In recent decades there has been increased interest in programs and psychological interventions based on mindfulness (Eberth & Sedlmeier, Reference Eberth and Sedlmeier2012). A search of the PsycInfo database using the keyword “mindfulness” gives us a total of 485 items – if the search is restricted to the period 1965 (the year the first indexed article appears in PsycInfo with this keyword) to 2004. However, the same search between January 2005 and December 2010 provides us with 1,842 items. One reason for this dramatic increase in scientific production related to mindfulness is the effect psychological techniques based on mindfulness have had on a variety of psychological variables, such as brain activity, in both clinical and non-clinical contexts. Thus, for example, in the clinical context, a meta-analysis by Grossman, Niemann, Schmidt and Walach (Reference Grossman, Niemann, Schmidt and Walach2004) found mean weighted effect sizes of .50 for mental health problems (depression, anxiety, sleep disorders etc.) and .42 for physical health problems (medical symptoms, physical pain or physical disability). In the non-clinical population, Sedlmeier et al. (Reference Sedlmeier, Eberth, Schwarz, Zimmermann, Haarig, Jaeger and Kunze2012) found a mean weighted effect size of .28. In another similar study, also conducted on a non-clinical population, Eberth and Sedlmeier (Reference Eberth and Sedlmeier2012) found mean weighted effect sizes ranging from .21 to .40. The categories of the dependent variables evaluated in these studies include aspects as diverse as anxiety, attention, memory, regulating emotion, verbal fluency, emotional stability, self-fulfillment and personal well-being.
With respect to changes in brain activity, studies incorporating neuroimaging techniques by means of functional Magnetic Resonance Imaging (MRI) show that the practice of mindfulness produces greater activation of brain regions involved in sustained attention, response inhibition and attention to distracting sounds and less brain activation in regions related to emotions and discursive thought (Brefczynski-Lewis, Lutz, Schaefer, Levinson, & Davidson, Reference Brefczynski-Lewis, Lutz, Schaefer, Levinson and Davidson2007), as well as changes in grey matter concentration in brain regions related to learning, regulating emotions and self-referential processing (Hölzel et al., Reference Hölzel, Carmody, Vangel, Congleton, Yerramsetti, Gard and Lazar2011). The correlation between these changes and the cumulative amount of hours of meditation suggests a possible plasticity of such mechanisms (Brefczynski-Lewis et al., Reference Brefczynski-Lewis, Lutz, Schaefer, Levinson and Davidson2007). It has also been observed that mental training in cultivating positive emotions through compassion meditation alters the activation of circuits related to empathy and theory of mind in response to stimulation. That is, it improves empathic responses to social stimuli (Lutz, Brefczynski-Lewis, Johnstone, & Davidson, Reference Lutz, Brefczynski-Lewis, Johnstone and Davidson2008). In addition to this, meditation increases the activity of the Anterior Cingulate Cortex (ACC) involved in the development of the self (Tang et al., Reference Tang, Lu, Geng, Stein, Yang and Posner2010) and has a positive effect on the neuro-endocrine and immune system in the face of stressful situations (Davidson et al., Reference Davidson, Kabat-Zinn, Schumacher, Rosenkranz, Muller, Santorelli and Sheridan2003, Pace et al., Reference Pace, Negi, Adame, Cole, Sivilli and Raison2009). Studies conducted with children and adolescents show similar results to those observed in adults, including reduced anxiety, improved care, reduced behavioral problems, etc. (Meiklejohn et al., Reference Meiklejohn, Phillips, Freedma, Griffin, Biegel and Saltzman2012). In addition, mindfulness-based interventions in the school context favour increased cognitive ability in paying attention and learning, coping and resilience (Zenner, Herrnleben-Kurz, & Walach, Reference Zenner, Herrnleben-Kurz and Walach2014).
