See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/11346057

Working Alliance in Online Therapy as Compared to Face-to-Face Therapy:

Preliminary Results

Article in Cyberpsychology & behavior: the impact of the Internet, multimedia and virtual reality on behavior and society · May 2002
DOI: 10.1089/109493102753770480 · Source: PubMed

CITATIONS READS

412 9,674

2 authors:

Jonathan Cook Carol Doyle

Pennsylvania State University Lewis & Clark College
37 PUBLICATIONS 1,516 CITATIONS 9 PUBLICATIONS 485 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Clinical clerkship timing revisited View project

All content following this page was uploaded by Jonathan Cook on 03 June 2014.

The user has requested enhancement of the downloaded file.

CYBER PSYCHOLOGY & BEHAVIOR
Volume 5, Number 2, 2002
© Mary Ann Liebert, Inc.

Working Alliance in Online Therapy as Compared

to Face-to-Face Therapy: Preliminary Results

JONATHAN E. COOK, M.S.,1 and CAROL DOYLE, Ph.D.2

ABSTRACT
Online therapy, defined as the provision of mental health services through the Internet, is a
growing field that has sparked an abundance of interest and controversy. A primary concern
in the practice of online therapy is whether a working alliance, considered a central compo-
nent of successful therapy, can develop when participants are geographically separated.
Working alliance scores were compared between a small, primarily female sample of online
therapy consumers and a representative sample of traditional face-to-face therapy clients. Re-
sults revealed significantly higher means on the goal subscale and composite score of the
Working Alliance Inventory in the online sample, suggesting that a working alliance can be
adequately established in therapy delivered online. No significant differences in the level of
working alliance were found within the online therapy sample with respect to modality of
communication, client presenting problem, or therapist. Themes from comments suggest the
importance for participants of the disinhibiting effects of the medium.

INTRODUCTION have focused on individual counseling, im-

provements have also been noted in online

R ECENT RESEARCH3 indicates that more than

14% of American adults with Internet ac-
cess go online to find mental health informa-
group settings.7
Despite growing evidence of effectiveness,
potential advantages and disadvantages to on-
tion. Since at least 1995,4 many have also used line therapy continue to be vigorously debated
the Internet as a medium through which to re- and discussed11–13 and include a range of ethi-
ceive mental health services. Although empiri- cal, legal, and clinical issues. One of the
cal research regarding the effectiveness of primary clinical concerns is the possible diffi-
online therapy is nascent, preliminary results culty, or even inability, for therapists and
are encouraging, with several published studies clients to establish a strong therapeutic rela-
finding similar outcomes between traditional, tionship in the absence of nonverbal cues.
face-to-face therapy and online therapy.5–7 Posi- This emphasis on the relationship as an im-
tive outcomes in both audio/video and text- portant outcome variable is grounded in a
based computer-mediated therapy have been well-established body of research. Lambert’s
found for a number of specific presenting prob- common factors research14 suggests that rela-
lems, including anxiety,5 posttraumatic stress tionship factors are the single largest variable
disorder,8 eating disorders,9 and panic disorder over which therapists have some control, ac-
with agoraphobia.10 While most of these studies counting for approximately 30% of therapeutic

1 Graduate School of Social Work, Regional Reseach Institute for Human Services, Portland State University, Port-

land, Oregon.
2 Department of Counseling Psychology, Lewis & Clark College, Portland, Oregon.

