Assessment of professional self‐efficacy in psychological interventions and psychotherapy sessions: Development of the Therapist Self‐Efficacy Scale (T‐SES) and its application for eTherapy

Abstract

Objective

This study aimed to develop the Therapist Self‐Efficacy Scale (T‐SES), and test its validity in a sample of Italian mental health therapists, to assess their professional self‐efficacy concerning their practice of eTherapy in a synchronous video‐based setting.

Methods

A sample of 322 Italian mental health professionals (37.6% psychologists, 62.4% psychotherapists; M _age  = 38.48, SD = 8.509) completed an online survey.

Results

The T‐SES showed a clear, one‐factor structure with good psychometric properties. Significant associations were found with insight orientation, general self‐efficacy, self‐esteem, and personality traits of openness, conscientiousness, and agreeableness. The results showed no differences between psychologists and psychotherapists, or differences based on years of experience.

Conclusion

The T‐SES is an agile and versatile self‐report measure for mental health professionals to assess their self‐efficacy concerning their therapeutic activity, which can provide information for tailoring training for eTherapy.

1. INTRODUCTION

In recent decades, the Internet has become increasingly accessible and used, leading to a hyper‐connected world, reshaping many aspects of daily life, and promoting the accessibility and availability of many services for all ages (Mihajlov & Vejmelka, 2017). Several online methods of digital mental health interventions have been implemented, promoting the rapid expansion of eTherapy. It is known by a range of different terms (e.g., teletherapy, cyber‐counseling, online therapy, web counseling, and others; Chester & Glass, 2006; Richards & Viganó, 2013), and is defined as “any type of professional therapeutic interaction that makes use of the Internet to connect qualified mental health professionals and their clients” (Rochlen et al., 2004; p. 270). In eTherapy, the professional‐patient interaction could be synchronous (i.e., in real‐time, such as using telephone or videoconferencing) or asynchronous (i.e., by exchanging messages or emails) (Gainsbury & Blaszczynski, 2011); in any case, it takes place through online technologies, with each person being in a different place, without physically meeting face‐to‐face (Zainudin & Yusop, 2018). Therefore, the online mode has the advantage of reducing significant barriers to accessing specialized mental health services, such as physical distance or difficulty in moving, making therapy more flexible and accessible (Leykin et al., 2012; Stoll et al., 2020). A useful option is the application of online treatment in addition to standard onsite care (e.g., Schuster et al., 2020; Zwerenz et al., 2019). In some cases, eTherapy may serve as an independent alternative to onsite treatment (e.g., Karyotaki et al., 2017). Given the ease of access to the Internet, the demand for eTherapy is growing (Paterson et al., 2017), and this trend has sharply increased with the spread of COVID‐19, declared by the World Health Organization (WHO) as a global pandemic on March 11, 2020 (World Health Organization, 2021). Indeed, prevention measures for COVID‐19 have led to the online mode as the only way to provide or access mental health services in some phases of the pandemic (Luca & Calabro, 2020). Previously, several studies (see Poletti et al., 2020 for a review) have focused on the outcomes of eTherapy; although the rate of improvement may be slightly slower (Egede et al., 2015; Zerwas et al., 2017), these studies show an overall effectiveness comparable to on‐site treatment for many mental health problems (e.g., Catarino et al., 2018; Egede et al., 2015; Zerwas et al., 2017), as well as an equivalent quality of life and treatment satisfaction for clients (Egede et al., 2016).

Despite this, surveys on attitudes toward eTherapy showed that many mental health professionals still have concerns about this modality because of multiple drawbacks that could emerge due to the integration of technology into their daily practice (Connolly et al., 2020; Gordon et al., 2015; Mendes‐Santos et al., 2020; Perle et al., 2013; Schulze et al., 2018). The issues raised were related to both ethical and technical problems, as well as concerns related to the therapeutic process (see Connolly et al., 2020 for a review). Overall, technology functionality is fundamental to providing satisfactory online interventions, and many professionals are concerned about the impact of insufficient digital literacy (Feijt et al., 2020; Topooco et al., 2017). However, encouraging data highlight that technical problems (e.g., low bandwidth or poor camera resolution) do not appear to hinder clinical outcomes or patient satisfaction (Richardson et al., 2015). Indeed, several precautions can be taken to limit these problems as much as possible, such as the use of adequate platforms, the use (or not) of headphones to improve audio quality, or attention to the camera placement at both ends (e.g., Waller et al., 2020). Online therapeutic practice requires specialist skills and knowledge to address unique ethical, privacy, and legal challenges (see Stoll et al., 2020 for a review). In this regard, several international professional organizations have developed guidelines and indications on the requirements for the practice of safe eTherapy (e.g., Joint Task Force for the Development of Telepsychology Guidelines for Psychologists, 2013; Shore et al., 2018). Regarding concerns about the therapeutic process, the primary issues related to digital mental health interventions were associated with the difficulty in having empathic communication (Roesler, 2017), access to visual cues and nonverbal behaviors (Alleman, 2002), and building a strong therapeutic relationship online (Roesler, 2017).

