Examining Our Tears: Therapists’ Accounts of Crying in Therapy

Introduction

The majority of therapists have cried during therapy with a client. A survey on ethical issues in psychological practice (Pope, Tabachnick, & Keith-Spiegel, 1987) found that 56.5% of 456 respondents had cried in the presence of a client. An informal survey of 19 psychodynamic therapists (Nelson, 2007) found that “about two thirds” (p. 1) of therapists reported having cried in therapy. More recently, Blume-Marcovici, Stolberg, & Khademi (2013) found that 72% of 684 psychologists and psychology trainees reported that they had cried during therapy with a client. Those who cry do so in an average of 7% of therapy sessions (Blume-Marcovici, et al., 2013). A study conducted by Trezza, Hastrup and Kim (1988) on patient crying found that patients cry in approximately 21% of therapy sessions. Based on these statistics, therapists cry approximately one third as frequently as their clients. A more recent study of client crying (Robinson, Hill, & Kivlighan, 2015) found client tears in approximately 14% of sessions. Compared to this figure, therapists experience tearfulness about half as often as their clients (of course, not taking into account intesity or longevity of crying). However, very little research exists regarding this most basic expression of emotion by the therapist in therapy, i.e., therapists’ crying in therapy (hereafter, TCIT). When searching for the terms “psychologist crying” or “therapist tears” in the PsycInfo database, there is only one result (Blume-Marcovici, et al., 2013). The words “therapist crying” produce one additional match: a 1995 qualitative dissertation by Jane Waldman focused on the crying experiences of ten psychodynamic psychotherapists.

A more in-depth search turns up three case studies on the topic of TCIT (Counselman, 1997; Owens, 2005; Rhue, 2001). Based on the case studies, along with Waldman’s dissertation (1995), several basic themes regarding TCIT can be discerned. It appears, for instance, that TCIT may be most likely to occur when therapy session content focuses on grief and loss. In addition, each study highlights the important role of therapist genuineness in treatment, and the ways in which their own tears acted as messengers of authenticity and compassion. Aside from these few case examples, however, very little is known about the clinical contexts in which TCIT tends to occur, or about therapists’ understanding of their own tears.

While TCIT specifically has received little research attention, patient crying in psychotherapy has been the topic of some theoretical writings (Labott, 2001; Nelson, 2008; Trezza, et al., 1988; Van Heukelem, 1979). Trezza et al., (1988) conducted the only study to date which directly asked therapists about their experiences with client crying. In their study of clinical psychologists and social workers, respondents reported clients “watery eyes” in 21% of sessions, “some tears” in 15%, “many tears” in 9%, and “sobbing” in 3%. Labott (2001), in a chapter on patient crying in psychotherapy, reviews the scant research on client crying and theories of crying in psychotherapy, as well as describes a small amount of research devoted to the importance of patient crying to therapeutic outcome. Van Heukelem (1979), a pastoral counselor, wrote an article called “Weep with Those Who Weep,” in which she encourages therapists to support and facilitate a client’s crying. In so doing, she briefly suggests that one possible response to client crying is a therapist’s own tears. In a chapter on patient crying in psychotherapy, Nelson (2008) presents an attachment-based perspective on crying and discusses how therapists may use their clients’ tears as a point of clinical assessment and intervention. Like Van Heukelem, Nelson acknowledges that therapists may have the urge to cry themselves when working with a crying client, and goes on to explain that the therapist’s own internal response to patient crying, which may lead to TCIT, provides valuable information about the patient’s grief and attachment orientation. In this way, the therapists’ understanding of and reaction to the clients’ tears—including the therapists’ own urge to cry—have important diagnostic and treatment-planning implications. However, no research to date has investigated the clinical contexts in which TCIT specifically tends to occur, or therapists’ understanding of their own tears.

Given that the majority of psychologists and psychology trainees will experience TCIT in their careers (72% in Blume-Marcovici, et al., 2013; 57% in Pope, Tabachnick, & Keith-Spiegel, 1987), it is important to have information regarding typical scenarios in which TCIT occurs, as well as psychologists’ experience of TCIT when it does occur. The present study sought to investigate psychologists’ and psychology trainees’ experiences of TCIT by reviewing data from 411 accounts of respondents’ most recent experience of crying in therapy. Three broad research questions were posed:

In answering these exploratory questions, the authors intend to present a framework for understanding the therapeutic situations in which TCIT may be most likely to occur (i.e., client demographics, when in treatment TCIT happens, session content), therapists’ experiences of their own tears (i.e., emotions felt, comfort/discomfort with tears, regret), and how clinicians work with their own tears in session (i.e., regarding therapeutic rapport, discussing TCIT with clients).

Method

Design and Ethical Issues

The study employed a survey design, asking psychologists and psychology trainees retrospectively about their experiences with TCIT. The survey was completed online. The study was approved by the Research Ethic Board of Alliant International University.

