How Between-Session Data Changes the Way You Do Therapy

Key takeaways

• The Research on Sudden Gains: Change Happens Between Sessions
• Homework and Between-Session Activities: What the Meta-Analyses Show
• The Recall Bias Problem: Why Session Openings Mislead

The Research on Sudden Gains: Change Happens Between Sessions

In 1999, Tony Tang and Robert DeRubeis published a landmark study in the Journal of Consulting and Clinical Psychology that introduced the concept of sudden gains — large, stable symptom improvements that occur between consecutive therapy sessions. Their analysis of cognitive-behavioral therapy for depression revealed that many patients experienced dramatic drops in Beck Depression Inventory scores during a single between-session interval, with the average sudden gain being 11 BDI points. These gains accounted for roughly 50 percent of the total improvement for patients who experienced them.

What made Tang and DeRubeis's findings so compelling was not just the magnitude of these between-session shifts, but their durability. Patients who experienced sudden gains were significantly less depressed at posttreatment, and they remained so 18 months later. The researchers proposed a three-stage model: preparation in the critical session preceding the gain, the sudden gain itself occurring between sessions, and an upward spiral of improved alliance and continued cognitive change in subsequent sessions. This model fundamentally challenged the assumption that therapeutic progress is gradual and linear.

Since that seminal paper, over 100 examinations of sudden gains have been published across diverse disorders and treatment modalities. A meta-analysis by Aderka and colleagues found that sudden gains consistently predict better treatment outcomes across anxiety disorders, PTSD, and depression, regardless of the specific therapy approach used. The phenomenon appears to be ubiquitous in psychological treatments, suggesting that between-session intervals are not downtime — they are where critical therapeutic change crystallizes.

For clinicians, the practical implication is striking: if roughly half of meaningful symptom improvement occurs between sessions, then the data generated during those intervals is at least as important as what happens in the room. Yet most therapists have no systematic way to capture what occurs during these critical windows. We rely on client self-report at the start of the next session, which is subject to recency bias, mood-congruent memory, and the simple compression of a week's experience into a five-minute check-in.

Homework and Between-Session Activities: What the Meta-Analyses Show

Nikolaos Kazantzis and colleagues at the University of Melbourne have conducted the most comprehensive meta-analytic work on between-session homework in psychotherapy. Their 2000 meta-analysis, published in Clinical Psychology: Science and Practice, examined 27 studies and found a weighted mean effect size of d = 0.36 favoring therapy with homework over therapy without homework. Their 2010 replication and extension, analyzing 46 studies with over 1,000 participants, found that therapy conditions with homework produced effect sizes of d = 1.08 compared to d = 0.63 for control conditions — a clinically meaningful difference.

The research makes clear that what clients do between sessions substantially amplifies treatment effects. Between-session exercises crystallize skills learned in treatment, promote self-efficacy, and reinforce therapeutic gains. Kazantzis and colleagues found that both the quantity and quality of homework completion predict outcomes, with a Hedges' g of 0.79 for homework quantity and 0.78 for homework quality in relation to posttreatment outcomes. Along with therapeutic alliance, practice assignments are among the strongest predictors of posttreatment outcomes in cognitive behavioral therapies.

However, homework compliance rates remain stubbornly inconsistent. Research suggests that 20 to 50 percent of clients do not complete between-session assignments as prescribed, and many more complete them partially or modify them without therapist awareness. The gap between prescribed and actual homework behavior represents a significant data blind spot. When a client reports that a behavioral experiment "didn't work," was the experiment itself flawed, or was the execution different from what was planned? Without detailed between-session data, clinicians are often guessing.

A 2024 review published in the journal Psychotherapy highlighted that between-session homework in clinical training and practice has become a transtheoretical concern — not limited to CBT. Psychodynamic, humanistic, and integrative therapies are all increasingly recognizing that what happens outside the therapy room is where the majority of living occurs, and that therapeutic effectiveness depends on the transfer of in-session learning to real-world contexts. The challenge is not whether between-session processes matter — the evidence is overwhelming that they do — but how to capture and use this information clinically.