However, an important contribution to overcoming these issues is the use of synchronous web modality via videoconference (Glasheen et al., 2018), which offers an online environment that makes eTherapy not a form of therapy in itself, but a different modality of treatment delivery. Reese et al. (2016) compared face‐to‐face therapies with some forms of eTherapy (including the synchronous video‐based setting), highlighting similar levels of empathetic accuracy between the different modalities. This is also corroborated by previous research, which showed that therapists who use videoconferencing are generally able to develop a good therapeutic alliance (e.g., Békés & Aafjes–van Doorn, Prout, et al., 2020; Norwood et al., 2018; Simpson & Reid, 2014), comparable to those of onsite therapies (Ruwaard et al., 2009). Furthermore, evidence supports the efficacy level of eTherapy in a synchronous video‐based setting on par with the face‐to‐face modality in terms of patient satisfaction (e.g., Backhaus et al., 2012), symptom reduction (e.g., Backhaus et al., 2012), and overall treatment outcome (e.g., Kingsley & Henning, 2015), with positive results in different patient populations (see Barnett et al., 2021 for a review). Therefore, the concerns raised by mental health professionals regarding eTherapy are often not based on empirical findings. The reluctance of clinicians toward eTherapy may be due to a lack of the right level of training and experience, as well as low levels of self‐confidence and professional self‐efficacy (Pierce et al., 2020; Poletti et al., 2020; Roesler, 2017). On one hand, the scientific literature has consistently identified a trend toward an increase in positive feelings toward online therapy conducted via videoconferencing after repeated use (Adler et al., 2014; Lindsay et al., 2017), suggesting that with adequate training and knowledge, professionals tend to adapt effectively to this modality of treatment delivery (Simpson & Reid, 2014). On the other hand, when therapists are not specially trained or do not have enough experience in the use of technology, they feel uncomfortable using it (Glueckauf et al., 2018; Lustgarten & Elhai, 2018; Russell, 2018; Vincent et al., 2017), and those who feel less competent and confident in their professional skills concerning their online performance report a more negative attitude towards eTherapy (Békés & Aafjes‐van Doorn, 2020). These reports suggest the importance of considering the construct of professional self‐efficacy in therapists, with a particular focus on their experience related to digital mental health interventions.

1.1. Professional self‐efficacy

In general, professional self‐efficacy can be defined as a form of self‐confidence and, more specifically, as an individual's confidence in their ability to obtain quality outcomes in the professional tasks of their specific occupation (Fraser et al., 2018; Yoo & Cho, 2020). Higher levels of professional self‐efficacy have been associated with greater control over work activity (Jones & Fletcher, 2003), higher engagement in challenging job demands (Ventura et al., 2015), more optimistic thinking, higher quality in decision‐making processes and job satisfaction (Lu et al., 2020). Furthermore, studies have explored this construct in job contexts that include significant relational challenges (e.g., teachers, nurses, etc.), highlighting its protective role against burnout, a relevant problem in such work situations (e.g., Kong et al., 2021; Ventura et al., 2015). Similar data have also been found in mental health professionals (Gunduz, 2012), which is a category of workers at high risk of psychological exhaustion (Gam et al., 2016). In this context, clinicians' professional self‐efficacy refers to beliefs and attitudes about their ability to effectively counsel patients (Larson et al., 1992) and contributes to a sense of confidence in managing the fundamental aspects of therapeutic activity (Lent et al., 2003). It is associated with higher job satisfaction (Lent et al., 2009), lower anxiety levels, better session management (Daniels & Larson, 2001; Mehr et al., 2015), more realistic definitions of clinical goals, and better performance (Reese et al., 2009).

In light of this, some studies focusing on the antecedents of professional self‐efficacy in the field of mental health have identified the central contribution of clinical supervision (Bernard, 2006), specifically, of the working alliance during supervision (e.g., Marmarosh et al., 2013; Mehr et al., 2015). Furthermore, positive correlations have also been identified with the number of courses, internship hours, clinical instruction, and greater adherence to best practices and evidence‐based therapies (Kozina et al., 2010; Reese et al., 2009; Tang et al., 2004). Therefore, given its effect in favor of clinical performance and the evident key role of supervision and training as precursors for its incremental development in the field of mental health (Lent et al., 2003; Reese et al., 2009), clinicians' professional self‐efficacy should be a necessary aim of clinical education, considering it as a set of expectations related to a varied set of skills that structure therapy performance (Bernard, 2006; Kozina et al., 2010). In this regard, the development of a measure that enables the evaluation of therapists' self‐efficacy in a transtheoretical way seems to be of great use for self‐monitoring and evaluating any training program.

1.2. Personality and individual differences

In addition to adequate training, supervision, clinical education and experience, scientific literature concerning the antecedents of self‐confidence in professional activity also identifies significant effects of individual factors (Holland et al., 2012), such as the ability to manage one's own feelings of anxiety, coping strategies, reflexivity, and self‐awareness (Brown et al., 2003; Crooks et al., 2005). Personality is a promising element in the therapeutic context. Indeed, previous evidence considers it as one of the elements that may influence how the clinician approaches their professional activity and their style in the course of therapy (Casari et al., 2019; Peters‐Scheffer et al., 2013), and which may therefore be partly responsible for the so‐called “therapist effects”, that is, a portion of variability in the treatment outcomes in the field of mental health explained by therapist' components (Castonguay & Hill, 2017; Norcross & Lambert, 2019; Wampold & Imel, 2015). Furthermore, Thériault and Gazzola (2006, 2010) showed that some personality traits characterized by greater self‐criticism and perfectionist tendencies could be associated with higher feelings of incompetence. In other words, these dispositions were associated with negative and self‐devaluating subjective evaluations of their professional performance, which may be harmful to the therapist's well‐being and can adversely affect the therapeutic process (Thériault & Gazzola, 2006, 2010).

Given this framework and the fact that personality dimensions show different correlations with the key elements of acceptance of change (Di Fabio & Gori, 2016), exploring the associations between personality traits and professional self‐efficacy in reference to the practice of eTherapy in a synchronous video‐based setting (i.e., a different setting with respect to the traditional one) could provide important insights for training that is as effective and personalized as possible based on the characteristics of the trainees.

1.3. The present study

Although some scales for the assessment of mental health professionals' self‐efficacy exist and have been validated in different contexts, such as the School Counselor Self‐Efficacy Scale (Bodenhorn & Skaggs, 2005) or the Multicultural Counseling Self‐Efficacy Scale‐Racial Diversity Form (Sheu & Lent, 2007), thus far, no scale has been conceived and applied to web‐based treatments. Furthermore, the therapeutic process involves a series of transtheoretical key components for which the clinician can experience different levels of self‐confidence, which would be useful for a complete assessment of professional self‐efficacy in the field of mental health.

Therefore, the central aim of the present research was to develop a new self‐report measure, the Therapist Self‐Efficacy Scale (T‐SES), to enable the assessment of the self‐perception of mental health therapists for the core and trans‐theoretical factors of therapeutic activity. The secondary objectives were as follows:

• Analysis of the psychometric properties of the T‐SES and its validation in an Italian sample of mental health therapists to assess their professional self‐efficacy concerning their eTherapy practice in a synchronous video‐based setting.

