Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2012 Oct 1.
Published in final edited form as: Cultur Divers Ethnic Minor Psychol. 2011 Oct;17(4):444–455. doi: 10.1037/a0025456

Assessment of Awareness of Connectedness as a Culturally-based Protective Factor for Alaska Native Youth

Nathaniel V Mohatt 1, Carlotta Ching Ting Fok 2, Rebekah Burket 3, David Henry 4, James Allen 5
PMCID: PMC3193160  NIHMSID: NIHMS324129  PMID: 21988583

Abstract

Research with Native Americans has identified connectedness as a culturally-based protective factor against substance abuse and suicide. Connectedness refers to the interrelated welfare of the individual, one’s family, one’s community, and the natural environment. We developed an 18-item quantitative assessment of awareness of connectedness and tested it with 284 Alaska Native youth. Evaluation with confirmatory factor analysis and item response theory identified a 12-item subset that functions satisfactorily in a second-order, four-factor model. The proposed Awareness of Connectedness Scale displays good convergent and discriminant validity and correlates positively with hypothesized protective factors such as reasons for living and communal mastery. The measure has utility in the study of culture-specific protective factors and as an outcomes measure for behavioral health programs with Native American youth.

Keywords: American Indian and Alaska Native youth, assessment, indigenous psychology

This paper describes the development of the Awareness of Connectedness Scale (ACS). We support the premise that scientific measurement tools for assessing risk, resiliency, and change among Native Americans can and should be based on cultural notions of disorder, wellness, and healing. Research with various Native American populations has identified a holistic sense of connectedness of the individual with their family, community, and natural environment as an important element of Native American world views (Hazel & Mohatt, 2001; Bowen, 2005; Hill, 2006;), a protective factor against substance use and suicide (Allen et al., 2006; Allen, Mohatt, Fok, Henry, & The People Awakening Team, 2009; Mohatt et al., 2004b), and an aid in recovery from substance use disorders (Gone, 2009; Mohatt et al., 2007)).

First, we review the role of culture, colonization, and community-identified cultural factors in health research among Native Americans to contextualize and frame the development of the ACS. Then we address the main scale development objectives, which include: 1) analysis of the underlying structure of awareness of connectedness as measured by the ACS, 2) investigation of the item pool characteristics to assess how each item functions, and 3) assessment of the convergent and discriminant validity of the scale. In conclusion, we discuss the utility of the instrument in terms of potential contributions to the fields of indigenous and cross-cultural psychology. We argue that assessment of awareness of connectedness is a valuable tool for research and practice, and that the identification and measurement of protective culture-based variables can help clarify the relationship between Native American cultural identification and behavioral health outcomes.

The ACS development is an example of identifying a culturally specific protective factor within the epistemology of a Native American culture that can be measured and verified. Oetting and Beauvais (1990), Dana (2000), Duran (2006), Mohatt et al. (2004a) and Gone (2009) raise culture and community as corner-stone considerations when conducting research or interventions in Native American communities, with particular attention to the diversity of cultural world views represented within this broad population. Duran (2006) further recommends transcendence of notions such as “cross-cultural” and “cultural sensitivity” by identification and validation of the epistemology or “life-world” of the people with whom we are working (p.14). Mohatt et al. (2004a) and Mohatt et al. (2007) describe how community based participatory research (CBPR) with Alaska Native (AN) peoples has lead to important insights into the resiliency and strengths of AN cultures, such as culture-specific ways of thinking about alcohol use, sobriety, and recovery from substance abuse, as well as ways of framing and understanding the relationships between the individual, community, and holistic wellness.

Culturally grounded protective factors are important for study precisely because the history of colonization has disrupted the connections to traditional values among Native American cultures. Durkheim (as cited in Hill, 2009) describes how colonization leads to a disruption of life and cultural system, which, in turn, leads to increased suicide risk. Today, Native American adolescents experience some of the highest substance abuse rates in the country (Hawkins, Cummins, & Marlatt, 2004) and suicide rates nearly twice as high as the national average (Alcantara & Gone, 2007). Hawkins et al. (2004) link the heightened risk for substance abuse in Native American youth to cultural dislocation, acculturative stress, and alienation.

