|Home | About | Journals | Submit | Contact Us | Français|
This paper describes the Spanish adaptation of PACE—Parenting Our Children to Excellence. Successfully offered in preschools and daycare centers since 2002, PACE is a research-based preventive intervention to support families in their parenting task through discussions and activities that address practical childrearing issues and promote child coping-competence. Developed in response to community calls, the new program is known as CANNE –Criando a Nuestros Niños hacia el Éxito. The paper makes the processes linking original and adapted versions explicit by accounting for the conceptual and practical decisions that were made as CANNE was being developed. We begin by summarizing the challenges of adapting and translating a behavioral intervention, and by describing the coping-competence model that informs both versions of the program. We turn then to a detailed account of the adaptation itself and of its results. Specifically, we describe: (a) the consultation process at the origin of this adaptation, (b) the adaptation of the manual and the steps taken to establish the extent to which the English and Spanish versions correspond (adaptation fidelity), and (c) the translation of the manual and the cross-language comparison of measures to demonstrate that they yield comparable data when administered in English and Spanish.
Interventions to strengthen families have long been recognized as critical in any comprehensive approach to the prevention of emotional and behavioral problems in youth, as strong families are essential to the development of competent children (e.g., Maccoby & Martin, 1983; Szapocznik et al., 1986). Consequently, many programmatic efforts to promote healthy child development focus on the family. This focus has resulted in the creation and refinement of a sophisticated intervention approach, known generically as behavioral parent training. Used successfully in different contexts, this approach offers a time-limited, cost-effective means of fostering positive parent-child interactions, especially in families in which the “fit” between parent, child, and environment is poor (Dumka, Garza, Roosa, & Stoerzinger, 1997; García Coll et al., 1996; Kazdin, 2005; Kumpfer, Alvarado, Smith, & Bellamy, 2002; Nixon, 2002; Serketich & Dumas, 1996). In recent years, this approach has also been applied to the promotion of coping-competence and the prevention of adjustment problems in young children at risk because of adverse socioeconomic circumstances (e.g., CPPRG, 1999; Dumas, Prinz, Smith, & Laughlin, 1999; Webster-Stratton, 1998), and adapted to meet the cultural and language needs of ethnic minority families (e.g., Matos, Torres, Santiago, Jurado, & Rodriguez, 2006; Reid, Webster-Stratton, & Beauchaine, 2001).
Availability and widespread use of effective interventions are rarely synonymous (Biglan, 1995). A major obstacle to promoting parenting effectiveness and child coping-competence through behavioral interventions stems from the fact that most research-based programs have been developed for and tested with nonminority participants. They reflect European American childrearing beliefs and preferences, and are only available in English. This does not invalidate such interventions, obviously, but it limits their relevance. This is critical when most minority groups in the U.S. are growing faster than the European American population, and when the largest one is not only growing very rapidly but speaks Spanish and shares distinct cultural values and expectations about children and families (Calzada & Eyberg, 2002; Varela et al., 2004).
The need to improve the range of health services for persons with limited English proficiency in the U.S. is a demographic imperative. This is particularly true for the Latino population, which is not only growing rapidly but also more severely affected by health disparities than most minority groups (Office of Minority Health, 2006). From 1990 to 2000, Latinos increased by almost 60%, and from July 2003 to July 2005, 1 out of every 2 persons added to our nation's population was Latino. Today, approximately 1 in 7 persons in the U.S.—and 1 in 5 children—are of Latino origin, two-thirds from Mexico. This growth is largely due to immigration.
As did countless immigrants before them, most Latino immigrants settle in metropolitan areas (U.S. Census Bureau, 2003, 2006). Often poor and lacking basic resources such as adequate housing and healthcare, many live in neighborhoods where services to families are limited or inaccessible, schools struggle to meet the needs of many students, and employment is scarce and unstable—in short, in neighborhoods “where creating effective family life is a formidable challenge” (Coatsworth, Pantin, & Szapocznik, 2002, p. 113). This challenge is particularly acute for parents of young children. Often with larger households and with a higher concentration of preschoolers than any other race or ethnic group (Cauce & Rodriguez, 2001; U.S. Census Bureau, 2006), many immigrant Latino parents: (a) lack extended family support when their children are young; (b) are unfamiliar with the “way things work” in the U.S. and see many of their cultural values and practices questioned by the majority culture; and (c) do not speak English and may be suspicious of people who do (especially if they are in the U.S. illegally). When such risk factors converge and persist for extended periods, they expose children to a host of emotional and behavioral problems, as many parents become overwhelmed in their ability to nurture and guide their development (Szapocznik & Williams, 2000).
