Empirical Evaluation of Brief Group Therapy Conducted in an Internet Chat Room

Azy Barak and Michal Wander-Schwartz
University of Haifa, Israel



 The purpose of this study was to examine group therapy procedures in an anonymous, password-protected Internet chat room. Participants were college students from several Israeli universities and community who elected to join either an Internet chat-room therapy group (n=6) or a standard face-to-face group (n=9). Both groups met for seven consecutive weekly sessions of ninety minutes each. These groups were compared to a no-treatment control group, made up of seven individuals who were referred to group therapy but who were unable to participate. Comparisons of the groups showed that both therapeutic groups had a small, statistically insignificant positive improvement in participantsÕ self-image, social relations, and well-being, with a trend in favor of the virtual group. Participants in the no-treatment control group generally remained unchanged. In addition, several group processes were found to be similar between the two therapy groups: cohesiveness, personal exposure, expression of feelings, independence, and order and organization. The Internet group, however, reported higher levels of aggression, action orientation, and therapist support and control than the face-to-face group. Participants in both therapy groups expressed general satisfaction with their respective group therapies.


Computer-mediated interpersonal communication has become widespread by Internet technology. This revolutionary channel of communication has opened a new window of opportunity for psychologists interested in interactive therapeutic interventions. The new way of therapeutic communicationÑhaving prospects and promises as well as risks and hazards (Childress, 1998)Ñhas attracted a growing number of service-providing mental health professionals (Barak, 1999). The asynchronous communication optionÑmade possible by e-mail and e-mail-based discussion groups and web-based forums for individual and group communication, respectivelyÑhas a major advantage: time and place elasticity. This technology, however, lacks a key feature of human interpersonal communication characterized by spontaneity, authenticity, immediacy, and directness. This feature is directly related to the commonly accepted and generally appreciated therapeutic factor of Òhere and now.Ó Numerous schools of psychotherapy argue that this immediacy factor may be responsible to a great degree for dramatic therapeutic developments. In addition, defense mechanisms or cognitive distortions (depending on school of therapy) are less likely to take place in a Òhere and nowÓ therapeutic situation.

Nevertheless, cumulative evidence, from anecdotal reporting as well as from empirical studies, supports the use of asynchronous, Internet-based communication for both individual and group mental help (e.g., Barak, 1999; Fink, 1999; King & Moreggi, 1998; Murphey & Mitchel, 1998). Online group interventions have focused primarily on support groups in various areas of need, such as sexual abuse of women, cancer patients, and single parents. The clinical proposals for interventions, and actual interventions provided, have primarily been asynchronous. An exception are the attempts made by Suler (1996a, 1996b), who viewed and analyzed psychological opportunities at the Palace, an Internet-based multimedia, live chat environment. Therapy-focused groups, using synchronous (i.e., real-time) communication, however, have hardly been offered and, therefore, have not been subject to research to date.

There might be two explanations for this: first, Internet chat rooms are relatively newer than e-mail distribution lists (listservs), and hence have had fewer opportunities to be used for therapeutic purposes. Second, from a technical point of view, web-based chat rooms have changed from the less reliable and less friendly HTML-based technology to advanced and efficient Java applications. In addition, the technology of password-protected chat rooms is relatively new. Suler (1999) listed several advantages offered by synchronous Internet communication: clear scheduling and definition of Òmeetings,Ó a feeling of presence, more spontaneous interactions resulting in more revealing disclosures, perceived commitment and dedication, and the inclusion of specific cues not present in asynchronous sessions (e.g., pauses in communication). He listed a few disadvantages as well: less convenience, and reduced time for responding and reflecting. Sempsey (1998), too, advocated the use of synchronous Internet chats (basically referring to MUD communities). To date, however, only a few attempts have been made to exploit synchronous Internet communication for therapeutic and/or consultation purposes. With the exception of Cohen and Kerr (1998), those attempts that have been reported (Colon, 1996; Quimby, 1999) lack close research, and mainly refer to practical considerations. The purpose of this study was to examine group therapy procedures that took place in an Internet, Java-based, password-protected chat room, and to compare its usefulness to a standard therapy group as well as to a no-treatment control group. We employed both quantitative and qualitative methods in examining the impact of the new intervention.



Participants were college students from several Israeli universities and community colleges who were recruited through newspaper and bulletin board ads that offered free group therapy to interested individuals. The ads mentioned both face-to-face and Internet chat room options, to be selected by the participant as preferred. Referrals were interviewed on the phone in order to screen out severe pathological cases and non-authentic referrals. The phone interviews inquired into a callerÕs motivation for group therapy, past therapeutic experience, and general mental health parameters. After screening, six of the participants (three men and three women) selected were assigned to the Internet chat-room therapy group, and nine (three men and six women) to the standard face-to-face group. The allocation of participants to the therapy groups was solely based on their preference. The size of the two groups was predetermined on the basis of an optimal group size desired for each intervention. Although this step put limits on the power of statistical analyses, we preferred to sacrifice this for therapeutic considerations. A no-treatment control group of seven individuals (three men and four women) was made up of those who were deemed eligible but were unable to participate in either therapy group because of scheduling or other technical problems. We had no reason to believe that this latter group was different in any essential psychological factor from the two therapy groups.