Participants

Four hundred and eleven (N = 411) U.S. psychologists, postdoctoral psychology fellows and psychology graduate students who reported that they had previously cried in therapy at least one time were asked to complete an additional survey about their most recent experience of crying in therapy.

In terms of sex, 80% (n = 327) of respondents were female, 20% (n = 84) male (N = 411). Ethnically, 82% (n = 333) of respondents identified as non-Hispanic white, 5% (n = 21) as Asian or Asian American, 5% (n = 21) as Hispanic or Latino, and 2% (n = 9) as Black or African American. The rest of other ethnicities made up 6% (n = 23) of respondents. Respondents’ age ranged from 24 to 72 years, with a mean of 37 years and a median of 26 years (N = 409). At least 40 U.S. states were represented (N = 411).

In terms of professional status, 49% (n = 202) of respondents were graduate students in doctoral programs in psychology, 47% (n = 193) were licensed psychologists, and 4% (n = 16) were post-doctoral fellows or license-eligible psychologists (N = 411). Regarding theoretical orientation, 27% (n = 117) of respondents identified as cognitive-behavioral therapists, 25% (n = 102) as eclectic/integrative with a psychodynamic emphasis, 22% (n = 89) as eclectic/integrative without a psychodynamic emphasis, 12% (n = 48) as psychodynamic, 12% (n = 48) as “other,” and 2% (n = 7) as psychoanalytic (N = 411).

Measures

The present study was conducted as part of a larger research project on TCIT (Blume-Marcovici, 2012): how often and when it occurs, demographic and clinical factors pertaining to which therapists cry, and perceived clinical impact of TCIT (Note: crying was defined following the Adult Crying Inventory (Vingerhoets & Cornelius, 2001): “tears in one’s eyes due to emotional reasons”). Participants from the larger study (N = 684) completed the TCIT Survey 1, a 40-item survey created by the research team to gather information about whether or not—and which—therapists had cried in therapy (results reported elsewhere; Blume-Marcovici, et al., 2013). Of these respondents, 72% reported previously having cried at least once in therapy. Those 72% of participants who reported having previously cried in therapy and who consented to complete a second survey (N = 411), also completed TCIT Survey 2, a 49-item survey created by the research team to gather information about past experiences of TCIT. The TCIT Survey 2 had three primary subsections—Clinical Context of TCIT, Therapists’ Experiences of TCIT and TCIT’s Impact on Treatment—and yielded both quantitative and narrative data. The TCIT Survey 2 was piloted on a group of 20 psychology graduate students prior to the start of the study, and feedback was incorporated into final survey.

The TCIT Survey 2 included items utilizing 7-point Likert-type scales (e.g., “The client I worked with was aware that I was crying,” where 1 = strongly disagree, 4 = neither agree nor disagree, and 7 = strongly agree), and nominal scales (e.g., “Did you discuss your tears with your client?” to which the respondent could select “Yes” or “No”). Several items were in an unstructured, narrative format to which respondents could write in open-ended responses (e.g., “Describe non-identifying session content of the therapy session”). Two items used a checklist format (e.g., “Which emotion(s) did you feel when you cried in therapy? Twenty emotions were listed and the respondent could check as many as necessary. “Which emotion(s) do you believe your client felt when you cried?” which listed the same 20 emotions).

Procedure

All data was collected through a single, cross-sectional, retrospective, self-report internet survey. A recruitment email, with a hyperlink to the survey, was sent to university program directors, psychology training sites, and psychology associations throughout the United States. Participants were asked to give their consent to participating by checking “yes” or “no” on a consent form. Participants were not compensated for their participation in the study, but could choose to donate five dollars. to one of two non-profit organizations (UNICEF or The Nature Conservancy) upon completion of the survey. Those participants who responded on TCIT Survey 1 that they had previously cried in therapy were prompted to complete TCIT Survey 2, which asked participants to answer questions about their most recent experience of TCIT. Quantitative and qualitative data was collected.

Data Analysis

Descriptive statistics were used to determine frequency of responses regarding characteristics of clients with whom respondents cried in therapy, other aspects of the clinical context of TCIT, therapists’ experiences of TCIT and the impact of TCIT. Pearson’s Chi Square tests were run to analyze response trends in items related to the clinical context of TCIT (e.g., significant differences in sex of client, and therapy format in which TCIT occurred) and to determine any significant differences in whether respondents discussed their TCIT with their clients. One-way ANOVAs were calculated to determine any significant trends regarding groups of therapists who were more (or less) likely to discuss their TCIT with their clients. In exploratory analyses, Pearson correlations and Spearman rank-order correlations were calculated to determine significant relationship trends between variables regarding clinicians’ experiences of TCIT.

Content analysis was performed to analyze narrative responses to a question regarding session content during the respondents’ most recent TCIT episode. The responses were coded by the primary researcher for session content themes. These codes were reviewed by a psychology graduate-student research group to reach consensus about any unclear content codes and multiple/overlapping codes. Based upon the groups’ decision, themes were updated and divided into overarching content categories by the primary researcher. These overarching categories were reviewed and revised by two psychologists acting as advisors to the project. Frequency of content categories, as well as content themes within categories, was analyzed and reported.

