Intersession Experience Questionnaire Short (IEQ-S): psychometric properties and longitudinal associations with symptoms of an 8-item short version of the Intersession Experience Questionnaire

Keywords

intersession experiences - psychotherapy process - routine outcome monitoring

Schlüsselwörter

Intersession Prozesse - Psychotherapieprozess - Routine Outcome Monitoring

Introduction

The systematic assessment of therapeutic processes and outcomes is increasingly recognized as a cornerstone of evidence-based clinical practice, as routine outcome monitoring (ROM) and feedback systems have shown benefits for the course and success of psychotherapy [1]. Economic measurement tools can help increase the already good acceptance of progress tracking. This facilitates treatment personalization by enabling therapists to make timely, evidence-informed decisions. However, brief instruments suitable for repeated measurements are not yet available for all relevant dimensions of psychotherapy. One such factor that lends itself to being harnessed in high-frequency monitoring and feedback systems but is scarcely represented within the international research landscape, are intersession experiences. Intersession experiences are thoughts, feelings, memories, and fantasies about therapy or the therapist that occur between sessions [2]. The concept is grounded in attachment theory, object relations theory, and mentalization/reflective functioning, but it is relevant across theoretical orientations and diagnoses to understand the translation of therapy contents into daily life. This aligns with the Generic Model of Psychotherapy [3], which emphasizes that lasting change arises from ongoing micro-outcomes that extend beyond sessions: As patients internalize therapeutic insights and apply them to real-life situations, they gain new experiences that contribute to personal growth and problem-solving capacities.

Intersession experiences have shown associations with patient characteristics, the current in-session process, the intensity of the therapeutic process, and the patient’s emotional involvement [4]. Positive intersession experiences are correlated with the therapeutic relationship [5] [6] [7] as well as treatment success [4] [5] [8]. While frequent assessments and longitudinal investigations are still scarce [8] [9], emerging evidence indicates that the emotional quality of intersession experiences is decisive for psychotherapeutic progress and outcome, also in the sense of tightly coupled temporal dynamics [10]. Questionnaires suitable for repeated measurements have been developed [2] [11] [12], among which the Intersession Experience Questionnaire (IEQ) [2] is the most widely used instrument. The original English version has been revised multiple times and includes 42 items. Hartmann, Orlinsky, Geller, et al. [12] validated a German translation that encompasses 52 items.

A more brief assessment would not only reduce patient burden but also enable rapid evaluations (to inform ROM), the implementation of intersession processes in more dense ambulatory assessment protocols, and use as a process variable, a perspective that is currently underrepresented [13].

To this aim, we examine the psychometric properties of a new, shorter instrument we call the Intersession Experience Questionnaire-Short (IEQ-S; German: Intersession-Fragebogen-kurz, ISF-K) and test expected positive associations with the longer, original IEQ, other constructs such as the working alliance, defense mechanisms, and symptoms (to support convergent validity); and comparatively weak associations with global assessments of mental and physical health (in terms of discriminant validity). Furthermore, we explore its sensitivity to changes over time, a prerequisite for effective monitoring, and how it relates to symptoms in a longitudinal assessment.

Methods

Design and Participants

Samples 1–5 were collected to support the initial validation of the IEQ-S. Participants filled out the full IEQ, the shorter IEQ-S, and further questionnaires once, approximately three weeks after therapy start. This was to ensure comparability between patients, as intersession experiences may shift over the course of psychotherapy [8]. Sample 6 differs regarding the design as this was a longitudinal assessment of only the IEQ-S and a symptom measure session by session.

Across samples and settings, patients were included in the study if they were at least 18 years old, had an ICD-diagnosis F. XX (mental and behavioral disorders), had participated in at least one past individual/group psychotherapy session. Patients were excluded if they had acute psychotic symptoms or did not speak German.

The Ethics Committee of the University of Klagenfurt approved the study in whose context samples 1–5 were collected (nr. 2018–084). The collection and use of the data for sample 6 for research purposes were approved by the ethics committee of the German Psychological Society (DGPs) in 2022. All assessments were conducted in accordance with the principles of the Declaration of Helsinki and all participants provided informed consent.