Participants in both intervention groups received time-limited, dynamically oriented therapy (Hudson-Allez, 1997; Wells, 1993). Therapy was provided by two female group therapists. Both therapists were similarly experienced professionals, in their 30s, and skillful in brief dynamic therapy. We preferred having two different therapists over a single therapist to prevent contaminating biasing effects. The therapist for the Internet chat-room group, in addition to her therapeutic skills, had reasonable personal experience in using the Internet and in participating in Web chats. Participants in the Internet-based group therapy met in a Java application, password- protected chat room for seven sessions of 90 minutes each. All sessions were held at a predetermined timeÑthe same day and time every week. Graphically, the chat room had three scrolling screen windows: a vertical side window on which the nicknames of participants were shown, a large window on which all group message exchanges appeared, and on top of this one a smaller window on which the therapistÕs messages appeared. Under the windows was a text box in which participants typed their messages. All participants in the chat room used nicknames, whether real or fabricated. At the beginning of the first session, the therapist provided a list of several specific rules in regard to communication and behavior (e.g., stay alone by the computer; use of certain code symbols to denote certain messages). Then the therapeutic communication began. Upon completion of each therapy session, we electronically saved the transcript of the group communication. The therapist used the transcripts after each session to analyze the group and to prepare for the next session. After the completion of the project, the transcripts were used to evaluate group processes.

Participants in the standard, face-to-face therapy group also met at a predetermined time in a convenient room on the campus of the University of Haifa.

Dependent Measures

Measures of Therapy Impact: Self-Esteem, Social Relationships, and Well-Being

These variables were measured by three scales taken from HudsonÕs (1982) clinical measurement package. Each of these scales includes 25 items, to which respondents respond on 5-point scales. The Self-Esteem scale includes items such as: ÒI feel that others get along better than me.Ó The Social Relationships scale includes items such as ÒMy friends are a source of joy for me.Ó The Well-Being scale (originally called Satisfaction Scale) includes items such as ÒI feel wonderful in the morning.Ó Cumulative empirical evidence highly supports the reliability and validity of measurement of these scales as efficient clinical measures.

Group Process Variables

We used a measure developed by Moose (1982) that includes 10 factors in the therapeutic group process: cohesiveness, action orientation, personal exposure, expression of feelings, expression of anger and aggression, independence, order and organization, newness, perceived therapist support, and perceived therapist control. Each scale is represented by nine items, to which respondents respond on a 6-point scale. The measure was found to be reliable and valid in previous research.

Evaluation Questionnaire

We constructed this questionnaire for the purpose of the study. It contained 11 items, each of which the participants had to rate on a 6-point scale. For example: ÒTo what degree would you recommend participation in such a group to other people?Ó Total score was calculated by summing up all items. Four specific items were added to the Evaluation Questionnaire used with the chat-room group; these referred to their specific experience (e.g., ÒTo what degree did anonymity influence your personal exposure to the group?Ó and ÒTo what degree would you be interested to meet the other group members face-to-face?Ó).


Participants were given the research questionnaires approximately two weeks before the intervention started. The questionnaires were sent to and returned from the Internet and standard intervention group participants online and through regular mail, respectively. Participants continued with their regular lifestyle during the seven weeks of therapy; they were invited to contact us in cases of technical difficulties. Participants in the Internet-based therapy group commonly used their home computers, but were allowed to use any location they wished for connecting to the chat room. In several cases, participants in this group contacted us for technical reasons (e.g., difficulties in connecting to the Internet). In all cases, problems were solved satisfactorily. Upon completion of therapy, the research questionnaires were re-administered to the participants. Members of the no-treatment control group were administered the questionnaires at the same points in time as were the members of the therapy groups. Later, we independently interviewed the therapists in regard to their respective professional experiences.


Quantitative Analyses of Dependent Measures

Comparisons of the groups showed that both therapeutic groups had a small, statistically insignificant positive improvement in participantsÕ self-esteem, social relations, and well-being, though the virtual group showed slightly more improvement. Participants in the no-treatment control group generally remained unchanged. Slight differences were found between the two therapeutic groups after intervention, with an advantage to the Internet chat-room group. In terms of group processes as reported by participants, the two groups were found to be mostly similar in perceptions of group cohesiveness, personal exposure, expression of feelings, independence, and order and organization. Members of the Internet group, however, reported higher levels of aggression, action orientation, and therapist support and control.