The Recall Bias Problem: Why Session Openings Mislead

Cognitive psychology has documented extensively how unreliable human memory is, and this unreliability affects both clients and therapists. Research published in the journal Memory for Patient Information examined how clinicians recall client data and found that experienced clinicians actually recalled fewer specific facts than intermediate trainees, instead retaining more abstracted, schema-driven impressions. While clinical expertise brings valuable pattern recognition, it also means that specific between-session details are filtered through the clinician's conceptual framework rather than preserved in their raw form.

For clients, the recall challenges are equally significant. Peak-end theory, established by Daniel Kahneman, demonstrates that people tend to remember the most intense moment and the most recent moment of an experience, rather than the experience as a whole. When a client walks into session and you ask how their week went, they are not giving you an accurate summary of seven days. They are giving you their emotional peak and whatever happened most recently — often the drive to your office. Entire sequences of meaningful between-session events are compressed, distorted, or omitted.

Research on memory in CBT specifically has found alarming levels of information loss. Studies show that clients forget up to 80 percent of treatment recommendations between sessions, and when they do think about CBT skills between sessions, less than half of those thoughts accurately reflect the content presented during treatment. This is not a failure of client motivation — it is a predictable feature of how human memory works under conditions of emotional distress, divided attention, and time delay.

The clinical consequence is that therapists are making treatment decisions based on incomplete and systematically biased data. We adjust interventions, modify treatment plans, and assess progress based on a distorted sample of between-session experience. It is analogous to a physician adjusting medication based on a patient's memory of their blood pressure rather than actual readings. The treatment may still work, but the decision-making process is operating with unnecessary handicaps.

This is where continuous between-session data capture becomes clinically meaningful — not as surveillance, but as a corrective for the inherent limitations of retrospective self-report. Tools like Empath that capture mood, journaling, and behavioral data in real time provide therapists with a fuller picture of the between-session landscape, reducing the distortions introduced by recall bias and enabling more precise clinical decision-making.

What Between-Session Data Actually Looks Like in Practice

When we talk about between-session data, we are not talking about a single measure or a single data point. Clinically useful between-session data includes mood trajectories across the week, sleep quality and patterns, social engagement or withdrawal, skill practice attempts and their perceived effectiveness, journaling reflections that capture thoughts and emotions in the moment, and behavioral indicators like activity level and substance use. Each of these data streams tells a story that a 50-minute session cannot capture on its own.

Consider a client being treated for major depression who reports in session that they "had an okay week." Standard practice is to take this at face value and move into the session's agenda. But continuous mood tracking might reveal that Monday through Wednesday involved a significant depressive episode that resolved by Thursday after a social interaction and exercise — a pattern invisible to the session-opening check-in. That pattern contains actionable clinical information: the social interaction and exercise were likely behavioral activation successes that deserve reinforcement, and the early-week pattern may correlate with work stressors or sleep disruption that warrant exploration.

Between-session data also provides what researchers call ecological momentary assessment — capturing experiences in the natural environment, in real time, rather than relying on retrospective summaries. Ecological momentary assessment studies in psychotherapy research have consistently shown that real-time reports of mood, behavior, and symptoms differ significantly from retrospective reports given days later. Stone and colleagues demonstrated that retrospective pain reports, for example, correlated only moderately with real-time diary entries, with patients systematically overestimating or underestimating their experiences based on current mood state.

For therapists, the practical shift is from asking "How was your week?" — a question that invites a compressed, biased narrative — to reviewing actual data patterns and asking targeted questions: "I noticed your mood dipped significantly on Tuesday and rebounded on Thursday. What was happening during those transitions?" This transforms session openings from recall exercises into collaborative data interpretation, and it typically surfaces more clinically relevant material in less time.

How Continuous Data Capture Transforms Clinical Decision-Making

Michael Lambert's extensive body of research on routine outcome monitoring has demonstrated that systematic data collection during treatment significantly improves outcomes. His work at Brigham Young University showed that providing therapists with patient progress feedback reduced deterioration rates and nearly doubled clinically significant change in clients who were predicted to have poor outcomes. The mechanism is straightforward: data enables earlier detection of problems and more timely clinical adjustments.