• Exploration of the association between the therapist's professional self‐efficacy scores and scores related to self‐esteem, general self‐efficacy, insight, and personality traits.

• Investigation of the differences in the levels of therapist's professional self‐efficacy in relation to the professional qualification (psychologists or psychotherapists) or years of clinical experience.

2. METHOD

2.1. Participants and procedure

This study included a sample of 322 Italian mental health professionals, predominantly women (89%), with an age range of 25–71 years (M = 38.48, SD = 8.509). They declared themselves to be psychologists (37.6%) or psychotherapists (62.4%) of different theoretical orientations (Table 1), employed in independent practice. Most respondents stated that they had been practicing professionally for more than 10 years (33.5%); among the others, 26.4% had been practicing for 5–10 years, 21.4% for 2–5 years, 11.2% for 1–2 years, and only 7.5% for less than a year (Table 1). Concerning the problems that clinicians were treating during eTherapy, 50.0% reported a predominance of anxiety disorders. Other issues included relational problems/conflicts (24.2%), trauma‐ and stress‐related disorders (7.5%), mood disorders (6.8%), economic/work/school issues (5.3%), personality disorders (3.7%), sleep‐wake disorders (0.9%), feeding and eating disorders (0.6%), obsessive‐compulsive and related disorders (0.6%), somatic symptoms and related disorders (0.3%). Participants were volunteers recruited on the Internet from various large and closed social groups, where it is necessary to report information that certifies one's qualification and registration on the professional register to be registered. They completed the survey on Google Forms after providing informed consent electronically. Participants were specifically asked to refer to their eTherapy activity in a synchronous video‐based setting when completing the T‐SES. The research protocol was approved by the Ethical Committee of ***** (*masked for peer review*) (ethical approval number 002/2021).

Table 1.

Demographic and professional characteristics of the sample (N = 322)

Characteristics	M ± SD	n	%
Age	38.48 ± 8.51
Sex
Females		288	89.4
Males		34	10.6
Professional qualification
Pychologists		121	37.6
Psychotherapists		201	62.4
Years of clinical practice
Less than a year		24	7.5
1–2 years		36	11.2
2–5 years		69	21.4
5–10 years		85	26.4
More than 10 years		108	33.5
Theoretical orientation (for psychotherapists only)
Psychoanalytic		16	8.0
Psychodynamic		33	16.4
Cognitive		22	10.9
Cognitive behavioral		34	16.9
Humanistic		18	9.0
Integrated		42	20.9
Systemic		36	17.9

2.2. Development of the T‐SES

The T‐SES was constructed to develop an agile measure with good psychometric properties. Furthermore, the T‐SES was conceptualized with the intent of:

• (1)being useful for mental health professionals from various theoretical orientations;
• (2)understanding the therapists' level of confidence in their resources experienced by them during their professional activity;
• (3)allowing monitoring of the clinician's professional self‐efficacy beliefs (at the beginning, in the course of, and at the end of training, as a measure able to capture changes in these beliefs).

To develop the T‐SES, an integrative perspective, as conceptualized in the Transtheoretical Approach (Prochaska & DiClemente, 2005; Prochaska & Norcross, 2018), was adopted to select the dimensions that should be covered by the scale. This framework is guided by the following key objectives: to preserve the valuable insights of the major approaches; to provide practical implications and applicable answers to clinical activity‐related questions; to find a guiding thread that allows an orderly integration and is not just a chaotic assortment of techniques; to offer a perspective that is based on scientific evidence and testable by further research activity; and to generate a systematic approach, including a complete structure of the core dimensions, which is also “adequately flexible to promote collaboration, creativity, and choice” (Prochaska & DiClemente, 2005; p. 148).

Following these principles, an integration, expansion, and deepening of the specific components of the effectiveness of therapy that were identified in the scientific literature were elaborated, and items were therefore built to assess the clinician's self‐efficacy belief in different relevant therapeutic factors. These aspects are:

• Communicative effectiveness: This dimension includes the clinician's ability to formulate effectively and intervene in a timely manner, as well as to interpret the client/patient nonverbal signals. Indeed, existing evidence agrees that a therapeutic modality is appropriate in which interventions are adapted according to the changing nature of the situation (Hatcher, 2015), and are formulated at the right time, according to the needs and requirements of the client/patient at that moment (Stiles, 2013). This requires attention not only in terms of verbal feedback, but also nonverbal feedback (Bonanno & Burton, 2013).

• Clinical competence: This refers to the clinician's skills with respect to the therapeutic contract, using clinical sensitivity and reasoning, understanding the client/patient's narratives and vicious circles, and favoring his or her expression of self. Specifically, the client/patient's self‐expression may be a process of growth as well as a source of information (Arnold, 2014; Rogers & Wallen, 1946). In addition to favoring this, an effective therapist should understand the significant elements in the client/patient's narratives in depth, identify and organize the crucial aspects and perpetuated dynamics, as well as reason and elaborate a useful mental representation of the problems, with sensitivity to the client/patient's context and therapeutic situation (Hill et al., 2017). Simultaneously, the clinician should protect the setting, by respecting the established boundaries and ground rules; the management of these aspects by the professional is the basis of an effective treatment experience (Langs, 2019).

• Intrapsychic competence: This dimension describes the clinician's ability to understand the dynamics of transference and countertransference, as well as favor the client/patient's processes of insight and mentalizing. In reality, the concepts of transference and countertransference were developed within psychodynamic traditions and, demonstrate relevance and diffusion from different perspectives (e.g., Cartwright, 2011), giving rise to transversal evidence of how the quality of the management of these processes has a significant effect on treatment outcomes (see Parth et al., 2017 for a review). Furthermore, insight (Castonguay & Hill, 2007; Gori et al., 2015) and mentalizing (Allen et al., 2008; Bateman & Fonagy, 2013; Bateman et al., 2009; Gori et al., 2021) were found to be complex constructs linked to a better awareness of oneself and others, functional in promoting changes and favoring positive treatment dynamics.