Duran (2006) indicates that historical and present day trauma has ruptured the physical, mental, and spiritual relationships between people and their holistic life-world. Duran suggests this trauma manifests in behavioral health issues such as family violence, suicide, and depression as well as in dysfunction in community-based support systems. Duran emphasizes the need to construct a socio-cultural narrative that is grounded in the Native life-world. Such a narrative would rebuild awareness of connections between people and their culture, community, and life-world. We developed the ACS because it allows us to assess success at promoting and reconstructing values grounded in the Native life-world and quantify the relationship between these efforts and other behavioral health outcomes.

The ACS development took place within a larger CBPR project that seeks to identify Alaska Native (AN) pathways to sobriety and strengthen culturally-based protective factors (Allen et al., 2006). This project, The People Awakening (PA) project, has engaged AN communities for over 15 years to address substance use disorders and suicide in rural Alaska. PA grew out of a grassroots response to change the deficit-focused discourse surrounding substance use issues in Native communities (Mohatt et al., 2004a). In an approach aligned with Duran’s (2006) call for positive cultural narratives, community members sought to highlight narratives of sobriety, resilience, and wellness. People Awakening began with a qualitative study that identified AN pathways to sobriety and concluded with the development of an heuristic model describing the processes leading to sobriety (Mohatt et al., 2004b). Investigators with the Center for Alaska Native Health Research (CANHR) and the Yup’ik Regional Coordinating Council1 (YRCC) community research partner have furthered the PA work with the development of quantitative measures of PA protective factors and testable models (Allen et al., 2006; Allen, Mohatt, Fok, Henry, & Burket, 2008). The ACS was refined in Cuqyun2, the portion of PA that focused on the development of culturally-appropriate outcomes measures for use in suicide and substance use research and intervention with Yup’ik youth.

People Awakening researchers and community members identified an awareness of one’s connectedness with family, community, and natural environment as a central factor that protected individuals from alcohol problems (Hazel & Mohatt, 2001; Mohatt et al., 2004a; Mohatt et al., 2004b; Allen et al., 2006). Awareness is embodied in the Yup’ik concepts of Ellam-iinga, the eye of awareness, ellanaq, the process of becoming aware, and ellangneq, awarenesss of consequences (Hazel & Mohatt, 2001; Mohatt et al., 2004a; Mohatt et al., 2004b). Yuuyaraq, or “the way of the human being,” is a related Yup’ik concept that prompts living in harmony with the natural environment because land, water, heavens, animals, and plants are viewed as spiritual entities (Napolean, 1996, p.5). This notion of holistic connectedness and interdependence is also present in many other Native American cultural epistemologies. While PA identified and employed the Yup’ik terms for awareness, they identified a very similar idea across all Alaska Native (AN) cultural groups (Mohatt et al., 2004b). Similar also, is the Lakota Sioux people’s notion of relatedness embodied in the term mitakuye oyas’in, which means “all of my relatives.” This word, which is central to Lakota spirituality and is used to close every prayer, refers to everything that is, has been, or ever will be created (Bowen, 2005). These concepts suggest sense of connectedness as a prosocial value associated with Native American cultural worldviews.

Many authors have more broadly identified cultural identity3 and enculturation4 as protective factors for Native youth (Oetting & Beauvais, 1990; LaFromboise, Hoyt, Oliver, & Whitbeck, 2006; Whitbeck, McMorris, Hoyt, Stubben, & LaFramboise, 2002). However, Hawkins et al. (2004) reported that cultural identity and engagement research among Native youth has produced mixed results in terms of identifying engagement in traditional cultural practices as protective against substance use. Chandler, Lalonde, and Sokol (2003) demonstrated how cultural continuity, the degree to which a culture preserves a sense of identity and meaningfulness from the past through to the present and a foreseeable future, is inversely related to suicide rates in Native American communities. Alcantara and Gone (2007) reported mixed results between connection with cultural practices and suicidality, but concluded that cultural continuity remains a useful construct in understanding Native youth suicide. Bates, Beauvais, and Trimble (1997) were unable to identify a direct relationship between Native cultural identification and substance use, but pointed out that culture communicates values, beliefs, and norms, whether positive or negative. They recommended that substance use prevention programs for Native American youth focus on building youth’s relationship to cultural values and traditions that promote positive behavior.

