Perceptions of Portuguese parents about the acceptability of a multicomponent intervention targeted at behavioral inhibition during early childhood

High and stable behavioral inhibition during early childhood is a risk factor for later anxiety disorders. The few available interventions targeted at behavioral inhibition have not yet been implemented in European countries. Evalu-ating intervention acceptability is essential when introducing interventions in new cultures. This study aimed to explore the perceptions of parents about the acceptability of the multicomponent Turtle Program in Portugal. Participants were 12 parents (from seven families) of children with a positive screening on the Behavioral Inhibition Questionnaire and no diagnoses of developmental disorders/selective mutism. Children’s mean age was 55.86 months and most children were female and first-born. Parents and children participated in the eight-sessions Turtle Program. After each session, parents completed weekly satisfaction checklists. Following completion of the full intervention, parents were invited to participate in individual qualitative in-depth interviews. The thematic analysis revealed that both parents perceived the intervention objectives and contents as relevant. Both parents suggested the introduction of follow-up sessions, the discussion of practical experiences, the need to be sensitive to cultural differences in positive language, and the provision of more feedback about children’s activities. These findings support prior research on the acceptability and cultural tailoring needed for parenting and child socioemotional learning interventions.

displaying high and stable BI and AW across time, children have been found to be at greater risk of developing later anxiety disorders (Chronis-Tuscano et al., 2009), to be less socially competent than children of similar ages, and, thus, more vulnerable to peer exclusion, rejection, and victimization (Rubin, Barstead, Smith & Bowker, 2018).
Nevertheless, Chronis-Tuscano and colleagues (2009) have shown that only one third of highly inhibited children during early childhood developed an anxiety disorder during adolescence. Recent research has focused on the modifiable parenting and peer factors that can intensify or decrease the strength of the associations between BI/AW and later anxiety and that need to be targeted in early intervention programs .
In line with the developmental and transactional framework (Rubin et al., 2009), research showed that parents often respond to children's inhibited behaviors in an overprotective way that did not foster independence and emotion-regulation skills (Hane, Cheah, Rubin, & Fox, 2008;Lewis-Morrarty et al., 2012). Given that emotionregulation skills are associated with socially competent behaviors, inhibited/withdrawn children who grow up in such family environments were found to refrain from engaging in peer interactions at preschool (Smith, Hastings, Henderson & Rubin, 2019). When withdrawing from the peer group, research showed inhibited children loose important opportunities to acquire age-appropriate social skills and to be more likely to experience peer difficulties, which, in turn, maintain their self-imposed isolation in the presence of peers and increase the risk of developing anxiety disorders .
These transactional influences of parent, peer, and child behaviors on the developmental pathways of inhibited children provide a rationale to develop multimodal interventions . This rationale sustained the development of the eight-sessions Turtle Program (Chronis-Tuscano et al., 2015), composed of parallel parent and child groups (oriented by two trained leaders in each) that address simultaneously parenting behaviors and peer group interactions to redirect inhibited preschoolers to healthier developmental pathways .

The Multi-Component Turtle Program
The parent group of the Turtle Program follows the principles of the Parent Child Interaction Therapy (PCIT; Eyberg, Nelson, & Boggs, 2008) adapted for anxiety problems in children aged 2-6 years (Pincus, Eyberg, & Choate, 2005;Comer, Busto, Dick, Furr, & Pulifiaco, 2018). In fact, PCIT is a transdiagnostic intervention that strengthens a core element of healthy child development (parent-child relationship) and has shown robust effects that generalize across contexts and display good maintenance (Niec, 2018). As in traditional PCIT for early externalizing problems, the Turtle Program (Chronis-Tuscano et al., 2015) starts with a first phase (Child-Directed Interaction [CDI]), during which parents learn the "do" and "don't" skills to follow child's lead during a 5-min special time of play. Unlike traditional PCIT, parents are then involved in a second phase (Bravery-Directed Interaction [BDI]), during which parents learn the principles of gradual exposure, using hierarchies of feared social situations (bravery ladders) and contingent rewards for social approach behaviors. The third phase (Parent-Child Directed Interaction [PDI]) is shorter than in traditional PCIT and teaches parents to distinguish between anxious and oppositional behaviors and implement effective discipline strategies (effective commands and time-out) for the latter. Sessions include not only psychoeducational activities based on direct instruction, role plays, and discussion of written handouts, but also in vivo therapist coaching of the parent and child together . These coaching activities provide additional benefits for both parents and children, because research has found that interventions involving parenting skills practice with children showed larger effects than purely psychoeducational ones (Kaminski, Valle, Filene, & Boyle, 2008).
The child group of the Turtle Program extends the Social Skills Facilitated Play Program (Coplan, Schneider, Matheson, & Graham, 2010). In each session, group leaders teach briefly specific social, social problem-solving, and emotion-regulation skills, using puppets and storytelling (Chronis-Tuscano et al., 2015). Group leaders also facilitate free play and group activities, using systematic modeling and reinforcement, to scaffold child peer interactions in an equipped playroom and enhance child gradual exposure to feared social situations . In the United States, the pilot study revealed that families assigned to the Turtle Program displayed significant decreases in child BI, social anxiety symptoms, and increases in observed parent warmth and sensitivity (Chronis-Tuscano et al., 2015) when compared to families assigned to a wait-list condition. Furthermore, children participating in the Turtle Program displayed a significant increase in the frequency of classroom social initiations and peer interactions at preschool than children from the wait-list condition .
Notwithstanding its promising outcomes, the Turtle Program is not yet available in European countries, such as Portugal. The high rates of anxiety disorders in Portugal (Caldas-Almeida & Xavier, 2013) justify the need to introduce novel interventions targeted BI/AW.