Communities throughout the U.S. are increasingly aware of the need to provide quality parenting services to Latino families with young children. As an otherwise supportive Head Start director angrily asked the first author in fall of 2004: “When will your program be available in Spanish? I am upset because I have to go back to my Policy Council and my teachers to tell them again that it is only for parents who speak English. But we don't only have parents who speak English. In fact, most don't!” The program this director was referring to is PACE—Parenting Our Children to Excellence. Successfully offered in more than 50 preschools and daycare centers in Indianapolis since 2002, PACE is a research-based preventive intervention to support families in their parenting task through discussions and activities that address practical childrearing issues and promote child coping-competence. The program is delivered in groups of 10 to 15 parents/caregivers, over eight 2-hour sessions. An outcome study involving more than 400 families shows that PACE has immediate beneficial effects for parents and children that are maintained at a 1-year follow-up (Begle & Dumas, 2010).
Calls for a Spanish version of PACE reflects, at the local level, the demographic imperative outlined earlier. From 2000 to 2005, the Latino student population rose by 220% in the Indianapolis Public Schools (IPS Planning Dept., April 21 2006). Most families accounting for this growth are recent immigrants from Mexico and other Central American countries, many with young children. They usually have little or no English proficiency and therefore only limited access to the services schools, preschools and daycare centers offer parents in the area.
This paper describes the Spanish adaptation of the PACE program, known as CANNE—Criando a Nuestros Niños hacia el Éxito. The paper makes the processes linking original and adapted versions explicit by accounting for the conceptual and practical decisions that were made as the new program was being developed. We begin by summarizing the challenges of adapting and translating a behavioral intervention, and by describing the coping-competence theoretical model that informs both versions of the program. We turn then to a detailed account of the adaptation itself and of its results. Specifically, we describe: (a) the consultation process at the origin of this adaptation, (b) the adaptation of the manual and the steps taken to establish the extent to which the English and Spanish versions correspond (adaptation fidelity), and (c) the translation of the manual and the cross-language comparison of measures to demonstrate that they yield comparable data when administered in English and Spanish.
Adapting and translating a parenting program faces two competing demands: The new program must be relevant and culturally appropriate for the target population and it must be faithful to the program from which it derives. These challenging requirements are regularly debated in cross-cultural psychology and other social sciences (Dumas, Rollock, Prinz, Hops, & Blechman, 1999; Zayas & Rojas-Flores, 2002). Some researchers argue that adaptations should be kept to a minimum to maintain fidelity to the original program, limit the amount of work required to implement and evaluate the new program, and reduce costs (Kazdin, 2005). This perspective, which argues for strong conceptual and methodological correspondence between the original and new programs, is supported by the fact that there is little empirical evidence for the superiority of culturally adapted prevention efforts. Research is sparse but cultural adaptations of well-validated programs have not been shown to be clearly superior to original or more generic versions at equivalent dosages (Harachi, Catalano, & Hawkins, 1997; Kumpfer & Alvarado, 1995).
While acknowledging this state of affairs, others argue for in-depth consultation with members of the target population and for a willingness to integrate their input when adapting existing interventions (APA, 2003; Bernal & Sáez-Santiago, 2006; Kumpfer et al., 2002). Ethically, they point to the need to insure that interventions developed in one cultural context are acceptable when transposed in another, and do not have unintended harmful effects because of differences in cultural values and practices, or language (Sanders, 2000). This is especially important in the parenting area. Childrearing is as much a cultural as it is an educational task, making it necessary to consider how culturally prescribed variables influence socialization practices in the target population (Dumas, Rollock, et al., 1999). Practically, advocates of cultural adaptations point to evidence that, if they do not yield better outcomes, adaptations may result in better adherence (Catalano et al., 1993; Kumpfer & Alvarado, 2003). For example, in the last study, a Spanish adaptation of the Strengthening Families Program facilitated recruitment and resulted in higher completion rates than a generic version of the program. Given that the impact of any program, no matter how good it might be, depends on high adherence, this is obviously a major consideration.
Our research attempts to strike a balance between these two perspectives. Translating the program into Spanish without changing its contents or presentation is an important first step. In addition to having the direct benefit of being understood by Spanish-only speakers, a translation can have secondary benefits, such as conveying to parents that the program developers respect their background and appreciate the challenges of learning new skills in a language with which they are unfamiliar. However, the in-depth consultation we conducted (see below) showed quickly that a Spanish translation alone was unlikely to be successful. This was the message of preschool directors, teachers, and community workers serving Latino families in Indianapolis and familiar with the PACE program, and of expert consultants from different regions of the U.S. They pointed out that the program's contents and methods of presentation would have to be modified for both cultural and practical reasons. Culturally, some of the parenting advice and practices promoted in PACE reflect assumptions and priorities that differ from those of Latino families, and the original program assumes a level of familiarity with American society that many Latinos are unlikely to have. Practically, the Latino families targeted by the new program in Indianapolis and surrounding areas are, for the most part, not highly acculturated to mainstream American culture. They struggle with English and have lower incomes and lower rates of functional literacy than the typical English-speaking families for whom the program was initially developed.