Identity and enculturation encompass cultural values and participation without distinguishing the values and worldviews endorsed by the culture that are most salient in health, resiliency, and well-being. Research with specific cultural components such as awareness of connectedness should yield clearer results. We developed the ACS to measure youth’s awareness of a culturally defined experience of connectedness with family, community, and nature in response to the identification of a sense of connectedness as both a central factor in a cultural theory of protection among AN cultures and as a potential bridge between theories of cultural identification and substance use outcomes. Accordingly, the objectives of this study were to (1) test the multidimensional nature of connectedness as measured by the ACS and its fit with our theory of the construct from which we developed the measure, (2) investigate the item characteristics functioning of the ACS item pool, and (3) assess the evidence for validity of ACS score interpretations.

Based on the PA research and other literature on the protective nature of a sense of connectedness, we anticipated that awareness of connectedness would be positively associated with experiences of purpose and meaning in one’s life (H1). As this sense of connectedness is a culture-based value, we also anticipated that it would be positively associated with Alaska Native cultural identification (H2), while unrelated with identification with a European American or White American lifestyle (H3). Similar to H2, we expected that the ACS would be positively associated with higher levels of mastery with a communal focus, defined as problem solving to overcome challenges through the assistance of family and friends (H4), as a communal approach to problem solving should be closely aligned with a heightened sense of connectivity to family and community. Conversely, we predicted the ACS would not display associations with self-focused mastery, defined as problem solving to overcome challenges on one’s own (H5). We expected that connectedness levels would not correlate with age (H6).

Method

Participants

Participants included 284 Alaska Native 12- to 18-year-olds from rural, remote Alaska communities. Of this total, 194 were recruited from a boarding school serving rural residents located in a regional hub community in Southeast Alaska. The school attracts students from all the cultural-linguistic groups indigenous to Alaska. Yup’ik was the largest cultural group within the student body. The other 90 participants lived in a predominantly Yup’ik regional hub community in Southwestern Alaska. Gender distribution included 120 males and 164 females. The mean participant age was 15.5 years with no significant age difference between males and females. Participants described their parent’s marital status as 62% married, 7% single, 19% divorced, and 12% separated. When living at home, 78% participants reported living with mom, 68% living with dad, 13% living with a grandparent, and 11% living with another relative. Self-identified Alaska Native cultural linguistic group membership was largely Yup’ik, with 72% identifying as Yup’ik, 21% as Inupiat, 11% as Athabaskan, 4% as Tlingit/Haida, and 6% as Aleut/Alutiq (some individuals identified with two or more ethnic categories).

Procedures

The University IRB, school boards, and local AN advisory school boards at each school approved all procedures. Parents of youth were contacted through their children's schools and through direct contact with research staff. Parents were invited to allow their child to participate in a study to help develop measures of the effectiveness of programs designed to enhance sobriety and reasons for life among rural AN youth. Youth were offered an honorarium of $15 for completing the survey. Following parental consent, youth were informed of the purposes, risks, and benefits of the study and completed an assent to participate.

Participants completed an online survey in small groups, ranging from 2 to 12 individuals. Computer administration was via a secure web server. The response format used a continuous “slider” scale with a salmon5 icon that the respondent clicked and moved across a horizontal blue background with three semantic anchors. At the suggestion of our linguistic advisors, most anchors read: “Not at all,” “Somewhat,” and “A lot.” The continuous analog scale was automatically segmented according to 20 response scores. Responses were transmitted via encryption from the remote locations in rural Alaska back to the secure server housed at the University of Alaska Fairbanks. Prior to analysis, we transformed the data into 5-point scores for ease of interpretation by evenly dividing the original 20-point scores (derived directly from the continuous “slider” assessment scale) by 4. In this way, responses that had been automatically coded as 1–4 were re-coded as a 1, 5–8 as a 2, 9–12 as a 3, 13–16 as a 4, and 17–20 as a 5. The resulting 5-point scale retained the approximately normal distribution of the 20-point scale and made it possible for us to conduct interpretable Item Response Theory (IRT) analysis.