1.2
The influence of culture on the acceptability of evidence-based interventions Within an ecological framework, culture is a major context of development (Bronfenbrenner & Morris, 2006) that shapes how parents perceive and respond to inhibited/withdrawn behaviors (Rubin et al., 2006) and, ultimately, evaluate available interventions. Portugal continues to be regulated by collectivist and family-oriented values that establish a high interdependence between the family members and an emphasis on respect for authority (Aboim, 2013;Wall & Gouveia, 2014). However, these family-oriented norms and values coexist with an increasing diffusion of individualistic values related to sociability and emotional expressiveness (Wall & Gouveia, 2014). This balance between collectivist and individualistic values may influence how Portuguese parents perceive the interventions targeted at inhibited children, in specific ways. Examining treatment acceptability, that is, the judgments of consumers about the appropriateness of intervention procedures (Kazdin, 1980), is an important preliminary step when introducing novel interventions in new cultural contexts (Barrera, Castro, Strycker, & Tooberg, 2013).
Consistent with research on other behavioral parent training interventions (Dumas, Arriaga, Begle, & Longoria, 2010;Mejia, Calam & Sanders, 2015;Parra-Cardona et al., 2009, few available evidence has shown that LatinX mothers and fathers living in the United States (McCabe, Yeh, Garland, Lau, & Chavez, 2005;Niec et al., 2014) or Central and Southern America (Matos, Torres, Santiago, Jurado, & Rodriguéz, 2006) perceived traditional PCIT for early externalizing problems as acceptable and culturally relevant. Consistently with common concerns that arose in non-LatinX samples, LatinX parents reported discomfort toward ignoring minor misbehaviors and timeout (Calzada, Basil, & Fernandez, 2013;Matos et al., 2006;McCabe et al., 2005;Niec et al., 2014). Both mothers and fathers expressed a preference for therapist-guided discussion with other parents and active skills training rather than for psychoeducational activities (Mejia et al., 2015;Niec et al., 2014;Parra-Cardona et al., 2009 and recommended the inclusion of opportunities to watch videos of parent-child interactions (Matos et al., 2006;McCabe et al., 2005). Parents also expressed concerns toward in vivo coaching and homework (Niec et al., 2014). In particular, fathers expressed more discomfort toward the PCIT laboratory-like setting than mothers (McCabe et al., 2005) and recognized that practical responsibilities take them away from time with their children (Niec et al., 2014). To our knowledge, the perceptions of LatinX parents about the acceptability of PCIT interventions adapted for anxiety problems have not yet been examined.
With respect to child interventions, research has found that Mexican (Gallegos-Guardado, Ruvalcaba-Romero, Garza-Tamez, & Villegas-Guinea, 2013) and Portuguese (Pereira, Marques, Russo, & Barros, 2014) parents reported favorable views toward the FRIENDS for Life Group Program targeted at preventing anxiety problems in school-aged children. Cognitive behavioral contents and activities were perceived as useful and enjoyable for children (Pereira et al., 2014), especially those related to emotional recognition and expression (Gallegos-Guardado et al., 2013). However, parents were only moderately engaged in the intervention, especially in out-session exposure practice (Pereira et al., 2015).
To date, research has only focused on the perceptions of practitioners about the acceptability of the multicomponent Turtle Program (Guedes, Alves, et al., 2019;Guedes, Coelho, et al., 2019). These qualitative studies have shown that Portuguese practitioners perceived the objectives and the contents of the parent and child groups as acceptable but recommended modifications in how the intervention is presented to families, such as giving more time to families (adding more and/or follow-up sessions); broadening the focus on social interaction and emotional expressiveness; and introducing videos and interactive materials (Guedes, Alves, et al., 2019;Guedes, Coelho, et al., 2019). New qualitative studies are needed to extend prior knowledge on the perspectives of practitioners and shed light on the perceptions of parents. Thus, the present study aimed to explore the perceptions of Portuguese mothers and fathers about the acceptability of the objectives, structure, format, contents, activities, and materials of the Turtle Program.

