Group Psychotherapy

Abstract

Group psychotherapy is a form of socioenvironmental therapy in which a small group of patients, guided by one or more therapists, becomes the instrument of treatment: members change through their interactions with one another as much as through the therapist. Jacob Moreno coined the term in the 1930s, but it was Irvin Yalom who gave the field its organizing framework — eleven therapeutic factors, from universality and altruism to interpersonal learning, and the idea of the group as a social microcosm in which each member's habitual ways of relating reappear and can be examined. Cohesion, the group analogue of the therapeutic alliance, is the precondition that lets those factors operate. Modern meta-analyses find group and individual formats roughly equivalent in efficacy, at markedly lower cost per patient.

Keywords: group psychotherapy, therapeutic factors, group cohesion, social microcosm, group development

What Group Psychotherapy Is

Group psychotherapy is the treatment of psychological difficulty in a small group — typically six to ten patients meeting regularly with one or two therapists — organized so that the interactions among members are themselves the medium of change. In the MeSH definition it is a form of psychotherapy in which several patients are treated together, and MeSH files it as a kind of socioenvironmental therapy, a treatment that works through the patient's social surroundings. What distinguishes it from individual therapy delivered to several people at once is that the other members are not an audience but the active ingredient: they offer feedback, model alternatives, and reproduce among themselves the very relationship problems that brought each of them to treatment (Yalom & Leszcz, 2020).

The organizing idea is that a therapy group becomes a social microcosm. Given enough time and freedom, members come to behave in the group as they behave in the world outside it — the domineering member dominates, the withdrawn member withdraws, the placating member placates — so that each person's characteristic way of relating is displayed in the room where it can be observed, named, and revised. The group is thus both a sample of the member's social life and a laboratory for changing it, and the therapist's task is less to interpret from outside than to help the group use what it generates (Yalom & Leszcz, 2020).

Because the mechanism is interpersonal, group psychotherapy poses a question of its own: what is it about being in a group that heals? Most of the field's history is the working-out of that question — first as a practical discovery that patients treated together improved, then as an articulated account of the specific factors a group supplies, and more recently as an empirical programme testing which of those factors carry the effect (Rosendahl et al., 2021).

Origins: From Pratt's Classes to Yalom's Framework

Group treatment began as a practical expedient. In 1907 the Boston physician Joseph Pratt gathered tuberculosis patients too poor for sanatorium care into weekly classes, where he found that the shared instruction and mutual encouragement of the group improved not only their adherence but their morale — the first documented use of a group as a therapeutic instrument, though Pratt understood it as education rather than psychotherapy (Pratt, 1907). The insight that the group itself did something a lecture could not was the seed of everything that followed.

The word came from Jacob Moreno, who used group psychotherapy in the 1930s and built around it the methods of psychodrama and sociometry — the mapping of the attractions and rejections that structure any group (Moreno, 1934). At almost the same time Kurt Lewin was founding the field of group dynamics, arguing that a group is a dynamic whole with properties not reducible to its members and that behaviour is a function of the person within a social field — the theoretical substrate on which later group therapists would draw (Lewin, 1947).

Two psychoanalytic traditions then gave the group depth. At the Tavistock Clinic, Wilfred Bion described the unconscious life of groups, distinguishing the work group that pursues its task from the basic-assumption states — dependency, fight-flight, and pairing — into which a group retreats when anxiety overwhelms it (Bion, 1961). Independently, Siegmund Foulkes developed group analysis, conceiving the group as a matrix of communication in which the individual is a nodal point and treatment proceeds through the network of relationships rather than the therapist alone (Foulkes, 1948). It was Irvin Yalom, writing from the 1970s onward, who synthesized these strands into the framework that organizes the field today: an interpersonal account in which the group is a social microcosm and its healing is carried by a specifiable set of therapeutic factors (Yalom & Leszcz, 2020).

The Therapeutic Factors

Yalom's central contribution was to decompose the vague notion that groups help into a list of therapeutic factors — the distinct mechanisms through which a group produces change. Eleven are conventionally named, among them the instillation of hope (seeing others improve); universality (the relief of discovering one's troubles are shared); the imparting of information; altruism (being useful to others); the corrective recapitulation of the primary family group; the development of socializing techniques; imitative behaviour; interpersonal learning; group cohesiveness; catharsis; and the existential factors of confronting mortality, freedom, and responsibility (Yalom & Leszcz, 2020).