95
96 COOK AND DOYLE

outcome across therapist theoretical orienta- For instance, in one of the earliest studies of
tions. Although the therapeutic relationship online relationships, Walther and Burgoon26
has been operationalized in a variety of ways, found that relationship development in a
most research in the past 20 years has favored computer-mediated sample approximated that
the working alliance construct, which can of a face-to-face sample. More recently, Parks
be broadly defined as collaboration between and Roberts27 conducted a study of relationship
therapeutic participants to facilitate healing.15 development in on-line, real-time, text-based
Horvath and Symonds16 conducted a meta- virtual environments called MOOs (Multi-User
analysis of 24 studies relating the quality of Dimensions, Object Oriented). They found that
working alliance to outcome. Across a variety 93.6% of their sample had formed ongoing per-
of studies using different treatments, therapist sonal relationships that they identified as either
theoretical orientations, and alliance measure- close friendships, friendships, or romantic rela-
ment devices, they found a combined effect tionships. A comparison of members’ off-line
size (Pearson’s r) of 0.26, showing a small to relationships to their on-line relationships re-
moderate association between alliance and vealed no differences in terms of the levels of
outcome. breadth and depth that were achieved.
Working alliance can be measured from the Miller and Gergen,28 in a study of exchanges
perspective of clients, therapists, and indepen- occurring on an electronic bulletin board de-
dent observers, but research16–19 has consis- voted to the topic of suicide, found the most
tently shown the strongest correlation between frequent forms of discourse (in order of fre-
clients’ view of working alliance and outcome. quency) to be empathic understanding, sup-
While it appears that type and severity of port, and gratitude for responses. Similarly
clients’ pretherapy symptomatology is unre- supportive communities have also been found
lated to establishment of alliance,20,21 quality of in a sports injury support group29 and an
past relationships does seem to have an im- online women’s mailing list.30 Finn31 studied
pact, as clients who have had a history of diffi- online self-help groups for individuals with
culties in interpersonal relationships are less disabilities and found that participants consis-
likely to develop strong alliances.21,22 Similarly, tently provided supportive, empathic, and
therapists’ ability to foster a strong emotional therapeutic responses, enabling members to
bond in therapy is mediated by the extent to connect and form a community that would not
which their own interpersonal relationships have otherwise been possible.
and skills are well established.23 In addition, In one of the most comprehensive research
highly motivated clients, who enter counsel- efforts studying relationship formation on the
ing with the expectation that they will need to Internet, McKenna32 conducted four experi-
assume personal responsibility for doing the mental and descriptive studies and consis-
work of therapy, are more likely to form a tently found evidence of intimate relationship
strong working alliance.24 Research is mixed, development online. In one project, McKenna
but it appears that experienced therapists may randomly selected ten newsgroups on the In-
be slightly more skilled at cultivating high lev- ternet and solicited participants for in-depth
els of working alliance.23,25 interviews. Of the 30 participants (57% fe-
Given its importance, it is understandable male), 51% of males and 60% of females tran-
that there is concern about how well alliance scended their exclusively online relationships
can be established online. Indeed, if the ab- and met face-to-face, 23% of males and 21% of
sence of nonverbal cues impairs relationship females had an affair, and 3.8% of males and
development, it is likely that the future of the 7.5% of females eventually married partners
Internet as a medium for delivery of psycho- they initially met online, demonstrating that
logical services will be limited to information significant relationship formation occurred
provision, at least until videoconferencing is in this sample. Like Parks and Roberts,27
more widely available. Yet both research and McKenna found no significant differences in
anecdotal data suggest that strong and lasting the depth or breadth of interactions between
relationships are being formed online. participants’ online and offline friendships.
WORKING ALLIANCE IN ONLINE THERAPY 97

Although a thorough description is beyond liance levels in a sample of participants who

the scope of this paper, McKenna32 hypothe- were receiving counseling online would not be
sized that the lack of gating features (i.e., nor- significantly different from a representative
mal barriers to friendship due to a variety of sample of face-to-face therapy recipients.
elements such as physical appearance and geo-
graphic location), accompanied by anonymity
of participants, allow many Internet-based re- MATERIALS AND METHODS
lationships to form very quickly and become
closer and deeper than “real life” friendships. Participant recruitment
Anonymity, or perceived anonymity, may fos-
ter intimacy by increasing the amount of per- Participants were recruited with the assis-
sonal, self-disclosure in friendships on the tance of online therapists in one of two ways:
Internet, where the fear of rejection that may through a link posted on therapists’ web pages
prevent disclosure in face-to-face relationships or through an e-mail announcement sent by
does not exist. In short, people are often more therapists. Therapists were found by browsing
frank when they feel anonymous, as many do lists of online therapists, performing Internet
over the Internet, and this leads indirectly to searches, and through referrals. One hundred
greater intimacy. individual therapists and group practices were
In an intriguing finding that may provide a contacted by e-mail and asked to inform
clue to the temperament of some who seek clients about this research opportunity. In ad-
counseling online, McKenna provides evi- dition, several posts explaining the research
dence to counter the claim that Internet use in- project and asking therapists’ help in recruit-
creases loneliness and isolation. Instead, she ing clients were made to e-mail lists, including
found that those who are already lonely and the general and members-only groups hosted
socially anxious and who have difficulties es- by the International Society for Mental Health
tablishing face-to-face relationships, are more Online (ISMHO). A total of 16 therapists
likely to develop relationships on the Internet. agreed to inform their clients, but participant
Rather than causing loneliness and isolation, responses were limited to one male and four
the Internet seems to provide an outlet that al- female therapists (one Ph.D., three master’s
lows those who are lonely and socially anxious level, and one master’s student) operating
to connect with others in a safe, controlled en- from a range of theoretical orientations. Clients
vironment. In addition, the friendships formed recruited for participation were given a link to
over the Internet are frequently integrated into the home page for the research project
users’ real lives, decreasing loneliness and so- (www.lclark.edu/,jecook) and a user name
cial anxiety and enhancing social skills. and password unique to their therapist.
The cumulative results of research on Inter-
net relationship development indicate that re-
Participants
lationships are frequently formed online and
that they are as strong and viable as their off- Participants were one male and 14 female
line counterparts. In addition, many, if not recipients of online therapy ranging in age
most, of the relationships developed on the In- from 19 to 80 years (M = 41.40; SD = 15.99).
ternet occur through text-based mediums, While the majority were United States citizens
often considered the least expressive mode of (53.3%), one-third (33.3%) were Canadian, due
communication online. It is likely that many of to the presence of a Canadian counselor. One
these same elements (e.g., empathy, support, participant listed the Philippines as the coun-
and a readiness to self-disclose) that develop try of citizenship, and one did not indicate citi-
informally and foster deep and intimate rela- zenship other than not being an American
tionship formation in online environments, citizen.
will enable therapists and clients to establish a Participants were primarily white and
robust working alliance in therapy. Conse- tended to be well-educated; all participants
quently, we hypothesized that working al- had at least a high school diploma or GED,
98 COOK AND DOYLE