Mindfulness or “full attention” is defined as a state of consciousness that involves paying conscious attention to experience, moment by moment (Brown & Ryan, Reference Brown and Ryan2003). It involves focusing one’s attention in a particular way, with intention, on the present moment, and without judgment (Kabat-Zinn, Reference Kabat-Zinn1994). This definition implies the presence of three axioms: a purpose or intention, paying attention and an attitude (Shapiro, Carlson, Astin, & Freedman, Reference Shapiro, Carlson, Astin and Freedman2006). The practice of mindfulness has its origins in Eastern contemplative traditions, being most explicitly and systematically articulated and developed within the Buddhist tradition, which dates back over 2,500 years (Kabat-Zinn, Reference Kabat-Zinn2003). From a psychological point of view, it has also come to be regarded as a personality construct (Vallejo, Reference Vallejo2006). In this respect, when the practice of mindfulness, which results in a transient state of mindfulness, is maintained repeatedly over time it can lead to a trait, or dispositional mindfulness (Garland, Gaylord, & Fredrickson, Reference Garland, Gaylord and Fredrickson2011). However, as noted by Barnhoher, Duggan, and Griffith (Reference Barnhofer, Duggan and Griffith2011), people may also differ in their natural tendency to be aware from moment to moment in an open and non-judgmental way. These levels of mindfulness, in the absence of training, remain relatively stable over time (Baer, Smith, & Allen, Reference Baer, Smith and Allen2004), and can therefore be regarded as dispositional mindfulness skills. We understand then, that dispositional mindfulness refers to the skills of mindfulness, both natural and those acquired through practice, that remain stable over time.
While there are several instruments for measuring the construct of mindfulness in adults, until the appearance of the CAMM, developed by Greco, Baer and Smith (Reference Greco, Baer and Smith2011), no instrument had been specifically designed to measure mindfulness in children and adolescents. This type of tool is essential in understanding the nature of mindfulness and its components (Baer, Smith, Hopkins, Krietemeyer, & Toney, Reference Baer, Smith, Hopkins, Krietemeyer and Toney2006, Dimidjan & Linehan, Reference Dimidjian and Linehan2003), conducting empirical research into how mindfulness influences health (Bergoni, Tschacher, & Kupper, Reference Bergoni, Tschacher and Kupper2013), reducing symptoms and increasing well-being (Baer, Reference Baer2007) and understanding its relationship with other psychological constructs (Baer et al., Reference Baer, Smith, Hopkins, Krietemeyer and Toney2006).
The CAMM was developed using the Kentucky Inventory of Mindfulness Skills (KIMS, Baer et al., Reference Baer, Smith and Allen2004). Of the 4 dimensions (observation, conscious action, acceptance and description) evaluated by the KIMS, only the first three were used. The fourth, which relates to the ability to put internal experiences into words (Baer et al., Reference Baer, Smith and Allen2004), is not included due to the probable impact of the participants’ developmental level on their responses to such items (Greco et al., Reference Greco, Baer and Smith2011). Cognitive and verbal abilities vary widely among young people and continue to evolve during adolescence. Therefore, as Greco et al. (Reference Greco, Baer and Smith2011) point out, items asking about the ability to label or covertly apply words to internal phenomena are likely confounded with the current level of verbal– cognitive abilities and language development. This gave rise to a first version of 25 items, which were eventually reduced to 10 evaluating a single factor. Items on the CAMM are reverse scored and reflect present-moment awareness and nonjudgmental acceptance of thoughts and feelings.
The aim of this study is to translate the CAMM and adapt it to Catalan so as to analyze its internal consistency and construct and convergent validity using measures of well-being, self-esteem and effortful control on a sample of secondary school students who had not followed a prior training programe in the practice of mindfulness.
In accordance with the empirical evidence, the Catalan version of the CAMM is expected to correlate positively with subjective well-being (Eberth & Seldmeier, Reference Eberth and Sedlmeier2012, Greco et al., Reference Greco, Baer and Smith2011), self-esteem (Brown & Ryan, Reference Brown and Ryan2003, Thompson & Waltz, Reference Thompson and Waltz2008) and effortful control (Eberth & Sedlmeier, Reference Eberth and Sedlmeier2012) and no differences are expected to be observed by sex or age (Cunha, Galhardo, & Pinto-Gouveia, Reference Cunha, Galhardo and Pinto-Gouveia2013, Greco et al., Reference Greco, Baer and Smith2011).
Method
Participants
This study involved 696 Secondary Education students, aged 11 to 16, of whom 56.5% were girls (see Table 1). Students were randomly selected by means of random cluster sampling, in which the classroom was the cluster, stratified by schools (state and subsidized) and academic years. A total of 12 schools located in the province of Girona (Catalonia, north-east Spain) took part, all of them with students from mixed socioeconomic backgrounds.
Table 1. Sociodemographic characteristics of the sample
The sample was randomly divided into two subsamples of 348 participants each. The first comprised 149 boys and 199 girls with a mean age of 12.96 years (Sd = 1.09) and the second 154 boys and 194 girls with a mean age of 12.91 years (Sd = .98). The two subsamples are similar in age, t(694) = .693, ns, and sex, χ2(1, N = 696) = .15, ns.
