Mutuality of Rogers's therapeutic conditions and treatment progress in the first three psychotherapy sessions

Abstract Objective: Research on the effects of Rogers's therapeutic relationship conditions has typically focused on the unilateral provision of empathy, unconditional positive regard, and congruence from therapist to client. Method: This study looked at both client and therapist mutuality of the Rogerian therapeutic conditions and the association between mutuality and treatment progress in the first three psychotherapy sessions. Clients (N = 62; mean age = 24.32; 77% female, 23% male) and therapists (N = 12; mean age = 34.32; nine female and three male) rated one another using the Barrett-Lennard Relationship Inventory after the first and third session. Results: Both clients and therapists perceived the quality of the relationship as improved over time. Client rating of psychological distress (CORE-OM) was lower after session 3 than at session 1 (es = .85, [95% CIs: .67, 1.03]). Hierarchical multiple regression was used to test the predictive power of mutually high levels of the therapeutic conditions on treatment progress. The association between client rating of therapist-provided conditions and treatment progress at session 3 was higher when both clients and therapists rated each other as providing high levels of the therapeutic conditions (R2 change = .073, p < .03). Conclusions: The findings suggest mutuality of Rogers's therapeutic conditions is related to treatment progress.

The therapeutic relationship is a broad umbrella term referring to the interpersonal aspects of the client-therapist dyadic relationship. Arguably, when taken as a single treatment variable, it is the factor of psychotherapy most reliably associated with both progress and outcome. Two key approaches to researching and conceptualizing the therapeutic relationship are the therapeutic alliance construct (Martin, Garske, & Davis, 2000) and the Rogerian (1957) conditions of client perception of therapist experience of empathic understanding, unconditional positive regard, and therapist congruence (Gurman, 1977;Lambert & Ogles, 2004;Norcross, 2002Norcross, , 2012Zuroff & Blatt, 2006). The therapeutic alliance is a complex tripartite construct that included client and therapist collaboration on the goals and tasks for therapy and the client and therapist emotional bond (Bordin, 1979). The therapeutic alliance theory posits that a direct effect on outcome comes through the collaborative efforts of both client and therapist. The alliance construct has received significant support and is considered an empirically supported relationship variable (Norcross, 2012). Similarly, the Rogerian components of the therapeutic relationship are shown to be reliably associated with outcome in a range of settings within diverse clinical samples; this includes adult out-patient services for depression in clinical trials (Ablon & Jones, 1999;Blatt & Zuroff, 2005;Zuroff & Blatt, 2006), treatment studies for depression (Watson & Geller, 2005;Watson, Gordon, Stermac, Kalogerakos, & Steckley, 2003), youth and family therapy (Karver, Handelsman, Fields, & Bickman, 2006), severe psychosis (Hewitt & Coffey, 2005;Rogers, Gendlin, Kiesler, & Truax, 1967), and within general counseling (Archer, Forbes, Metcalfe, & Winter, 2000).
Research into the association between the three Rogerian conditions on outcome has tended to posit a unilateral structuring of the therapeutic relationship. That is, the extent the client perceives the therapist's empathy, unconditional positive regard, and congruence is important for good therapeutic outcome. However, contemporary theory supports a bi-directional structuring to the Rogerian conditions (Murphy, Cramer & Joseph, 2012). The bidirectional view also implies mutuality of Rogers's conditions; that is, client and therapist experience of each other will be associated with outcome and treatment progress. The term mutuality refers to the bi-directional, reciprocal experience of empathy, unconditional positive regard, and congruence of each person within the dyadic relationship. Mutuality of feelings between client and therapist has been previously researched through the idea of the real relationship (Gelso, 2002(Gelso, , 2009Gelso & Carter, 1994). The real relationship refers to the genuine and authentic client and therapist feelings towards one another that is thought to be separate from the transference relationship. Along with this, Murphy et al. (2012) have theorized the process of therapy involves the mutual experiencing of the Rogerian conditions empathy, unconditional positive regard, and congruence. Support for a bi-directional view also comes from research into relational depth events (Tudor, 2010;Wiggins, Elliott & Cooper, 2012) that has considered client and therapist experiences of deep reciprocal connection, and relational connectedness (Cooper, 2012), involving the synchronous client and therapist perception of connectedness.