Results

Clinical Context for TCIT

Client Characteristics

Based on respondents’ answers to questions about the demographics of the client with whom they most recently cried, we found that TCIT occurred with clients of all ages—from young children to older adults at the end of their lives, and with a range of diagnoses, from posttraumatic stress disorder and depression to schizophrenia and personality disorders. The clients were female in 64% (n = 193) of the most recent crying episodes, and male in 35% (n = 106). In approximately 1% (n = 1) the client identified as genderqueer (only gender of individual therapy clients are reported here). In several accounts, gender of client was not included. A Chi Square-Goodness-of-Fit analysis was conducted to determine any significant differences between reported gender of client. All expected cell frequencies were greater than five. There was a statistically significant difference in reported gender of client, with TCIT occurring more frequently with female clients than male clients (χ²(1, n = 299) = 25.31, p < .001).

Table 1 shows the gender of client as well as the gender of therapist. Female therapists cried more with female clients (65%) in their most recent TCIT episode than they did with male clients (31%), while male therapists were quite split, with 51% crying with female clients and 48% crying with male clients. A Pearson Chi Square analysis shows this gender difference to be statistically significant (χ²(1, n = 299) = 5.72, p = .017), supporting the finding that female therapists cried more with female clients while male therapists have an equal distribution of the gender of client with whom they cry.

Therapy Format

Respondents reported that of their most recent tcit episodes, 74% (n = 304) occurred with individual clients, while 4% (n = 15) were with family, 3% (n = 13) were with groups, and less than 1% (n = 3) were with couples (the remainder of accounts (18%; n = 76) did not report specific therapy format). A Chi Square-Goodness-of-Fit test was conducted on this response set and all expected cell frequencies were greater than five. There was a statistically significant difference with respondents reporting crying in individual therapy significantly more frequently than other therapy formats (χ²(3, n = 335) = 773.29, p < .001).

When Do Tears Occur?

Respondents were asked to report at what point in treatment their most recent TCIT episode occurred (n = 365). Table 2 presents results. A Chi Square-Goodness-of-Fit analysis was run on this response set; all expected cell frequencies were greater than five, and there was a statistically significant difference between the phases of treatment in which TCIT occurred, with respondents reported crying significantly more often in “mid-treatment,” as well as (though to a lesser extent) “late in treatment,” than in other phases of therapy (χ²(4, n = 355) = 73.52, p > .001).

Intensity and Duration of TCIT

Respondents were asked to indicate the intensity of their crying in their most recent TCIT episode. Table 3 presents results. A Chi Square-Goodness-of-Fit analysis was run on this response set and all expected cell frequencies were greater than five. There was a statistically significant difference between expected and occurring frequency of responses such that respondents reported having “tears in eyes only” significantly more often than having tears on cheeks or sobbing (χ²(2, n = 361) = 516.80, p > .001).

Respondents were asked to report the duration of their crying in their most recent TCIT episode. Table 4 presents results. A Chi Square-Goodness-of-Fit test was conducted on this response set and all expected cell frequencies were greater than five. There was a statistically significant difference found between duration of TCIT responses, such that respondents reported “instantaneous,” as well as (to a lesser extent) “less than one minute,” TCIT episodes significantly more frequently than other TCIT durations (χ²(4, n = 362) = 489.74, p > .001).

Did the Client Know?

Respondents rated the extent to which they believed their client was aware that they were crying in their most recent TCIT episode on a Likert-type scale. Just over half (55%, n = 198) of respondents indicated that, from their perspective, their client was aware of their (i.e., the therapist’s) tears (by selecting 5 (somewhat agree), 6 (agree) or 7 (strongly agree)). Approximately one quarter (28%; n = 101) indicated that their client was not aware of the therapist’s tears (by selecting 1 (strongly disagree), 2 (disagree) or 3 (somewhat disagree), while 17% (n = 61) reported that they did not know whether or not their client was aware that they were crying (i.e., selected 4 (neither agree nor disagree)). A Spearman rank order correlation test was conducted in exploratory analyses to determine any relationship between the intensity of a therapists’ tears and client awareness of TCIT, finding that more intense TCIT (i.e., sobbing over tears on cheeks over tears in eyes only) correlated with increased client awareness of therapists’ tears (as reported by the therapist) (r = .26, p > .001, n = 357).

Was the Client Crying?

In 73% of respondents’ most recent TCIT episodes (n = 262), the client was crying when therapist tears occurred. However, 27% (n = 96) of respondents stated that their client was not crying when they cried. A Chi Square-Goodness-of-Fit test was conducted on this response set and all expected cell frequencies were greater than five. There was a statistically significant difference between expected and occurring cell frequencies such that respondents reported their client to be crying during TCIT significantly more than they reported their client not to be crying during TCIT (χ²(1, n = 358) = 76.97, p > .001).