and the majority had a college or graduate de- TABLE 1. PARTICIPANT CHARACTERISTICS
gree. Participant income was spread across all n %
response choices with most reporting an an-
nual income of $25,000–39,999. Weekly time Sex
Male 1 6.7
spent online ranged from less than 1 hour to Female 14 93.3
more than 20 hours, with the majority (60%) Race
indicating that they spent between 1 and 10 White 14 93.3
Filipino 1 6.7
hours online per week. Nationality
All participants in this sample received indi- United States 8 53.3
vidual counseling and communicated with Canada 5 33.3
Philippines 1 6.7
therapists through a text-based medium— Other 1 6.7
either e-mail or chat. A third of the sample also Primary modality used
reported communicating with therapists E-mail 12 80.0
through a second modality, three using e-mail, Chat 3 20.0
Secondary modality used
one using chat, and one using audio confer- None 10 66.7
encing. All participants who indicated that E-mail 3 20.0
chat was their primary mode of communica- Chat 1 6.7
Audio conference 1 6.7
tion also used e-mail as a secondary method. Presenting problem
Most participants were seeking counseling for Depression 3 20.0
relationship issues (40%) and depression (20%; Anxiety 1 6.7
Relationship issues 6 40.0
Table 1). Family issues 1 6.7
Clients were asked to complete the WAI im- School issues 1 6.7
mediately after the third session, but prelimi- Grief/bereavement issues 1 6.7
nary results indicated a ceiling effect for Other 2 13.3
Weekly time spent online
number of sessions, with all clients answering Less than 1 hour 1 6.7
“more than three,” the highest possible answer 1–5 hours 5 33.3
choice. As a result, we added answer choices 6–10 hours 4 26.7
11–20 hours 3 20.0
to this question after data collection had More than 20 hours 2 13.3
begun, making it impossible to accurately de- Education
termine the average number of sessions clients High school diploma/GED 1 6.7
Some college 5 33.3
had received, although the range was from College degree 6 40.0
one session to more than five. Some graduate school 1 6.7
Graduate degree 2 13.3
Income
Measures Under $25,000 3 20.0
$25,000–$39,999 5 33.3
Working Alliance Inventory. Although there $40,000–$54,999 4 26.7
are a variety of alliance measurement instru- $55,000–$70,000 2 13.3
Over $70,000 1 6.7
ments, only the Working Alliance Inventory1,2
(WAI) was designed to avoid any specific the-
oretical bias and to apply across theoretical
orientations. Based on the work of Bordin,33 to the human relationship between therapist
the WAI consists of a measure of overall al- and client in which trust and attachment are
liance, as well as three subscales: tasks, bonds, formed so that the intimate work of therapy
and goals. The tasks dimension refers to the can progress.
collaboration between therapeutic partners on The WAI is a 36-item self-report question-
specific, technical, in-session behaviors and naire with 12 questions in each subscale that
techniques (e.g., guided imagery, role plays, respondents answer on a fully anchored seven-
exposure therapy). Goals refers to the degree point Likert scale from never to always. Using
that therapist and client agree on the desired a multitrait-multimethod analysis, Horvath
outcomes of therapy. The concept of bonds and Greenberg1 found good construct validity
is similar to the empathy construct. It refers and high internal consistency on the compos-
WORKING ALLIANCE IN ONLINE THERAPY 99