Participants
The sample consisted of 12 parents (six mothers and six fathers) who participated in the multicomponent Turtle Program with their children. Inclusion criteria were: (1) child age between 3.5 and 5 years at the time of screening; (2) a positive screening for BI; (3) ability of parents and children to understand Portuguese, assessed during the pre-intervention interview; (4) preschool attendance; and (5) parent consent and child assent to participate in the study. Exclusion criteria were a diagnosis of developmental disorders or selective mutism. The sample size was determined based on code saturation criteria, considering the use of homogeneous samples focused on study objectives and semistructured interview guides (Guest, Bunce, & Johnson, 2006). Mothers and fathers had a mean age of 38.14 years (SD = 3.02) and 39.14 years (SD = 3.19), respectively. All parents lived in the Metropolitan Lisbon area. Half of the couples (n = 3) were married, whereas the other half (n = 3) were cohabitating. Most mothers (n = 5) and all fathers (n = 6) had college education. All mothers (n = 6) and most fathers (n = 5) worked full time. None of the parents reported any emotional and/or behavioral problem. Children had a mean age of 55.28 months (SD = 11.86) at the beginning of the intervention. Most children were female (n = 5), were first-born (n = 5), and had siblings (n = 6). All parents reported that children's development was the expected for age. Social anxiety problems were previously identified in two of the children, although they were not involved in a medical or psychological treatment at the beginning of the Turtle Program.

Procedures
This study based on a mixed-method design is part of a research project, approved by the ISPA Ethics Committee. From January to November 2018, the intervention program was presented to families by pediatricians or preschool teachers from the contact network of the research group. Primary caregivers (the parent who demonstrated interest in participating in the Turtle Program) were contacted by the research group. In the present sample, the primary caregiver was the mother in all families. During the first contact, mothers were informed about the study aims and procedures. Mothers who agreed to participate signed an informed consent and completed the pre-assessment, which was conducted by a trained researcher. After the pre-assessment, families who met the inclusion criteria were invited to participate in the Turtle Program. After each session, parents completed weekly satisfaction checklists. Following completion of the full intervention, parents were invited to participate in individual qualitative in-depth interviews to explore their perceptions of acceptability about the Turtle Program. The postintervention assessment was conducted by a blinded and trained researcher who did not conduct the groups with the families. For triangulation purposes, parents completed a self-report satisfaction questionnaire. The flowchart of recruitment and retention data is presented in Figure 1.

Instruments
During the pre-intervention assessment, the following instruments were used: Sociodemographic and clinical form: Parents provided information on their child (age, sex, birth order, and number of siblings) and own (age, education, and employment status) sociodemographic data. With respect to clinical data, parents were asked to report if they and/or their child were experiencing any developmental, emotional, and/or behavioral problem. If they responded affirmatively, parents reported the type of developmental, emotional, and/or behavioral problem that they and/or their child were experiencing and whether they were receiving any intervention for the reported problems.
Behavioral Inhibition Questionnaire (Bishop, Spence, & McDonald, 2003;Fernandes, Santa Rita, Martins, & Faísca, 2017): This self-report questionnaire consists of 30 items that assess parent perceptions about child's BI, considering six contexts that reflect three domains: Social Novelty (14 items), which refers to child's inhibited behaviors toward unfamiliar adults, unfamiliar peers, and performance situations in front of others; Situational Novelty (12 items), which refers to child's inhibited behaviors during separation and at preschool and unfamiliar situations; and Physical Activities (four items), which refers to child's inhibited behaviors when there is a minor possible risk of injury. For each item, parents are asked to report how frequently their children displayed inhibited behaviors, using a Likert scale ranging from 1 (Almost Never) to 7 (Almost Always). Higher total scores in the Behavioral Inhibition Questionnaire indicate higher levels of child BI. Children whose mothers reported mean total scores higher than the reference mean scores plus one standard deviation (Fernandes et al., 2017) were considered eligible. Cronbach's alpha was .67.
Selective Mutism and Additional Childhood Disorders Supplementary Modules -Anxiety Diagnostic Interview Schedule (ADIS) for DSM-IV -Parent Version (Albano & Silverman, 1996;Russo, Marques, Pereira, & Barros, 2011): These modules of the ADIS-IV-P allow the screening evaluation of selective mutism and additional childhood disorders, according to the Diagnostic and Statistical Manual of Mental Health Disorders, Fourth Edition (DSM-IV). The selective mutism module consists of eight yes/no questions assessing diagnostic criteria related to the child's persistent inability to speak at school or in social situations (although being able to speak at home), the interference of child behavior at school, and the length of the reported difficulties. Additional childhood disorders module includes seven yes/no questions assessing child social interaction, communication and ritualistic behaviors, and their interference in four relevant areas of child life (school, friendships, family life, sleep, eating, and concentration). Strong reliability for ADIS-IV-P was reported in prior research with preschool children (Kennedy, Rapee, & Edwards, 2009).
During the postintervention assessment, the following instruments were used:

F I G U R E 1 Flowchart of the recruitment and retention data
• Semistructured interview guide: This interview guide was developed in accordance with the recommendations of Daly (2007) for qualitative in-depth interviews. After a brief social conversation, the research objectives and procedures were introduced. Then, general issues concerning parental motivations and expectations about the intervention were discussed. The interview guide explored if parents perceived that the participation in the Turtle Program was in line with their initial expectations and why, before focusing on the key questions concerning the acceptability of the intervention. First, parents were asked to describe their general impressions about the Turtle Program and whether they perceived it as acceptable for Portuguese families. Then, more specific questions were used to explore specific parental perceptions about the acceptability of the intervention objectives ("do you feel that the intervention objectives were targeted to your child difficulties? why?"), structure (e.g., "what is your perspec-tive about the number of sessions?"), format ("was the group format appropriate for parents/children?"), contents ("which contents were more/less meaningful for you?"), activities ("which activities did you appreciated more/less?"), and materials ("which materials were more/less useful?"). For each of the previously described topics, parents were also given the opportunity to suggest modifications. The interview guide closed with questions to examine the perceived improvements in child and parent behaviors and the satisfaction with the intervention. At the end of the interview, parents were given the opportunity to make additional comments. For study aims, we only examined parental responses to key questions.  behaviors; (3) perceived changes in child's behaviors; and (4) parental satisfaction and suggestions of improvements. For triangulation purposes, we only examined parental responses to three questions from the first section: the yes/no question about the appropriateness of the number of sessions ("did you feel the program was sufficient?") and the two questions assessing the perceived usefulness of home practice ("how helpful do you find the homework assignments?") and relevance of intervention contents/strategies ("how well did the information and strategies presented in the group apply to your child's challenges?"). For each question, parents were asked to respond, using a 7-point Likert scale ranging from 0 (Not at All) to 6 (A Lot or Very Much).
Higher mean scores in these questions indicate more favorable parental perceptions about the usefulness of home practice and the relevance of intervention contents/strategies. • Weekly Sessions Satisfaction Checklists: Each weekly checklist consists of three sections assessing the perceived relevance of the session objectives, the usefulness of the intervention procedures, and the satisfaction with the group leaders, respectively. For triangulation purposes, we only examined parental responses to the first two sections. With respect to the perceived relevance of the objectives, parents were asked to indicate how much true each of the presented statements were for them concerning each session (e.g., "I feel that the first session provided me important information about anxiety"), using a 5-point Likert scale (1 = Not True at All to 5 = Completely True). With respect to the usefulness of the intervention procedures, parents were asked to classify how useful each of the presented procedures were for them in each session (e.g., "discussion with other parents" or "role plays"), using a 5-point Likert scale (1 = Not Useful at All to 5 = Extremely Useful). Item scores are averaged to yield a mean score about the perceived relevance of the intervention objectives and usefulness of the intervention procedures for each session.
Higher mean scores indicate more favorable parental perceptions.

Intervention
The general features of the multicomponent Turtle Program were described in Section 1 (Introduction). The structure and contents of the eight weekly parent and child sessions are summarized in Table 1.

Data analysis
Qualitative data analysis was performed using QSR NVivo Pro 12. Data analysis was based a continuous process of data collection, reduction, display, and verification (Huberman & Miles, 1994). Each interview was transcribed verbatim, using fictional names. Following an existentialist/realistic approach, a deductive thematic analysis was conducted to identify, analyze, and report patterns (themes) within data (Braun & Clarke, 2006). Themes were meaningful units (references), defined as set of phrases about the same topic. Data were initially coded and collated for each code; thereafter, codes were collated into themes and subthemes, which, in turn, were reviewed, analyzed, and interrelated. Average Cohen's Kappas (ĸ) were calculated and interpreted as poor (below 0.40), fair to good (0.40 to 0.75), and excellent (over 0.75) agreement. Table 2 summarizes the 46 subthemes, grouped into five main themes that were identified. Below, we describe each of the five main themes.