TCIT Session Content

Participants were asked to describe the session content of their most recent TCIT episode. The 331responses to this question were coded for common themes regarding session content. These themes were grouped into categories which were, in order of frequency: grief and loss, trauma, termination, patient suicidality, and client gratitude toward the therapist.

The following excerpt from a response to the question about session content in the respondent’s most recent TCIT episode provides an example of the most frequent content area described by respondents: grief.

Overall, 30.8% (n = 102) of responses were coded within the “grief” category, i.e., 30.8% of all most recent TCIT episodes occurred when grief was the content area of focus. Of these, 82% (n = 84) had a session content theme of loss of a loved one, while 18% (n = 18) were coded with the theme of patient describing grief over his/her own mortality (i.e., end-of-life issues). After grief, trauma was the next most common content category of the session, with 27.1% (n = 90) of the most recent TCIT episodes occurring when therapy focused on trauma work. Of these, 22% (n = 20) were coded with the theme of sexual trauma and 10% (n = 9) with the theme of combat-related trauma (the remainder were other types of trauma or were unclear about the specific nature of the trauma). Termination was the content category of the therapy session in 11.5% (n = 38) of most recent TCIT episodes, with 31.6% (n = 12) of these cases coded with the theme of forced termination (for instance, due to the therapist leaving a placement versus due to a “natural” ending of the treatment). The client expressing suicidality was the content category in 3.3% (n = 11) of most recent TCIT episodes. Client gratitude for the therapist was the content category of the therapy session for 3.3% (n = 11) of the most recent TCIT episodes.

In response to this question, 8.9% of respondents (n = 29) did not provide thematic information about session content, but instead described their client’s emotionality as the primary backdrop for TCIT (i.e., these responses did not fit into any of the above-mentioned content categories). In these particular narrative accounts, therapists discussed the client’s emotional expression (or lack thereof) as a primary focus of the session in which TCIT occurred. In 62% (n = 18) of these accounts, respondents described intense and uncharacteristic emotional expressions by the client (i.e., intense, loud sobbing) as the clinical context for the therapist’s tears. In the remaining 38% (n = 11), respondents described their own TCIT occurring in the context of a client’s apparent lack of emotion while talking about something that would typically engender an affective display. These clinicians described crying when their client failed to express appropriate and/or congruent emotion during a session. In these cases, some clinicians explicitly described feeling “hit” by their client’s emotion, feeling an emotion “come on suddenly” or “out of nowhere,” almost as though it were “not [their] own” (quotations taken verbatim from participants’ responses). Some of these participants conceptualized this process as a transference of emotion from the client to the therapist; a “projection of emotion that [the] client did not allow himself [or herself] to feel” (quotation from participant). The below excerpt is an example of the latter such clinical context for TCIT.

In the sizable remainder of respondents’ most recent TCIT episodes (15.1%; n = 50), the session content did not clearly fit into any of the above categories and was coded “other”.

Therapists’ Experience of TCIT

Emotions Felt During Most Recent TCIT Episode

Respondents were asked to indicate which emotion(s) they felt during their most recent TCIT episode. They were provided a list of 20 possible emotions (they could indicate more than one emotion). Table 5 portrays the seven emotions that more than 10% of respondents checked (n = 369).

Respondents were also asked to indicate which emotions they believed their client was feeling when they (i.e., the therapist) cried during the most recent TCIT episode. Respondents were provided with the same list of 20 possible emotions (and could indicate more than one emotion). Table 6 portrays the 13 emotions that more than 10% of respondents checked (n = 366). For both therapists and clients, sadness was the most common emotion reported, from the therapists’ perspective.

Comfort with TCIT

Respondents were asked to rate their level of comfort with TCIT in their most recent episode on a Likert-type scale. The majority of respondents indicated that they felt some level of comfort with their tears (59%; n = 212). On the other hand, approximately one third of respondents expressed some level of discomfort with crying in therapy (31%; n =113). The remainder of respondents (10%; n = 35) remained neutral regarding this question. Exploratory analyses found that comfort with TCIT significantly and positively correlated with one’s level of experience as a clinician, though this correlation was small. The more years a clinician was in practice, the more comfortable he/she was with his/her tears (r = .178, p = .001, n = 328. A curious relationship was found through exploratory analyses in which the more a respondent agreed with the statement, “My client was aware of my tears,” the more they also agreed with the statement, “I felt comfortable with my tears” (r = .324, p > .001, n = 359).

Authenticity and TCIT

Respondents were asked to indicate the degree to which they felt their TCIT was authentic on a Likert-type scale. The majority of respondents (87%, n = 313) indicated that their tears were authentic, with only 5% (n = 20) reporting that their tears were not authentic. The remainder (8%; n = 28) remained neutral regarding the authenticity of their tears.

To Whom Did the Tears Belong?