ite score (0.93 on the client form) as well as was established within the first three sessions.
good internal consistency estimates for the Accordingly, most alliance research has used
subscales (0.85–0.88 on the client form). Koko- the completion of the third session as the
tovic and Tracey21 found even higher subscale time to have participants complete alliance
reliability estimates in their sample ranging measurements.
from 0.88 to 0.91 on the client version. Reliabil-
ity coefficients from the current study were Demographics questionnaire. In addition to the
slightly lower than those found elsewhere, WAI, participants completed 11 demographic-
with a composite alpha of 0.86 and subscale type questions in order to obtain general,
scores of 0.59, 0.70, and 0.76 for tasks, bonds, descriptive information. Examples of these
and goals respectively. Studies that have com- questions included participants’ sex, age, in-
pared the WAI with other working alliance come level, education level, presenting problem,
instruments 17,19,34 have found high intercorre- and method of communication with therapists
lations among the measures, providing addi- (e.g., chat, e-mail, videoconferencing).
tional support for its construct validity.
Although the WAI attempts to measure Additional comments. Upon completing the
three distinct constructs (tasks, bonds, and demographics questionnaire and the WAI,
goals) as hypothesized by Bordin,33 factor participants were given the opportunity to
analysis has not yielded clear evidence of dis- type comments about their experience of re-
crete delineations between the subscales. Both ceiving counseling online.
confirmatory 35 and exploratory36 factor analy-
sis have revealed that the composite score is
Procedure
the most salient measurement of working al-
liance. In particular, Hatcher and Barends36 At the home page, participants had the
found that the goals and tasks subscales choice of beginning the study, or viewing sev-
grouped together, indicating a large degree of eral other pages with background and contact
overlap between these two constructs. This re- information, as well as a link where results of
search suggests that interpretation of the WAI the research were posted when the study was
yields more meaningful results from the com- complete. Those who chose to participate were
posite score than from the individual subscale directed to the informed consent page and
scores. provided with general information about the
In a meta-analysis of eight studies that nature of the study, a definition of working al-
used the WAI to investigate the relationship liance, and a description of the questionnaires.
between working alliance and outcome, Hor- Information in the informed consent ensured
vath 34 found an average weighted effect size participants of their anonymity, unless they
of 0.33 for the client-based measures, but es- entered an optional $50 raffle, in which case
timated that the actual size may be closer to they were asked to provide an e-mail address
the upper bound of the 95% confidence inter- that was treated confidentially and deleted
val (0.43) due to the use of a conservative upon completion of the study. Participants
procedure to estimate effect size. These fig- were given information about potential risks
ures suggest a moderate link between clients’ and benefits to participating in the study and
perception of the working alliance as mea- informed of their freedom to withdraw con-
sured by the WAI and therapeutic outcome, sent at any time and immediately exit the
similar to alliance results obtained else- study by using a link at the bottom right of
where. 14,16,36 each page. Disagreeing to the informed con-
Meta-analysis has also revealed that early sent returned individuals to the home page,
alliance is slightly more predictive of outcome while agreeing allowed participants to con-
than alliance measures taken as an average tinue, redirecting them to a Microsoft Win-
across all sessions or taken towards the middle dows 2000 Professional system using an
of treatment. Eaton et al.37 found that level of Internet Information Systems 5 web server. A
alliance, regardless of the length of therapy, security dialogue box (based on Windows NT
100 COOK AND DOYLE

File System [NTFS] technology) would now initially validated.1,2 To assess the representa-
open, asking participants for their user name tiveness of the scores from this sample, we
and password. The site was password pro- computed the standard error of measurement
tected both to reduce the possibility of errone- (SEM) and created a confidence interval with
ous responses from individuals not eligible to which to compare scores from four other stud-
participate and to track participant data based ies that used the WAI (combined N of 182).
on referring therapist. After entering a correct Comparing the results of the four studies to the
user name and password, the demographics initial validation sample revealed that 11 of
questionnaire was displayed and participants the 12 subscale scores fell within one SEM of the
could begin. In order to ensure complete re- initial validation scores. The other subscale,
sponses, a JavaScript application was used on from the study with the smallest sample
this page and on the 3 pages of the WAI to in- (N = 22), was slightly outside the 68% confi-
form participants if they had left any questions dence interval, but was within the 95% confi-
unanswered. Unless they chose to exit the dence interval, or approximately two SEMs.
study, participants could not continue without These results indicate that despite the lack of a
answering all the questions. large sample, scores on the initial validation of
Upon completion of the demographics page, the WAI nevertheless seem to be representative.
participants were directed to the instructions Single sample t tests were used to conduct
for the Working Alliance Inventory. Consistent the primary analysis, in which working al-
with the suggestions of Rosenfeld and col- liance composite and subscale scores from the
leagues, 38 the instructions and WAI questions online sample were compared to scores from
were transcribed online as accurately as possi- the representative sample of face-to-face ther-
ble from the paper and pencil form. Partici- apy clients (for scores, see Table 2). Subscale
pants were presented with the 36 questions of and composite scores were all higher for the
the WAI across three web pages. When they online therapy sample, although only the goal
had finished, participants were able to enter (t(14) = 3.039; p < 0.01) and composite (t(14) =
comments about online therapy in a text box, 2.307, p < 0.05) scores were significantly differ-
although completion of the comments page ent from the scores in the face-to-face sample.
was optional and users were not prompted to Given the small sample and corresponding
enter information if they left it blank. lack of statistical power, the effect sizes for
In order to reduce or eliminate data entry er- the significant t scores were relatively large at
rors, participant responses were automatically d = 0.79 and d = 0.60 for the goal and compos-
coded into number values (where appropriate) ite scales respectively. Post-hoc power analy-
and entered into a Microsoft Access 2000 data- ses were conducted on the scales that did not
base, using Active Server Page (ASP) tech- reach significance, using G*Power.42 These re-
nology. Participants’ Internet Protocol (IP) sults ranged from 0.22 to 0.33, indicating a
address and the date and time of data entry large potential for type II error.
were also entered into the database in order to
check for multiple submissions by the same re-
spondent, a possible disadvantage to conduct-
TABLE 2. SAMPLE MEANS AND T TEST COMPARISONS
ing research over the Internet noted by several BETWEEN O NLINE THERAPY AND FACE-TO-FACE SAMPLES
researchers. 39–41 Upon completion of the study,
Sample means
results were imported from Access into the
Statistical Package for the Social Sciences Online Face-
(SPSS 10) for statistical analysis. therapy to-face
WAI scale (N = 15) (N = 25) ta p