Although we invited and incorporated feedback on how to make the program culturally meaningful and practically relevant for Latino families, we decided from the outset that CANNE would reflect the coping-competence model, which is the conceptual cornerstone of the original program and which has been empirically supported in earlier research with PACE (Begle & Dumas, 2010; Moreland & Dumas, 2007). This was done to insure that original and adapted versions would be guided by the same theoretical perspective, and that research conducted in the two languages would yield comparable data of relevance to parent and child adjustment in the preschool years. As such, our approach—one of relying on a theoretically based and empirically supported program, yet incorporating cultural and practical changes for the target population—is consistent with guidelines to develop interventions for diverse families (e.g., Bernal, 2006).
Coping-competence theory (CCT) assumes that three types of competencies—social, affective, and instrumental—are essential to cope with the many challenges children encounter daily, and that the family is the setting providing most children with the emotional support and guidance they need to develop those competencies (Blechman, Prinz, & Dumas, 1995; Dumas, 1997, 2005; see also Garcia Coll et al., 1996). Challenges are very varied. In the preschool years, they range from small frustrations, such as awaiting one's turn, learning to use scissors, and eating vegetables, to developmental tasks and major life events, such as learning to walk, welcoming a sibling, starting school, facing discrimination, and more. Whatever their nature, challenges put children to the test: they are beyond what children can do easily or entirely new; they tend to cause strong emotions; and they are stressful. CCT assumes further that there are primarily three ways of coping—prosocial, antisocial, and asocial—and that parents and other socializing agents spend considerable time teaching young children to cope prosocially. In other words, much of early socialization consists of providing children with support and guidance to approach challenges constructively: In all cultures, children must learn to listen and negotiate with others, and to respect others' needs and preferences, as they learn to express their own wishes and develop their interests and skills. Such ways of coping are incompatible with crying, hitting, and temper tantrums (antisocial coping), or with being fearful and withdrawing (asocial coping), which are frequent in the early years.
Applied to a preventive intervention such as CANNE, CCT asserts that it is most likely to succeed if it centers on parental strengths rather than on deficit remediation, and helps parents: (a) deal effectively with the tendency of some young children to respond to challenges in an antisocial or asocial manner, and (b) teach children prosocial solutions to challenges at home and beyond. In other words, it is important that the intervention gives parents who need them tools to deal with a young child who regularly refuses to cooperate and becomes aggressive, or withdraws when frustrated. But this is not sufficient. The program must also empower parents to provide the child with emotional support and guidance to anticipate and respond to challenges in ways that are respectful of family values and expectations and effective in the broader social environment (Dumas, 2005; Gorman-Smith, Tolan, Henry, & Florsheim, 2000). This bicultural competence results in a parenting profile that may be most adaptive for children as it promotes early their own ability to navigate effectively within and across cultures (Bornstein & Cote, 2006; LaFromboise, Coleman, & Gerton, 1993). Specifically, evidence shows that positive developmental outcomes for Latinos are best fostered by a balance of acculturation and enculturation, rather than by the maintenance or adoption of one culture over the other (Coatsworth et al., 2002; Gonzales, Knight, Morgan-Lopez, Saenz, & Sirolli, 2002; Rogler, Cortes, & Malgady, 1991). For example, bicultural adolescents have higher social competence (Bautista de Domanico, Crawford, & De Wolfe, 1994) and self-esteem (Phinney, Horenczyk, Liedkind, & Vedder, 2001), and lower involvement in delinquency and substance use (Brook, Whiteman, Balka, & Cohen, 1997).
This approach to effective parenting and socialization in Latino families is not new, as Szapocznik's programmatic research on bicultural effectiveness training illustrates (e.g., Santisteban, Suarez-Morales, Robbins, & Szapocznik, 2006; Szapocznik et al., 1986). This approach is also consistent with ongoing research with African American and European American participants in PACE (Moreland & Dumas, 2007), which shows that firm but responsive and affectionate parenting fosters coping-competence in children and protects them from problems, as it nurtures a sense of efficacy in parents themselves.
Following repeated community requests for a Spanish version of PACE, we consulted with three experts conducting parent training research with Latino families and over 30 individuals working with Latino families in the Indianapolis area. The 12 agencies/organizations these individuals represented included childcare and family agencies, schools, churches, and community and neighborhood organizations. Some issues rapidly emerged as central from the richness of their input.