Instruments
Child and Adolescent Mindfulness Measure (CAMM) (Greco et al., Reference Greco, Baer and Smith2011). Self-reported questionnaire consisting of 10 items, rated on a Likert scale of 0–4 points, measuring mindfulness skills in children and adolescents. The authors found an internal consistency of .81 (Cronbach’s alpha) and correlations with quality of life, social skills and academic performance. To obtain the total score for the scale, values should be inverted and all items summed together. The questionnaire was translated into Catalan and then back-translated to assess the degree of semantic and conceptual equivalence of the two versions.
Early Adolescent Temperament Questionnaire (EATQ-R) (Ellis & Rothbart, Reference Ellis and Rothbart2001). Self-reported questionnaire consisting of 64 items, rated on a 5-point Likert scale, which assesses 10 temperamental and two behavioral dimensions. Of the 10 dimensions, only the three relating to effortful control were adopted: activity control, attention and inhibitory control. Scores for the EATQ-R dimensions represent the mean of all items for each dimension. Compared with the original version, the Catalan version used in this study presents minor differences in the structure of temperament (Ellis, González, & Viñas, Reference Ellis, González and Viñas2009). The internal consistency found for each of the assessed dimensions is as follows: activity control (α = .65), attention (α = .55) and inhibitory control (α = .34). Internal consistency for the Effortful Control scale in this study is adequate (α = .70). In studies carried out on adult populations, it has been observed that the dimensions of inhibitory control and attention predict enhanced performance on the Stroop color interference task (Yamagata, Takahashi, Shigemasu, Ono, & Kijima, Reference Yamagata, Takahashi, Shigemasu, Ono and Kijima2005).
Personal Wellbeing Index (PWI; Cummins, Eckersley, van Pallant, Vugt, & Misajon, Reference Cummins, Eckersley, Pallant, van Vugt and Misajon2003). Participants were asked to what extent they felt satisfied with different aspects of their lives included in the Personal Wellbeing Index (PWI). The PWI comprises 7 items relating to satisfaction with the following life domains: standard of living, health, life achievements, relationships with other people, present security, relationships with the community and future security. Items are scored on a 0 (Completely dissatisfied) to 10 (Completely satisfied) scale, in accordance with Cummins and Gullone’s (Reference Cummins and Gullone2000) recommendations. The final score is obtained by calculating the mean score for items and adjusting this to a scale of 0 to 100. The authors report good psychometric properties for the instrument (Cummins et al., Reference Cummins, Eckersley, Lo, Okerstrom, Hunter and Davern2004). The translation of the original version of the domains included in the PWI into Catalan and its back-translation from Catalan to English was the work of Casas et al. (Reference Casas, Coenders, Cummins, González, Figuer and Malo2008). Internal consistency of the PWI in the present study was good (α = .81).
Multidimensional Self-Concept Scale, AF-5 (García & Musitu, Reference García and Musitu2001). 30-item scale that evaluates five dimensions related to self-concept within the following contexts: social (perception of own sociability, such as the ability to make and keep friends), academic (perception of own ability to work, speed and efficiency in the workplace, liking school, image with teachers and similar other aspects), emotional (perception of own emotional stability and self-control), family (perception of how your family sees you and ability to control oneself in situations where you have a lot of confidence) and physical (perception of own appearance and physical condition). It consists of 6 items for each dimension and is designed for ages 10–16. We used the Catalan version created by Malo, Bataller, Casas, Gras, and González (Reference Malo, Bataller, Casas, Gras and González2011), whose psychometric properties are very good and similar to the original version. The internal consistency found in this study ranges from .64 (social) to .90 (academic). With the exception of the social dimension, all dimensions have a Cronbach’s alpha greater than .77.
Procedure
After requesting permission from the authors, the English to Catalan translation was done by the research team. The CAMM was then applied to a small group of pupils to check their level of understanding of each item. The Catalan version was back-translated into English by an independent, native English professional translator with a good knowledge of Catalan. The back-translated version was then compared to the original, assessing the degree of conceptual and semantic equivalence of the two versions in English, and a satisfactory level of equivalence was found. After translating the CAMM into Catalan, it was administered to the sample. After the corresponding permission was requested from school heads and the Autonomous Government of Catalonia’s Department of Education, participants were informed of the general aims of the research and that the confidentiality of their data was guaranteed. The protocol for administering the questionnaire was submitted to schools and parents for their approval. Once consent had been obtained, questionnaires were administered to groups of students in the classroom during class time. After reading the items on the questionnaire carefully, all children received standardized instructions on how to respond to them. The adolescents were accompanied by researchers during the questionnaire administration process in case they needed help or clarification.