• Relational competence: This refers to the clinician's skills in maintaining an interaction style based on acceptance, favoring the therapeutic alliance and its maintenance, managing stalemate phases and repairing any breakdowns in the relationship with the client/patient. In this regard, previous research highlighted the centrality of the therapeutic alliance as: an important nonspecific factor, transversal to various approaches, having the greatest role in influencing client outcomes (Crits‐Christoph & Gibbons, 2002; Horvath et al., 2011; Messer & Wampold, 2002; Nienhuis et al., 2018; Parth et al., 2017). Similarly, general agreement was found in considering relational expertize as a necessary element to favor positive treatment results (Heinonen & Nissen‐Lie, 2020; Hill et al., 2017).

• Affect regulation: This dimension describes the clinician's competence in effectively tolerating and managing emotions during professional activities. The client/patient's emotional dysregulation may elicit strong emotional reactions from therapists; therefore, their ability to regulate affect may be a core element in developing and sustaining a genuinely therapeutic relationship based on coregulation (e.g., Murphy & Joseph, 2013). This transversal dimension is supported and corroborated by evidence from the field of neuroscience (e.g., Dana, 2018; Hill, 2015; Porges, 2011; Schore, 2015; Van der Kolk, 2014).

• Diagnostic skills: These refers to the clinician's sensitivity to psychopathological signals, as well as their effectiveness in activating a diagnostic process. Indeed, adequate diagnostic skills allow for the correct classification of symptoms according to the main international reference systems (e.g., DMS‐5, ICD‐10, PDM‐2; American Psychiatric Association, 2013; Lingiardi & McWilliams, 2017; World Health Organization, 2019), but also the evaluation of aspects central to the functioning of the individual (e.g., affects, relational patterns, beliefs, defenses and others; Gabbard, 2014; McWilliams, 1999; McWilliams, 2011; Weiss, 1993; Westen & Gabbard, 2002a; Westen & Gabbard, 2002b), his/her personality organization (Caligor et al., 2007; Kernberg, 1993), and his/her structure (Millon & Davis, 1997). These skills fit within the broad framework of psychological assessment, a key factor for facilitating, orienting, and evaluating treatment (Bornstein, 2017), that may be potentially beneficial in itself in specific applications (Finn, 2007).

The development of the scale came from the collaboration of a team of experts: psychologists, psychometrists, and psychotherapists. The questionnaire items were generated using a twofold process. During the first stage, 24 items were developed by the researchers based on an analysis of the clinical literature and their professional experience. After conceptually identifying the therapeutic factors that should have been covered by the scale, a list of associated statements was drawn up. The statements were selected and perfected by avoiding terms of frequency (e.g., never, often, sometimes, etc.), limiting ambiguity, keeping the items as simple and short as possible, and favoring clear and direct language. This phase was also implemented by organizing focus groups with a pool of eight researchers and clinical experts to make this step more effective and obtain a satisfactory level of agreement regarding the content of the items. Specifically, each researcher and clinical expert were asked individually if there were any redundant or similar items, and seven out of eight participants indicated the same three statements, which were excluded. To evaluate the quality and representativeness of the 21 retained items of the construct, a second step was implemented, wherein items were reviewed by a second panel of four experts. The protocol that was implemented in this stage involved the evaluation of the items in relation to: (1) appropriateness and clarity of language with respect to mental health professionals across several theoretical orientations; and (2) correspondence between the content of each item and the aspect it was proposed to evaluate (content validity). None of the items were removed based on this review, as all the experts unanimously voted for their appropriateness. A response format on a 5‐point Likert scale (1 = “not at all,” 2 = “a little,” 3 = “somewhat,” 4 “much,” 5 = “a great deal”) was adopted. The global T‐SES score was calculated by adding the item scores.

2.3. Measures

2.3.1. Therapist Self‐Efficacy Scale

The T‐SES is a self‐report questionnaire designed to measure mental health therapist's professional self‐efficacy. It consists of 21 items on a 5‐point Likert scale, ranging from 1 (not at all) to 5 (a great deal) and assesses the clinician's self‐confidence in specific components of therapy effectiveness: Communicative effectiveness, Clinical competence, Intrapsychic competence, Relational competence, Affect regulation, Diagnostic skills. The global T‐SES score was calculated by summing the item scores. In the present sample, the scale showed good internal consistency (Cronbach's alpha and McDonald's omega coefficients are reported in the Results section).

2.3.2. Rosenberg Self‐Esteem Scale (RSES)

The RSES (Rosenberg, 1965) is a self‐report questionnaire designed to measure global self‐esteem. The Italian version of Prezza et al. (1997) was used in this study. It consists of 10 items scored on a 4‐point Likert scale, ranging from 0 (strongly agree) to 3 (strongly disagree). The present sample good internal consistency, with a Cronbach's α of 0.83 and a McDonald's ω of 0.85.

2.3.3. General Self‐Efficacy Scale (GSE)

The GSE (Schwarzer & Jerusalem, 1995) is a self‐report questionnaire designed to measure self‐efficacy. The Italian version of Sibilia et al. (1995) was used in this study. It consists of 10 items scored on a 4‐point Likert scale, ranging from 1 (not at all true for me) to 4 (very true for me). The present sample showed an excellent internal consistency, with a Cronbach's α and McDonald's ω of 0.90.

2.3.4. Insight Orientation Scale (IOS)

The IOS (Gori et al., 2015) is a self‐report measure designed to assess some characteristics of insight, including behaviors, feelings, and opinions about this construct. It consists of 7 items scored on a 5‐point Likert scale, from 1 (not at all) to 5 (a great deal). In the present sample, it showed good internal consistency, with a Cronbach's α of 0.88 and a McDonald's ω of 0.90.

2.3.5. Ten‐Item Personality Inventory (TIPI)

The TIPI (Gosling et al., 2003) is a self‐report scale designed to assess personality traits, in line with the Big Five model (Costa & McCrae, 1992a). The Italian version of Di Fabio, Gori, and Giannini (2016) was used in this study and showed an acceptable internal consistency. It consists of 10 items scored on a 7‐point Likert scale, ranging from 1 (disagree strongly”) to 7 (agree strongly), grouped into five dimensions: extraversion (α = 0.73; ω = 0.73), agreeableness (α = 0.50; ω = 0.60), conscientiousness (α = 0.69; ω = 0.71), neuroticism (α = 0.66; ω = 0.67), and openness (α = 0.60; ω = 0.64).