Respondents indicated the extent to which they felt their most recent TCIT episode was related to their own personal life circumstances on a Likert-type scale. The majority of respondents (77%, n = 277) indicated that they did not feel that their tears were related to their own personal issues, while 16% (n = 56) reported that their tears were personal in nature (i.e., had to do with their own life circumstance). The remainder (7%; n = 27) were neutral in response to this question. Exploratory analyses found that the more a respondent agreed that their TCIT was related to their own personal circumstances (versus unrelated to personal circumstance), the more that respondent also tended to report feeling uncomfortable about his/her tears (e.g., agreeing with the statement “I was uncomfortable with my tears”; r = .211, p > .001, n = 358).

In the optional “comments” section at the end of the survey, respondents were allowed to leave comments for the research team; several participants emphasized the importance of TCIT not distracting the therapy from the focus on the client. Some participants expressed concern that a clinician’s tears may represent a professional – even an ethical – breach, if it takes the focus off of the client’s own material. One participant, for instance, wrote (emphasis in the original quotation): “If a therapist cries, it should be the therapist crying the client’s tears, versus his/her own. Otherwise, the therapist has some countertransference issues to work out in his/her own personal therapy.” Another participant commented (emphasis in the original quotation), “Crying with the client can communicate empathy. Crying for the client or for yourself is unhelpful.” Yet another participant stated, “Crying is bad…if therapy was for the therapist, I could understand. They should be fired for crying and making it about their own needs. Shame on them for being inappropriate.”

TCIT and Regret

Respondents were asked to agree or disagree on a Likert-type scale with the statement, “I wish I had not cried” in their most recent TCIT episode. The majority of respondents (81%, n = 291) did not express remorse or regret at having cried (i.e., wishing they had not cried). However, 9% (n = 31) expressed some regret, wishing that they had not cried. The remaining 10% (n = 36) indicated that they were not sure, one way or the other, if they regretted their tears. Respondents were asked to indicate the extent to which they felt that their most recent TCIT had been a therapeutic error (N = 359) by agreeing or disagreeing with the statement, “My tears were a mistake” on a Likert-type scale. The majority of respondents (83%, n = 299) indicated that they did not feel their tears were a mistake. In fact, only 4% (n =15) of respondents indicated that, to some extent, their crying had indeed been a therapeutic misstep. The remainder of respondents (13%; n =45) remained neutral regarding this question.

Exploratory analyses regarding regret found that the more a respondent agreed with the statement, “I was uncomfortable with my tears,” the more he or she tended to endorse regret about TCIT, both in the form of feeling their tears were a mistake (r =.53, p >.001, n =357) and in the form of wishing they had not cried (r = .60, p > .001, n = 356). In addition, a significant, positive correlation was found between feeling that one’s TCIT was related to one’s own personal life circumstances and agreement with the statements, “my tears were a mistake,” (r =.23, p > .001, n =359) and “I wish I hadn’t cried” (r =.29, p >.001, n =357). On the other hand, the more a respondent agreed with the statement, “My client was aware of my tears,” the less he or she tended to report regret about crying, either in the form of feeling tears were a mistake (r = −.25, p >.001, n =358) or in wishing tears had not been shed (r = −.26, p > .001, n =357).

Impact of TCIT

TCIT and Insight

In 27.9% (n =100) of most recent crying episodes, respondents agreed with the statement, “my tears allowed me to realize something new about my client.” Similarly, when describing the session content of their most recent TCIT episodes, several participants spontaneously alluded to TCIT allowing the clinician to become aware of new relationship dynamics. As a case example, a female respondent/therapist from the present study described, in her narrative account of session content, tearing up while feeling highly protective, and angered on behalf, of a male client when he talked about the grief and loss he had suffered as a child. This clinician felt her TCIT made her aware of her strong protective response, and allowed her to recognize that this response was one likely elicited by the client in many aspects of his life. The clinician described feeling that her TCIT provided her with “first-hand experience of the way in which [her] client related to people in his life” (quotation from respondent).

TCIT and Therapeutic Rapport

As reported elsewhere (Blume-Marcovici, et al., 2013), just over half of respondents (53%; n = 191) answered the question “Do you think your crying changed your relationship with your client?” by indicating that they believed nothing to have changed in their relationship due to their tears, while just under half (46%; n = 163) reported that they felt their relationship with their client improved due to TCIT (less than 1% (n = 3) reported that they believe their relationship worsened due to TCIT). When therapists of different self-reported theoretical orientations were compared in their responses to this question during exploratory analysis, more than half of respondents who identified as dynamically-oriented reported that their crying had changed the therapeutic relationship for the better, i.e., reported improvement (“psychoanalytic” category (83%), “psychodynamic” category (63%) and “eclectic/integrative with a dynamic emphasis” category (55%)). Conversely, more than half of clinicians who identified as “cognitive-behavioral” (70%) or “eclectic/integrative without a dynamic emphasis” (59%) felt that their tears had no impact on the relationship, i.e., reported that nothing changed.