Despite the growing recognition of the therapeutic relationship as a bi-directional phenomenon there is little available research evidence from empirical studies supporting the notion of mutual experiencing of the Rogerian therapeutic conditions of empathy, unconditional positive regard, and congruence. However, the tripartite process variables of the therapeutic alliance of goals, task, and bond (Bordin, 1979) have received significant empirical support. The alliance has been measured using a number of different scales, across a large number of research studies, and has been described as an increasingly broadening concept (Baldwin, Wampold, & Imel, 2007). This has led researchers to conclude it is difficult to ascertain the precise contribution of specific alliance components to successful psychotherapy outcome (Elvins & Green, 2008). Notwithstanding the methodological problems for psychotherapy process-outcome research, the alliance construct posits both clients and therapists contribute to the therapeutic relationship. However, studies of the Rogerian conditions and Ablon and Jones (1999) measures of the alliance have tended to ask either the therapist or the client to report on their view of the alliance. Consequently research findings are restricted to reporting on the therapeutic relationship as it is experienced from one side only.
A study by Saunders (1999) that looked at the therapeutic bond component of the alliance construct found the level of reciprocal intimacy was related to positive changes in functioning. Positive functioning was assessed by a measure of remoralization, a term referring to decreasing levels of hopelessness, helplessness, isolation, and self-esteem. Clients reported emotions that matched their reports of emotions identified in therapists. Positive client emotions were associated with positive therapist emotions and likewise with negative emotions. The research concluded that reciprocal intimacy was significantly correlated with client ratings of session quality and with treatment effectiveness when fewer sessions in treatment were received. Reciprocal intimacy refers to client feelings towards the therapist and the feelings a client perceives are experienced by the therapist. Like much alliance-outcome focused research Saunders's (1999) study claimed to measure the actual level of reciprocal intimacy. However, the unilateral method of measurement employed made this impossible as only the perceived level of reciprocal intimacy was measured. Differentiating between actual and perceived aspects of the alliance is an issue for alliance research more generally. A way to resolve this issue is by developing dyadic research designs that would enable researchers to capture agreement between client and therapist on their perceived relationship.
The level of commitment within the therapeutic alliance has been termed role investment. The mutual emotional and affirmative feelings of the client and therapist towards one another refer to the emotional bond within the therapeutic dyad. Empathic resonance refers to the level both of feeling understood and of understanding the other person. A second study by Saunders (2000) reported that after session 3 role investment and mutual affirmation were both more strongly associated with client ratings of session quality than empathic resonance. However, hierarchical regression showed empathic resonance to be the only significant contributor to the relief of distress when controlling for distress at intake. Saunders (2000) concluded that perceived client understanding of the therapist and client feeling of being understood were more likely to be related to change earlier rather than later within the course of therapy. These studies attempted to measure the feelings that both client and therapist have towards each other and the effect of time on their association to progress. However, once again the measures were taken from factors as they were unilaterally reported. For a truer test of the mutuality of feelings within the therapeutic relationship both client and therapist needed to report independently upon both their own and their perception of the other's feelings. Support for the bi-directional view of the therapeutic relationship can also be found in research from the wider field of applied social psychology. Research in the area of social support can be helpful and instructive for the field of psychotherapy research (Barker & Pistrang, 2002). Social support refers to the behavioral and emotional support available within a relationship. The availability and accessibility of social support to individuals have been associated with both better psychological functioning and lower incidence of mental illness (Lindorff, 2000;Sarason, Sarason, & Gurung, 2001). However, the association might not be so straightforward. It is often assumed that access to social support alone results in better psychological health, yet this is not always the case. Receipt of social support has resulted in feelings of inadequateness, indebtedness, and inequity (Rafaeli & Gleason, 2009), leaving the recipient often feeling worse off within their social relationships. Interestingly, the concept of bi-directional support, meaning one gives and one gets, has been significantly related to perceived helpfulness in social support relationships where one partner has a disabling medical condition (Rafaeli & Gleason, 2009). Within the social