Acceptability of the objectives
The overarching idea of both mothers' and fathers' thoughts on this theme was that the intervention objectives were relevant to manage their child's inhibited behaviors. Table 3 shows that both mothers (mean scores ranging from 4.07 to 4.50) and fathers (mean scores ranging from 3.78 to 4.53) reported the intervention objectives of each session as very relevant.

Acceptability of the structure
Common among two subthemes on this theme were parents' mixed attitudes toward the number of sessions and time interval between sessions. Most fathers agreed that the number of sessions was enough but considered that a 2-weeks interval or a pause at mid-intervention would give more time to families to practice and notice changes in child behaviors. Half of the mothers considered that additional sessions would give families more time for group discussion and coaching, but all of them perceived that weekly sessions were the best option. In the PSSSQ, 60% of the mothers and 75% of the fathers reported that the number of sessions was enough. Also captured in this theme was both parents' agreement that sessions were too lengthy for young children and that the introduction of follow-up sessions would be useful to monitor child progress. Lindsey stated: "An individual follow-up would make sense. But a group format would be also important, so that we can understand the progress of other children and parents' difficulties after the end of the intervention sessions."

Acceptability of the format
The overarching idea of both parents' thoughts on this theme was the advantages of the group format for parents and, to a lesser extent, for children. Almost all mothers and fathers highlighted that having a sense of group belonging and sharing experiences with other parents allowed them to feel supported and understood in a nonjudgmental way. For example, Liliane explained: "This was the first place where I was able to talk about my child's difficulties and I didn't feel that others were judging me. Because I always feel that other people judged me and think that I am exaggerating my child's difficulties. And this didn't happen in the group." Half of the mothers and fathers considered that the peer group format allowed children to understand that there are other children with similar difficulties and practice social approach behaviors.

Acceptability of the contents
Common to four subthemes on this theme were parents' thoughts on the relevance of most parent (psychoeducation on BI, CDI, and BDI) and child contents (bravery and social interaction skills). In the PSSSQ, mothers (M = 5.40, SD = 0.89) and fathers (M = 4.25, SD = 1.50) also classified the presented information and strategies as very useful. Specifically, both mothers and fathers explained that understanding the origins of BI reduced parental blame and enhanced awareness on its parenting maintenance factors. Most mothers and fathers agreed that following child's lead during special time play and using labeled praises for child appropriate behaviors (CDI "do" skills) were among the most important learned skills and that bravery ladders (BDI skills) were useful to promote child gradual exposure to feared social situations. However, three subthemes arose that highlighted cultural dissonance toward the definition of small bravery steps and the provision of contingent material rewards, the counterintuitive order of the intervention contents (CDI TA B L E 2 Themes and subthemes concerning the acceptability of the objectives, structure and format of the Turtle Program Also captured by this theme was parents' uncertainty toward the relevance of PDI. Both mothers and fathers acknowledged that distinguishing anxious/oppositional behaviors and (in)effective commands were useful contents, but that time-out was excessively punitive, because most inhibited children are usually compliant.

Acceptability of the activities
Five subthemes that arose highlighted that half of the mothers and fathers acknowledged the appropriateness of parent psychoeducational activities (with a particular emphasis on group discussion) but recommended the introduction of more active learning methods, such as the discussion of culturally tailored videos of parent-child interactions, concrete examples, and hypothetical problematic child behaviors. For example, Jason explained: "It would be useful to give us more practical examples (. . . ) of situations that parents may experience (. . . ) with their children and how they can manage these situations (. . . ) to help us to assimilate the contents. . . " Another overarching idea in this theme was both mothers' and fathers' perspectives on the appropriateness of coaching activities to receive immediate therapist feedback and understand how to manage parent-child interactions. Lindsey stated: "For me, the CDI coach was the most useful activity. It was very important for the parent to apply the learned skills and to understand how these skills were really effective, in practical situations. We were able to receive feedback and to understand what we were doing well and not so well." Table 3 shows that mothers (mean scores ranging from 4.10 to 4.65) and fathers (mean scores ranging from 3.61 to 4.15) perceived all the intervention procedures of the Turtle Program (i.e., psychoeducational and "coaching" procedures) as moderately to very useful. However, four subthemes that arose highlighted that mothers and fathers diverged on their perspectives about coaching-related difficulties. Half of the mothers reported experiencing discomfort, due to video-recording and observation by other parent group members during CDI coach. Conversely, fathers focused mainly on the need to receive more corrective feedback on what they did not do so well (instead of labeled praises for appropriate parenting behaviors), the boredom related to the repeated construction of hierarchies of feared social situations (bravery ladders), and the artificiality of the laboratory-like setting used to practice graduated exposure during the BDI coach sessions.
Also captured by this theme was mothers' and fathers' thoughts on the usefulness of CDI and BDI home practice but also on the demands related to these experiences due to time restrictions and work-family balance; a  In the PSSSQ, mothers (M = 5.40, SD = 0.89) and fathers (M = 4.50, SD = 0.58) also described home practice as very useful but reported that they only completed some of them (mothers: M = 4.40, SD = 0.55; fathers: M = 3.50, SD = 0.58).
Three subthemes that arose highlighted that mothers and fathers diverged in their perspectives on how to overcome practical difficulties in home practice. Mothers recommended the introduction of home practice in pre-existing routines, such as practice special time play during bath. Fathers suggested the introduction of interactive (video-recording) and interpersonal (cell phone reminders) tools to keep motivated.
The last overarching idea in this theme was both parents' thoughts on the usefulness of child free play in a peer context and on collective games for child difficulties. In fact, parents considered that these types of activities were important to promote the gradual exposure to feared social situations and to reinforce children's social approach behaviors.