First, consultants confirmed that there was considerable interest in a Spanish program and community willingness to support its development and implementation. For example, individuals and agencies offered to help bring community groups together, collaborate to recruit and retain families, and host the program. Second, consultants stressed that the new program would need to reflect the fact that many Latinos who are likely to be served by this program are poorer and less educated than more acculturated Latinos, as well as non-Latinos (African Americans and non-Latino White Americans). They also share very different life histories. Many come from rural areas of Mexico and have little or no formal education; they may not read or write, and speak little or no English. As one consultant put it: “Imagine that Indianapolis is all new to you. Many Latinos here are struggling daily with basic routines, like turning on the lights, running a vacuum cleaner, driving… They are still learning how the place works and how they and their kids must adjust, so your program must reflect this by being practical and offering lots of useful information.” Consultants also commented that non-Latinos evoke fear in some Latino families, who do not want to appear “stupid” when they do not know something or think that they may be denounced when they are in the U.S. illegally. Third, consultants stressed that the new program would need to keep key features of the original program intact, such as its focus on participants' strengths and on positive parenting, and its practical, present-oriented approach. From the first session, PACE (a) asks parents what brings them to the group and integrates their concerns and priorities throughout, and (b) promotes consistent discipline and supportive parenting. As one consultant put it: “Discipline is the biggest topic parents want to talk about. In Mexico, there are several family members disciplining their child, and they want to know how parents and schools do it in America.” Finally, regarding translation, consultants recommended that program materials be easy to follow and presented in an idiom of Spanish readily understood primarily by Mexican and possibly other Central American immigrants. Written material should be kept to a minimum, discussed in session so that parents who do not read understand, and provide information on which parents can act (e.g., a list of preschools with bilingual classes).
On the strength of this initial consultative process, we sought to determine the relevance of each module component to Latino parents and, whenever issues arose, to obtain suggestions as to how the program may need to be adapted. To that end, we obtained input from members of the target population and from a second set of consultants.
To obtain input from members of the target population, we recruited a small group of Latino parents at a Head Start parent meeting at the beginning of the school year. The three individuals who provided feedback on the program were mothers of Mexican origin, each with one to three young children. We held three 90-minute sessions, each 1 week apart, to describe and seek input on the program; all three participants attended all three sessions. Modules 1 through 4 were discussed at the first meeting, modules 5 through 8 at the second meeting, and recruitment and retention procedures at the last meeting. A Head Start community coordinator also attended the first meeting. The meetings were run in Spanish; the second author, who speaks Spanish fluently, served as the moderator. Participants were paid $25 for each of the three sessions ($75 total).
At the first meeting, the moderator provided an overview of the program and stressed the need to be honest in making suggestions, to respect differences of opinion, and to maintain confidentiality of the discussion. The moderator then answered questions and distributed consent forms. Once participants had signed the consent form, the moderator distributed a 45-page document that included summaries of the PACE modules, sample exercises and worksheets completed by program participants, and sample items from scales administered to evaluate program outcomes. The packet served as a discussion guide.
All three mothers were active in the discussion and expressed much interest in the program. Their suggestions and comments converged on four major issues. One concerned the need to be sensitive to cultural differences in the use of praise and rewards with children. Rather than use verbal praise or give tangible rewards (e.g., a prize), participants indicated that Mexican parents are more likely to use physical affection as a form of praise. Participants were also concerned that praising and rewarding too much might cause children to become spoiled, and suggested that group leaders explain how much praise is too little or too much.
A second issue that led to considerable discussion concerned cultural differences in the use of physical punishment as a disciplinary practice, which is less acceptable in the U.S. than in other countries. Participants were reluctant to share their own feelings about spanking and other forms of physical punishment, but they recognized that such disciplinary practices are not uncommon among Mexicans and are not tolerated as much in the U.S. They were concerned that program participants might resist being told not to use physical punishment, but they also felt strongly that parents needed to adopt U.S. standards and eliminate this practice.
A third issue was whether Mexicans and Latinos more generally may be reluctant to set up child-centered routines. Participants suggested that several routines advocated in the English program may meet with resistance—such as making a routine of reading and playing with the child, and having the child go to bed at a regular time. This resistance may be partly cultural, as Latino parents tend to be more focused on the family as a whole than on individual members, such as children. This resistance may also reflect practical issues that are not unique to Latinos, such as the fact that many parents have to work long hours and have schedules that are not conducive to child-centered routines (e.g., getting young children to bed early).