The Evaluation Questionnaire revealed that participants in both groups expressed general satisfaction with their respective group therapies; there was no statistical difference here between the two groups. A review of the specific items administered to the participants of the chat-room-based therapy found that anonymity was a major factor in their readiness to open up. Furthermore, although participants in this group felt emotionally close to one another, they were not particularly interested in actually meeting the other group members.

Qualitative Analysis of Chat Room Transcripts

Because of the very small number of participants in each group and the idiosyncratic and phenomenological nature of group therapy, standard quantitative psychological research methods have limited validly in detecting real processes (Heppner, Kivlighan, & Wampold, 1999). Hence, we chose to analyze the chat roomÕs transcripts qualitatively in an attempt to identify cues to group therapeutic factors, processes and impact (Yalom, 1995). Hill, Thompson, and WilliamsÕ (1997) consensual qualitative research system was employed for a qualitative analysis of the transcripts of the seven chat room sessions. Generally, we found evidence of growing messages of positive support, personal disclosures, interpersonal sensitivity, and group cohesiveness made by the participants during the course of therapy.

TherapistsÕ Evaluation

The therapists of the two treatment groups were generally satisfied with their respective groups. Both thought the groups had positive, constructive processes, generally similar to previous therapy groups that they had led. Both therapists, however, noted the very brief length of the group intervention and thought it was too time-limited to bring about a substantial and permanent change in participants. The chat-room therapist reported that her experience was similar in many factors, and different in others, to standard therapy groups that she had led. In addition to common therapeutic group factors that she noticedÑsuch as cohesiveness, catharsis, leadership, disclosure, support, and aggressionÑshe recognized much faster processes of interpersonal relationships, reflected in such various aspects as love and dislike, aggression, and support. She emphasized that her professional experience with the group was unique and very valuable; she felt much emotional involvement, and was very frustrated over the groupÕs early dissolution.


Although this study should be considered preliminary, and its findings should be handled carefully, it seems that an anonymous, Internet-based, chat-room group therapy is a legitimate method of psychological intervention and has a positive impact on interested individuals in need. The quantitative measures showed a trend toward positive change in terms of participantsÕ self-esteem, interpersonal relations, and well-being in comparison to the standard therapy and the no-treatment control groups. We believe that the statistical non-significance of these findings should be attributed mainly to the very small group size and apparentlyÑas emerged from both participants and therapistsÕ feedbackÑfrom the limited number of sessions, as well. In terms of group dynamics, as reported by the participants, some of the group processes evinced by the two groups were similar, and others were different. Interestingly, the factors found to be different between the groups are consistent with those that characterize online group behavior (King & Moreggi, 1998). In terms of satisfaction with therapy, both groups were similarly partially satisfied. From a qualitative perspective, moreover, the chat-room group experienced therapeutic factors that typically exist in therapy groups. Taking all this evidence together, we may cautiously conclude that an Internet chat-room could be used for group-therapy purposes.

The current research has several limitations that warrant attention. First, as mentioned above, the groups were very small therefore one should be cautious with generalizations. That is, we do not know how individuals in other age groups, level of education, or problem areas would react to such an online intervention. Moreover, as indicated, the small size of the groups prevented us from detecting what seemed to be significant differences. Only future research, replicating and extending our preliminary investigation may offer answers to these problems. Second, while we preferred using two different therapists in the face-to-face and the online groups for methodological reasons (preventing a therapistÕs biasing effects), this choice could have resulted in yet other erroneous findings. That is, differences in the therapistsÕ personality, motivation, or skill level could be responsible for differences between the groups rather than the different therapeutic modality. This possibility, too, should be examined in future investigation. Intensive research is needed to further examine numerous other professional and scientific questions, such as the preparation of therapists for this type of professional conduct, the rules by which groups ought to operate, and so on. The findings of this study, however, provide initial empirical support for the use of an Internet chat room as a legitimate therapeutic mode.

Much has been said and written in recent years about the growth and change mental health professions are going through in relation to the Information Revolution and opening opportunities in cyberspace. For instance, Grohol (1998) listed seven significant reasons why online therapy has prospects for becoming an effective and accepted line for provision of mental help: perceived anonymity of users, computer-mediated communication disinhibiting effects, accustomed Net use, response immediacy, trying out new behaviors in a secured online situation, providing of help to people who cannot access help otherwise, and the resemblance of online situations to real-world situations. Finfgeld  (1999) added a few other reasons for providing mental help online, emphasizing institutional aspects: general professional moving from inpatient to outpatient clientele; the change toward short-term, crisis-oriented care; availability of computers and Internet communication; and the resourcefulness and ingenuity made possible by the new technologies. Fink (1999) extensively demonstrated how the Internet, in addition to the exploitation of personal computers, can effectively be used for psychotherapeutic applications. The current research, though limited in scope and empirical examination, is taking the field a step forward in proposing and testing what only a short while ago seemed to be a science fiction. If we can maneuver our professional conduct wisely, profiting from the advantages while avoiding the pitfalls of online-based mental help (Barak, 1999; Childress, 1998; Lebow, 1998), we may advance psychotherapy to a new and much improved realm.