Data analysis
The statistical analysis was performed using version 19.0 of the SPSS and version 18.0 of AMOS. With a first half of the sample (n = 348), which was randomly selected, an exploratory factor analysis (EFA) was conducted using the principal axis extraction method. To extract the number of factors, we applied Kaiser’s (Reference Kaiser1958) criteria (eigenvalues greater than 1), a sedimentation test (Cattell’s Scree-test, Reference Cattell1966) and parallel analysis (Horn, Reference Horn1965) using O’Connor’s syntax program for SPSS (Reference O’Connor2000). We calculated Cronbach’s alpha coefficients to analyze the internal consistency of the factors. Confirmatory factor analysis (CFA) was performed on the second half of the sample (n = 348), using the method for estimating maximum likelihood (ML). To test the fit of the model, we used the χ² index and its associated level of likelihood, the ratio χ²/df (degrees of freedom) and the following indices: CFI (Comparative Fit Index), GFI (Goodness of Fit Index), AGFI (Adjusted Goodness of Fit Index) and RMSEA (Root Mean Square Error of Approximation). In line with the objective of the study, we calculated the internal consistency of the CAMM and analyzed CAMM scores by age and sex. Of the 696 adolescents initially evaluated, it was possible to make contact with 531 of them after one year to obtain a second CAMM score. In addition, linear correlations were calculated between the CAMM scores and the scores for the PWI, AF-5 and the EATQ-R effortful control scale.
The minimum required level of significance in all statistical tests was p < .05.
Results
Structural validity
Exploratory Factor Analysis (EFA)
Both the Kaiser-Meyer-Olkin measure of sampling adequacy, with a value of .849, and Bartlett’s test of sphericity indicate the suitability of the data for the EFA, χ2(45) = 806.78, p < .001. While the Kaiser criterion extracted two factors, the Scree plot and parallel analysis point to extraction of a single factor (see Figure 1). This single factor explains 36.95% of the variance.
Table 2 shows the saturation matrix with the weight of each of the items. With the exception of item 5, whose weight on the scale is .25, the other items obtain weights ranging from .53 to .77. Internal consistency, Cronbach’s alpha, is .80. Corrected correlations range from .44 (item 10) to .65 (item 8), except for item 5, whose corrected correlation is .20. If said item is removed, the internal consistency remains virtually unchanged (α = .81).
Figure 1. Scree plot and parallel analysis.
Table 2. Factor structure of the CAMM (Catalan Version)
Confirmatory Factor Analysis (CFA)
Table 3 shows the goodness of fit indices for four models. In the first and second models, all 10 items are considered, with and without error covariances, respectively, and in the third and fourth without item 5, with and without error covariances, respectively. Of the four models, the one with the best fit is model 2 (see Figure 2).
Figure 2. Confirmatory factor analysis for the 10-item CAMM.
Table 3. Fit indices for factores of the CAMM
Model 1: 10 items without error covariances
Model 2: 10 items with error covariances
Model 3: 9 items without error covariances
Model 4: 9 items with error covariances
Convergent validity
The Mindfulness scale correlated positively with the PWI (r = .23, p < .001), activation control (r = .35, p < .001), attention (r = .36, p < .001) and inhibitory control (r = .37, p < .001), and with the overall score for effortful control (.46). It also correlated positively with PWI scores and the dimensions of social, family, academic and emotional self-concept on the AF-5 (see Table 4).
Table 4. CAMM correlations with measures of validity
Note: **p < .001.
Reliability
Internal consistency, Cronbach’s alpha, has a value of .80 and the test-retest correlation, one year later, a value of .47 (n = 531). The second time the CAMM was administered, the Cronbach’s alpha had the same value.
Descriptive data
Table 5 gives the mean values for the CAMM according to sociodemographic variables. No differences were observed by sex, t(694) = –.93, ns, age, F(5, 690) = .81, ns, or school type, t(694) = .72, ns. Also, when comparing the two occasions when the CAMM was administered, a year apart, no differences were detected between the two means, t(530) = –1.01, ns.