The objectives for this study are to assess the properties, functionality, and utility of the ACS through: 1) confirmatory factor analysis (CFA) to assess the factor structure of the original 18 items of the ACS and establish the unidimensionality of the proposed sub-scales and; 2) IRT modeling to examine each item’s properties; and 3) exploration of the associations between the ACS and other scales to assess convergent and discriminant validity.

Measures

Awareness of Connectedness Scale (ACS)

The ACS consists of 18 items developed by students from the University of Alaska, Center for Alaska Native Health Research (CANHR) researchers, and the Yup’ik Regional Coordinating Council (YRCC). The items are targeted at assessing awareness of self as a member of a broader human and natural community including an awareness of connections between one’s own well-being and the wellbeing of other entities in the various ecological spheres that one occupies. The ACS assesses the degree to which a person endorses the concept of interrelatedness between self, family, community, and natural environment. In this case, interrelatedness refers to a relationship of reciprocal well-being or ill effect.

Initial development of the ACS began in a graduate research methods class and focused on developing a brief scale for assessing awareness of connectedness among adults. Three students6, two of whom are co-authors on this paper, drew upon their cultural backgrounds and an interview with the CANHR cultural consultant to develop a preliminary 30-item pool. The original scale authors independently rated the preliminary items for relevance, representativeness, clarity, and specificity in order to evaluate each item’s ability to tap the target construct. The items rated as most relevant, representative, clear, and specific across evaluators were retained. The student project produced a final 12-item scale grounded in indigenous philosophies such as the Yup’ik concept of ellanaq, the Lakota concept of mitakuye oyas’in, and general knowledge of pan-Indian concepts such as the medicine wheel. CANHR researchers agreed to include the scale for pilot testing in the Cuqyun study due to the scale’s conceptual connection to the Alaska Native youth-focused measurement development project.

The ACS includes items deliberately selected to maintain balance between the domains of self, family, community, and natural environment. The readability of the assessment is at the 4th grade level to accommodate participants for whom English is a second language and to ensure appropriateness for use with youth. We solicited the YRCC’s critique of the items for cultural validity and included an additional six questions based on their input. The resultant scale includes 4 subscales, Awareness-Individual and Awareness-Family (4 items each) and Awareness-Community and Awareness-Natural Environment (5 items each), for a total of 18 items. Whereas the Awareness-Family, Awareness-Community, and Awareness-Natural Environment subscales are comprised of items to assess one’s awareness of connectedness to the respective domain, the Awareness-Individual subscale consists of items assessing awareness of how one’s individual actions can impact oneself.

Alaska Native Cultural Identification (ANCI)

The ANCI is an eight-item scale adapted from the Orthogonal Cultural Identification scale (Oetting & Beauvais, 1990–1991) by Center for Alaska Native Health Researchers (Walsko, Mohatt, Lardon, & Burket, 2009). The ANCI uses two item stems: “How much you live by or follow the [Native/White American] way of life?” and “How much do you speak [Native language/English].” We asked these question stems under the heading “When you are at [home/school]” because cultural behavior and identification can differ based on setting (Okazaki & Sue, 1995). This resulted in two subscales, Alaska Native Identification (ANI) and White American Identification (WAI), which tap elements of identity with the culture of origin and the dominant culture.

Reasons for Life (RFL)

The RFL scale is a new 13-item scale for use with AN adolescents (Allen et al., 2008). The RFL is modeled after Osman et al.’s (1996) Brief Reasons for Living-Adolescent (BRFL-A) scale. Whereas the BRFL-A assesses the reasons that a person would choose not to end their life if they felt suicidal, the RFL explores beliefs and experiences that contribute to making life enjoyable, worthwhile, and meaningful. The RFL includes four subscales: Other’s Assessment of Me, Cultural and Spiritual Beliefs, Personal Efficacy, and Family Responsibility.