The factors are not equal, nor independent. Yalom placed interpersonal learning and cohesiveness at the centre: cohesion is the condition under which the others operate, and interpersonal learning — the cycle in which a member enacts a habitual pattern, receives feedback, and experiments with something new — is the group's distinctive engine. The others enter the work at different times and for different members; universality tends to matter most early, when a member first discovers they are not alone, while interpersonal learning does its work later, once cohesion is secure enough to bear confrontation (Yalom & Leszcz, 2020).

That the factors can be named means they can be measured. The Therapeutic Factors Inventory operationalized Yalom's constructs into a self-report scale, letting researchers ask which factors members actually experience and how those experiences relate to outcome (Lese & MacNair-Semands, 2000). Contemporary studies use such measures to test the account in specific populations — for example, tracing which factors high-risk adolescents in an intensive mentalization-based programme report as most helpful, and finding cohesion and interpersonal learning prominent among them (Hauber et al., 2019). The first demonstration below builds the factors into a simple model in which cohesion gates the effect of the rest.

Types of Group Psychotherapy

MeSH files group psychotherapy beneath socioenvironmental therapy and gives it four narrower descriptors — the direct children of the Psychotherapy, Group heading in the thesaurus. These are not competing schools so much as formats distinguished by who is in the room: whether the members are strangers assembled for treatment, a couple, a family, or a training group convened to learn about group process. The classification is a MeSH indexing scheme for retrieving literature, not a theory of group therapy, and the categories can overlap in practice — a family therapy may draw on psychodramatic techniques, and a sensitivity-training group shares much with an unstructured therapy group.

Of the four, psychodrama and sensitivity training groups are treated as pillar topics here; couples and family therapy are large enough fields to stand on their own and are noted for completeness. What unites all four under the group heading is the premise that treatment happens through a structured set of relationships rather than a single dyad (Yalom & Leszcz, 2020).

Cohesion and Group Development

If one factor has emerged from the research as central, it is cohesion — the group's analogue of the individual therapeutic alliance. Cohesion is the sense of solidarity, belonging, and engagement that binds members to the group and to one another, and it is both an outcome the group must achieve and a condition for everything else it does. A meta-analysis pooling decades of studies found a robust positive relationship between cohesion and treatment outcome across group types and disorders, comparable in strength to the alliance-outcome relationship in individual therapy — the clearest quantitative evidence that the quality of the group bond is not a byproduct of treatment but a mechanism of it (Burlingame et al., 2018).

Cohesion is not present at the start; it is built, and it develops in a characteristic sequence. Bruce Tuckman's synthesis of the small-group literature named four stages every group tends to pass through: forming, the tentative, dependent early phase in which members orient themselves and look to the leader; storming, a phase of conflict and resistance as members test the group and one another; norming, in which the group settles shared norms and cohesion consolidates; and performing, the mature phase in which the group works productively on its task (Tuckman, 1965). The sequence matters clinically because the same intervention lands differently at different stages: confrontation that a performing group metabolizes into learning can fracture a group still forming, and the conflict of the storming phase, often mistaken for failure, is a normal and necessary passage toward cohesion. The second demonstration traces this developmental curve, and the third turns to the format question that group therapy's efficiency raises.

The Group in Motion

The three demonstrations below make the group's dynamics manipulable. The first builds Yalom's therapeutic factors into a model in which cohesion gates the rest. The second traces Tuckman's developmental sequence, showing how a group's productive capacity dips through storming before rising to performing. The third makes the group-versus-individual comparison concrete, weighing roughly equal efficacy against the therapist time each format consumes.

Set the strength of cohesion and of the other therapeutic factors. Therapeutic potency is P = cohesion × mean(other factors): cohesion multiplies the rest, encoding Yalom’s claim that it is the precondition under which the other factors work.

Potency P = 0.637 (cohesion 0.85 × factor mean 0.75). Mature, potent group — cohesion lets the whole set of factors operate. Because cohesion multiplies rather than adds, no strength of the other factors can substitute for it: with cohesion near zero the group does little, however rich its factors.

The factors demonstration builds Yalom's account into a model of therapeutic potency. Adjusting cohesion and the strength of the other factors shows how the first gates the rest — a group without cohesion gains little from catharsis or interpersonal learning — so that potency rises steeply only once the group bond is in place.

Move through the life of a group. Cohesion rises steadily, but productive capacity falls as the group enters the conflict of storming and recovers as norming consolidates the bond — which is why the storming phase is a passage, not a failure.