A fourth issue, which received the most discussion, concerned the need for the program to be sensitive to difficulties Latino parents may have in becoming involved in their child's school. As in the case of child-centered routines, there are cultural and practical barriers to implementing the practices suggested in the English program. Participants emphasized that, more than their non-Latino counterparts, Latino families share a cultural norm that delegates academic learning to teachers. The adapted program will have to make clear how and why learning begins at school and ends at home (empieza en la escuela y termina en la casa). Practically, regular communications with a child's teacher and/or involvement in school activities can be difficult for parents specifically targeted by the CANNE program (although not necessarily for more acculturated Latinos), as many do not speak Spanish fluently, have demanding work schedules, and may not have finished high school, thus creating a perception that they lack sufficient education to have a positive influence on their child's education.
The moderator devoted the third meeting to discussing recruitment and retention efforts that might work best with Latino families. Participants made several concrete suggestions that stand to work with all populations targeted by the program. For example, they suggested emphasizing that the program stands to improve a child's performance in school, being persistent in trying to contact potential participants in person or by telephone, and/or sending a letter home that describes the program. One suggestion that addressed the issue of recruiting parents lacking formal education was to emphasize that the program is not a course. Specifically, participants recommended describing the program as a discussion group where one can exchange opinions and converse among friends (como una conversación entre amigos), and where there will not be materials or questions that are difficult to understand.
In parallel to the input just described, we sought additional comments on the program and suggestions from six professional consultants. Three have published extensively on parenting practices among Latino families. They had an average of 20.6 (range 7-46) peer-reviewed articles and chapters on topics relevant to parenting among Latinos and several funded projects requiring firsthand knowledge of behavioral parent training with Latino families. One resided in Southern California, one in New York City, and one in St. Louis, Missouri, thus reflecting knowledge of various Latino groups. The other three consultants have established records of community service with Latinos from the target population. They had spent an average of 11 years (range 4-22) working with Latino children or as advocates for Latino families, primarily with Mexican families in the Midwest. All six consultants spoke Spanish fluently and reviewed portions of the program in Spanish.
We sent each consultant two or three of the eight program modules, along with a program overview and a series of questions to guide their written feedback. Specifically, for each module, they were asked to comment on how the program would work and on which issues might be problematic for the target population; on how problematic program components may need to be modified; and on which program components might be controversial and/or met with resistance.
All six reviewers had suggestions, many of which overlapped with suggestions raised by members of the target population. The first one concerned the need for the program to reflect the strong value Latinos place on family ties, in contrast with the emphasis the original program puts on child autonomy and assertiveness. This led us to tailor several examples accordingly. For instance, whereas the original program suggests that parents reward young children for dressing themselves without assistance, the Spanish program suggests that parents reward children for helping siblings or other family members.
A second suggestion was to train group leaders to recognize the value placed on the presence of a clear hierarchy in Latino families. Consultants suggested that Latino parents frequently establish themselves as authority figures within the family and place a high priority on having children show respect (respeto) toward them and other elders and authority figures. Maintenance of a clear family hierarchy has several implications for the program. First, as target group members also indicated, consultants warned that some Latino parents might resist the idea of providing too much praise and rewards to children out of concern that it might undermine this hierarchy. Second, Latino parents may be reluctant to allow children to direct play activities when they are together, or to explain commands or rules that they expect children to follow without question (again, under the dictate of respeto). In each case, we incorporated these cultural considerations by adding specific discussion points on respect and authority, and by describing these issues in relevant sections of the group leader CANNE manual. For example, training guidelines and specific modules have been modified to emphasize bicultural competence. Group leaders are trained to anticipate topics that might elicit resistance because of cultural or practical differences, and to suggest ways of incorporating new ideas without displacing existing values. One exception to this general approach was in the use of physical punishment as a means of discipline. Given the higher acceptance of corporal punishment among many Latinos, the program was modified to include explicit statements (with rationales) against this practice as well as a warning that, if abusive, physical punishment can have major legal implications in the U.S.
A corollary of holding the value that one should respect authority figures is that Latino program participants might be motivated to please group leaders that they perceive as authorities, and present themselves well. This increases the likelihood that participants might verbalize understanding and agreement with issues even when they misunderstand or disagree. To address this concern, we added role-play activities to several modules, in which group leaders coach parents in the practices advocated by the program and give opportunities to all participants to comment on what is being role-played.
Another issue that received much attention from our consultants was that many parents targeted by CANNE may not be functionally literate. This has implications for several program activities, such as summarizing take-home points in writing, completing worksheets during sessions, and implementing the recommendation to read to young children regularly. We added text to the manual to make group leaders sensitive to this issue and to offer practical suggestions to address it. For example, parents might complete worksheets by thinking about their responses instead of writing them down and might read to their children by telling stories from picture books.