support literature the concept of mutual support has been shown to be particularly beneficial with regard to empathy (Cramer & Jowett, 2010;Pistrang, Picciotto, & Barker, 2001). These studies point towards the need for researching the area of mutuality in relationship factors between client and therapist. Murphy et al. (2012) proposed the therapeutic relationship involves the development of mutual experiencing of the therapeutic conditions empathy, unconditional positive regard, and congruence. Perceived mutuality has been found to be associated with lower levels of bulimic symptoms, fewer symptoms of depression, and greater levels of positive therapeutic change (Tantillo & Saftner, 2003;Sanftner et al., 2006;Sanftner, Tantillo, & Seidlitz, 2004). Further to this, mutuality was also associated with less depression and better health outcomes for patients with rheumatoid arthritis (Kasle, Wilhelm, & Zautra, 2008), better wellbeing among couples dealing with multiple sclerosis (Kleiboer, Kuijer, Hox, Schreurs, & Bensing, 2006) and less negative mood when giving support (Gleason, Iida, Bolger, & Shrout, 2003). Mutuality as a core conceptual process within psychotherapy is beginning to emerge as a useful variable both for understanding the causes of psychological distress and as a facilitative factor in improving psychological wellbeing. A number of studies have researched client perception of their therapists on the impact of therapeutic progress (Hill, 1989;Hill, Thompson, Cogar, & Denman, 1992;Hill, Thompson, & Corbett, 1992;Regan, & Hill, 1992;Rennie, 1990;Thompson & Hill, 1991), yet no research has specifically addressed the presence of mutuality of Rogers's (1957) core conditions of empathy, unconditional positive regard, and congruence within the therapeutic relationship. An additional step would also be to consider the contribution of perceived mutuality towards psychotherapy progress.
It was hypothesized, first, that the data would support the common assumption that client rating of therapist empathy, unconditional positive regard, and congruence would be associated with positive treatment progress; second, that the association between client positive rating of therapist conditions and treatment progress would be stronger when client and therapist perceived each other as mutually experiencing high rather than low therapeutic conditions. We have termed the second hypothesis the "mutuality hypothesis."

Method Participants
Clients. Data were collected from a group of 72 clients, 65 of whom (90% of intended to treat) received at least three sessions of therapy for mixed presenting problems in routine clinical practice in a university mental health service. All clients (77% female, 23% male) were allocated to therapists on a rota system. Sixty-two clients provided pre-therapy scores that were above clinical cut-off on the Clinical Outcomes in Routine Evaluation -Outcome Measure (CORE-OM) and formed the cohort for analysis. Clients ranged in age from 18 to 54 years (M = 24.32, SD = 8.61) and 85% identified themselves as White/European, 3% as Asian or Black, 3% as Chinese and 6% as other. A majority of clients (82%) were in full-time higher education during the study and the remaining (18%) educated to secondary level or were high-school graduates in work either in or outside the home. Thirty-seven (60%) clients received a humanistic (personcentered experiential, integrative, Gestalt) therapy and 25 (40%) received a cognitive and behavioral therapy. Of the 62 clients included in the study sample 35 (56%) continued for up to five sessions of therapy. No significant differences were found for demographic, process or outcome measures between the group that received fewer than five sessions and those who received five or more sessions.
Therapists. A total of 12 (nine female and three male) therapists participated in the study. Therapist approach can be divided along humanistic/experiential and cognitive-behavioral theoretical perspectives. Of the humanistic/experiential oriented therapists, seven identified as client-centered, two integrative/ experiential, and one as gestalt. Two therapists offered a cognitive-behavioral therapy. Therapists ranged in age from 28 to 55 years (M = 34.32, SD = 7.54). Ten therapists identified themselves as White/European and two as other. Two therapists were in training at Master's level, the remaining therapists having post-qualification experience ranging between 1 and 15 years.

Barrett-Lennard
Relationship Inventory (BLRI; Barrett-Lennard, 1964). This measure assesses the level of the therapeutic conditions of empathy, positive regard, unconditionality of regard, and congruence. The full measure consists of 64-items with 16 items on each of the four subscales and can be completed for own feelings towards other (myself to other: MO) or from other to self (OS). Each subscale has an equal number of positively and negatively valenced items that are rated from 1 to 6 on a Likert-type scale for the extent a respondent feels the phrase is true regarding the client-therapist therapeutic relationship. Internal reliability coefficients for the four subscales of the revised BLRI OS version were: empathy, .84; regard, .91; unconditionality of regard, .74; and congruence, .88. Testretest reliability also showed stable results with mean correlations for empathy, .83; regard, .83; unconditionality of regard, .80; and congruence, .85 (Gurman, 1977).
