Relational processes in face-to-face and videoconferencing psychiatric interviews with young people: a multimodal analysis

Introduction: Videoconferencing (VC) interviews are often judged clinically acceptable and comparable with face-to-face (FTF) encounters when overall alliance scores are considered [1]. However, similar alliance ratings may mask differences in how therapeutic relatedness is produced, maintained and experienced. This within-subject crossover study examined whether FTF and VC interviews differ in working alliance, interactional microprocesses and subjective relational experience among adolescents and emerging adults during early clinical encounters. We also explored whether attachment characteristics, which are associated with working alliance, could inform person-centred modality selection [4].

Methods: Twenty participants aged 15–29 years completed one FTF and one VC interview with the same therapist, with modality order counterbalanced. Working alliance was assessed immediately after each session using the Working Alliance Inventory-Short Form (WAI-SF) [2]. Quantitative analyses included Wilcoxon signed-rank tests, linear mixed models controlling for period effects, Bayesian analysis of modality differences, and the Jacobson–Truax Reliable Change Index to examine individual-level change [3]. Multimodal process analyses combined speaker-diarised acoustic measures, including fundamental frequency (F0) variability and interactional silence, with linguistic analyses of transcripts and large language model-assisted qualitative coding of participants’ modality-specific relational accounts [5].

Results: Total working alliance did not differ significantly between FTF and VC sessions (standardised mean difference=0.06, p=.732). Bayesian analysis provided moderate evidence favouring the absence of a modality effect on total alliance (BF01=4.15). At the subscale level, however, linear mixed modelling showed a significant FTF advantage for the Task dimension after adjustment for period effects (β=-1.27, p=.002). A robust period effect was observed: alliance ratings were substantially higher in the second session regardless of modality (standardised mean difference=0.88, p<.001), suggesting the importance of relational familiarisation across repeated encounters. Process analyses indicated modality-specific interactional demands. VC sessions showed a higher proportion of interactional silence than FTF sessions (27% vs 16%, standardised mean difference=1.46, p<.001). Therapist F0 variability was also significantly greater in VC (standardised mean difference>1.0, p<.001), consistent with increased vocal modulation when visual and embodied cues were reduced. Attachment avoidance showed a trend towards relatively stronger alliance in VC (Spearman’s ρ=-.43, p=.061). Qualitative coding indicated that participants with higher avoidant tendencies sometimes experienced VC as psychologically safer and less exposing, whereas several participants described an abrupt sense of disconnection after VC sessions ended. Subjective preference favoured FTF in 70% of participants, but preference did not consistently correspond to alliance scores.

Conclusion: In this small crossover study, FTF and VC interviews yielded comparable overall alliance ratings but differed in relational process. VC appeared to require greater therapist vocal adaptation and involved more interactional silence, while FTF showed an advantage in perceived task collaboration. These findings suggest that modality selection should not rely solely on aggregate outcome equivalence, but should consider session phase, therapeutic goals, therapist responsiveness and individual attachment-related comfort with interpersonal distance.

References

1. Seuling PD, Fendel JC, Spille L, Göritz AS, Schmidt S. Therapeutic alliance in videoconferencing psychotherapy compared to psychotherapy in person: a systematic review and meta-analysis. J Telemed Telecare. 2024;30(10):1521-1531.
2. Hatcher RL, Gillaspy JA. Development and validation of a revised short version of the Working Alliance Inventory. Psychother Res. 2006;16(1):12-25.
3. Jacobson NS, Truax P. Clinical significance: a statistical approach to defining meaningful change in psychotherapy research. J Consult Clin Psychol. 1991;59(1):12-19.
4. Notsu H, Blansfield RE, Spina DS, Levy KN. An updated meta-analysis of the relation between adult attachment style and working alliance. Psychother Res. 2025;35(5):721-734.
5. De Paoli S. Performing an inductive thematic analysis of semi-structured interviews with a large language model: an exploration and provocation on the limits of the approach. Soc Sci Comput Rev. 2024;42(4):997-1019.