2.4. Data analysis

All statistical analyses were performed using SPSS 21.0, AMOS 24.0, and JAMOVI 2.0. Due to the diffusion of the survey on the Internet, it was not possible to detect the view rates of the anonymous link. The completion of the survey was voluntary, and attendance was not formally recorded. Therefore, response rate were not calculated. There were no missing data in the data response set, as the Google Form platform was set up in such a way that all fields had to be completed to allow the survey to be submitted by the participants.

Descriptive statistics were calculated. Item analysis was conducted, by exploring the mean, standard deviation, skewness, and kurtosis for each item of the T‐SES. The dimensionality of the T‐SES was also investigated. The Kaiser‐Meyer‐Olkin (KMO) measure of the sampling adequacy and Bartlett's test of sphericity were used to assess whether the data are suited for factor analysis: appropriateness is indicated if the KMO value is more than 0.7 and Bartlett's test is statistically significant (p < 0.001) (Mulaik, 2009). Exploratory factor analysis (EFA) with a principal axis factoring extraction method (Promax rotation with Kaiser normalization) was performed, using three approaches to evaluate the number of components: Kaiser Greater‐Than‐One Rule Criterion, for which eigenvalues greater than one indicate the factors that could be retained for interpretation (Kaiser, 1960); the Scree Test, which include a visual examination of the scree plot in which the slope of the curve changes most abruptly in proximity of the point dividing the relevant number of factors from the trivial ones (Cattell, 1966); and the variance extracted, wherein the number of factors is sufficient when the variance percentage explained is at least 50% of the total variance (Streiner, 1994). Then, a confirmatory factor analysis (CFA) was implemented, considering the following fit indices: the model Chi‐square (χ ²) model, indicating a good model fit when the probability value was statistically nonsignificant (p > 0.05) (Hooper et al., 2008); the Non‐Normed Fit Index (NNFI), indicating a reasonable value above of 0.90 (Kline, 2015); the comparative fit index (CFI), indicating a recommended value of >0.95, although those between 0.90 and 0.95 were recognized as a reasonable fit (Kline, 2015); the root mean square error of approximation (RMSEA), indicating a recommended value of less than 0.05, although values up to 0.08 represent reasonable errors of approximation (Marsh et al., 2004); and the standardized root mean square residual (SRMR), indicating a reasonable fit with values less than 0.08 (Hooper et al., 2008). The reliability of the scale was calculated using Cronbach's alpha, McDonald's omega coefficients, and item‐total correlation indices. Pearson's r correlation was used to investigate the association between the variables, to assess some aspects of concurrent validity and the associations with the Big Five personality traits. Discriminant validity was explored in greater depth by implementing the heterotrait‐monotrait ratio of correlations (HTMT; Henseler et al., 2015) using an AMOS plugin (Gaskin & James, 2019). To interpret the HTMT, a recommended threshold of <0.85 was considered, although values up to 0.90 can be considered acceptable (Henseler et al., 2015). The T‐SES scores of psychologists and psychotherapists were compared using an independent samples t test. Finally, to explore differences in T‐SES scores based on years of experience, an analysis of variance (ANOVA) was implemented.

3. RESULTS

Descriptive statistics of the sample were reported in Table 1.

As indicated in Table 2, the skewness and kurtosis values for each T‐SES item were between −1 and +1, showing an approximately normal distribution. The mean T‐SES items scores ranged from 3.11 to 4.21.

Table 2.

Descriptive statistics and item‐total correlations of each T‐SES item (N = 322)

Items	Minimum	Maximum	Mean	Standard deviation	Skewness	Kurtosis	Item‐total correlation
1	1	5	4.21	0.872	−0.856	0.121	0.797
2	1	5	3.77	0.903	−0.393	−0.336	0.817
3	1	5	3.81	0.906	−0.401	−0.475	0.818
4	1	5	3.97	0.952	−0.543	−0.594	0.832
5	1	5	3.67	0.968	−0.241	−0.813	0.837
6	1	5	3.78	0.905	−0.319	−0.559	0.823
7	1	5	3.54	0.970	−0.219	−0.683	0.802
8	1	5	3.11	1.054	0.090	−0.794	0.653
9	1	5	3.41	0.998	−0.232	−0.551	0.797
10	1	5	3.67	0.936	−0.277	−0.575	0.835
11	1	5	3.91	0.956	−0.513	−0.414	0.803
12	1	5	3.67	0.916	−0.281	−0.487	0.814
13	1	5	3.53	0.964	−0.309	−0.386	0.823
14	1	5	3.84	0.943	−0.457	−0.566	0.832
15	1	5	3.88	0.929	−0.501	−0.366	0.860
16	1	5	3.58	1.036	−0.327	−0.610	0.857
17	1	5	3.62	1.023	−0.506	−0.301	0.801
18	1	5	3.93	0.931	−0.550	−0.339	0.867
19	1	5	3.64	0.989	−0.471	−0.191	0.820
20	1	5	3.76	0.912	−0.394	−0.376	0.850
21	1	5	3.80	0.916	−0.404	−0.290	0.833

The significance of Bartlett's test of sphericity (p < 0.001) and the KMO index of 0.97, suggested the adequacy of the data for factor analysis. The EFA with the principal axis factoring extraction method (Promax rotation) showed a factor structure with one principal dimension with 70% of the total variance explained (eigenvalue = 14.712), as confirmed in the scree plot (Figure 1).

Figure 1.

Scree plot

The factor structure matrix shows one independent factor of the scale (Table 3).

Table 3.