Between-session data extends this principle beyond standardized outcome measures to the rich, contextual information that emerges in daily life. When a therapist can see that a client's anxiety spikes reliably on Sunday evenings, that sleep disruption precedes depressive episodes by two to three days, or that journaling entries become progressively more ruminative during a specific period, treatment planning shifts from reactive to proactive. Instead of waiting for problems to manifest in session, clinicians can identify emerging patterns and intervene earlier.

This data also supports what Barry Duncan and Scott Miller have termed "practice-based evidence" — using client outcome and process data to inform real-time clinical decisions. The distinction from evidence-based practice is subtle but important: rather than applying a standardized protocol and hoping it fits, practice-based evidence uses the individual client's data to customize and refine the approach. Between-session data makes this individualization far more granular and responsive.

There are legitimate concerns about data overload — therapists are already managing heavy caseloads and extensive documentation requirements. The key is not raw data volume but clinically actionable summaries. Tools that aggregate between-session data into pattern-level insights, highlight significant changes, and surface material that is most relevant to the current treatment focus respect the therapist's time while providing richer clinical context. Empath's AI-generated session preparation summaries represent this approach: not replacing clinical judgment, but providing the therapist with a more complete informational foundation from which to exercise that judgment.

Between-Session Processes Across Theoretical Orientations

The importance of between-session processes is not limited to cognitive-behavioral therapy. Psychodynamic therapy has long recognized that therapeutic change occurs between sessions through processes like internalization of the therapeutic relationship, working through of transference patterns in daily life, and dream processing. Contemporary psychodynamic researchers like Jonathan Shedler have argued that the internalization of the therapist's empathic stance — which continues to operate between sessions as a kind of internal working model update — is a central mechanism of change in long-term therapy.

In humanistic and person-centered approaches, between-session reflection is understood as part of the self-actualization process. Carl Rogers's concept of the "fully functioning person" implies that therapeutic conditions catalyze growth processes that operate continuously, not just during the therapy hour. Research by Robert Elliott on experiential therapy outcomes suggests that clients process and integrate emotional experiences raised in therapy over days and weeks, with key insights often emerging long after the session ends.

Emotion-focused therapy, developed by Leslie Greenberg, explicitly acknowledges that emotional processing is an ongoing process that extends beyond session boundaries. Clients who engage in between-session emotional reflection — whether through journaling, mindfulness, or simply allowing themselves to stay with feelings that emerged in therapy — show more complete emotional processing and better outcomes. The therapy session activates emotional schemas, but the between-session period is where those schemas are gradually reorganized.

Even in brief, solution-focused approaches, between-session change is formally built into the model. The "first session formula task" — asking clients to notice what is already working in their lives — is designed to shift attention and generate new data between the first and second sessions. Insoo Kim Berg and Steve de Shazer built their entire approach around the assumption that clients are constantly changing, and that the therapist's job is to notice and amplify those changes. Having actual data about between-session observations, rather than relying on recollection, would make this approach even more powerful.

Privacy, Trust, and the Clinical Frame

Any discussion of between-session data collection must address the legitimate concerns about privacy, boundaries, and the therapeutic frame. Clients need to feel that the therapy relationship is a safe space, and the introduction of data-sharing tools requires thoughtful clinical conversation about what will be shared, who has access, and how the data will be used. Informed consent should be explicit and ongoing — not a one-time checkbox.

Research on technology in therapy consistently shows that client receptivity depends on how the technology is framed by the therapist. A study by Clough and Casey found that clients who perceived technology as an extension of the therapeutic relationship, rather than a replacement for it, were significantly more likely to engage with digital tools and report positive experiences. The key is positioning between-session data capture as a collaborative tool that serves the client's goals, not as a monitoring or compliance mechanism.

Confidentiality is paramount. Between-session data often includes some of the most vulnerable material a client produces — real-time emotional reactions, private reflections, and unfiltered thoughts captured outside the structured safety of the therapy room. Any tool that handles this data must meet the highest standards of security and HIPAA compliance. Therapists should evaluate data platforms with the same scrutiny they would apply to any clinical tool that handles protected health information.

The therapeutic frame is strengthened, not weakened, when clients feel that their between-session experiences are seen and valued. Many clients report that sharing between-session data makes them feel more connected to the therapeutic process and more confident that their therapist understands the full scope of their experience. When implemented with sensitivity and clear boundaries, between-session data collection becomes an extension of the empathic attunement that defines effective therapy.