A Pearson’s Chi Square analysis was performed to determine if there was a statistically significant difference between the responses of dynamic clinicians as a group (all respondents who identified as psychoanalytic, psychodynamic and eclectic with a dynamic emphasis were combined into one “dynamic” group for this analysis) and cognitive behavioral therapists (CBT) clinicians with regards to the impact of TCIT on the therapeutic relationship. Because less than 1% (n = 3) of respondents stated that the therapeutic relationship worsened due to TCIT, these responses were filtered out to look at differences in reporting relationship improvement versus no improvement between the two groups. The analysis showed a significant difference in reporting improved relationship ((χ²(1, n = 239) = 19.71, p > .001)) when therapists practicing psychodynamic and CBT were compared, with psychodynamic clinicians significantly more likely to report relationship improvement than CBT clinicians, while CBT clinicians were more likely to report that nothing changed.

TCIT as a Technique

Respondents were asked to report the extent to which they agreed or disagreed with the statement, “the act of crying in therapy was a therapeutic technique that I used consciously.” Of the 363 respondents, 84.5% (n = 307) disagree, while 7.5% (n = 27) agreed (8% (n = 29) neither agreed nor disagreed.

Talking About TCIT with Clients

Respondents were asked to report whether or not they had discussed their tears with their clients during their most recent TCIT episode. Of the 354 who answered this question, 61% (n = 215) indicated that they had not discussed their tears with their client(s) (i.e., checked “no”), while 39% (n = 139) indicated that they had discussed their tears with their client (i.e., checked “yes”). A Chi Square-Goodness-of-Fit test was run on this response set and all expected cell frequencies were greater than five. There was a statistically significant difference found, respondents reported not discussing TCIT significantly more frequently than they reported discussing TCIT with clients (χ²(1, n = 354) = 16.32, p > .001). When respondents were asked the extent to which they agreed or disagreed (on a Likert-type scale) with the statement, “My tears communicated something to my client that words could not express,” 65.7% (n = 237) of respondents agreed, while 16.9% (n = 61) disagreed (17.5% (n = 63) remained neutral). These respondents were significantly less likely to report discussing their tears with their clients (F = 34.33, df = 1, p > .001, d = .65; “yes” discussed tears: n = 139, M = 5.44, SD = 1.29; “no” did not discuss: n = 214, M = 4.48, SD = 1.64).

Whether or not a respondent reported discussing his/her TCIT with the client was analyzed in relation to whether or not the respondent indicated a change in the relationship due to TCIT (i.e., relationship improvement, worsening, or no change). Table 7 portrays results from the 349 respondents who answered both questions. As shown, the majority of those who discussed their TCIT reported improvement in the therapeutic relationship (66%), while the majority of those who did not discuss their TCIT (66%) reporting that nothing changed in the relationship. A Pearson Chi-Square analysis found this difference to be significant (χ²(1, n = 346) = 36.07, p > .001), with respondents who discussed TCIT with their client significantly more likely to report relationship improvement due to TCIT, while those who did not discuss TCIT were more likely to report no change in the relationship (note: because less than 1% (n = 3) respondents reported that their relationship worsened due to TCIT, these responses were filtered out for this analysis).

The reported intensity of crying was analyzed in relationship to discussing TCIT with one’s client. A Pearson Chi-Square analysis found a statistically significant relationship between the intensity of the clinicians’ tears and discussing TCIT (χ²(1, n = 351) = 16.63, p > .001), with respondents who reported more intense tears (tears on cheeks or sobbing versus tears in eyes only) also significantly more likely to report having discussed TCIT with their client. In addition, clinicians whose clients were aware of their tears (by the clinicians’ report) were significantly more likely to have reported discussing their tears with their client (F = 142.06, df = 1, p > .001, d = 1.35; “yes” discussed tears: n = 138, M = 5.7, SD = 1.22; “no” did not discuss: n = 214, M = 3.65, SD = 1.77) than clinicians who reported that their client was not aware of their tears.