Instruments

The German version of the IEQ [12] was the basis for the development of the IEQ-S. Its items are organized into five item groups, three of which have subfactors (Suppl. Table 1). The items are rated on a five-point Likert scale from 0=not at all to 4=very often. The authors of the original IEQ suggest that the analysis be performed at the factor level.

For the IEQ-S item selection, we chose each item of every factor with the highest factor loadings based on the published psychometric values of the factor analyses [6] [12]. The original item group “significant others, sharing intersession experiences” showed poor psychometric properties. Further, it was not considered as directly relevant to the therapeutic relationship as other intersession experiences and could furthermore be confounded by the quality and quantity of patients' relationships outside of psychotherapy, so it was omitted. As a result, the IEQ-S encompasses eight items (Suppl. Table 2).

In addition to the IEQ and IEQ-S, patients in samples 1–5 filled out the following questionnaires:

The Working Alliance Inventory – Short revised (WAI-SR) [14] is an instrument for the assessment of therapeutic alliance and comprises a total of 12 items. Items are being answered on a five-point Likert scale (1–5). The WAI-SR assesses the three dimensions with four items each, and scale scores are computed as the mean of the respective items: agreement on the tasks of therapy, agreement on the goals of therapy, and development of an affective bond.

The Symptom-Checklist Short 9 (SCL-K-9) [15] is a one-dimensional 9-item short version of the SCL-90-R and assesses the symptom burden during the past 14 days. Items are answered on a five-point Likert scale (0–4), mirroring the extent to which individuals were burdened by the respective symptom, and then summed up. The sum score thus ranges from 0 to 36.

The Short Form 12 is a self-reported outcome measurement and the short version of the SF-36. It captures health on eight dimensions: physical functioning, role-physical, bodily pain, general health, vitality, social functioning, role-emotional, and mental health. It was also validated based on a representative German population sample [16]. In the present work, we concentrated on the scales physical and mental health (to investigate the IEQ-S' discriminant validity), which each contain two items. The scale values determined can range from 0 to 100 points, with low values reflecting poorer health and higher values reflecting better health.

The Defensive Style Questionnaire (DSQ-40) [17] is a self-report measurement for the assessment of one's own defense mechanisms. It is composed of a total of 40 items, which are responded to on a nine-point Likert scale, loading on a total of three factors: maladaptive defense, intermediate neurotic defense, and adaptive defense.

In sample 6, patients completed the depression module of the Patient Health Questionnaire (PHQ-9) [18] which assesses the DSM-IV diagnostic criteria for major depressive disorder. Respondents indicate how often they have been bothered by each symptom over the past two weeks, using a 4-point Likert scale (0=not at all to 3=nearly every day), yielding a total score between 0 and 27.

Statistical Analyses

For the investigation of the psychometric properties of the IEQ-S, we first used the cross-sectional data of samples 1–5. Exploratory factor analyses (EFA) were conducted on the combined data. All information regarding the EFA is provided in the Supplemental Material (Suppl. Table 3).

Internal consistency estimates were calculated as McDonald’s omega (ω), which we preferred over Cronbach’s α as it does not assume tau-equivalence and provides a more accurate reliability estimate. Values of≥.70 are generally interpreted as acceptable [19].

Based on sample 6 that included longitudinal assessments, we calculated intraclass correlations (ICCs) to quantify the proportion of variance attributable to between-person differences versus within-person fluctuations over time, using unconditional random intercept models. Higher ICC values indicate that a larger proportion of the variance is stable across persons, whereas lower ICC values suggest greater within-person variability (i. e., state-like fluctuations). To further capture temporal dynamics, we calculated mean squared successive differences (MSSD), which quantify the magnitude of short-term variability by examining the squared differences between consecutive data points. MSSD values were first computed at the person level and then averaged across the sample. Higher MSSD values suggest comparatively more dynamic variation in these intersession experiences, whereas lower values indicate more temporal stability. We also calculated repeated-measures correlations of the IEQ-S and the PHQ-9 sum score. The repeated-measures correlation coefficient quantifies this association while accounting for the non-independence of observations within individuals.

