Patients’ Self-presentational Tactics as Predictors of the Early Therapeutic Alliance

Agenda setting is a means for patients to present themselves as motivated and cooperative. In medical environments patients are trained to use Agenda setting as a self-management skill (Frankel, Salyers, Bonfils, Oles, & Matthias, 2013). Researchers have found that the use of Agenda setting was associated with better treatment outcomes, reduced treatment costs (Bodenheimer, Lorig, Holman, & Grumbach, 2002), higher satisfaction of patient (Michie, Miles, & Weinman, 2003) and physician (Beckman, Markakis, Suchman, & Frankel, 1994; Roter, Hall, Blanch-Hartigan, Larson, & Frankel, 2011), less premature acceptance of hypotheses by the physician (Beckman et al., 1994), and higher treatment adherence (Michie et al., 2003). It is possible that Agenda setting will produce similar outcomes with psychotherapy patients where they may have a clearer idea about what they want to achieve in therapy and practitioners in turn would appreciate this. Conversely, an overly concrete conception about what should be accomplished in a therapy could render patients less open towards the therapists’ suggestions. Nevertheless, results of the aforementioned studies seem to provide more support for Agenda setting to have a positive effect. Therefore we proposed the following hypothesis:

Similar to Agenda setting, the tactic of Provoking a response from the therapist is indicative of an active communication between patient and therapist. Studies in medical settings have found that patients who asked more questions had better outcomes as measured by level of anxiety (Thompson, Nanni, & Schwankovsky, 1990), role limitations, physical limitations (Greenfield, Kaplan, Ware Jr, Yano, & Frank, 1988; Greenfield, Kaplan, & Ware, 1985; Kaplan, Greenfield, & Ware Jr, 1989), functional status (Greenfield et al., 1985; Kaplan et al., 1989), and physiological status (Greenfield et al., 1988; Kaplan et al., 1989).

Asking questions is a means for patients to actively participate in the treatment, which should, similarly to Agenda setting, be welcomed by practitioners. We therefore hypothesize that it is positively associated with the therapeutic alliance in psychotherapy.

The tactic of Self-promotion has been investigated most often in organizational settings. Barrick, Shaffer, and DeGrassi (2009) found that people who used Self-promotion in job interviews were evaluated more positively by interviewers than those who did not. Because people use this tactic to highlight their competencies (Jones & Pittman, 1982), it makes sense for it to have a positive impact on the person’s evaluation. In psychotherapy, the use of Self-promotion signals the patients’ awareness of their resources and thereby draws the therapists’ attention to these, so that they can be used for therapy. Being focused on one’s resources activates the approach mode (Gassmann & Grawe, 2006), which makes patients more open for therapeutic interventions. The use of this tactic should thus be positively associated with the early alliance.

The burden of suffering is generally viewed as a prerequisite for therapy motivation. Klauer, Maibaum and Schneider (2007) found that a high burden of suffering at the beginning of therapy was associated with fewer dropouts. The explicit expression of suffering, Supplication, is not necessarily perceived as positive for the therapeutic alliance. Caspar (2007) sees Supplication as a problematic behavior, possibly driven by motives that are in the way of straightforward change unless one deals with them appropriately. He recommends therapists not respond to it on the behavioral level to avoid reinforcement. Instead, he advises therapists to try to comprehend the motives behind it, trace them back to unproblematic higher motives, and behave complementarily to these. These goals are associated with a prescriptive concept referred to as motive oriented therapeutic relationship, which has been shown experimentally to improve therapies (e.g. Berthoud, Kramer, Roten, Despland, & Caspar, 2013). Establishing such a relationship in the intake interview seems to be challenging for therapists. The intake interview is a situation in which the therapist and patient do not know each other, and yet, the patient is expected to talk about his/her problems. Finding a balance between responding empathically to patients’ reports of difficulties, but at the same time avoiding reinforcement of Supplication might be difficult for the therapist to achieve. Furthermore, a large amount of Supplication might indicate avoidance in addressing problems (Sachse, Fasbender, & Sachse, 2011), which is a suboptimal basis for psychotherapy in general and also for the therapeutic alliance specifically. We, therefore, suggest:

Participants (n = 60; 30 women, 30 men; age M = 39.05, SD = 14.55 years) were randomly selected from a pool of patients (n = 3,863) who were treated at the psychotherapeutic outpatient clinic of the University of Bern, Switzerland. Patients are treated according to an integrative form of cognitive behavior therapy (CBT) which is based on a detailed case conceptualization formulated using Plan analysis (Caspar, 2007; Grawe, 1999). The clinic accepts patients who have a wide range of problems and disorders, with the exception of psychotic disorders and substance dependency. The most frequent disorders in this sample were affective, anxiety, and somatoform disorders.

Twelve therapists conducted the interviews (five men, seven women). Half of the therapists were trainees at various stages of their four-year psychotherapy training course. The other half were experienced therapists involved in the weekly supervision of the trainees. The combination of patients’ and therapists’ sex was chosen so that 15 dyads were male-male, 15 male-female, 15 female-male, and 15 female-female.

The ratings were performed by four advanced masters students (one man, three women), who had at least four years of psychology and psychotherapy education. They had undergone systematic training to learn to apply the rating manual.

We defined patients’ tactics based on previous studies (Friedlander & Schwartz, 1985; Howard, Blumstein, & Schwartz, 1986; Kipnis, Schmidt, & Wilkinson, 1980; Pratkanis, 2007). We refined the preliminary list of tactics after discussing with ten practicing psychotherapists and testing the tactics by using them to assess five videotaped interviews. The final list consists of 12 tactics. For the present study, four tactics were chosen based on the quality of their inter-judge reliability.

The Bern Post-Session Report (Flückiger, Regli, Zwahlen, Hostettler, & Caspar, 2010) is a process measure. It is comprised of a patient form with eight scales and a therapist form that includes 11 scales. At the psychotherapeutic outpatient clinic both patient and therapist complete the Post-Session Report immediately after each session. For this study we examined the “therapeutic alliance” scale, which is comprised of three items. The patient items are:

The items were measured on a 7-point Likert scale where negative three (–3) = “not at all“ and plus three (+3) = “exactly.” The internal consistency for the present study included the values from the first three sessions and was .82 for the patient form and .83 for the therapist form, indicating good internal consistency.

All interviews were conducted during the years 2001 to 2012. The intake interview had the purpose of gathering information on the patient’s problem(s), previous treatments, family circumstances, demographic information, current living conditions, personal resources, and expectations towards psychotherapy. Although the interview was structured, it left enough freedom so that interactional distinctiveness of patients could unfold. The duration of the interview was flexible, and in this sample ranged from 50 minutes to 100 minutes.

Judges completed systematic training lasting approximately 12 weeks; training was completed when inter-judge agreement no longer improved.

The interview was assessed on an utterance-to-utterance basis using a structured manual. (The German manual is available from corresponding author upon request). Videos were randomly allocated to judges, who were blind to the research questions. Each judge rated 18 videos. The judges were instructed to distinguish instrumental from reactive behaviors using cues of the patients’ behavior (Brunner, 1996; Caspar, 2007, p. 159). We defined instrumental behavior as being purposeful, even if the patient was not aware of it being so; all other behaviors were categorized as reactive. In the rating manual, each tactic was outlined with an illustrative example to give the judges more confidence in detecting the tactics (see Table 1). Of course, the list of possible behaviors was not comprehensive, and thus judges were told to use their intuition when judging behaviors that were not defined. During the rating, judges watched each video three times. In the first run, they noted the beginning and the end times of the patients’ utterances. In the second run, they judged each utterance, determining if the patient was trying to influence the therapist or not. In the third watch, only the utterances in which influencing behavior had been detected, were inspected. For these utterances, judges were told to note a maximum of three different tactics. For each tactic, an intensity rating was given on a 1 to 10 scale with one being very low and 10 high. Tactics that appeared either very frequently or with a low intensity could be coded as meta-tactics.