Factor loadings of Therapist Self‐Efficacy Scale (T‐SES)

Itema	Content	Factorb
1. Be welcoming	Relational competence	0.808
2. Express verbal interventions effectively	Communicative effectiveness	0.828
3. Stimulate insight processes	Intrapsychic competence	0.826
4. Promote therapeutic alliance	Relational competence	0.842
5. Facilitate self‐expression	Clinical competence	0.846
6. Properly manage emotions	Affect regulation	0.833
7. Choose the correct moment to speak	Communicative effectiveness	0.809
8. Pick up nonverbal signals	Communicative effectiveness	0.658
9. Overcome setbacks in the relationship	Relational competence	0.802
10. Understand the deeper meaning of narratives	Clinical competence	0.844
11. Respect the therapeutic contract	Clinical competence	0.814
12. Understand psychopathological signs	Diagnostic skills	0.823
13. Repair fractures in the relationship	Relational competence	0.831
14. Tolerate negative emotions	Affect regulation	0.843
15. Use clinical reasoning	Clinical competence	0.872
16. Understand transference	Intrapsychic competence	0.866
17. Activate the diagnostic process	Diagnostic skills	0.809
18. Use clinical sensitivity	Clinical competence	0.879
19. Understand countertransference	Intrapsychic competence	0.829
20. Encourage mentalizing	Intrapsychic competence	0.862
21. Understand vicious circles	Clinical competence	0.845

^aInstructions: “During psychological meetings or psychotherapy sessions, I am able to:”.

^bExtraction method: principal axis factor.

Concerning the CFA, the goodness‐of‐fit indices indicated a good fit between the one factor model and the data. Although the Chi‐square was statistically significant (p < 0.001), the other indices showed acceptable values (NNFI = 0.949, CFI = 0.956, RMSEA = 0.075, and SRMR = 0.029).

Furthermore, the T‐SES showed very good reliability: Cronbach's alpha (α = 0.93) and McDonald's omega (ω = 0.94) coefficients were excellent. The item‐total correlations (Table 2) ranged from 0.653 (item 8) to 0.867 (item 18).

Pearson's r analysis indicated statistically significant and positive correlations between the T‐SES scores and RSES (r = 0.191, p < 0.01), GSE (r = 0.307, p < 0.01), and IOS (r = 0.474, p < 0.01) scores, indicating good convergent validity (see Table 4). Furthermore, T‐SES was also positively and statistically significantly associated with agreeableness (r = 0.129, p < 0.05), conscientiousness (r = 0.145, p < 0.01), and openness (r = 0.170, p < 0.01). Furthermore, the HTMT inference did not indicate problems of discriminant validity for the T‐SES, as its associations with the other scales considered were below the threshold value of 0.85 (Table 4).

Table 4.

Correlations, means and standard deviations of the variables (below the diagonal), and Heterotrait‐Monotrait (HTMT) correlation ratio for discriminant validity (over the diagonal)

	1	2	3	4	5	6	7	8	9
1. T‐SES	1	0.209	0.326	0.514	0.001	0.206	0.179	0.106	0.221
2. RSES	0.191 a	1	0.485	0.245	0.379	0.465	0.434	0.612	0.414
3. GSE	0.307 a	0.419 a	1	0.568	0.225	0.259	0.200	0.408	0.451
4. IOS	0.474 a	0.198 a	0.509 a	1	0.107	0.192	0.197	0.148	0.225
5. Extraversion	−0.004	0.272 a	0.170 a	0.086	1	0.245	0.012	0.166	0.423
6. Agreeableness	0.129 b	0.356 a	0.161 a	0.106	0.103	1	0.376	0.870	0.383
7. Conscientiousness	0.145 a	0.367 a	0.151 a	0.152 a	0.015	0.204 a	1	0.522	0.127
8. Neuroticism	−0.081	−0.474 a	−0.319 a	−0.105	−0.102	−0.561 a	−0.357 a	1	0.384
9. Openness	0.170 a	0.299 a	0.330 a	0.160 a	0.256 a	0.166 a	0.087	−0.248 a	1
M	78.1	24.38	30.30	26.80	17.75	21.47	23.61	12.64	21.49
SD	16.69	4.49	4.81	4.88	4.23	3.09	3.48	3.90	3.44

Note: Bold values indicate significant p‐values. Agreeableness = Ten Item Personality Inventory (Agreeableness subscale); Conscientiousness = Ten Item Personality Inventory (Conscientiousness subscale); Extraversion = Ten Item Personality Inventory (Extraversion subscale); Neuroticism = Ten Item Personality Inventory (Neuroticism subscale); Openness = Ten Item Personality Inventory (Openness subscale).

Abbreviations: GSE, General Self‐Efficacy Scale; IOS, Insight Orientation Scale; RSES, Rosenberg Self‐Esteem Scale; T‐SES, Therapist Self‐Efficacy Scale.

^aCorrelation is significant at the 0.01 level (two‐tailed).

^bCorrelation is significant at the 0.05 level (two‐tailed).

The independent‐samples t test did not find statistically significant differences in T‐SES scores between psychologists (M = 77.23, SD = 16.78) and psychotherapists (M = 78.62, SD = 16.65): t(320) = −0.724, p = 0.470. Parallelly, the ANOVA showed that there was not statistically significant effect of the years of experience on T‐SES score at the p < 0.05 level for five conditions (less than a year, 1–2 years, 2–5 years, 5–10 years, and more than 10 years): F(4, 317) = 0.695, p = 0.596.

4. DISCUSSION

Given the opportunity of overcoming many logistic and displacement barriers through the Internet (Leykin et al., 2012; Stoll et al., 2020), eTherapy for mental health is growing in popularity. Therefore, it is important that clinicians who decide to adopt it feel able to adapt to this new web‐based modality and maintain a good level of confidence in their professional skills. The aim of the present study was to develop the T‐SES to assess the professional self‐efficacy of mental health therapists in its core trans‐theoretical components. Since supervision and training can be precursors to a clinician's self‐concept (Reese et al., 2009), a specific tool can be extremely useful for evaluating and tailoring these activities.

The process of construction and development of the T‐SES was implemented by adopting an integrative perspective as conceptualized in the Transtheoretical Approach (Prochaska & DiClemente, 2005; Prochaska & Norcross, 2018), which aims to preserve the richness of the facets of the different approaches while offering useful implications for clinical practice in a conceptually ordered form, based on scientific research, and in a systematic view, as comprehensive and flexible as possible. This process resulted in a self‐report scale consisting of 21 items, covering the clinician's self‐perception in six dimensions that emerged as central in the scientific literature regarding therapeutic efficacy: communicative effectiveness, clinical competence, intrapsychic competence, relational competence, affect regulation, diagnostic skills (see Appendix A and Table 3 for the original version and English translation of the items, respectively).