Acceptability of the materials
Common to three subthemes were both parents' thoughts on the usefulness of written handouts and comfort with the coaching materials but also on the relevance of introducing interactive tools during (e.g., media adjuncts to convey the contents) and between (e.g., WhatsApp group to discuss home practice among the parent participants) the sessions. Also captured in this theme was the desire of half of the mothers and fathers to be provided more specific and detailed feedback on child behaviors, difficulties, and progress (biweekly meetings with child therapist or final individualized written report) to enhance the practice of the learned skills in naturalistic contexts.

DISCUSSION
This study extends prior research on the perspectives of practitioners (Guedes, Alves, et al., 2019;Guedes, Coelho, et al., 2019) and examines the perceptions of parents about the acceptability of the multicomponent Turtle Program targeted at inhibited preschoolers. Our findings support prior research, showing that the objectives and contents of PCIT (Matos et al., 2006;McCabe et al., 2005;Niec et al., 2014) and child cognitive behavioral interventions involving socioemotional learning and gradual exposure to feared situations (Gallegos-Guardado et al., 2013;Pereira et al., 2014) were perceived as acceptable by both mothers and fathers from more familyoriented cultures. Specifically, psychoeducation on the etiology of BI and parent influences on the developmental pathways of inhibited children that sustained the development of the Turtle Program (Rubin et al., 2009) was considered beneficial to reduce parental blame  by Portuguese mothers and fathers. Nevertheless, main emphasis was given to the relevance of CDI, namely, special time and labeled praise. Previous studies on the acceptability of behavioral training (Calzada et al., 2013;Dumas et al., 2010;Parra-Cardona et al., 2017) and traditional PCIT (Matos et al., 2006;McCabe et al., 2005;Niec et al., 2014) evidenced that LatinX parents expressed a preference for family intervention components that promote warmth and closeness in parent-child relationships. Family-oriented norms valuing the interdependence between family members in Portugal (Wall & Gouveia, 2014) may have influenced the favorable attitudes of parents toward CDI. However, we cannot ignore that our sample was collected in an urban context, where parents may be less tied to traditional family norms.
Consistent with prior research on traditional PCIT in LatinX populations (Matos et al., 2006;Niec et al., 2014), both mothers and fathers reported mixed attitudes toward PDI. On the one hand, our findings support previous empirical studies, showing that LatinX parents appreciated learning how to give good commands (Matos et al., 2006) or good directions to children (Parra-Cardona et al., 2009. Several researchers have found that parents from LatinX cultures typically place a higher value on proper demeanor and respectfulness when compared with European American parents, so that child problematic behaviors may represent a social burden for them (see Harwood, Leyendecker, Carlson, Asencio, & Miller, 2002 for a review). The importance of respect for authority (Aboim, 2013) and sociability (Wall & Gouveia, 2014) in the Portuguese society may have influenced parents' favorable views toward effective commands. On the other hand, the perspectives of Portuguese parents are also in line with common concerns about time-out that arose in non-LatinX and LatinX samples (Matos et al., 2006;McCabe et al., 2005;Niec et al., 2014). The type of child difficulties targeted in the Turtle Program may explain why Portuguese parents perceived time-out as excessively punitive for inhibited children. In fact, Pincus, Santucci, and Ehrenreich (2008) also found that non-LatinX parents involved in PCIT for separation anxiety felt that their children were typically compliant and that PDI skills were more relevant to manage outbursts and noncompliance, in the presence of comorbid oppositional behaviors.
The insights of Portuguese mothers and fathers support prior research, showing that both LatinX (Niec et al., 2014;Parra-Cardona et al., 2009 and non-LatinX parents (see Barnett & Niec, 2018 for a review) appreciated the cohesive experiences that are fostered in therapist-guided parent groups. Nearly half of mothers and fathers also valued the benefits of the child group format and all of them highlighted the appropriateness of child play activities. Parents' thoughts on these subthemes are consistent with the developmental and transactional framework (Rubin et al., 2009), which establishes that the promotion of child social approach behaviors toward the peer group in the context of play can contribute for the development of age-appropriate social and interpersonal negotiation skills and redirect inhibited children to healthier developmental pathways .
Consistent with this idea, both parents highlighted the relevance of child contents related to social skills and bravery. Children's young age may explain why our parents focused on modifying children's behavioral manifestations of internalized anxiety during social interactions (Rubin et al., 2009) rather than children's emotional skills, like Mexican parents involved in the FRIENDS for Life Program (Gallegos-Guardado et al., 2013). However, fathers considered these contents more relevant than mothers. This may be attributable to societal expectations, according to which inhibition/anxious withdrawal is typically less acceptable for men than for women (Rubin et al., 2009).