Session 6 — Storming: conflict and testing; productivity dips but the work is real. Cohesion = 0.24, productive capacity = 0.08. The conflict of storming lowers what the group can accomplish without lowering the value of passing through it.

The development demonstration traces Tuckman's four stages across the life of a group. Moving through the sessions shows the group's productive capacity fall as it enters the conflict of storming and rise again as norming consolidates cohesion and the group reaches performing, making visible why the storming phase is a passage rather than a failure.

Meta-analysis finds the two formats roughly equivalent in outcome. Vary the group size to see efficacy hold nearly constant while the therapist time spent per patient falls as 1 / n — the source of the group’s efficiency.

Individual: 90 therapist-minutes per patient (d = 0.58). Group of 8: 11.3 minutes per patient (d = 0.55). At roughly equal efficacy, the group serves 8 patients in the therapist time an individual session gives to one — equivalence, not compromise.

The format demonstration weighs the two ways of delivering therapy. Varying the group size holds efficacy roughly constant — the meta-analytic finding of equivalence — while the therapist hours spent per patient fall as the group grows, showing why the group format is an efficient use of scarce clinical time rather than a dilution of care.

Worked Example

Take the therapeutic-potency model the first demonstration draws from Yalom's account. Score each of a set of therapeutic factors on a 0-to-1 scale for how strongly the group supplies it, and let cohesion act as a gate on the rest: potency P = cohesion × mean(other factors). The gate encodes Yalom's claim that cohesion is the precondition under which the other factors operate — without it, catharsis and confrontation do little.

Consider a newly formed group in its second session. The members feel some hope from seeing others show up and some relief of universality, but the group is not yet a group: cohesion is low. Say the other factors average to a moderate 0.55 — hope, universality, and information are present, but interpersonal learning and catharsis have barely begun — while cohesion sits at 0.20. Then P = 0.20 × 0.55 = 0.11. The group is doing a little good, but the low cohesion caps it: the factors are present in principle and inert in practice.

Now take the same group six months on. Cohesion has been built through the storming and norming it has passed, and now stands at 0.85; with a secure bond, interpersonal learning and catharsis have come alive, raising the mean of the other factors to 0.75. Then P = 0.85 × 0.75 = 0.6375. The single number captures the change: not because any one factor was maximized, but because cohesion rose enough to let the whole set operate. Because cohesion multiplies rather than adds, no amount of the other factors can substitute for it — a group with average factors of 1.0 but cohesion of 0.10 would still score only 0.10, which is Yalom's point that cohesion comes first (Burlingame et al., 2018).

Discussion

Group psychotherapy occupies an unusual place among treatments because its active ingredient is a set of relationships rather than a technique a therapist applies. That gives it a distinctive reach: a member is treated not only by the therapist but by the whole group, and the treatment touches capacities — being useful to others, tolerating confrontation, seeing oneself as others see one — that a private session reaches only indirectly (Yalom & Leszcz, 2020). The corresponding difficulty is that a group is a complex, multiply-determined system, so isolating what within it produces change has been the field's central problem. Yalom's therapeutic factors, the cohesion research, and Tuckman's developmental stages are successive attempts to decompose the whole group into nameable, measurable parts (Rosendahl et al., 2021).

For cognitive psychology the interest of group therapy is that it treats social behaviour as situated and revisable — as arising from the continuous interaction between a person and a structured social field rather than from the individual alone. The social-microcosm idea is, in effect, a claim about social cognition: that a person's model of how relationships work is enacted rather than merely reported, that it becomes visible in behaviour, and that feedback within a valued group is the condition under which it can be updated (Yalom & Leszcz, 2020). The finding that cohesion must come first sharpens the same point: the group is an engineered environment for interpersonal learning, and cohesion is the precondition for a member's willingness to expose and revise how they relate (Burlingame et al., 2018).

Current Directions

The contemporary evidence base has moved from asking whether groups help to specifying where, how much, and by what mechanism. On efficacy, the durable finding is equivalence: a meta-analysis directly comparing individual and group formats found no reliable overall difference in outcome, a result that reframes the group format as a clinically equivalent and markedly more efficient use of therapist time rather than a compromise (McRoberts et al., 1998). Disorder-specific reviews have since tested the claim in controlled designs: a meta-analysis of randomized trials of group psychotherapy for social anxiety disorder found it efficacious against control conditions, giving the format a firm evidence base in one of the conditions for which it is most often used (Barkowski et al., 2016).