In the present study the full-length measure was shortened to include 32 of the original 64 items to avoid overloading participants. An example of an item for the empathy scale is "They understand me," the positive regard scale "They respect me," the congruence scale "They express their true feelings and thoughts to me," and the unconditionality scale "Whatever I say or do makes no difference to the way they feel about me." A number of factor analyses have been carried out on the original and revised BLRI measures that suggest consistency in identifying the four-factor structure (Cramer, 1986a(Cramer, , 1986bLietaer, 1974;Walker & Little, 1969). However, these studies suggest no overall pattern of consistently high-loading items, i.e., item loadings vary according to the sample. Accordingly, eight items (four positively and four negatively valenced) from each of the four subscales were selected for inclusion. Items were selected by taking every other item as they appeared in order on the full 64-item version. Cronbach's alpha reliability for scores of the total measure was high and this was the case for both the client's perception of therapist conditions (.91) and for the therapist's perception of clients' feelings towards them (.90). As the study aimed to use a 32-item version of the BLRI a principal component analysis (PCA) with the Varimax method of extraction was carried out. A Kaiser-Meyer-Olkin test of sampling adequacy gave a value of .78 and Bartlett's test of sphericity a significance level of p < .001. This suggested the data were satisfactory for factor analysis. Specifying an eigenvalue greater or equal to 1, a single-factor solution accounted for approximately 43% of the total variance explained in clients' perception of therapist-experienced conditions after session 3. The PCA was run a second time specifying a four-factor solution but the results provided an unsatisfactory mixture of items loading on each of the four factors (see Murphy, 2010, for detailed PCA results). As the new measure provided a clear singlefactor structure, a score for the total relationship inventory was used in subsequent analyses, and correlations for the four separate subscales and treatment progress are not reported.
Clinical Outcome in Routine Evaluation -Outcome Measure (CORE-OM). . The CORE-OM is a 34-item self-report measure that can be used as a pre, interim and post therapy outcome measure. In the present study the measure was used to assess pre-therapy levels of psychological distress and treatment progress at session 3 and session 5. Participants completed the measure based on how they have felt "over the last week." The CORE-OM consists of high and low intensity items in three areas; subjective well-being (four items, e.g., "I have felt like crying"), problems (12 items, e.g., "I have felt tense, anxious or nervous"), and functioning (12 items, e.g., "I have felt able to cope when things go wrong"). In addition the measure also assesses a fourth factor, risk to self and other (six items, e.g., "I have thought of hurting myself" or "I have been physically violent to others"). The CORE-OM is made up of positively and negatively worded questions. It uses a 5-point Likert scale ranging from 0 to 4 corresponding to the verbal responses of "Not at all," "Only occasionally," "Sometimes," "Often," and "Most or all of the time." The psychometric properties of the CORE-654 D. Murphy & D. Cramer OM show that it correlates highly with the BDI (.85) and has high internal consistency of .75 to .94 with 1-week test retest reliabilities of .60 to .91 (Evans et al., 2002).
Taking an aggregated sample (n = 10,761) from previous research studies a mean cut-off score of .99 was obtained when compared with a general population sample (Connell et al., 2007). The index for assessing reliable and clinically significant change was calculated for the CORE-OM using the method proposed by Jacobson and Truax (1991). The clinical cut-off score for the CORE-OM was rounded to a mean of 1.00 and reliable change was calculated as ± .48 . For reliable and clinically significant change to occur, a score must have moved from a clinical to nonclinical score and must have been reduced by at least a score of .48.

Procedure
All clients and therapists were informed that the study was interested in researching the quality of their relationship by looking at how they viewed each other. All clients and therapists provided written informed consent and an information sheet containing contact details of the researcher and they were under no obligation to take part or remain in the study.
The CORE-OM was completed by clients at pretherapy and after the third session of psychotherapy. The third session was selected as the second time point for analysis to maximize the number of respondents in the treatment as usual sample. The main site for data collection was a student counseling center where the average number of sessions attended was four. With only 30 of the 62 clients who reached session 3 providing further data at session 5, session 3 proved to be the most efficient set of data to assess early treatment progress and mutuality within the therapeutic relationship. The BLRI was completed at the end of the first and third sessions by both clients and therapists and each completed the measure rating how they felt towards the other person and how they perceived the other person felt towards them. All responses were collected anonymously by sealing response sheets and posting in envelopes at specific collection boxes situated in the clinical setting. The response sheets were collected by the researcher at regular intervals. Clients and therapists never looked at one another's response sheets for the BLRI although therapists did see the client's CORE-OM response sheets for both pre-and post-therapy as these forms were completed as part of treatment as usual within the clinical setting.