Calculations were completed with the program SPSS version 21.0. Nonparametric tests were used because the majority of the data was not normally distributed.

The calculation of inter-judge agreement and reliability have been described in detail elsewhere (Frühauf et al., 2015). Twenty percent of the videos (12 out of the 60 videos) were double-coded to calculate inter-judge agreement, as suggested by Wirtz and Caspar (2002). Both inter-judge agreement and inter-judge reliability were calculated. Cohen’s kappa is an appropriate measure if marginals (i.e. the number of cases for which a judgment must be made) are fixed a priori. It can, however, underestimate agreement ratings if marginals are free to vary. The corrected kappa (Brennan & Prediger, 1981) is more appropriate for free marginals and is reported here in addition to Cohen’s kappa. The McNemar-Test elucidates if raters have different thresh olds (i.e., one judge detects an object more often than another judge). Yule’s Y is an adjusted measure which is independent of the judges’ thresholds. Cohen’s kappa yielded a mean value of .55, which refers to acceptable inter-judge agreement. McNemar-Test yielded significant results in six of the 12 videos. This suggests that Yule’s Y might produce higher agreement rates. Yule’s Y yielded a mean value of .87, which refers to good to very good inter-judge agreement. Inter-judge agreement referring to the different tactics was calculated using the category specific kappa by Fleiss (Fleiss & Cohen, 1973). Values can be found in Table 1.

Correlation coefficients and p-values for the associations between the tactics and the therapeutic alliance can be found in Table 4. There was a significant positive and moderate association between the tactic Agenda setting and the therapeutic alliance, as judged by the therapist in session 1 and3(r =.421, p = .002 and r=.298, p=.026, respectively), but not for the judgements in session 2 (r=–.001, p=.497). There was a significant negative and moderate association between Agenda setting and the therapeutic alliance as judged by the patient for session 1 and session 2 (r=–.392, p=.007 and r=–.268, p=.043, respectively). For session 3 there was also a negative and small to moderate association, which was moderately significant (r=–.238, p=.056). Hypothesis 1 can thus be accepted for the association between Agenda setting and the therapeutic alliance as judged by the therapist. However, the association between Agenda setting and the therapeutic alliance from the patient’s perspective was in the opposite direction from the predicted.

No significant correlations were found between the tactic Provoking a response from the therapist and the therapetic alliance as judged by the therapist (r=–.022, p=.443; r=–.046, p=.381 and r=–.002, p=.494, respectively) and by the patient (r=.004, p=.491; r=–.163, p=.151 and r=–.059, p=.347, respectively). Therefore, hypothesis 2 is rejected.

The association between the tactic Self-promotion and the therapeutic alliance as judged by the therapist was significantly positive and moderate in all three sessions (r=.462, p=.001; r=.364, p=.006; and r=.296, p=.027, respectively), whereas the strength of the association decreased in the course of the sessions. The association between Self-promotion and the therapeutic alliance as judged by the patient was significantly positive and small to moderate only in session 2 (r=.274, p=.040), but there were no significant associations in session 1 (r=.107, p=.262) and session 3 (r=-.081, p=.296). Hypothesis 3 can thus be accepted for the therapists’ perspective and partly for the patient’s perspective.

The associations between the tactic Supplication and the therapeutic alliance as judged by the therapist were negative and small to moderate in all sessions, with significant correlations in sessions 2 and 3 (r=–.287, p=.026 and r=–.411, p=.003, respectively), but not in session 1 (r=–.186, p=.110). There was no association between Supplication and the therapeutic alliance as judged by the patient (r=.151, p=.183; r=.018, p=.455 and r=.144, p=.169, respectively). Hypothesis 4 can thus be accepted regarding the relation between Supplication and the therapeutic alliance as judged by the therapist, but not for the therapeutic alliance as judged by the patient.