Data were analyzed using R statistics version 4.5.0 [20] using the packages dplyr, esmpack, ggplot2, nlme, psych, remotes, and rmcorr.

Results

Sample characteristics

The combined sample for the psychometric analyses included 237 patients from six clinical settings in Austria and Germany. [Table 1] depicts the separate samples. Sample 1 is composed of n=37 inpatients from an Austrian private clinic. Patients are admitted for six weeks and receive multimodal treatment. The items were answered in relation to the individual therapy. Sample 2 (n=58) was collected at an Austrian inpatient rehabilitation clinic where patients were also admitted for six weeks and underwent a multimodal therapy program. The questionnaires also referred to the individual therapy. Sample 3 is composed of n=27 outpatients from an Austrian outpatient psychotherapy center.

Table 1 Characteristics of the combined sample and the six subsamples.

Sample		Combined samples 1–6	1: Private Clinic	2: Rehabilitation Center	3: Psychotherapy Center I	4: Student Counseling Center	5: Day Clinic	6: Psychotherapy Center II
Country			Austria	Austria	Austria	Austria	Germany	Austria
Setting			inpatients	inpatients	outpatients	outpatients	outpatients	outpatients
Therapy			individual	individual	individual	individual	group	individual
n		237	37	58	27	20	23	72
Age	M (SD)	43.59 (11.62)	51 (9.00)	46.65 (9.71)	42.39 (8.23)	28.75 (7.51)	47.17 (12.54)	40.69 (13.21)
	Minimum – Maximum	18–81	32–68	25–64	27–55	21–53	23–65	18–81
Gender (n,%)	Women	133 (56.12)	15 (40.54)	37 (63.80)	20 (74.07)	–	13 (56.52)	48 (66.7)
	Men	75 (31.65)	18 (48.65)	21 (36.20)	5 (18.52)	–	10 (43.84)	21 (29.2)
Primary Diagnosis (ICD-10) (n,%)	F3	129 (54.4)	25 (67.57)	40 (68.97)	19 (70.37)	2 (10.00)	21 (91.30)	22 (30.6)
	F4	52 (21.9)	7 (18.92)	15 (25.86)	2 (7.41)	13 (65.00)	2 (8.70)	13 (18.1)
	Other	40 (16.9)	1 (2.70)	1 (1.72)	2 (7.41)	–	–	36 (50.0)
	Not available	16 (6.8)	4 (10.81)	2 (3.50)	4 (14.81)	5 (25.00)	–	1 (1.4)

Note. Diagnoses of International Statistical Classification of Diseases and Related Health Problems 10^th Revision (ICD-10): F3: affective disorders; F4: neurotic, stress-related, and somatoform disorders.

The fourth sample comprised n=20 outpatients from a psychological counseling center for students. On average, students are in individual treatment for six months. Sample 5 consists of n=23 patients from a German day clinic. They answered the items referring to their group therapy.

Sample 6, which was collected longitudinally, is based on routine clinical data from the Psychotherapeutic Research and Teaching Outpatient Centre (Psychotherapeutische Forschungs- und Lehrambulanz der Universität Klagenfurt, PUK) of the University of Klagenfurt, which offers psychodynamic psychotherapy to adults free of charge. The dataset included 2,119 sessions from 72 patients (M=29.43; SD=27.35 sessions/patient).

IEQ and IEQ-S descriptives

For each variable within the cross-sectional dataset (samples 1–5), the rate of missing values was between 0% and 4.2%. The missing completely at random test (MCAR-Test) shows that the missing values were missing completely at random (χ ²(5287)=4484.34, p=0.99). Therefore, we used an iterative maximum-likelihood method, the expectation maximization (EM) algorithm, to impute missing values. The descriptive data for each questionnaire are shown in Suppl. Table 4. The mean reports on the IEQ-S items are shown in [Fig. 1], indicating variation in how frequently different intersession experiences were reported.