Task 70.33 68.6 1.260 0.228

RESULTS Bond 72.47 69.6 1.627 0.126
Goal 72.27 67.3 3.039 0.009
Composite 215.07 205.5 2.307 0.037
The face-to-face comparison group was the
small sample (N = 25) on which the WAI was a df = 14.
WORKING ALLIANCE IN ONLINE THERAPY 101

Additional comparisons were conducted to ten communication (three), and convenience

determine whether there were differences in and flexibility (two).
working alliance based on modality of com-
munication. Although participants who used Viability. This theme refers to participants’
chat for the primary mode of communication feeling that online therapy is a viable method
reported consistently higher means on the for receiving mental health assistance. Six par-
WAI, no significant differences were found. ticipants explicitly stated that they believe on-
Results among those who used a secondary line therapy to be effective.
modality also revealed a trend of higher (but
again nonsignificant) means on all WAI sub- Disinhibition. This theme was discussed in
scales and the composite score (Table 3). the most depth. Participants described the
Given the small numbers in each category, sense of freedom they felt to express them-
presenting problems were combined into three selves online without embarrassment or fear
groups for statistical analysis: depression and of judgment from therapists. Many expressed
anxiety, relationship issues, and other (school the stress they typically feel in a face-to-face
issues, family issues, grief/bereavement, and therapy situation and indicated that, for the
dependency issues). One-way analysis of vari- first time, they were able to be completely hon-
ance revealed no significant differences or est and open with a therapist.
trends on WAI subscale or composite scores
based on type of presenting problem (Table 4). Cost. Participants who discussed cost were
Additional analyses, conducted to test for uniformly pleased with the affordability of on-
differences based on other variables, revealed line therapy.
no significant differences in working alliance
between therapists, participant weekly time Travel. These comments focused on the ad-
online, participant education, participant in- vantage of online therapy for those who have
come, or participants’ nationality. mobility challenges or who live in an isolated
region.
Comments
Client/counselor relationship. Two participants
Nine of the 14 participants entered addi- discussed the strength of the relationship with
tional comments at the end of the question- their therapists, noting the care and respect
naire. Seven themes emerged from an analysis their therapists demonstrated. Another partic-
of the comments. These themes are described ipant indicated the importance of establishing
in greater detail below along with the number an honest relationship with the therapist.
of participants making that type of comment
(in parentheses): viability (six), disinhibition Advantages of written communication. These
(five), cost (three), travel (three), client/coun- comments focused on participants’ percep-
selor relationship (three), advantages of writ- tions of the unique advantages to text-based

TABLE 3. WORKING ALLIANCE MEANS AND STANDARD DEVIATIONS BASED ON MODALITY

Primary modality Secondary modality

E-mail Chat None Othera

(n = 12) (n = 3) (n = 10) (n = 5)

WAI scale Mean SD Mean SD Mean SD Mean SD

Task 69.75 4.94 72.67 7.37 68.80 4.59 73.40 5.86

Bond 70.92 6.52 78.67 4.51 70.50 7.11 76.40 4.51
Goal 71.00 5.89 77.33 6.43 70.70 6.11 75.40 6.15
Composite 211.67 14.47 228.67 17.50 210.00 14.82 225.20 14.70
aOther consists of e-mail (n = 3), chat (n = 1), and audio conferencing (n = 1).
102 COOK AND DOYLE

TABLE 4. WORKING ALLIANCE MEANS AND STANDARD DEVIATIONS BASED ON CLIENT

PRESENTING PROBLEM
Depression and anxiety Relationship issues Othera
(n = 4) (n = 6) (n = 5)

WAI scale Mean SD Mean SD Mean SD

Task 74.25 8.42 69.17 3.54 68.60 3.05

Bond 74.00 8.29 74.00 2.28 69.40 9.34
Goal 76.00 4.69 72.33 6.95 69.20 6.14
Composite 224.25 20.79 215.50 11.64 207.20 15.72
a Other includes school issues, family issues, grief/bereavement, and dependency

issues.