A second strand returns to process, asking which of the group's factors carry the effect. The cohesion meta-analysis established the group bond as the strongest process predictor of outcome and pressed the field to treat cohesion as a target of intervention rather than a happy accident (Burlingame et al., 2018). Newer work extends the therapeutic-factors programme into specific and difficult populations — tracing, for instance, which factors high-risk adolescents in intensive programmes report as most helpful (Hauber et al., 2019) — while broad reviews of the field's research front consolidate these strands and map the open questions: how to measure group processes well, how to match format to disorder, and how to train group therapists in the skills the process research identifies (Rosendahl et al., 2021).

Common Misconceptions

Group therapy is just individual therapy done with several people to save money.

The economy is real but incidental. The defining feature of group psychotherapy is that the members treat one another — through feedback, modelling, and the enactment of relationship patterns in the group — so the other members are the active ingredient, not an audience for a therapist working one patient at a time (Yalom & Leszcz, 2020).

Conflict in a group means the group is failing.

Conflict is usually a developmental stage, not a breakdown. Tuckman's storming phase — a period of testing and friction — is a normal and often necessary passage on the way to the cohesion of norming and the productive work of performing; groups that never storm may simply be avoiding the interpersonal honesty the work requires (Tuckman, 1965).

Group therapy is less effective than individual therapy.

Meta-analytic evidence finds them roughly equivalent. A direct comparison of the two formats found no reliable overall difference in outcome, which makes the group an equally effective and more efficient option for many conditions rather than a second-best one (McRoberts et al., 1998).

The therapist does all the therapeutic work.

In a group the therapist is a facilitator of a process the members largely sustain. Much of the change is carried by group-level factors — cohesion, universality, altruism, interpersonal feedback — that the members generate among themselves, with the therapist shaping conditions rather than dispensing all the treatment (Burlingame et al., 2018).

Glossary

Basic-assumption group.

Bion's term for the unconscious states — dependency, fight-flight, and pairing — into which a group retreats from its task when anxiety overwhelms it, as opposed to the work group that pursues its aim.

Cohesion.

The solidarity, belonging, and engagement that bind members to a group; the group analogue of the individual therapeutic alliance and the strongest process predictor of outcome.

Forming, storming, norming, performing.

Tuckman's four stages of small-group development: initial orientation and dependence, conflict and testing, the settling of shared norms, and mature productive work.

Group analysis.

Foulkes's tradition of group psychotherapy, which conceives the group as a matrix of communication in which the individual is a nodal point and change proceeds through the network of relationships.

Group dynamics.

Lewin's field studying the group as a dynamic whole with properties not reducible to its members; the social-psychological substrate of group treatment.

Interpersonal learning.

One of Yalom's therapeutic factors and the group's distinctive engine: the cycle in which a member enacts a habitual pattern, receives feedback, and experiments with a new way of relating.

Psychodrama.

Moreno's method in which members enact scenes from their lives, using role-play and role reversal to explore and revise how they relate; a MeSH child of group psychotherapy.

Sensitivity training group.

A group convened chiefly to heighten members' awareness of themselves and of group process, on the boundary between treatment and personal development; a MeSH child of group psychotherapy.

Social microcosm.

Yalom's idea that, given time and freedom, members come to behave in the group as they do in the world outside it, so that each person's characteristic way of relating is displayed and can be worked on.

Socioenvironmental therapy.

The broad MeSH class of treatments that work through a patient's social and physical environment, of which group psychotherapy is a member.

Sociometry.

Moreno's method for mapping the attractions and rejections that structure a group, quantifying the pattern of interpersonal choice among its members.

Therapeutic Factors Inventory.

A self-report scale developed by Lese and MacNair-Semands that operationalizes Yalom's therapeutic factors, allowing them to be measured and related to outcome.

Therapeutic factors.

The distinct mechanisms through which a group produces change — among them universality, altruism, instillation of hope, interpersonal learning, cohesion, and catharsis — as named by Yalom.

Universality.

One of Yalom's therapeutic factors: the relief a member feels on discovering that their troubles are shared rather than uniquely theirs, typically most powerful early in a group's life.