Consultants also stressed that many target parents might live with extended family. This creates a challenge if the disciplinary practices advocated by the program differ from those used by other members of the household (e.g., grandparents, other relatives). Consultants suggested encouraging all members who have childcare duties to participate in the program. Although already in place in PACE, we underscored the suggestion in the CANNE manual to make it salient. Other issues mentioned by consultants were also raised by the focus group mothers who provided input, namely difficulties in maintaining child-centered routines and challenges in getting more involved in school. In each case, we revised the manual to alert group leaders and added discussion points in relevant modules.
The eight PACE sessions or modules cover the following topics:
As the structure of the program was not called into question in the consultation process, we retained it for the Spanish adaptation. However, we reviewed each module systematically and adapted its contents whenever necessary, as the examples given above illustrate. To keep track of the changes we made, we developed a set of adaptation fidelity procedures. These enabled us to describe each change and where it occurred, and thus to quantify the extent to which the new program corresponds to the original. These procedures are based on the methodology used to assess group leaders' adherence to protocol in PACE. Specifically, each PACE module has a checking form detailing its content. When assessing protocol adherence, group leaders are audiotaped during sessions (via directional microphones that capture their voice only). Trained coders later listen to the tapes and complete these forms to monitor the extent to which group leaders implement the program with fidelity. For example, in the first half of Module 2, leaders must cover five topics: (1) Welcome and discussion of previous week's topic and home activities; (2) Why does everybody need limits, even adults? (3) Rules and limits participants enforce already; (4) Rules and limits participants would like to enforce; (5) Ineffective limit setting.
To assess adaptation fidelity, two trained coders reviewed the manual after it had been modified based on the community and expert input described above. Working independently, each coder completed purpose-made fidelity forms to indicate whether each of the topics covered in a module remained the same in the new program or was modified (see Appendix for an example). Ratings were made on a 7-point scale (1 = no change to 7 = complete change). Whenever a topic was modified, the forms asked coders to specify whether this was mainly for cultural or practical reasons (see Matos et al., 2006). Cultural adaptations were defined as changes made to provide a different “frame” or rationale for a parenting practice, or to reflect differences between Latinos and non-Latinos in values, academic goals or disciplinary practices. Practical adaptations were defined as changes made to increase the relevance and applicability of the information and guidance the new program provides. For example, CANNE provides information about agencies and organizations that offer services in Spanish, whereas PACE offer similar information but of relevance to English-speaking families. Although some adaptations could be defined as either cultural or practical, coders showed good agreement in their judgments. Specifically, their interrater reliabilities were high, ranging from 0.77 to 1.00 (M = 0.89; p<.001) for cultural adaptations and 0.80 to 1.00 (M = 0.92; p<.001) for practical adaptations.
Translation of the PACE manual and questionnaires into Spanish relied on procedures that have become standard practice in the adaptation of a product in another language (e.g., http://www.mapi-research.fr/). An initial translation of the PACE manual was conducted by two bilingual, native Spanish speakers. One translated each module and the other reviewed the translated material with the aim of improving clarity and checking for appropriate word usage. This method was designed to insure that the initial translation did not reflect idiosyncratic language tendencies of one translator and to produce the new manual in a language and style that are meaningful and relevant to families, and clear and easy to follow for group leaders.
A pilot study of the efficacy of the new program is ongoing at the time of writing. The study will rely on three measures of parental and two measures of child adjustment, all available in English and Spanish. These measures will be completed by parents before the program begins, at program completion, and at a 3-month follow up:
The Revised Acculturation Rating Scale for Mexican Americans (ARSMA-II; Cuéllar, Arnold, & Maldonado, 1995) is a 20-item, Likert scale that evaluates language familiarity and use, ethnic interaction, ethnic identity and pride, and generational ties to Mexico. This well-validated measure is internally consistent and its factor structure corresponds well with that of the original instrument. The Parenting Scale (Arnold, O'Leary, Wolff, & Acker, 1993) is a 30-item scale on which respondents describe their approach to discipline on a continuum of polar opposites (e.g., “When my child misbehaves: I do something right away to I do something about it later”). The scale yields three factorially derived dimensions of effective-ineffective parenting that are internally consistent and stable, and that correlate with observational measures of parenting in discipline encounters. The Parenting Practices Interview (Webster-Stratton, 1998) is an adaptation of the Oregon Social Learning Center's Discipline Questionnaire for use with parents of preschoolers. Its scales, which focus on discipline (effective and harsh) and emotional support and reinforcement, have adequate internal consistency and are sensitive to intervention effects when used with low-income parents of preschoolers from diverse ethnic backgrounds, including Latino groups (Brotman et al., 2005; Webster-Stratton, 1998).