Fig. 1 Participants’ responses on the eight Intersession Questionnaire – Short items. Legend: Each box represents the distribution of responses for a single item. The horizontal line inside each box indicates the median, while the box boundaries reflect the interquartile range (IQR). Whiskers extend to 1.5×IQR. Items 1, 2, 5, and 7 exhibit relatively high median values and narrow interquartile ranges, indicating these types of experiences were reported more frequently and consistently. In contrast, items such as 3, 4, 6, and 8 show lower median scores and broader or more skewed distributions, suggesting they occurred less frequently or were more variable between individuals. Responses to items 3 and 6 are strongly skewed toward the lower end of the scale, with the majority of participants reporting they occurred rarely or never.

Factor analyses and internal consistency

Exploratory factor analyses supported a three-factor solution that explained 42% of the variance and showed excellent fit, χ²(7)=5.18, p=0.639. A screeplot also supported a one-factor solution, but this explained less variance (28%) and showed worse fit, χ²(20)=38.72, p=0.007.

The correlations between the IEQ-S items are shown in Suppl. Table 5. The internal consistency of the IEQ-S is ω=0.72, which can be seen as acceptable and allows for the calculation of a sum score. We also calculated the internal consistencies of the longer, original IEQ factors (Suppl. Table 1).

Convergent Validity

We examined the correlations between the overall mean score of the IEQ and IEQ-S. It was r=0.70, p<0.001 (Suppl. Figure 1). A Bland-Altman Plot visualizes their agreement as well (Suppl. Figure 2). All IEQ-S items were correlated with their original factors (item 1 with item group A: r=0.592; p<0.001; item 2 with B1: r=0.638; p<0.001; item 3 with B2: r=0.489; p<0.001; item 4 with C1: r=0.561; p<0.001; item 5 with C2: r=0.671; p<0.001; item 6 with C3: r=0.755; p<.001; item 7 with D1: r=0.530; p<0.001; and item 8 with D2: r=0.551; p<0.001). The correlation pattern showed that each item correlated most strongly with its conceptually intended factor (Suppl. Table 6).

The sum score of the IEQ-S showed correlations with the WAI-SR subscales (bond: r=0.26, p=0.002; task: r=0.57, p<0.001; goal: r=0.45, p<0.001), the SCL-K-9 (r=0.23, p=0.002), and neurotic defense mechanisms (r=0.26, p=0.002).

Discriminant Validity

There were no correlations between the IEQ-S and patients' health (physical health r=−0 .03, p=0.76; mental health r=0.002, p=0.98).

Longitudinal analyses including the association with depression symptoms

The ICCs indicated that a substantial proportion of variance in the IEQ-S sum score and items was due to within-person changes over time, but the proportion varied (Suppl. Table 7).

Lastly, we checked repeated-measures correlations of the PHQ-9 sum score with the IEQ-S sum score (r=0.130, p<0.001) and the IEQ-S single items (item 1: r=0.192, p<0.001; item 2: r=0.111, p<0.001; item 3: r=0.086, p<0.001; item 4: r=0.105, p<0.001; item 5: r=− 0.109, p<0.001; item 6: r=0.131, p<0.001; item 7: r=− 0.171, p<0.001; item 8: r=0.210, p<0.001). The largest correlation is visualized in Suppl. Figure 3.

Discussion

This study aimed to evaluate a short form of the Intersession Experience Questionnaire, the IEQ-S, for use in both psychotherapy research and routine clinical monitoring. The goal was to create a time-efficient instrument while preserving the breadth of the original IEQ.

Accordingly, the present results show that the items capture related but not redundant aspects of intersession activity. For example, some processes happened less often than others. Especially scarce were dreams about therapy or therapists (item 3). While they can be an indicator of how engaged a patient is with therapy; dreams are unconscious or at least preconscious, which is why they might not be accessible to conscious recollection. Further, in adapting the item for the short version, the original wording (“thinking about therapy/therapist in a dream”) was simplified to “dreaming of therapy or the therapist”, which changes the scope of the item.