Notwithstanding the acceptability of the intervention objectives, format, and contents, Portuguese mothers and fathers recommended modifications in the way how the intervention is presented to families. First, both parents' thoughts on follow-up sessions and active learning activities/materials during psychoeducational sessions are consistent with prior research on the acceptability of behavioral training and traditional PCIT, showing that Lat-inX parents usually appreciate to be given more time for building rapport and prefer the group discussion of concrete examples, videos of parent-child interactions, and simple written/digital materials instead of direct instruction (Matos et al., 2006;McCabe et al., 2005). Second, both mothers and fathers suggested that the CDI verbal "do" skills (namely, labeled praises) and BDI contingent rewards may need to be presented/provided in a more culturally tailored manner. These findings are consistent with prior research on parent behavioral training, showing that LatinX parents tend to rely more on nonverbal expressions of encouragement and social rewards than on verbal praises and material recompenses (Dumas et al., 2010). Third, both parents in our sample wanted to be provided with more opportunities to observe child activities, as well as detailed written and in-person feedback on what happened in child-only sessions. The type of child difficulties targeted in the Turtle Program may have also influenced the findings. In fact, the developmental and transactional framework (Rubin et al., 2009) establishes that parents often respond to children's inhibited behaviors in a more protective way (Hane et al., 2008;Lewis-Morrarty et al., 2012) that may increase the need of being provided more information about children's intervention activities. Although our sample was collected in an urban setting, the socialization toward child independence has been found to be lower in more collectivist family-oriented cultures than in more individualistic ones (Rubin, Oh, Menzer, & Ellison, 2011) and Portuguese parents appeared to be less prone to encourage child autonomy than parents from North European cultures (Aboim, 2013). The preference for healthcare providers who foster close personal relationships among LatinX parents also arose in prior research (see Christian-Brandt & Philpott, 2018 for a review) and seems to be consistent with Portuguese values related to social support (European Commission, 2012).
However, mothers and fathers reported divergent opinions about the remaining features of intervention structure and activities. Half of the mothers considered that additional sessions would give families more time for group discussion and skills practice compared to fathers, who suggested to lengthen the time interval between sessions. The benefits of interventions involving skills practice with children that was found in the meta-analysis of Kaminski and colleagues (2008) also arose in the perspectives of both mothers and fathers. Nevertheless, mothers and fathers experienced different difficulties during coaching activities. Consistent with the anticipated concerns of Portuguese practitioners (Guedes, Coelho, et al., 2019), half of the mothers experienced discomfort, due to the observation from other parent group members during the CDI.
Despite the increasing involvement of fathers in child rearing, Portuguese mothers have been found to continue to be the main responsible for child direct and indirect care (Monteiro et al., 2010), so that they may be particularly vulnerable to the social exposure of their parenting skills and children's behaviors.
On the other hand, half of the fathers experienced discomfort due to the artificialism of coaching during BDI and boredom with the preparation of bravery ladders for out-session exposure practice with the other parent group members. Concerns about the laboratory-like setting emerged in the perspectives of LatinX fathers toward traditional PCIT (McCabe et al., 2005) and may be even more apparent in PCIT adapted to anxiety problems, because BDI presupposes the development of hypothetical social scenarios to practice in-session bravery challenges . The preparation of bravery ladders for out-session exposure during BDI may not be consistent with father's preferences for intervention activities. In fact, Phares, Rojas, Thurston and Hankinson (2010) found that family intervention contents can be more interesting for fathers, when focusing on gender differences in emotional disclosure and social roles, framing the intervention as practical or using examples from recreational or sports activities. Fathers' negative attitudes toward parent group activities during BDI coach is also in line with the findings on the acceptability of home practice, showing that fathers were more prone to recommend the introduction of interactive tools (e.g., video-recording and cell phone messages) than mothers.