Therapeutic Conditions and Outcome
We hypothesized, first, client rating of therapist experienced congruence would be associated with treatment progress. The hypothesis was supported and client rating of therapist-experienced empathy, unconditional positive regard, and congruence (C-OS) at session 3 and treatment progress at session 3 were significantly correlated and remained so whilst controlling for distress at session 1, r(60) = −.27, two-tailed p < .05. Table I shows the intercorrelations, means, standard deviations, and alpha reliability coefficients for all variables at sessions 1 and 3.
Client rating of BLRI scores towards the therapist (C-MO) was also significantly correlated with CORE-OM at session 3 whilst controlling for session 1 CORE-OM scores, r(60) = −.26, p = .05. There was a significant positive correlation between client rating of therapist conditions and client rating of their own feelings towards the therapist (r(60) = .92, p < .0001).
Therapist rating of the quality of therapeutic relationship was a poor predictor of treatment progress. Session 3 client-rated CORE-OM was not significantly correlated with session 3 therapist rating of therapist-experienced conditions (T-MO), r(60) = −.11, p = .38. These findings support previous research that the client is the most reliable predictor of the association of therapist-experienced empathy, unconditional positive regard, and congruence with outcome.

Mutuality and Therapeutic Outcome
Our second hypothesis aimed to test the association of mutuality of Rogers's facilitative conditions of empathy, unconditional positive regard, and congruence with treatment progress. Relational mutuality in this instance was assessed by measuring client and therapist ratings of how they perceived the other as feeling towards them using the BLRI measure. First, scores on the BLRI for the two predictor variables, client perception of the therapist and therapist perception of the client, were standardized. Hierarchical linear multiple regression was used to test the moderating effect of therapist perception of the client on the association between client perception of the therapist and treatment progress. There are several steps to completing hierarchical linear multiple regression that were followed. First, an interaction term was created using the standardized scores for the predictor variables. Next, session 1 CORE-OM scores were controlled to take into account pretherapy levels of distress and giving a score for progress. All variables were entered into the regression and a significant proportion of variance in CORE-OM at session 3 was accounted for by the interaction between the two predictor variables. This was the case when all individual variables were controlled, R 2 change = .044, F-change = 4.458, p < .05. Table II shows the regression weightings and significance levels.
The next stage of analysis was to take the significant interaction term and interpret this by plotting two separate unstandardized regression lines. These were the standardized session 3 client rating of therapist-experienced conditions (C-OS), standardized therapist rating of the client-experienced conditions (T-OS) on the BLRI and the standardized level of session 3 client rated CORE-OM. The interaction chart is shown in Figure 1.
Session 3 therapist (OS) BLRI scores had a significant moderating effect on the association between session 3 client (OS) BLRI score and treatment progress in the first three sessions. From this we concluded that when both clients and therapists indicate high, rather than low, perceived levels of the therapeutic conditions, the association between client rating of the therapist-experienced conditions and treatment progress is strongest. Putting this another way, client rating of therapist conditions is a stronger predictor of treatment progress when therapists also perceive the client as experiencing higher rather than lower levels of the therapeutic conditions.

Further Analyses
Quality of the therapeutic relationship and treatment progress. The perceived quality of the therapeutic relationship conditions of empathy, unconditional positive regard and congruence increased significantly over the first three sessions. Client rating of therapist-experienced therapeutic conditions was significantly greater at session 3 than session 1, t(61) = −4.50, p < .001. Therapist rating of therapist conditions was not significantly different at session 3 compared with session 1. Client rating of their own therapeutic conditions experienced towards their therapist was significantly higher in session 3 than session 1, t(61) = −5.64, p < .001, and therapist rating of client-experienced conditions towards the therapist was also significantly greater in session 3 than session 1, t(61) = −5.85, p < .001. The effect size (es) for changes between sessions 1 and 3 in the perceived quality of the therapeutic conditions was calculated using Cohen's d using the pooled standard deviation for client rating of therapist conditions, es = .33 [95% CIs: .17, .49], and client rating of client conditions, es = .32 [95% CIs: Note. Coefficients appearing in brackets represent partial correlations controlling for CORE session 1 scores.