As predicted, there was a positive association between this tactic and the therapeutic alliance, as judged by the therapist. Therapists probably appreciate when patients participate actively in therapy by introducing their own ideals and goals. There was, however, a negative correlation between the patients’ ratings of the alliance and the use of Agenda setting, meaning that patients who judged the therapeutic alliance more negatively demonstrated this tactic to a greater extent. A possible reason for a frequent use of this tactic is these patients’ need for autonomy or control. Negatively put, patients who do not trust their therapists or doubt their ability to help them may feel the need to take matters into their own hands. One way of doing so is setting their goals for the therapy. Another possibility is that patients who come to therapy with already defined goals are more demanding and critical. These high standards may be reflected in their more critical evaluation of the therapeutic alliance, compared to other patients.

This tactic was not associated with the therapeutic alliance as judged by both therapist and client. The simplest explanation is that this tactic is not relevant for the therapeutic alliance. This would, however, be contradictory to the findings of studies from medical setting. Some of these studies have found that patients who asked more questions had better outcomes (Greenfield et al., 1988, 1985; Kaplan et al., 1989; Thompson et al., 1990). It is possible that these studies measured something different from what we did. Specifically, we included organizational and technical questions, whereas in the medical studies patients were explicitly instructed to ask questions pertaining to their medical condition. Another possibility is that the intake interview with a physician is fundamentally different from a psychotherapy intake interview. The traditional role of a patient in medical settings is a passive one-patients are not expected to get very involved. Asking more questions can thus be expected to lead to a better exchange between physician and patient. In psychotherapy, on the other hand, an active participation of the patient is understood and necessary. The tactic Provoking a response from the therapist might hence not be an appropriate operationalization for an active exchange between the therapist and the patient.

As predicted, this tactic was positively associated with the therapeutic alliance from the therapists’ perspective and partly also from the patients’ perspective. These results suggest that patients who use the tactic Self-promotion in the intake interview might build up a better therapeutic alliance. Patients who are capable of Self-promotion likely have more psychological resources and more confidence that they can benefit from psychotherapy. Being aware of their resources activates the approach mode, in which one is oriented towards positive goals. Being in the approach mode is beneficial for the development of the therapeutic alliance (Grawe, 2007). Additionally, it is conceivable that Self-promotion helps the therapist identify the patient’s resources better, and the therapist can more easily reinforce them. This would result in a further activation of the approach mode, and set up a positive feedback effect. It is likely that these patients have a better treatment outcome than patients who are less aware of their resources. The non-consistent positive association between Self-promotion and the therapeutic alliance as judged by the patient could indicate an unstable activation of the approach mode: It is likely that patients who use Self-promotion often are aware of their resources, but not consistently so.

As predicted, this tactic was negatively related to the therapeutic alliance, but only from the therapists’ perspective. Several reasons might explain the therapists’ negative evaluation of the alliance with these patients. First, it might be difficult for therapists at this early stage of therapy to identify the unproblematic higher motives of these patients that underlie the Supplication on the behavioral level, and respond complementary to them as the motive oriented therapeutic alliance approach suggests. Furthermore, excessive use of Supplication might indicate an avoidance of approaching the problem (Sachse et al., 2011). It is evident that therapists respond less positively to patients who do not seem to be motivated to change. Supplication might also point to relationship issues. For example, the excessive report of difficulties might play an instrumental role in these patients’ lives to fulfill goals like getting attention or not having to take responsibilities or fulfill a duty. It is probable that such behavioral patterns evoke negative reactions like anger or helplessness in the therapist, like they do in the other relationships of the patient.

Taken together, patients’ self-presentational behavior can be regarded as a promising predictor for the early therapeutic alliance: We found moderate associations between the tactics Agenda setting, Self-promotion, and Supplication and the therapeutic alliance.