Item 6 is about wondering whether the therapist is thinking about the patient, which involves reflective functioning, the capacity to consider the mental states of others [21]. This ability, however, may be limited, especially during times of acute distress, making such reflections more difficult or even inaccessible. Item 8 captures a negative connotation and differs from the other items already in terms of polarity. We opted to include it, as information about negative intersession experiences is also relevant, and uncomfortable feelings can be a natural part of the therapeutic process, which often involves working through painful experiences and tolerating temporary distress. Indeed, psychotherapy tends to be most effective when it maintains a balance between emotional challenge and a sense of safety [22] [23]. As items 7 and 8, capturing emotionally positive and negative experiences, were only weakly correlated, this indicates they are not just two sides of the same coin. This perspective is supported by the longitudinal findings from Sample 6: Particularly items capturing negative emotions or disruptions covaried more strongly with symptom severity, suggesting that the single items' clinical relevance is not uniform [8] [10]. Additionally, the ICCs for the IEQ-S sum score and individual items indicated both between- and within-person variability, and MSSD values showed item-level variation in within-person dynamics, meaning that some intersession experiences are more volatile than others.

Taking into account the comparisons with the full IEQ, the IEQ-S can be regarded as a pragmatic compromise that balances efficiency and conceptual coverage. Although it might not replace the long version in all respects, in many clinical settings, including those represented by samples 1–5 with their diverse inpatient and outpatient contexts, the short version may be particularly attractive, as it can be more easily embedded into routine clinical processes.

However, we recommend still focusing on the eight individual items rather than computing a sum score, as collapsing across conceptually distinct processes risks obscuring critical patterns of variation. Intersession experiences are, by definition, multidimensional; they include affective, cognitive, imaginative, and interpersonal phenomena that vary in form, function, and clinical relevance [2] [12]. In future research, item-/profile-based approaches may yield more fine-grained insights, particularly in applied settings where early detection of change is essential. One key direction is thus to examine the predictive validity of specific items regarding outcomes such as symptom reduction, alliance, or dropout. If only a subset consistently shows clinical relevance, further shortening may be warranted.

Limitations

The first and largest limitation in this study is the sample. Although the different settings and patient groups add to the investigation's external validity, individual subsamples were too small and too different to systematically investigate the effects of their characteristics, and there was very little information about patients and therapists. In future studies, a sample with more homogeneous patients and therapists should be examined. Second, it is necessary to consider recall and biases as intersession experiences might be fleeting and therefore not be (accurately) remembered. Third, it is difficult to identify an appropriate measurement point. Patients in samples 1–5 participated about three weeks after admission, when the therapeutic relationship may not be as evolved as at the end of therapy. This could explain the low correlations with the WAI-SR. Future studies should add more systematic comparisons of therapy phases.

Conclusion

The IEQ-S appears to be a psychometrically adequate and clinically promising tool for capturing the rich, dynamic field of intersession experiences. By preserving the heterogeneity of the original measure in a concise format, the IEQ-S provides both a practical solution for repeated assessment and an informative, complementary perspective on the therapy process. We suggest that clinical applications and future studies move beyond a sum-score-centric approach and leverage the diagnostic and prognostic value embedded in the diversity of intersession experiences.

Clinical implications

The IEQ-S offers a brief, reliable, and valid tool to monitor patients’ intersession experiences session-by-session. Tracking these thoughts and feelings between sessions can enhance understanding of patient engagement, detect ruptures or negative dynamics early, and inform timely therapeutic interventions. Its brevity makes it feasible for routine use, supporting feedback-informed care without adding undue burden.