Coping-competence and emotional and behavioral problems will be assessed with two measures. The Social Competence and Behavior Evaluation Scale, Short Form (SCBE-SF; LaFreniere & Dumas, 1996) is a 30-item Likert scale of social competence, anxiety-withdrawal, and anger-aggression in preschoolers. The Spanish version has satisfactory internal consistency and test-retest reliability, and a factor structure corresponding closely to the original measure (Dumas, Martinez, & LaFreniere, 1998; Dumas, Martinez, LaFreniere, & Dolz, 1998). The Behavior Assessment System for Children, 2nd ed.– Preschool (BASC-2; Reynolds & Kamphaus, 2005) is a 134-item Likert scale measuring social competence, and internalizing and externalizing behavior problems. Widely used, the Spanish and English versions have satisfactory internal consistency and test-retest reliability, and yield comparable pictures of child functioning.
To determine the appropriateness of these measures before we administer them, we conducted a cross-language comparison study in which we asked bilingual speakers to complete them in both languages. Results are presented in the next section.
In keeping with our primary aim to make explicit the adaptation of the PACE program, we provide data pertaining directly to the adaptation process. Specifically we describe how closely the adapted program matches the original, and compare the original (English) and adapted (Spanish) version of our measures. We also report preliminary data on parental engagement in the new program and on the program's social validity from our ongoing pilot work.
Results showed that, on average, only slight changes were made to each module in the adaptation process (range = 1.05 to 2.43; M = 1.77; SD = 1.28). Cultural adaptations were somewhat more extensive (M = 2.03; SD = 1.49) than practical adaptations (M = 1.77; SD = 1.07). These averages hide important variations in the magnitude of the changes made. Typical adaptations involved relatively minor modifications, such as providing examples that might resonate with Latino parents, as suggested by input from consultants and community members. In a few cases, adaptations were substantial (e.g., adding a section to explain the importance of praising children for what they are expected to do). The most significant cultural changes took place in Modules 2 (Setting clear limits for children), 3 (Helping children behave well at home and beyond), and 7 (Helping our children do well at school). The most significant practical changes occurring in Modules 7 and 8 (Anticipating challenges and seeking support).
To compare the English and Spanish versions of the measures, we recruited bilingual participants through local community centers (i.e., YWCA, Latino Cultural Center, and International Center), elementary schools, and university courses. Participants were asked to complete the Spanish or the English version (randomly assigned) during a first testing session and to complete the other version approximately one week later, during a second session. Participants were compensated $15 for each session. Fifty-two individuals completed one version of the measures and 43 completed both. The 43 individuals who completed both versions consisted of 32 females and 11 males; 77% were either parents or grandparents of young children, and the other 23% were teachers. They ranged in age from 21 to 72 years (M = 38.77, SD = 11.65) and reported an average of 15.76 years of education (SD = 1.9). Eighty-four percent were employed. Sixty-seven percent described their ethnicity as Hispanic, 28% as African American, and 5% as Other. Seventy-two percent were married or living with an adult partner and 28% were single. Boys were represented in comparable proportions to girls (48% vs. 52%) among the participants' children. Table 1 provides two statistics that compare the English and Spanish versions: Cronbach alphas assess internal consistency within version and Pearson rs report correlations between versions. Overall, these results suggest that the measures yield comparable data, whether they are administered in one language or the other.
A pilot study evaluating parental engagement in CANNE and acceptability of the new program is ongoing. Data show that, of the parents who expressed an interest in the program by completing a pre-intervention survey, 61% went on to enroll in CANNE in Fall 2008 and 76% in Spring 2009. Attendance in the Fall groups has been very encouraging and augurs well for attendance in the Spring groups, which are under way at the time of writing. Whereas 20% of parents who enrolled in the Fall groups attended 4 sessions or less (out of 8), 28% attended 5 or 6 sessions, and 52% attended 7 or all 8 sessions. This shows that more than half of the participants received 87% or more of the intended program “dose” and that 80% of the participants received 62% of that “dose.”
Parents provide social validity ratings of the program during Sessions 4 and 8. To date, results show that CANNE participants are highly satisfied, as average ratings in all areas ranged from 4.71 to 5.00 (SD = 0.00 to 0.56) on a 5-point Likert scale (1 = strongly disagree to 5 = strongly agree). Specifically, parents reported that group leaders respected their values and opinions and talked to them in ways they understood; that goals of the program were important to them; that the program fit their community and that they could use what they had learned; and that their family gets along better since they began the program. In keeping with these positive ratings, all parents reported that they would recommend CANNE to other parents like themselves (M = 5.00, SD = 0.00).