In fact, our findings support prior research, showing that both mothers and fathers recognized the relevance of home practice (especially, child-directed play), but reported difficulties in implementing the learned skills in naturalistic contexts (Dumas et al., 2010;Pereira et al., 2015). As found in previous studies on the FRIENDS for Life in Portugal (Pereira et al., 2015), both mothers and fathers considered the home practice based on exposure as being especially demanding. Difficulties toward home practice was more apparent for fathers, because nearly half of them reported a decrease in their motivation over time. Several explanations for these findings are possible. First, it is possible that parents did not to take the anxietyrelated problems of their children so seriously, due to the preventive focus of the intervention (Pereira et al., 2015) and children's young age. Second, the practice of gradual exposure in everyday social contexts and situations may be challenging for parents, due to the high number of working hours in Portugal (OECD, 2016). Niec and colleagues (2014) also found that LatinX fathers perceived that responsibilities take them away from time with their children. Last, the parental desire to skip the CDI phase and get right to changing child problematic behaviors that was reported in PCIT for separation anxiety disorders (Pincus et al., 2008) was also apparent in our sample, especially among fathers. According to a developmental and transactional framework (Rubin et al., 2009), it is possible that parents displayed overprotective and controlling behaviors toward child inhibited behaviors (Hane et al., 2008;Lewis-Morrarty et al., 2012), so that the goal in CDI to let the child take the lead can be difficult for them. However, it is also possible that these findings reflect parental expectations toward immediate child behavior changes, which were also identified in traditional PCIT with Lat-inX parents (McCabe et al., 2005). Consistent with this idea, our fathers shared similar perspectives with practitioners (Guedes, Coelho, et al., 2019) and considered that a 2-weeks interval between sessions would give them more opportunities to notice changes in child behaviors.
To the best of our knowledge, this is the first study that explored the perceptions of parents about the acceptability of the Turtle Program targeted at inhibited preschoolers. Strengths of the present study were the inclusion of both mothers and fathers and the conduction of in-depth qualitative interviews by a blinded and independent researcher. However, some limitations need to be acknowledged. The sample size was established based on code saturation criteria (Guest et al., 2006). However, the small sample size, the biased sex ratio of the children, the purposive sampling method, and the context of data collection limit the generalization of the findings. Furthermore, all parents in our sample lived in an urban area, so that their perspectives may be less tied to traditional family culture than those of parents from rural communities. Although all the interviews were conducted by the same research assistant (who was uninvolved in the treatment) using a semistructured interview guide, we cannot ignore that the obtained findings reflect a situated understanding, which depends on the characteristics of the interviewer. The measurement of treatment acceptability was based on a multimethod approach but only relied on parents' opinions about the Turtle Program. Despite their consistency with qualitative findings, the triangulation of the results was limited to selfreport questionnaires, developed by the research team.
Our findings should be considered in the context of communities with similar characteristics in Portugal. Future longitudinal studies with larger samples based on a multi-informant approach and mixed methods (qualitative in-depth interviews, direct observations, and self-report questionnaires) need to be conducted to explore the acceptability of the Turtle Program in rural communities and in other cultures. This may be useful to examine if culturally tailoring is needed when introducing the Turtle Program in other contexts. These studies also need to explore gender differences in intervention preferences and motivation in other contexts. Evidence on this topic can clarify if targeting fathers' preferences and developing add-on motivational modules may be useful to enhance treatment adherence, persistence, and, ultimately, effectiveness.

CONCLUSIONS
Overall, our findings revealed that Portuguese mothers and fathers perceived the intervention objectives and contents as relevant, with the exception of time-out. Despite the advantages of the group format, mothers and fathers described the demands related to home practice, namely, to gradual exposure outside the sessions and recommended modifications, such as the introduction of follow-up sessions; the discussion of more practical examples and culturally tailored videos; the need to be sensitive to cultural differences in the use of positive language during parentchild play and coaching; and the provision of more feedback to parents about children's activities.

A C K N O W L E D G M E N T S
This study is part of the research project "Adaptation and pilot evaluation for Portugal of the multi-component Turtle Program," conducted by the R&D William James Center for Research, ISPA and co-funded by the Programa Academias do Conhecimento of the Calouste Gulbenkian Foundation (process n • 222926). Maryse Guedes is supported by a scholarship from the Portuguese Foundation for Science and Technology (SFRH/BPD/114846/2016). The study was approved by the ISPA Ethics Committee (I/011/10/2017). We acknowledge the support of all the colleagues who contributed to the intervention implementation and data collection. We would like to also thank all the families who shared their time and experiences for the aims of the present study.