally distributed with a mild negative skew (skew = -.32) and spanned the full range of 5 to 25. Only 0.6% of administrations obtained the lowest possible score and 3.7% the highest. A ceiling effect is conventionally judged present when more than 15% of respondents obtain the highest possible score (Terwee et al., 2007). The ARM-5, administered on the same platform to an overlapping population, reached its maximum on 51.5% of the administrations in the matched subsample (Buchanan et al., 2026), well above that threshold; the SRS-5 sits well below it.
The four interpretive categories are derived by NovoPsych and anchored to the meaning of the SRS-5 response scale rather than to percentile landmarks. Each response point is verbally labelled, so the average item response carries direct meaning, and the category boundaries sit at the average-item values that coincide with the scale’s own descriptors of “somewhat” and “quite a bit”, and at the midpoint between “quite a bit” and “very much”.
The bands therefore describe what the client actually endorsed, and hold across revisions of the reference sample. Percentiles are reported alongside them as relative context.
Alliance measures ask how a client regards the therapeutic relationship, whereas the SRS-5 asks what the client actually took from the session, such as new insight, a sense of hope, greater clarity, a coping strategy, or emotional relief. The two are related but distinct: a session can feel warm and supportive yet leave the client with little to work with, or feel demanding yet produce real movement. Measuring impact alongside the alliance gives a fuller picture of the session process, and the SRS-5 is designed to sit comfortably beside an alliance measure such as the Agnew Relationship Measure-5 (ARM-5).
A low score is best treated as a conversation starter rather than a verdict. It signals that the client took relatively little from that particular session, which is valuable to know early. Exploring what was missing, whether the focus felt relevant, and what the client had hoped for can surface a mismatch in goals or approach while there is still time to adjust. Used this way, the SRS-5 acts as an early-warning signal that supports the kind of responsive course-correction that feedback-informed treatment is built around.
The SRS-5 is designed to be completed after every session. It takes well under a minute, so it can be used routinely without adding meaningful burden to the client or the clinician. Regular administration is what makes the session-by-session pattern visible and allows a run of low-impact sessions to be noticed early.
Client ratings of the alliance and of session quality are often clustered near the top of the scale, which limits a measure’s ability to tell sessions apart. In the NovoPsych reference data, 3.7% of SRS-5 totals reached the maximum, against 51.5% of ARM-5 administrations among clients who completed both measures (Buchanan et al., 2026). Score variability of this kind is a precondition for detecting differences between sessions.
Yes. The SRS-5 is intended as a progress-monitoring tool, and the report plots the average item score across administrations, with a smaller panel for each of the five items. Broader research on feedback-informed treatment indicates that giving clinicians structured feedback on client progress can improve outcomes and reduce drop-out (de Jong et al., 2021), and the SRS-5 adds brief, well-distributed feedback on the session itself. It is derived from the Session Reactions Scale-3-Brief (SRS-3-B), which measures both helpful and hindering reactions in a longer format when a fuller picture is desired.
Buchanan, B., Hegarty, D., & Řiháček, T. (2026). Development and initial validation of the Session Reactions Scale-5 (SRS-5): A brief client-report measure of in-session impact for feedback-informed treatment. NovoPsych.
Buchanan, B., Hegarty, D., & Řiháček, T. (2026). Development and initial validation of the Session Reactions Scale-5 (SRS-5): A brief client-report measure of in-session impact for feedback-informed treatment. NovoPsych.
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. https://doi.org/10.1080/10503307.2011.643253
de Jong, K., Conijn, J. M., Gallagher, R. A. V., Reshetnikova, A. S., Heij, M., & Lutz, M. C. (2021). Using progress feedback to improve outcomes and reduce drop-out, treatment duration, and deterioration: A multilevel meta-analysis. Clinical Psychology Review, 85, 102002. https://doi.org/10.1016/j.cpr.2021.102002
Delgadillo, J., de Jong, K., Lucock, M., Lutz, W., Rubel, J., Gilbody, S., Ali, S., Aguirre, E., Appleton, M., Nevin, J., O’Hayon, H., Patel, U., Sainty, A., Spencer, P., & McMillan, D. (2018). Feedback-informed treatment versus usual psychological treatment for depression and anxiety: A multisite, open-label, cluster randomised controlled trial. The Lancet Psychiatry, 5(7), 564–572. https://doi.org/10.1016/S2215-0366(18)30162-7
Duncan, B. L., Miller, S. D., Sparks, J. A., Claud, D. A., Reynolds, L. R., Brown, J., & Johnson, L. D. (2003). The Session Rating Scale: Preliminary psychometric properties of a “working” alliance measure. Journal of Brief Therapy, 3(1), 3–12.
Elliott, R. (1985). Helpful and nonhelpful events in brief counseling interviews: An empirical taxonomy. Journal of Counseling Psychology, 32(3), 307–322. https://doi.org/10.1037/0022-0167.32.3.307
Flückiger, C., Del Re, A. C., Wampold, B. E., & Horvath, A. O. (2018). The alliance in adult psychotherapy: A meta-analytic synthesis. Psychotherapy, 55(4), 316–340. https://doi.org/10.1037/pst0000172
Miller, S. D., Bargmann, S., Chow, D., Seidel, J., & Maeschalck, C. (2016). Feedback-informed treatment (FIT): Improving the outcome of psychotherapy one person at a time. In W. O’Donohue & A. Maragakis (Eds.), Quality improvement in behavioral health (pp. 247–262). Springer. https://doi.org/10.1007/978-3-319-26209-3_16
Řiháček, T., Elliott, R., Owen, J., Ladmanová, M., Coleman, J. J., & Bugatti, M. (2024). Session Reactions Scale-3: Initial psychometric evidence. Psychotherapy Research, 34(4), 434–448. https://doi.org/10.1080/10503307.2023.2241983
Terwee, C. B., Bot, S. D. M., de Boer, M. R., van der Windt, D. A. W. M., Knol, D. L., Dekker, J., Bouter, L. M., & de Vet, H. C. W. (2007). Quality criteria were proposed for measurement properties of health status questionnaires. Journal of Clinical Epidemiology, 60(1), 34–42. https://doi.org/10.1016/j.jclinepi.2006.03.012
Tryon, G. S., Blackwell, S. C., & Hammel, E. F. (2007). A meta-analytic examination of client-therapist perspectives of the working alliance. Psychotherapy Research, 17(6), 629–642. https://doi.org/10.1080/10503300701320611