Practical Steps for Integrating Between-Session Data Into Your Practice

If you are interested in incorporating between-session data into your clinical workflow, start small. Choose one or two clients who are stable enough to experiment with new tools and whose presenting concerns would benefit most from between-session tracking — clients with mood disorders, anxiety, or behavioral activation goals are often ideal starting points. Discuss the rationale with them: "I want to understand your experience more fully, not just what you remember at the start of session, but what actually happens during your week."

Decide what data streams are most relevant to each client's treatment goals. For a client working on behavioral activation, activity level and mood tracking may be most useful. For a client in exposure therapy, anxiety ratings before and after feared situations provide critical data about habituation curves. For a client in emotion-focused work, journaling entries that capture emotional experiences in real time offer rich clinical material. Matching the data to the clinical need prevents overwhelm for both you and your client.

Build a session structure that incorporates data review without consuming the entire session. A practical approach is to spend the first five to seven minutes reviewing between-session patterns, identifying one or two themes or data points that warrant deeper exploration, and then transitioning into the core therapeutic work. This replaces the often-unfocused "How was your week?" with a targeted, data-informed opening that gets to clinical material faster.

Finally, monitor the impact on your own clinical decision-making. Notice whether between-session data changes how you conceptualize cases, whether it surfaces material that would not have emerged through standard session openings, and whether your clients experience the process as enhancing or interfering with the therapeutic relationship. Like any clinical intervention, between-session data collection should be evaluated against outcomes — not assumed to be beneficial simply because it provides more information.

Frequently asked questions

What is between-session data in therapy?

Between-session data refers to any information about a client's experience, mood, behavior, or reflections that occurs between scheduled therapy appointments. This can include mood tracking, journaling entries, sleep data, social engagement patterns, skill practice logs, and behavioral observations captured in real time rather than recalled retroactively at the start of the next session.

How much therapeutic change actually happens between sessions?

Research by Tang and DeRubeis on sudden gains found that roughly 50 percent of total symptom improvement in clients who experienced sudden gains occurred during a single between-session interval. Kazantzis's meta-analyses show that between-session homework adds a medium effect size (d = 0.53) to treatment outcomes. While the exact proportion varies by client and treatment modality, the evidence strongly supports that between-session intervals are where much of the meaningful change occurs.

Does collecting between-session data create ethical concerns?

Any data collection in therapy requires informed consent, clear boundaries about what is shared and how it is used, and robust privacy protections including HIPAA compliance. The key ethical principle is that between-session data should serve the client's therapeutic goals, not function as surveillance. Discussing the purpose, scope, and limitations of data collection openly with clients is essential.

Will reviewing between-session data take too much session time?

When done effectively, reviewing between-session data actually saves session time by replacing the often-unfocused session opening ("How was your week?") with a targeted, data-informed check-in. Most therapists find that five to seven minutes of data review at the start of session surfaces more clinically relevant material more quickly than traditional recall-based openings.

What types of clients benefit most from between-session data tracking?

Clients with mood disorders, anxiety, behavioral activation goals, and those in CBT or DBT who use structured skill practice tend to benefit most from systematic between-session tracking. However, clients in psychodynamic, emotion-focused, and humanistic therapies also benefit from real-time journaling and mood data that captures experiences as they occur rather than as they are later remembered.

How do I introduce between-session data collection to clients without it feeling intrusive?

Frame it collaboratively: "I want to understand your experience more fully between sessions so our time together is more focused and useful." Emphasize that the client controls what they share, that the data is for therapeutic purposes only, and that it is an extension of the work you are doing together — not a monitoring tool. Start with one simple data stream and build from there based on the client's comfort.

Does between-session data collection work across different therapy modalities?

Yes. While the specific data types may differ — mood tracking for CBT, emotional reflection journals for psychodynamic or emotion-focused work, skill practice logs for DBT — the principle of capturing between-session experience applies across theoretical orientations. Research on between-session processes has been conducted across cognitive-behavioral, psychodynamic, humanistic, and integrative approaches, with all showing that between-session periods are clinically significant.