C-OS = client rating of the therapist conditions, T-OS = therapist rating of client conditions. .16, .48], therapist rating of therapist conditions, es = .18 [95% CIs: .04, .32], and therapist rating of client conditions, es = .45 [95% CIs: .29, .51]. The effect size change in the quality of the therapeutic relationship was moderate for client rating for other to self (OS) and for myself to other (MO). Additionally, therapist rating of other to self was a large effect. However, therapist myself to other effect size indicated only a small change, suggesting that therapists view themselves as providing consistently high levels of the therapeutic conditions.
Effectiveness of psychotherapy. The principal measure of treatment progress used in the present study was the CORE-OM. The data showed pretherapy client mean score as 1.95 (n = 62, SD = .55) which is above the clinical cut-off point. At session 3 results showed a mean score of 1.43 (n = 62, SD = .67). The session 3 mean score remained just above the clinical cut-off (1.00) for the CORE-OM. Of the 62 clients who completed at least the first three sessions of psychotherapy 55 (89%) clients showed at least some improvement whilst seven (11%) clients showed deterioration. Table III shows data for clinical and reliable change for all 62 participants. The overall pre-post effect size for therapeutic change as measured by the CORE-OM over the first three sessions was calculated as the difference between the first and third session mean divided by the pooled standard deviation (Cohen's d). The study produced an effect size (es) for change in the CORE-OM between sessions 1 and 3 of es = .85 (n = 62 [95% CIs: .67, 1.03]).

Discussion
The findings supported the mutuality hypothesis. Therapeutic progress was associated with mutuality of client and therapist experience of the therapeutic   relationship conditions as described by Rogers (1957). The association between treatment progress and client perception of therapist attitudes was stronger in dyads when therapists also gave their clients higher rating of providing therapeutic relationship conditions rather than a lower rating. These findings are the first to offer support to a "mutuality hypothesis" and for the experience of relational mutuality and its association with outcome. The present study provides evidence for the bi-directionality in Rogers's (1957Rogers's ( , 1959 therapeutic relationship conditions. Prior to the present study the relationship conditions proposed by Rogers were researched as unilateral therapistexperienced conditions. The findings reported here set out a new direction for research to understand the mutual generation and experiencing of the therapeutic conditions. The findings are also consistent with the positive effects of mutual communicative attunement and affirmation found in the Saunders (1999Saunders ( , 2000 studies. Mutual attunement and mutual affirmation refer to the degree of mutual understanding and positive affect experienced within the therapeutic dyad. Saunders (1999Saunders ( , 2000 found these to be related to outcome when assessed by either the therapist or the client. The present study has shown that mutual experiencing of Rogers's (1957) relationship conditions within the dyad is more strongly associated with outcome. Although not synonymous, mutual affirmation and attunement are related to and overlap with Rogers's (1957) conditions empathy and unconditional positive regard as they were measured in the present study using the BLRI. However, it is important to note that the current study used a measure of outcome looking at a generic model of distress. There are some difficulties and a measure of self-relatedness might have shown different findings. For example, it is possible that clients rated as high rather than low in mutuality of relationship conditions might also be clients more likely to improve in early sessions of therapy. Showing an improvement through lower levels of distress is different to demonstrating a significant shift in client self-relatedness. Future studies looking at mutuality would also benefit from including a measure of self-relatedness.
Recent studies carried out have looked at the concept of the real relationship and found that mutual genuineness was related to positive outcome (Gelso, 2002(Gelso, , 2009). The present study used a measure of the therapeutic relationship that included items from the congruence scale of the BLRI. Mutuality of the relationship scale was related to outcome, showing consistency with findings from studies of the real relationship. However, the potential for mutual experiencing of relationship conditions needs to be carefully considered. The tripartite therapist attitudes of empathy, unconditional positive regard, and congruence can also be construed as outcome as well as process variables. Rogers (1951) referred to therapy as starting a psychological chain reaction where the client receives the therapist's empathy and unconditional positive regard and becomes more congruent, selfunderstanding, and self-accepting. This activates a reciprocal process whereby the person is more likely to act genuinely and in an accepting and understanding way in relationships with others. The findings in the present study point towards changes in self-relatedness but no measure was included to assess this.