therapy. Participants appreciated the ability to sample to be lower than those from the face-to-
re-read messages from therapists and noted face sample. In fact, results revealed that all WAI
that this allowed for greater cognitive process- subscales and the composite score were higher
ing than verbal communication. Participants in the online sample. Exceeding expectations
also described the greater ease they felt to ex- and despite low power, results on the composite
press thoughts and feelings through writing. scale and goal subscale of the WAI were signifi-
cantly higher than those from the representative
Convenience and flexibility. Two participants sample of face-to-face clients. These findings
described the convenience and flexibility af- strongly suggest that, at least among those who
forded by online therapy, including the ab- find online therapy to be an appealing method
sence of scheduling issues, lack of travel time of receiving mental health assistance, working
and parking challenges, and not needing to be alliance, and perhaps most importantly, an em-
concerned with personal grooming. One par- pathic relationship, can be strongly established
ticipant noted the advantage to being able to regardless of modality of communication.
communicate intimate thoughts when most While all WAI subscales were higher in the
personally appropriate. online sample, only the therapeutic goals sub-
scale was significantly higher than the face-to-
face comparison group. Although subscale
DISCUSSION scores should be interpreted with caution, it
may be that online therapy can be a particularly
Results from this small, self-selected sample effective method for therapeutic participants to
of online therapy consumers should be inter- clearly establish and agree upon the outcomes
preted and generalized cautiously. Despite of therapy. Upon reflection this result may not
limitations, however, working alliance levels be surprising, given that the easily accessible,
demonstrate that participants felt a collabora- written record of agreed-upon goals in text-
tive, bonding relationship with therapists, and based therapy leaves less room for ambiguity
comments overwhelmingly indicated partici- than a verbal agreement in traditional therapy.
pants’ belief that online therapy was a positive Online participants’ relatively low rating of the
experience with unique advantages over face- task subscale in comparison to the other sub-
to-face counseling. scales from the online sample (although still
Based on previous research documenting higher than the face-to-face sample) may be ex-
the depth and breadth of online relationship plained by the smaller role of traditional in-ses-
development, 27,30,32 we predicted that working sion tasks in asynchronous therapy.
alliance levels from the online sample would Despite the absence of significant differ-
be comparable to the face-to-face population. ences in working alliance level between com-
If this was incorrect, and alliance is not able to munication modalities, a trend was observed
be adequately established online, we would indicating that participants who corresponded
have expected WAI scores from the online with therapists using more than one modality
WORKING ALLIANCE IN ONLINE THERAPY 103

consistently had higher composite and sub- and sense of control the Internet offers. These
scale scores on the WAI. It is possible that the findings are supported by participants’ com-
limited sample size in this study masked sig- ments that disinhibition was a major attraction
nificant effects about modality that may be to online therapy. The importance for partici-
found in a larger sample. The impact on rela- pants of the disinhibiting effects of online ther-
tionship development of communicating in apy also suggests that the population who
multiple formats, especially of a text-based na- seeks online therapy may differ from the gen-
ture, has the potential to be an important area eral therapy-seeking public. Those who go on-
for future research. line for therapy may, in fact, share more in
As expected, the type of problems for which common with the population described by
participants sought help did not significantly McKenna.32 Many in McKenna’s sample tended
impact working alliance scores. This finding is to be uncomfortable in traditional social situa-
similar to previous research on working al- tions and turned to the Internet to connect with
liance development in face-to-face therapy others. For the individuals in the current study,
showing that severity and type of client pre- it is conceivable that online therapy, in addition
senting problems are unrelated to working to frequently being more affordable, is more
alliance.20,21 conducive to strong therapeutic relationship
The overwhelming majority of participants development than face-to-face therapy. The in-
in this study were female, consistent with previ- teraction between online relationship develop-
ous research that has found that more women ment and personality type is a fertile area for
than men use the Internet for mental health in- future research.
formation and services.4,43,44 Interestingly, par- Large-scale research projects examining
ticipants were well-represented at all income both working alliance and treatment outcomes
levels, in contrast to surveys that have shown are necessary to increase the validity of re-
those with higher incomes to be over repre- search into online counseling. It will also be
sented in the Internet population.3 Participants important for future research to obtain a wider
did, however, tend to be highly educated. cross section of therapists willing to partici-
Lack of control presented difficulties in this pate. Nevertheless, this project adds to a
study that may have affected sample selection rapidly developing body of research indicat-
and results. For instance, many participants ing that the delivery of mental health services
had received more than three sessions when through the Internet is both plausible and ef-
they completed the questionnaires, even though fective and that the empathic bond, central to
they were asked to complete them after the working alliance and successful outcome, can
third. However, Horvath and Symonds’16 find- occur when therapeutic participants are geo-
ing that alliance is fairly stable when taken as graphically separated.
an average over the course of therapy, suggests
that results may not have been greatly affected.
It is also unclear how systematically participat- ACKNOWLEDGMENTS
ing therapists informed clients of this research
project. Bias may have been introduced by the We would like to acknowledge and thank the
method therapists used to notify clients, al- online therapists who informed their clients of
though this was probably less of a factor for this research project. Without their help, this
those who posted notifications on their web study could not have been conducted.
sites.
Online therapy is not for everyone and many
will continue to prefer and need face-to-face
counseling. Previous research has indicated REFERENCES
that individuals with certain personality char-
1. Horvath, A.O., & Greenberg, L.S. (1986). The devel-
acteristics, such as introversion,45 and particular opment of the Working Alliance Inventory. In: Pin-
problems, such as panic disorder and agora- sof, W.M., ed. The psychotherapeutic process: A research
phobia, 10 may prefer the perceived anonymity handbook. New York: Guilford Press, pp. 529–556.
104 COOK AND DOYLE