Discussion

Typical TCIT Clinical Context

Based on data from the 411 most recent TCIT episodes collected and analyzed in the present study, it appears that TCIT occurs with a range of clients (including children, adolescents, and adults) who have a range of diagnoses and in a range of session content areas. Similar to the few prior studies on TCIT (Counselman, 1997; Rhue, 2001; Waldman, 1995), content analysis of the present sample’s most recent TCIT episode session content determined that grief was the most common session topic in which TCIT occurred, followed by trauma and termination. However, the session topics that accompanied TCIT varied greatly (15% did not fit in any specific content category). The majority of most recent TCIT episodes were reported to be in individual therapy settings, with fewer cases reported in group, couples, and family treatment settings. Crying was most often described as instantaneous or brief (less than one minute in duration) and as occurring in the form of tears in the eyes (i.e., “tearing up”), though longer and more intense (i.e., tears on cheeks, sobbing) crying by the therapist was also reported. Therapist crying in therapy was reported to happen throughout the stages of treatment, from first meeting with a client (i.e., during an intake) to finally saying goodbye (i.e., during a termination session). However, TCIT was most likely to occur in “mid-treatment” or “late in treatment” than in other stage of therapy. This may be because of the inclusive nature of the categories; the labels “mid-treatment” and “late in treatment” encompassed more therapy sessions than the other more specific options (intake session, early in treatment, termination session) and, therefore by default, more respondents selected them. However, given that the label “early in treatment” was similarly inclusive, a more substantive hypothesis is that TCIT occurs more frequently once a strong bond between client and therapist has been established. In addition, as described above, termination was a session content that tended to accompany TCIT and discussion of termination likely occurs later in treatment, in preparation for saying good-bye. Therapist crying in therapy was least likely to be reported to occur in an intake session. This finding seems to support the results of an informal survey on TCIT conducted by Nelson (2007) in which respondents anecdotally reported that TCIT early in treatment would be more likely to have a detrimental impact on therapy.

Respondents overall reported crying more frequently with female clients. However, when taking into account the sex of the therapist, it appears that female therapists were more likely to report crying with female clients, while male therapists were equally likely to report crying with male and female clients. In a dissertation on self-disclosure, Hansen (2008) described a similar finding. She found that male therapists disclosed with both male and female clients while female therapists disclosed less with male clients. It is, thus, possible that the gender trends found in therapist self-disclosure may apply to trends in TCIT. Future research may benefit from investigating this further.

More often than not, the client was aware of the clinician’s tears (by the therapist’s report). It is important to note that this was not inherently negative or therapeutically detrimental. In fact, clinicians whose clients were aware of their tears reported feeling more comfortable with TCIT than clinicians whose clients were not aware of their therapist’s crying. It may be that the latter (i.e., clinicians whose clients did not notice TCIT) were preoccupied with hiding their tears and, thus, the experience of TCIT was more uncomfortable due to fear—and anticipated shame—of being exposed. It is also possible that clinicians who were more comfortable with TCIT were less likely to hide their tears and more likely to explicitly “use” their own tears in a therapeutic sense. Indeed, clinicians who reported that their client was aware of their tears were also more likely to have reported discussing their tears with their client.

Therapists’ Experiences of TCIT

The most common emotions that accompanied TCIT for the clinician were sadness and feeling “touched” or moved by emotion, with more than half of clinicians indicating that they felt these emotions in their most recent TCIT episode. Most of the time clinicians reported feeling comfortable with their own TCIT, their tears were authentic, and their tears were related to their client’s situation (as opposed to the therapist’s personal life circumstances). When the latter was true, however, (i.e., TCIT was related to therapist’s own personal situation, as it was 16% of the time in the present study) clinicians were more likely to report feeling uncomfortable with their tears and to regret TCIT. Regret about TCIT in general, however, was relatively uncommon, with less than 10% of respondents in the present study indicating that they wished they had not cried in their most recent TCIT episode.

Impact of TCIT

Most of the time (i.e., in 84.5% of cases) TCIT was not a clinical technique consciously utilized by a clinician, suggesting that TCIT typically occurs without clinical motivation. Based on research by Cornelius (1981; 1997), who wrote that the social purpose of tears is to communicate sympathy, and that they can only communicate sympathy if they are seen as involuntary, we surmise that the involuntary nature of TCIT may be important to transmitting genuine compassion. While TCIT appears not to be a conscious intervention or technique, it can lead to new insights, as was the case in 28% of most recent TCIT episodes reported. Further, TCIT itself may be the impetus for an intervention, such as discussing the new insight or discussing the TCIT itself.

To Discuss or Not to Discuss

When TCIT occurs, the clinician is in a position to decide whether or not to address it with the client. The majority of respondents in the present sample did not discuss TCIT with their clients (61%); however, 39% did discuss their tears, and those who did were significantly more likely to report improvement in the relationship with the client. It may be that discussion of TCIT leads to increased rapport because it allows the client to fully appreciate the therapists’ empathy. Orlinksky & Howard (1986), for instance, explain that only client-perceived empathy is related to therapeutic outcome. Whatever the mechanism, it appears that discussing TCIT with a client may be a way to make it productive and positive within the therapy.

The intensity of a clinician’s tears may have been a factor in whether or not a clinician discussed TCIT with their client. Clinicians with more intense tears (i.e., tears on cheeks and sobbing versus tears in eyes only) were significantly more likely to report discussing TCIT with their client. Since discussion of TCIT was related to reported improvement in rapport, and discussing TCIT occurred more often with more intense tears, it may be particularly advisable to discuss TCIT with a client when a clinician’s tears are more intense or more visible to the client.

The present study did not capture details regarding how clinicians discussed their tears with their clients. While discussing tears was associated with relationship improvement, numerous respondents who did not discuss their tears still reported relationship improvement. Indeed, in 66% of TCIT episodes, respondents reported that their tears communicated something to their client that words could not express.