Previous research into Rogers's therapeutic conditions has considered them from a unilateral perspective . Typically, studies have assessed the extent the relationship conditions are provided unilaterally from therapist to the client. The findings from the present study suggest that future studies should consider bi-directional and mutual experiences of the relationship conditions and a measure of self-relatedness. Rogers's (1959) statement of personality and behavior change posited that the conditions were mutual and reciprocal and this positioned clients as active partners in the therapeutic relationship. As active partners in the therapeutic relationship clients and therapists experience the relationship conditions from and for one another. The findings from the present study suggest that in order for the client to receive the therapistoffered conditions they engage in the mutual relational process. These mutual relational experiences are directly positively associated with outcome.
The findings concerning mutuality within this study do not reach far enough to explore the relation to clients' characteristics. However, the view that the therapeutic relationship acts in concert with a range of client variables was a primary conclusion noted by Norcross and Lambert (2011). Additionally, it is important to recognize that the therapeutic relationship cannot be understood in isolation from other important client variables. Further study of the contribution clients make in maximizing the therapeutic benefit of the therapeutic relationship conditions, their capacity for mutually experiencing these, and the association with outcomes is in line with the increasing recognition of clients' contribution to successful therapy outcome.

Limitations
There are several limitations to the present study. For example, the sample size is relatively small and data for a number of clients are nested within individual therapists. Whilst this is a common problem in psychotherapy research the issue of sample size prevents meaningful analyses for the effect of nested data to be tested using appropriate statistical strategies. Thus, the analyses here are indicative, rather than conclusive, of the effect of mutuality on early treatment progress. Related to this is that findings in the present study referred to progress in only the first three psychotherapy sessions. Caution must be taken when interpreting changes from the first three sessions as they are not evidence for lasting change and the study does not provide data from a follow-up. Early changes in psychotherapy may be lost over time and the session 3 scores are not representative of the score when therapy was ended. Thus, due to the small number of therapists involved and the focus on early progress, findings from the present study ought not to be generalized to longterm treatment outcome. The present study used a correlation design. Whilst the progress made during an early stage of therapy is of interest, caution should be taken when drawing conclusions about the predictive capability of the mutual experiencing of Rogers's core conditions for outcome.
This study has gone some way to being the first study assessing mutuality as a therapeutic concept. As such, additional work remains to be carried out in further refining the terms for researching mutuality of the therapeutic conditions. One such issue is whether the BLRI will, in the long run, remain the best method for measuring the effects of mutuality of the therapeutic conditions of empathy, unconditional positive regard, and congruence. As the full scale is long it might be prudent to develop a shorter scale with a reliable factor structure for use in future studies. Larger studies with more therapists would also be an informative next step together with observer-rated measures of perceived mutuality of the therapeutic conditions. It would also be informative to replicate the study including a measure of self-relatedness.

Implications for Practice and Training
Considering the findings of the present study there are some implications for psychotherapy practice and training. First, for therapists it seems the study provides some evidence to suggest that when clients mutually experience empathy and acceptance towards the therapist progress in therapy might be greater. Therapist congruence plays a significant role, enabling the client to trust their own experiencing and thereby creating opportunities for clients to experience empathy and acceptance towards the therapist. Therapists who are congruent, both inwardly and outwardly, can foster the development of mutual empathic understanding and unconditional positive regard.
As therapists recognize their client as actively involved in the reciprocal process of the mutual generation and experiencing of the therapeutic conditions, clients are repositioned as agents of change. Thus mutuality has implications for the distribution of power within the therapeutic relationship. Recognizing the client is actively co-creating the relational atmosphere brings a greater sense of equity to the structure of the therapeutic relationship, one that might make some therapists feel less comfortable and secure in their position. Training can help support therapists in their development to see greater equity as beneficial and indicative of progress.
Training therapists to become intersubjectively aware is a relatively straightforward task. Any training will typically provide the opportunity for experiential contemplative introspection. Becoming more aware of internal states and experiencing will enable therapists to be more congruent. Building on the development of self-awareness, training environments can add reflective process into dyadic exercises. For example, two trainees can reflect on their intersubjective experience of the relationship. Additionally, working in exercises giving direct feedback on the accuracy of empathic attunement and feelings of acceptance in the relationship can equip both parties with greater awareness and communication skills.