Table 5. Means and standard deviation (in parentheses) of the CAMM scores by sex, age and school type
Discussion
The results obtained using the Catalan version are very similar to those obtained by Greco et al. (Reference Greco, Baer and Smith2011) with the original version and Portuguese versions (Cunha, Galhardo, & Pinto-Gouveia, Reference Cunha, Galhardo and Pinto-Gouveia2013), both in terms of psychometric properties (internal consistency and construct validity) and the distribution of scores when considering sociodemographic characteristics. Also, the average levels of mindfulness found in the sample are almost identical to those observed by Greco et al. (Reference Greco, Baer and Smith2011) and Cunha et al. (Reference Cunha, Galhardo and Pinto-Gouveia2013). Confirmatory factor analysis indicates the existence of one single factor, although one of the items, namely item 5, does not perform very well given that its standardized loadings are very low. A similar result was observed for this item in the Portuguese version of the CAMM (Cunha et al., Reference Cunha, Galhardo and Pinto-Gouveia2013). However, when said item is retained, goodness of fit indices indicate a good fit of the model to the data, meaning its removal is not necessary.
It should be noted that the CAMM was designed for the same purpose as other tools developed for adults, namely, to measure the impact of programs and interventions based on the practice of mindfulness. From this perspective, the mindfulness skills evaluated by the CAMM are the result of training based on the practice of meditation. As previously mentioned, however, these skills may be present in the general population to a greater or lesser degree as a trait or disposition. In the present study, we have evaluated the latter aspect, as the adolescents who participated in the study had not followed a prior training program in the practice of mindfulness. The correlation between the two occasions on which the CAMM was administered, a year apart, indicates the relative stability of mindfulness skills over time and consequently its dispositional nature. Compared to the average correlation coefficient when analyzing the stability of personality traits, which stands at around .58 (Bermudez, Reference Bermúdez, Bermúdez, Pérez-García, Ruiz, Sanjuán and Rueda2011, Costa & McCrae, Reference Costa, McCrae, Hogan, Johnson and Briggs1997), the correlation obtained in this study is below average, but within normal intervals (between .34 and .77), and practically identical (.46) to that obtained in the Portuguese version with a one month interval (Cunha et al., Reference Cunha, Galhardo and Pinto-Gouveia2013).
Various studies have shown a relationship between mindfulness and attention or the set of cognitive processes grouped under the term executive functions (see, for example, Chambers, Lo, & Allen, Reference Chambers, Lo and Allen2008, or Eberth & Sedlmeier, Reference Eberth and Sedlmeier2012). Consequently, and as one would expect, the CAMM correlated with the Effortful Control dimension, which evaluates aspects related to the executive functions such as attention, activation control or inhibitory control (Yamagata et al., Reference Yamagata, Takahashi, Shigemasu, Ono and Kijima2005).
Furthermore, as several studies have shown (Eberth & Seldmeier, Reference Eberth and Sedlmeier2012), mindfulness is also associated with greater subjective well-being and a better self-concept. In this regard, the correlations found between the CAMM and the PWI are similar to those obtained by Greco et al. Reference Greco, Baer and Smith2011 with the Youth Quality of Life-Revised (YQOL–R; Edwards, Huebner, Connell, & Patrick, Reference Edwards, Huebner, Connell and Patrick2002; Patrick, Edwards, & Topolski, Reference Patrick, Edwards and Topolski2002).
Similarly, the observed association between mindfulness and self-esteem is consistent with results from other studies when analyzing the relationship between the two constructs (Brown & Ryan, Reference Brown and Ryan2003, Thompson & Waltz, Reference Thompson and Waltz2008). The five dimensions evaluated (academic, social, family, physical and emotional self-concept) are associated with a higher degree of mindfulness skills, although the correlation with physical self-concept is very low. It is worth highlighting the correlation with the dimension of emotional self-concept, which indicates an association between increased positive perception of one’s ability to control fear or anxiety or the perceived absence of emotional disturbance and greater mindfulness skills.
Studies conducted on the adolescent population in the context of the school classroom have the advantage of reaching a higher number of subjects, but the limitation, despite the best efforts of researchers, of not guaranteeing a reliable and valid response from subjects. However, results obtained using the Catalan version of the CAMM indicate that it has some acceptable psychometric properties, which make it a useful tool for evaluating mindfulness skills in adolescents in the Catalan context.