Rapid increases in the U.S. Latino population in recent years make efforts to deliver evidence-based interventions pressing in many health areas. The project described here represents the first step in an attempt to adapt a behavioral parent training program for Latino parents and, thus, to contribute in a modest way to the well-being of families in a growing sector of the population. This adaptation has confronted us with a major quandary facing public health efforts in the U.S. generally, and cognitive and behavioral practice more specifically: how to provide access to quality interventions to groups affected by major health, but without tailoring those interventions to every group that might benefit from them and straying away from what is known to “work.” To strike a balance between the need to be responsive to the concerns and values of Latino parents, and to remain faithful to an intervention that has been shown to be efficacious, we decided before we adapted the PACE program: (1) to consult widely with members of the target population as well as local and nationally recognized experts to guide the changes we would make to the Spanish version, but to frame all of those changes within the coping-competence model that provides the theoretical rationale for the original program; and (2) to center the new program on the promotion of competence in both Latino and U.S. cultures, in order to help target participants promote their own and their children's ability to navigate effectively within and across these cultures. This focus on bicultural effectiveness is found throughout the Spanish PACE program, both in the specific instructions that guide group leader training and in the many examples given to parents in every module.
The adaptation fidelity procedures we developed to track the changes we made show that the Spanish program about to be piloted is much more similar to its English counterpart than it is dissimilar. Results of our quantitative analysis indicate that, on average, cultural and practical changes were of the order of 2 on a 7-point scale ranging from (1) no change to (7) complete change. This suggests that the parenting challenges PACE is designed to address cut across target groups and that, with some important differences in priorities or emphases, Latino parents share many concerns with past recipients of the program.
More importantly, although the sections on bicultural competence added to each module for group leader training address cultural issues and values that are central to Latinos targeted by CANNE, much in these sections is applicable to any culturally sensitive behavioral approach to parent training. This has led the authors to appreciate that what started as an adaptation has progressively become a cultural broadening that stands to benefit different target groups. This broadening, which has proved to be much more stimulating than what we had initially conceived the task to be, is already feeding back to the original PACE intervention and strengthening it in an iterative process. In other words, if the new program stands to benefit from the rich feedback we received as we developed it, so does its original counterpart, both with respect to contents and group leader training.
We are encouraged by our efforts to date, especially by the finding that numerous parents have enrolled in CANNE and, in a majority of cases, attended sessions regularly, and by the fact that they find the program relevant and helpful. However, we are acutely aware that CANNE will have to be tested before any conclusion about its efficacy can be drawn. Our ongoing pilot study asks whether the new intervention matches what is at stake for the target population and yields promising findings about its short-term efficacy. We are confident that the measures we plan to use will provide comparable data when administered in Spanish as they do when administered in English. Results of the cross-language comparison we conducted to prepare for the pilot study show that these measures are internally consistent in both languages and that their English and Spanish versions yield scores that are highly correlated. However, to evaluate short-term outcomes, the pilot study will not rely on standardized measures only. It will also get testimonials from parents and group leaders in order to obtain qualitative feedback that may enable us to make final changes to the program before conducting an efficacy study.
To conclude, we see the results of the work described here essentially as a roadmap to address three questions:
The study was supported by grants R21MH077680 from the National Institute of Mental Health (NIMH) and R49/CCR 522339 from the Centers for Disease Control and Prevention (CDC).
Has the program been changed from its original English version?
1 = Not at all, 2 = Very slightly, 3 = Slightly, 4 = Somewhat, 5 = Moderately, 6 = Very Much, 7 = Extremely
|Cultural Change||Practical Change|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||Welcome and discussion of previous week's topic and home activities|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||II. Why does everybody need limits, even adults?|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||III. Rules and limits participants enforce already|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||IV. Rules and limits participants would like to enforce (or enforce better)|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||V. Ineffective limit setting|
|VI. Practical ways to set limits|
|Using effective commands to set limits|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||What makes commands effective?|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||Practicing giving effective commands|
|Using effective consequences to set limits|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||Natural consequences|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||Logical consequences|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||“When-then” statements|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||Applying consequences to a limit-setting plan|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||VII. Home activities|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||VIII. Summary: Take Home Points|
|1 2 3 4 5 6 7||1 2 3 4 5 6 7||IX. Closing|
An earlier version of the research described here was presented in March 2008 at the Center for Latino Family Research conference (Luis H. Zayas, Chair) on Developing Interventions for Latino Children, Youth, and Families, Washington University. St. Louis, Missouri (R13 MH077403: PI: Luis H. Zayas).
Jean E. Dumas, University of Geneva.
Ximena Arriaga, Purdue University.
Angela Moreland Begle, Medical University of South Carolina.
Zayra Longoria, Purdue University.