To achieve the first secondary goal proposed in this study, psychometric properties of T‐SES were evaluated in a sample of mental health professionals who performed eTherapy in a synchronous‐video‐based setting. The T‐SES showed good psychometric properties, demonstrating its validity as an internally consistent self‐report measure. EFA revealed a clear factor structure with a single dimension explaining a substantial percentage of variance, which was also supported by CFA. Furthermore, although T‐SES covered the perception of being skilled in several (but not all) key elements for favors patients' changes and therapy effectiveness (communicative effectiveness, clinical competence, intrapsychic competence, relational competence, affect regulation, diagnostic skills), it showed excellent internal consistency. This indicates that even if the items may cover different and apparently heterogeneous aspects, each of them effectively constitutes and defines the construct that the scale aims to measure, that is, the professional self‐efficacy of the clinician and, more specifically, in this research, the one referred to their eTherapy activity. Confirming this, the T‐SES was significantly associated with the measures used to assess convergent validity: it was positively related to self‐esteem, general self‐efficacy, and, to a greater extent, with insight orientation. Therefore, professional self‐efficacy seems to fit a more general framework of awareness and reflexivity. In this regard, the American Psychological Association (2012) identified "professionalism" as one of the competency benchmarks in professional psychology, including reflective practice, self‐assessment, and self‐care: in other words, professional clinical practice should be conducted with “personal and professional self‐awareness and reflection; with awareness of competencies; with appropriate self‐care” (ibidem, p. 4). This is in line with previous evidence that identifies self‐awareness as an important key to being an effective psychotherapist (Hatcher et al., 2013; Knapp et al., 2017). Given the presented evidence, the therapist self‐efficacy should be conceptualized in this wide‐ranging framework, as a kind of therapist awareness, reflection and perception of one's own ability in favor of the intervening dynamics and producing therapeutic changes in different clinical settings. Although there were some interesting associations supporting convergent validity, the T‐SES also showed statistically distinguishable scores compared to those obtained from the other scales. These data indicated the good discriminating validity of the scale. Furthermore, the results also highlighted the relationship between the scale and some of the personality traits included in the Big Five model (Costa & McCrae, 1992b). Although no relevant associations were found between extraversion and neuroticism, positive and significant correlations were found between the T‐SES and openness, conscientiousness, and agreeableness. Such data could therefore highlight personality traits that seem to favor a better adaptability and self‐efficacy beliefs in using this relatively new online modality of therapy. This is consistent with previous research, which states that openness, conscientiousness, and agreeableness were found to be satisfactory predictors of career adaptability (Li et al., 2015), defined as “the readiness to cope with the predictable tasks of preparing for and participating in the work role and with the unpredictable adjustments prompted by changes in work and working conditions” (Savickas, 1997; p. 254). The openness trait describes individuals as imaginative, creative, versatile, and open to change (Yong, 2007). This improves the perception of requests as challenges, resulting in greater involvement in tasks and a sense of self‐efficacy (Sanchez‐Cardona et al., 2012). These characteristics, therefore, could foster a sense of self‐confidence in the therapist when experimenting with their professional activities, even with new means, such as online. Conscientiousness, like openness, is also associated with a strong motivation to learn (Major et al., 2006). Conscientious people are persistent, disciplined, reliable, and hardworking (Costa & McCrae, 1992). This is associated with higher commitment and effort in tasks, promoting higher self‐efficacy beliefs (Brown et al., 2011; Chen et al., 2001) and a higher tendency to accept and use technologies (Lakhal & Khechine, 2017). Therefore, the Internet can be perceived as an additional effective tool, allowing access to mental health services for a greater number of people, breaking down barriers, and facilitating closeness with others (Leykin et al., 2012; Stoll et al., 2020). This aspect can also be a source of a better predisposition for participants with higher levels of agreeableness, described as confident, accommodating, indulgent, available, and motivated to achieve interpersonal intimacy (Lakhal & Khechine, 2017). These characteristics can also facilitate access to new activities, the mastery of which can lead to greater self‐efficacy (Caprara et al., 2009) as well as a greater acceptance of technology (Keeton, 2008). However, the low Cronbach's alpha of the agreeableness scale should be considered in interpreting these results. This value could be partly due to the small number of items composing the scale, even considering an acceptable, albeit low, omega value, which previous evidence highlighted as more appropriate than alpha (Dunn et al., 2014). It could also suggest lower levels of internal consistency, considering the scores close to the cut‐off of 0.90 obtained in the HTMT analysis, in association with the neuroticism scale. Therefore, further studies are required to confirm and investigate this data. Such findings indicate that clinicians who are curious and open‐minded, persistent, and self‐disciplined, or more altruistic and cooperative, were able to perceive themselves as more effective in fostering functional dynamics during eTherapy in a synchronous video‐based setting. Overall, the analyzed and discussed correlations offer an answer to the second secondary objective of this research, which aimed to explore the associations between the therapist's professional self‐efficacy scores and those relating to self‐esteem, general self‐efficacy, insight orientation, and personality traits. Finally, the investigation of the influence of the type of professional qualification or the years of experience on the levels of therapists' professional self‐efficacy (the third secondary aim of this research) also showed no significant differences between different kinds of mental health professionals or based on time of clinical exercise. This suggests the utility and versatility of the T‐SES for clinicians in general, regardless of any achieved level of specialization, and highlights the importance of a specific evaluation of the therapist's professional self‐efficacy in the online context, regardless of years of experience in face‐to‐face practice. This is consistent with previous research (Sucala et al., 2013), which found no association between years of experience and the therapeutic alliance in eTherapy, highlighting the need for specific supervision and training to develop confidence in the online setting (Poletti et al., 2020).