2. Horvath, A.O., & Greenberg, L.S. (1989). Develop- apy: a meta-analysis. Journal of Counseling Psychology,
ment and validation of the Working Alliance Inven- 38:139–149.
tory. Journal of Counseling Psychology, 36:223–233. 17. Tichenor, V., & Hill, C.E. (1989). A comparison of six
3. Rainie, L., Packel, D., Fox, S., et al. (2001). More on- measures of working alliance. Psychotherapy, 26:
line, doing more: 16 million newcomers gain Internet 195–199.
access in the last half of 2000 as women, minorities, 18. Hatcher, R.L., Barends, A., Hansell, J., et al. (1995).
and families with modest incomes continue to surge Patients’ and therapists’ shared and unique views of
online [On-line]. Available: www.pewinternet.org/ the therapeutic alliance: an investigation using con-
reports/pdfs/PIP Changing Population.pdf. firmatory factor analysis in a nested design. Journal
4. Powell, T. (1998). Online counseling: a profile and of Consulting & Clinical Psychology, 63:636–643.
descriptive analysis [On-line]. Available: http:// 19. Safran, J.D., & Wallner, L.K. (1991). The relative pre-
netpsych.com/Powell.htm. dictive validity of two therapeutic alliance measures
5. Cohen, G.E., & Kerr, B.A. (1998). Computer- in cognitive therapy. Psychological Assessment, 3:
mediated counseling: an empirical study of a new 188–195.
mental health treatment. Computers in Human Ser- 20. Horvath, A.O. (1995). The therapeutic relationship:
vices, 15: 13–26. from transference to alliance. In Session—Psychother-
6. Day, S.X., & Schneider, P. (2000). The subjective expe- apy in Practice, 1:7–17.
riences of therapists in face-to-face, video, and audio 21. Kokotovic, A.M., & Tracey, T.J. (1990). Working al-
sessions. In: Walz, G.R., ed. Cybercounseling and cy- liance in the early phase of counseling. Journal of
berlearning: strategies and resources for the millennium. Counseling Psychology, 37:16–21.
Alexandria, VA: American Counseling Association, 22. Horvath, A.O., & Luborsky, L. (1993). The role of the
pp. 203–218. therapeutic alliance in psychotherapy. Journal of Con-
7. Barak, A., & Wander-Schwartz, M. (2000). Empirical sulting & Clinical Psychology, 61:561–573.
evaluation of brief group therapy conducted in an 23. Dunkle, J.H. (1996). Contribution of therapist experi-
Internet chat room. Journal of Virtual Environments. ence and personal characteristics to the working al-
Available: www.brandeis.edu/pubs/jove/html/v5/ liance. Journal of Counseling Psychology, 43:456–460.
v5n1jove.html. 24. Tokar, D.M., Hardin, S.I., Adams, E.M., et al. (1996).
8. Lange, A., Schrieken, B., van de Ven, J.-P., et al. Clients’ expectations about counseling and percep-
(2000). “Interapy”: the effects of a short protocolled tions of the working alliance. Journal of College Stu-
treatment of posttraumatic stress and pathological dent Psychotherapy, 11:9–26.
grief through the Internet. Behavioural & Cognitive 25. Mallinckrodt, B., & Nelson, M.L. (1991). Counselor
Psychotherapy, 28:175–192. training level and the formation of the psychothera-
9. Celio, A.A., Winzelberg, A.J., Wilfley, D.E., et al. peutic working alliance. Journal of Counseling Psy-
(2000). Reducing risk factors for eating disorders: chology, 38:133–138.
comparison of an Internet- and a classroom- 26. Walther, J.B., & Burgoon, J.K. (1992). Relational
delivered psychoeducational program. Journal of communication in computer-mediated interaction.
Consulting & Clinical Psychology, 68:650–657. Human Communication Research, 19:50–88.
10. Bouchard, S., Payeur, R., Rivard, V., et al. (2000). 27. Parks, M.R., & Roberts, L.D. (1998). “Making
Cognitive behavior therapy for panic disorder with MOOsic”: the development of personal relationships
agoraphobia in videoconference: preliminary re- on line and a comparison to their off-line counter-
sults. Cyberpsychology & Behavior, 3:999–1007. parts. Journal of Social & Personal Relationships, 15:
11. Fenichel, M. (2001). Online psychotherapy: technical 517–537.
difficulties, formulations and processes [On-line]. 28. Miller, J.K., & Gergen, K.J. (1998). Life on the line: the
Available: www.fenichel.com/technical.shtml. therapeutic potentials of computer-mediated con-
12. Childress, C. (1998). Potential risks and benefits of versation. Journal of Marital & Family Therapy, 24:
online psychotherapeutic interventions [On-line]. 189–202.
Available: www.ismho.org/issues/9801.htm. 29. Preece, J. (1999). Empathic communities: balancing
13. Barak, A. (1999). Psychological applications on the emotional and factual communication. Interacting
internet: a discipline on the threshold of a new mil- with Computers, 12:63–77.
lennium. Applied & Preventive Psychology, 8:231–245. 30. Adams-Price, C.E., & Chandler, S. (2000). The star
14. Lambert, M.J. (1992). Psychotherapy outcome re- fleet ladies auxiliary: evolution of an online women’s
search: implications for integrative and eclectical mailing list. Cyberpsychology & Behavior, 3:811–816.
therapists. In: Goldfried, M.R., ed. Handbook of psy- 31. Finn, J. (1999). An exploration of helping processes
chotherapy integration. New York: Basic Books, pp. in an online self-help group focusing on issues of
94–129. disability. Health & Social Work, 24:220–231.
15. Bachelor, A., & Horvath, A. (1999). The therapeutic 32. McKenna, K.Y.A. (1998). The computers that bind:
relationship. In: Duncan, B.L., ed. The heart and soul of relationship formation on the Internet [Doctoral dis-
change: what works in therapy. Washington, DC: Amer- sertation]. Ohio University.
ican Psychological Association, pp. 133–178. 33. Bordin, E.S. (1979). The generalizability of the psy-
16. Horvath, A.O., & Symonds, B.D. (1991). Relation be- choanalytic concept of the working alliance. Psycho-
tween working alliance and outcome in psychother- therapy: Theory, Research & Practice, 16:252–260.
WORKING ALLIANCE IN ONLINE THERAPY 105