Key Researchers

Wilfred R. Bion. Psychoanalyst (1897-1979) at the Tavistock Clinic whose Experiences in Groups described the unconscious basic-assumption states — dependency, fight-flight, and pairing — that pull a group away from its task. Wikipedia - Wikidata

Gary M. Burlingame. Group-psychotherapy researcher at Brigham Young University whose meta-analytic work established cohesion as the strongest process predictor of group outcome and who co-authored the field's evidence-based practice guidelines. ORCID - Faculty page - Wikidata

Siegmund H. Foulkes. Psychoanalyst (1898-1976) who founded group analysis, conceiving the group as a matrix of communication in which the individual is a nodal point and treatment proceeds through the whole network of relationships. Wikipedia - Wikidata

Molyn Leszcz. Psychiatrist at the University of Toronto and past president of the American Group Psychotherapy Association who, as co-author of the 5th and 6th editions of Yalom's text, carried the therapeutic-factors framework into contemporary evidence-based practice. ORCID - Faculty page

Kurt Lewin. Psychologist (1890-1947) who founded the field of group dynamics, arguing that a group is a dynamic whole not reducible to its members and that behaviour is a function of the person within a social field. Wikipedia - Wikidata

Jacob L. Moreno. Psychiatrist (1889-1974) who coined the term group psychotherapy and founded psychodrama and sociometry, giving the field both a name and a set of methods for mapping and enacting group relationships. Wikipedia - Wikidata

Bernhard Strauss. Psychotherapy researcher at Jena University Hospital whose meta-analyses and reviews of group psychotherapy — spanning social anxiety, cohesion, and the field's research front — anchor much of its modern evidence base. Wikidata

Irvin D. Yalom. Emeritus professor of psychiatry at Stanford University (b. 1931) whose synthesis of the field defined the eleven therapeutic factors and the idea of the group as a social microcosm, giving group psychotherapy its organizing framework. Wikipedia - Wikidata - Faculty page

Frequently Asked Questions

What is group psychotherapy?
Group psychotherapy is the treatment of psychological difficulty in a small group, usually six to ten patients meeting with one or two therapists, organized so that the interactions among members are the medium of change. The other members are the active ingredient: they give feedback, model alternatives, and reproduce among themselves the relationship problems that brought each of them to treatment. MeSH files it as a form of socioenvironmental therapy (Yalom & Leszcz, 2020).

Who founded group psychotherapy?
Joseph Pratt ran the first documented group treatment in 1907, gathering tuberculosis patients into classes he found improved their morale as well as their care. Jacob Moreno coined the term group psychotherapy in the 1930s and founded psychodrama, and Irvin Yalom later gave the field its organizing framework of therapeutic factors (Pratt, 1907).

What are Yalom's therapeutic factors?
They are the distinct mechanisms through which a group heals; eleven are conventionally named, including the instillation of hope, universality, altruism, interpersonal learning, group cohesiveness, and catharsis. Yalom placed cohesion and interpersonal learning at the centre: cohesion is the condition under which the others work, and interpersonal learning is the group's distinctive engine (Yalom & Leszcz, 2020).

What is group cohesion and why does it matter?
Cohesion is the solidarity and sense of belonging that bind members to a group, the group analogue of the individual therapeutic alliance. A meta-analysis found a robust relationship between cohesion and outcome across group types, making it the strongest single process predictor of whether a group helps, and the precondition for its other factors (Burlingame et al., 2018).

What are the stages of group development?
Bruce Tuckman named four: forming, the tentative early phase of orientation and dependence; storming, a phase of conflict and testing; norming, in which shared norms settle and cohesion consolidates; and performing, the mature phase of productive work. The storming phase is a normal passage toward cohesion, not a sign of failure (Tuckman, 1965).

Is group therapy as effective as individual therapy?
Meta-analytic evidence says roughly yes. A direct comparison of the two formats found no reliable overall difference in outcome, and controlled trials support group therapy in specific conditions such as social anxiety disorder. This makes the group an equally effective and more efficient option for many patients rather than a second-best one (McRoberts et al., 1998).

What kinds of group psychotherapy are there?
MeSH lists four narrower types under the group heading, distinguished by who is in the room: couples therapy, family therapy, psychodrama, and sensitivity training groups. These are formats rather than competing schools, and they can overlap in practice; what unites them is that treatment happens through a structured set of relationships rather than a single therapist-patient dyad (Yalom & Leszcz, 2020).

How is group therapy used with difficult or high-risk populations?
It is increasingly applied in intensive, structured programmes for populations such as high-risk adolescents, where researchers use therapeutic-factors measures to track what members find most helpful. Studies of such programmes report cohesion and interpersonal learning prominent among the factors that promote recovery, extending Yalom's framework into demanding clinical settings (Hauber et al., 2019).

References