A shorter version of this article was presented in S. A. King (Chair), Internet Support Groups: Group Therapy by E-mail and Chat, Symposium conducted at the annual convention of the American Psychological Association, Boston, August 1999. The authors are indebted to NetVision, an Israeli Internet Service Provider, for the donation of the Internet chat room and related programming. Correspondence concerning this paper should be addressed to Prof. Azy Barak, Department of Psychology, University of Haifa, Mount Carmel, Haifa 31905, Israel. E-mail: azy@construct.haifa.ac.il


Barak, A. (1999). Psychological applications on the Internet: A discipline on the threshold of a new millennium. Applied and Preventive Psychology, 8, 231-246. Also available online: http://construct.haifa.ac.il/~azy/app-r.htm

Childress, C. (1998). Potential risks and benefits of online psychotherapeutic
interventions [on-line]. Available (December, 1999): http://www.ismho.org/issues/9801.htm

Cohen, G. E., & Kerr, B. A. (1998). Computer mediated counseling: An empirical study of a new mental health treatment. Computers in Human Services, 15, 13-26.

Colon, Y. (1996). Chatt(er)ing through the fingertips: Doing group therapy online. Women & Performance: A Journal of Feminist Theory, 17, 205-215.

Finfgeld, D. L. (1999). Psychotherapy in cyberspace. Journal of the American Psychiatric Nurses Association, 5(4), 105-110.

Fink, J. (1999). How to use computers and cyberspace in the clinical practice of psychotherapy. Northvale, NJ: Aronson.

Grohol, J. M. (1998). Future clinical directions: Professional development, pathology, and psychotherapy on-line. In J. Gackenbach (Ed.), Psychology and the Internet (pp. 111-140). San Diego, CA: Academic Press.

Heppner, P. P., Kivlighan, D. M., & Wampold, B. E. (1999). Research design in counseling (2nd ed.). Belmont, CA: Wadsworth.

Hill, C., E., Thompson, B. J., & Williams, E. N. (1997). A guide to conducting consensual qualitative research. The Counseling Psychologist, 25, 517-572.

Hudson, W. W. (1982). The Clinical Measurement Package: A field manual. Homewood, IL: The Dorsey Press.

Hudson-Allez, G. (1997). Time-limited therapy in a general practice setting. London: Sage.

King, S. A., & Moreggi, D. (1998). Internet therapy and self help groups - the pros and cons. In J. Gackenbach (Ed.), Psychology and the Internet: Intrapersonal, interpersonal and transpersonal implications (pp. 77-109). San Diego, CA: Academic Press. Also available online (December, 1999): Also available online: http://www.concentric.net/~Astorm/Chapter5/index.html

Lebow, J. (1998). Not just talk, maybe some risk: The therapeutic potentials and pitfalls of computer-mediated conversation. Journal of Marital and family Therapy, 24, 203-206.

Moose, R. H. (1981). Group Environment Scale Manual. Palo Alto, CA: Consulting Psychologists Press.

Murphy, L. J., & Mitchell, D. L. (1998). When writing helps to heal: e-mail as therapy. British Journal of Guidance and Counselling, 26, 21-32.

Quimby, G. R. (1999). Avatar process training group project. Self-Help & Psychology Magazine [on-line]. Available (December, 1999): http://www.shpm.com/ppc/viewpoint/avatar.html

Sempsey, J., III (1998). The therapeutic potentials of text-based virtual reality. Journal of Virtual Environments, 3, 2. Available: http://journal.tinymush.org/v3n2/sempsey.html

Suler, J. (1996a). Life at the palace: A cyberpsychology case study [on-line]. Available (December, 1999): http://www.rider.edu/users/suler/psycyber/palacestudy.html

Suler, J. (1996b). One of us: Participant observation research at The Palace [on-line]. Available (December, 1999): http://www.rider.edu/users/suler/psycyber/partobs.html

Suler, J. (1999). Psychotherapy in cyberspace [on-line]. Available (December, 1999): http://www.rider.edu/users/suler/psycyber/therapy.html

Wells, R. A. (1993). Clinical strategies in brief psychotherapy. In R. A. Wells & V. J. Giannetti, (Eds), Casebook of the brief psychotherapies. New York: Plenum.

Yalom, I. D. (1995). The theory and practice of group psychotherapy (4th ed.). New York: Basic Books.