The clinician may want to reflect on whether new information was learned about the client or the therapy due to TCIT when deciding whether to discuss it further. If new information or a new insight about the client or the therapeutic relationship was gleaned due to TCIT (as 28% of therapists in the present study reported), the therapist may want to consider whether this information would be beneficial to share with the client. A TCIT case example was described in the Results section (TCIT and insight) in which a clinician described crying when she felt protective and angered on behalf of her client. As the respondent noted, the TCIT may have provided her with new information about her client’s pattern of relating. The clinician was then in a position to decide whether or not to discuss her tears with her client as a means of sharing this new understanding of the client’s relational style (particularly if the therapist had other evidence to support this interpretation and believes it would be beneficial to the treatment).

As reported in the Results section, 27% of clients were not crying during the most recent TCIT episode, and content analysis revealed a session content theme of the client’s lack of emotion accompanying TCIT in a subset of cases. This suggests that therapists’ tears may, at times, be a sign of unprocessed or unexpressed emotion by the client. If the therapist believes TCIT has occurred due to “projection,” the clinician has, by virtue of this understanding, gained new information about the client’s defensive structure, as well as information about what emotions are not tolerable to the client. This may be useful information to share with the client. If the clinician becomes aware of a strong emotion (sadness, feeling touched, warmth, loss, powerlessness, gratitude and joy were the most commonly reported) causing the TCIT (possibly through projection), the clinician may consider interpreting or reflecting this emotion through discussion of his/her own TCIT, particularly if it appears that the client is not aware of feeling such an emotion.

Keeping TCIT Client-Centered

Based on optional comments, a theme emerged in which respondents expressed a concern that TCIT could shift the therapy to become therapist centered. Several participants cautioned that it is “important to keep the focus on the client.” We recommend that if a clinician chooses to discuss his/her tears with the client, it the focus should be on the client’s experience of TCIT, or on the clinicians’ understanding of how TCIT relates to the client’s treatment. The therapist may verbalize to a client who expresses no affect that his/her story is in fact sad, discuss the importance of emotional expression or containment, express the sense of connection the therapist feels to a client who may struggle to feel connected or validate an emotion to a client who has been consistently invalidated.

If a therapist believes that the cause of TCIT was caused by personal circumstances or emotions (e.g., grief about the therapist’s own loss, TCIT related to therapist burnout, etc.), the therapist be particularly thoughtful about the choice to discuss TCIT with the client. If the therapist chooses to discuss TCIT, it should be with the well-being and benefit of the client in mind, not in an effort to make an excuse, give a personal explanation, or to find further emotional release. Similar guidelines have been applied to therapist self-disclosure (Goldfried, Burckell, & Eubanks-Carter, 2003; Knox & Hill, 2003).

Limitations

The present study surveyed only psychologists and psychology trainees. No data was collected from clients or from outside observers, thus, the perspectives in this study are exclusively those of practitioners. Future research would do well to incorporate the important perspective of clients regarding TCIT. Additionally, the present study asked participants to focus on their most recent experience with TCIT as a means of increasing the likelihood of accurate recall and eliciting a “typical” scenario for TCIT; however, it may be that most recent TCIT experiences were not “typical.” They also may not have been as meaningful or powerful as clinicians’ other experiences of TCIT (as several participants commented at the end of the survey). Future researchers may consider focusing on eliciting respondents’ most profound experiences of TCIT, as a means of capturing the full potential positive and/or negative impact of TCIT on treatment.

Other important limitations are that this study was based on a sample that was largely Caucasian (82%) and was of a limited size (N = 411), both of which impact generalizability. In addition, the self-report survey method asked respondents to recall events from memory, and such autobiographical events may be poorly remembered; further, memory distortions are more pronounced when the events being recalled are emotional in nature (Bradburn, Rips, & Shevell, 1987).

An additional noteworthy study limitation (that is also a direction for future research) is the omission of details regarding how therapists discussed their TCIT with clients. This may include a switch from a retrospective study (such as the present study) to a more process-oriented one. In addition, the content analysis conducted in the present study was aimed specifically at determining session content (as this was one of the original research questions). More open-ended, qualitative research approaches to broader-based narrative accounts of TCIT and/or interviews with practitioners and clients may allow new knowledge about TCIT to emerge. Future research may also focus on therapist countertransference as well as understanding the specific stages of therapy in which therapists cry.

Conclusion

To our knowledge, no research to date has investigated the specific clinical scenarios in which TCIT tends to occur, the therapists’ experiences of TCIT, or ways in which clinicians manage their tears in treatment. As the majority of clinicians will experience TCIT at some point in their career (Blume-Marcovici, et al., 2013; Pope, et al., 1987), the purpose of the present study was to provide a framework for understanding these aspects of TCIT in hopes that having such a frame will help clinicians feel prepared for the moments in therapy when they may cry.

Special thanks to Dr. Matthew Porter and Dr. Sharon Foster for their valuable input on this project.