4.1. Limitations & directions for future research

This study had some limitations that need to be identified and discussed. First, we used a convenience sample of mental health professionals; that is, participants were self‐selected to participate in the study, which could imply that only those who were interested and motivated to participate in the study completed the survey. Replication in larger samples and different settings will be needed in future research. Furthermore, only general information was collected about the predominance of problems reported by patients encountered by clinicians in eTherapy. Although previous research has shown stability in therapists' features, and that those who are effective or ineffective in treating one type of issue have also been shown to be more effective or ineffective with another type of disease (Nissen‐Lie et al., 2016), it is also true that professionals may have problem‐specific skills (Kraus et al., 2016). Therefore, it could be useful for future research to investigate these aspects and verify the influence of patient features on the clinician's perception of effectiveness. Moreover, predictive validity was not tested in this study. This aspect may be of importance in future research by exploring the predictivity of T‐SES scores of therapists' attitudes toward online therapy. Finally, the use of self‐report measures exposes a series of biases (e.g., the desirability bias). To overcome this limitation, future research should use a multi‐method approach by integrating different kinds of instruments (e.g., semi‐structured interviews). Despite these limitations, this study presents the development of a new self‐report scale, the T‐SES, which provides evidence for the goodness of its psychometric properties. In addition, further food for thought pertinent to the current context is also offered, highlighting interesting associations between professional self‐efficacy in the practice of eTherapy and personality traits, insight orientation, self‐esteem, and general self‐efficacy in psychologists and psychotherapists. These results can provide useful insights for future research, which, through longitudinal designs, will be able to confirm and enrich these findings by establishing the directionality and causality of these relationships. Finally, the absence of differences in the levels of professional self‐efficacy in eTherapy based on type of professional qualification or years of experience supports the hypothesis that specific training activities are required to carry out one's professional activity online. This finding could be confirmed in future research by exploring the presence of significant differences in T‐SES scores among clinicians in pre‐and posttraining comparisons.

5. CONCLUSIONS

In summary, the T‐SES is a new, valuable, agile, and versatile self‐report measure for different mental health professionals to assess their self‐efficacy concerning their activity, which could be useful when adopted for eTherapy. The authors conceptualized the “therapist self‐efficacy” in a wide‐ranging framework, as a kind of therapist's awareness, reflection, and perception of their own ability in favor the intervening dynamics and produce therapeutic changes in different clinical settings. Indeed, it is important for the therapists to feel confident with their own personal and professional resources and to be able to implement them at their best, for therapy to be effective (Hatcher et al., 2013; Knapp et al., 2017). Therefore, the T‐SES can be functional in increasing awareness, possibly indicating the need for further training as well as supervision for clinical practice (Borgueta et al., 2018; Ladany & Inman, 2012; Lustgarten & Elhai, 2018), better addressing the new challenges that technological development offers and being able to make the most of its advantages. For example, the positive associations between self‐esteem, general self‐efficacy, insight, and T‐SES scores concerning eTherapy shown in this study provide important stimuli to enrich training for mental health professionals, suggesting the importance of integrating specific education on online therapy with specific interventions on these dimensions of self‐concept. This could foster a greater perception of confidence and mastery in online clinical practice. Furthermore, the presence of positive and significant associations between the levels of professional self‐efficacy in eTherapy and some personality traits (agreeableness, conscientiousness, and openness) and negative relationships, albeit insignificant, with others (extraversion and neuroticism), can be important information to support clinicians in a tailored way based on their characteristics, supporting them where necessary with more recurrent supervision and more training. In light of this evidence, owing to its good psychometric properties and theoretical relevance, the T‐SES can be adopted in both research and practice.

1. PEER REVIEW

The peer review history for this article is available at https://publons.com/publon/10.1002/jclp.23391

ETHICS STATEMENT

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The research was approved by the Ethical Committee of the Integrated Psychodynamic Psychotherapy Institute (Ethical Approval Number: 002/2021). Informed consent was obtained from all individual participants included in the study.

APPENDIX A. Therapist Self‐Efficacy Scale (T‐SES)

Nel corso degli incontri di trattamento psicologico o psicoterapia sono in grado di:

Completamente in disaccordo	Moderatamente in disaccordo	Né d'accordo né in disaccordo	Moderatamente d'accordo	Completamente d'accordo
1	2	3	4	5

1. Essere accogliente	1	2	3	4	5
2. Formulare interventi in modo efficace	1	2	3	4	5
3. Stimolare processi d'insight	1	2	3	4	5
4. Promuovere l'alleanza terapeutica	1	2	3	4	5
5. Facilitare l'espressione di Sé	1	2	3	4	5
6. Gestire adeguatamente le emozioni	1	2	3	4	5
7. Scegliere il momento adeguato per intervenire	1	2	3	4	5
8. Cogliere i segnali non verbali	1	2	3	4	5
9. Superare i momenti di stallo nella relazione	1	2	3	4	5
10. Comprendere il significato profondo delle narrazioni	1	2	3	4	5
11. Rispettare il contratto terapeutico	1	2	3	4	5
12. Comprendere i segnali psicopatologici	1	2	3	4	5
13. Riparare le rotture che si presentano nel rapporto	1	2	3	4	5
14. Tollerare le emozioni negative	1	2	3	4	5
15. Utilizzare il ragionamento clinico	1	2	3	4	5
16. Comprendere il transfert	1	2	3	4	5
17. Attivare il processo diagnostico	1	2	3	4	5
18. Utilizzare la sensibilità clinica	1	2	3	4	5
19. Comprendere il controtransfert	1	2	3	4	5
20. Favorire la mentalizzazione	1	2	3	4	5
21. Comprendere i circoli viziosi	1	2	3	4	5

Gori, A. , Topino, E. , Brugnera, A. , & Compare, A. (2022). Assessment of professional self‐efficacy in psychological interventions and psychotherapy sessions: Development of the Therapist Self‐Efficacy Scale (T‐SES) and its application for eTherapy. Journal of Clinical Psychology, 78, 2122–2144. 10.1002/jclp.23391

DATA AVAILABILITY STATEMENT

The data presented in this study are available on request from the corresponding author. The data that support the findings of this study are available from the corresponding author upon reasonable request.