• References

• 1 McAleavey AA, de Jong K, Nissen-Lie HA. et al. Routine outcome monitoring and clinical feedback in psychotherapy: Recent advances and future directions. Administration and Policy in Mental Health and Mental Health Services Research 2024; 51: 291-305
• 2 Orlinsky DE, Geller JD. Patients’ representations of their therapists and therapy: New measures. In Miller NE, Luborsky L, Barber JP, Docherty JP (eds.), Psychodynamic treatment research: A handbook for clinical practice. New York: Basic Books 1993: 423-466
• 3 Orlinsky DE. The “Generic Model of Psychotherapy” after 25 years: Evolution of a research-based metatheory. Journal of Psychotherapy Integration 2009; 19: 319
• 4 Gablonski TC, Herrmann PL, Lüdemann J. et al. Intersession experiences and internalized representations of psychotherapy: A scoping review. Journal of Clinical Psychology 2023; 79: 1875-1901
• 5 Andreas S, Gablonski T-C, Hiesberger S. et al. Intersession-Prozesse in stationärer psychotherapeutischer Behandlung: Ist die Einzel-oder die Gruppentherapie entscheidender für die therapeutische Beziehung und das Therapieergebnis?. PDP-Psychodynamische Psychotherapie 2016; 15: 206-218
• 6 Hartmann A, Orlinsky D, Zeeck A. The structure of intersession experience in psychotherapy and its relation to the therapeutic alliance. J Clin Psychol 2011; 67: 1044-1063
• 7 Owen J, Quirk K, Hilsenroth MJ. et al. Working through: In-session processes that promote between-session thoughts and activities. Journal of counseling psychology 2012; 59: 161
• 8 Hartmann A, Zeeck A, Herzog W. et al. The intersession process in psychotherapy for anorexia nervosa: Characteristics and Relation to Outcome. J Clin Psychol 2016; 72: 861-879
• 9 Gablonski TC, Senft B, Andreas S. Level of structural integration and its association with intersession experiences and outcomes: A pilot study. Int J Environ Res Public Health 2022; 19: 9254
• 10 Kaiser T, Laireiter AR. Daily dynamic assessment and modelling of intersession processes in ambulatory psychotherapy: A proof of concept study. Psychother Res 2019; 29: 1062-1073
• 11 Gablonski T-C, Pryss R, Probst T. et al. Intersession-online: A smartphone application for systematic recording and controlling of intersession experiences in psychotherapy. J 2019; 2: 480-495
• 12 Hartmann A, Orlinsky DE, Geller JD. et al. The Intersession Questionnaire--an inventory for the measurement of patients' processing of psychotherapy between sessions. Psychother Psych Med 2003; 53: 464-468
• 13 Mink F, Lutz W, Hehlmann MI. Ecological Momentary Assessment in psychotherapy research: A systematic review. Clinical Psychology Review. 2025: 102565
• 14 Wilmers F, Munder T, Leonhart R. et al. Die deutschsprachige Version des Working Alliance Inventory-short revised (WAI-SR)-Ein schulenübergreifendes, ökonomisches und empirisch validiertes Instrument zur Erfassung der therapeutischen Allianz. Klinische Diagnostik und Evaluation 2008; 1: 343-358
• 15 Klaghofer R, Brähler E. Konstruktion und Teststatistische Prüfung einer Kurzform der SCL-90–R. Zeitschrift für Klinische Psychologie. Psychiatrie und Psychotherapie. 2001 49. 115-124
• 16 Drixler K, Morfeld M, Glaesmer H. et al. Validierung der Messung gesundheitsbezogener Lebensqualität mittels des Short-Form-Health-Survey-12 (SF-12 Version 2.0) in einer deutschen Normstichprobe. Zeitschrift fur Psychosomatische Medizin und Psychotherapie 2020; 66: 272-286
• 17 Schauenburg H, Willenborg V, Sammet I. et al. Self-reported defence mechanisms as an outcome measure in psychotherapy: A study on the German version of the Defence Style Questionnaire DSQ 40. Psychology and Psychotherapy: Theory, Research and Practice 2007; 80: 355-366
• 18 Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med 2001; 16: 606-613
• 19 Hayes AF, Coutts JJ. Use omega rather than Cronbach’s alpha for estimating reliability. But…. Communication Methods and Measures 2020; 14: 1-24
• 20 R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing. Vienna: 2024. https://www.R-project.org
• 21 Bateman A, Fonagy P, Campbell C. et al. Cambridge guide to mentalization-based treatment (MBT). Cambridge University Press; 2023
• 22 Sønderland NM, Solbakken OA, Eilertsen DE. et al. Emotional changes and outcomes in psychotherapy: A systematic review and meta-analysis. J Consult Clin Psychol 2024; 92: 654-670
• 23 Ernst M, Krakau L, Hettich N. et al. Emotional awareness and processing in (psychodynamic) psychotherapy. In: Samson AC, Sander D, Kramer U (eds). Change in Emotion and Mental Health. London: Academic Press; 2024: 241-258