34. Horvath, A.O. (1994). Empirical validation of Bor- 41. Schmidt, W.C. (1997). World-Wide Web survey re-
din’s pantheoretical model of the alliance: the Work- search: benefits, potential problems, and solutions.
ing Alliance Inventory perspective. In: Greenberg, Behavior Research Methods, Instruments, & Computers,
L.S., ed. The working alliance: theory, research, and prac- 29:274–279.
tice. New York: John Wiley & Sons, pp. 109–128. 42. Erdfelder, E., Faul, F., & Buchner, A. (1996). GPOWER:
35. Tracey, T.J., & Kokotovic, A.M. (1989). Factor struc- a general power analysis program. Behavior Research
ture of the Working Alliance Inventory. Psychological Methods, Instruments, & Computers, 28: 1–11.
Assessment, 1:207–210. 43. Fox, S., Rainie, L., Horrigan, J., et al. (2000). The
36. Hatcher, R.L., & Barends, A.W. (1996). Patients’ view online health care revolution: how the web helps
of the alliance in psychotherapy: exploratory factor Americans take better care of themselves [On-line].
analysis of three alliance measures. Journal of Con- Available: www.pewinternet.org/reports/pdfs/
sulting & Clinical Psychology, 64:1326–1336. PIP_Health Report.pdf.
37. Eaton, T.T., Abeles, N., & Gutfreund, M.J. (1988). 44. Stubbs, P. (2000). Mental health care online [On-line].
Therapeutic alliance and outcome: impact of treat- Available: www.geocities.com/online form/.
ment length and pretreatment symptomatology. Psy- 45. Hamburger, Y.A., & Ben-Artzi, E. (2000). The rela-
chotherapy, 25:536–542. tionship between extraversion and neuroticism and
38. Rosenfeld, P., Booth-Kewley, S., & Edwards, J.E. the different uses of the Internet. Computers in
(1993). Computer-administered surveys in organiza- Human Behavior, 16:441–449.
tional settings: alternatives, advantages, and appli-
cations. American Behavioral Scientist, 36:485–511. Address reprint requests to:
39. Hewson, C.M., Laurent, D., & Vogel, C.M. (1996).
Proper methodologies for psychological and socio-
Jonathan E. Cook, M.S.
logical studies conducted via the Internet. Behavior Regional Institute for Human Services
Research Methods, Instruments, & Computers, 28: Portland State University
186–191. P.O. Box 751
40. Pasveer, K.A., & Ellard, J.H. (1998). The making of a Portland, OR 97207–0751
personality inventory: help from the WWW. Behavior
Research Methods, Instruments, & Computers, 30:
309–313. E-mail: [email protected]
View publication stats