Multicultural Mastery Scale (MMS)

The MMS was adapted from the Mastery scale (Pearlin, Lieberman, Menaghan, & Mullan, 1981) and the Communal Mastery scale (Jackson, McKenzie, and Hobfoll, 2000) for AN youth by Center for Alaska Native Health researchers (Fok, Allen, Henry, Mohatt & People Awakening Team, 2010). The multicultural mastery scale is comprised of three subscales: Mastery-Self, Mastery-Family, and Mastery-Friends. The Mastery-Self subscale measures a personal sense of control over goal achievement and a sense an individual can overcome life challenges though their own effort. The Mastery-Family subscale evaluates the young person’s belief that he or she can face life’s problems successfully through joining with family, while the Mastery-Friends subscale assesses this sense with regards to friends. Together the Mastery-Family and Mastery-Friends subscales represent a Communal-Mastery Scale that we expect to correlate positively with the ACS as an indicator of convergent validity.

Results

Objective 1: Assessing the Scale Structure of the ACS

Because IRT assumes that the items represent a single dimension, we examined the eigenvalues of the data matrices for each subscale to decide whether IRT analysis was appropriate. According to Lord’s (1980) criterion, if the ratio between the first and the second eigenvalues is greater than three, the scale is essentially unidimensional. In addition to indicating which subscales did not measure a single underlying dimension, principal components analysis identified six items that loaded strongly on secondary dimensions. Removing these items7, resulted in unidimensional subscales with ratios of the first to second eigenvalues approaching three. See the Appendix for the final 12-item scale, which includes 2-item Awareness-Individual and Awareness-Family subscales, and 4-item Awareness-Community and Awareness-Natural Environment subscales.

After establishing unidimensionality of the subscales, we used confirmatory factor analysis to compare three essentially tau-equivalent models of the 12 items to test which underlying structure best fit the data (Lee, Dunbar, & Frisbie, 2001). We conducted the CFA analyses using AMOS-16 (Arbuckle, 2006), using maximum likelihood estimation to appropriately model the items as ordered categories. Use of the slider for data collection provided us with truly continuous data, which we divided into equal intervals for analysis. Thus we avoided having to make assumptions about our response scale that often prove incorrect in practice (cf., Wirth & Edwards, 2007).

In the tau-equivalent models, factor loadings for all items were fixed at 1.0. Error variances were estimated, as were the covariances among factors that were appropriate to the model of factor structure being estimated. These models permitted us to compare factor structures with loadings held constant. We compared a unidimensional model, a first-order four orthogonal factor model, and a second-order four-factor model (see Figure 1). Although the unidimensional model is not suggested by the theory underlying the scale construction, it is the most parsimonious model and should be rejected before a more complex model is retained. In the first-order four orthogonal factor model each scale taps a separate unique dimension. The second-order four-factor model retains this multidimensionality, but structures the correlations among the dimensions in a single higher-order factor. In addition to constraining all factor loadings to 1.0, we left the unique variances of the items uncorrelated with each other.

Figure 1.

Figure 1

Open in a new tab

Confirmatory factor analyses comparison between the unidimensional, first-order four orthogonal factor, and second-order four-factor models.

We assessed model fit using four fit indices: 1) chi-square to degrees of freedom ratio (χ2/df; Hatcher, 1994); 2) comparative fit index (CFI; Bentler, 1990); 3) goodness-of-fit index (GFI; Bentler & Bonnett, (1980); and 4) root-mean squared error of approximation (RSMEA; Browne & Cudeck, 1993). Bentler (1990) and Hu and Bentler (1999) suggest that a χ2/df ratio of less than 2 and CFI and GFI of .90 or higher are indicators of good fit. RMSEA values lower than .08 are generally considered acceptable with values lower than .05 indicating good fit (Browne & Cudeck, 1993). We also compared models using the Bayesian information criterion (BIC) (Bollen, 1980; Guo, Morales, Schwartz, & Szapocznik, 2009). According to Raftery (1993), a BIC difference value of 9.2 or greater between two models indicates an important difference.

Table 1 reports fit statistics for all models. The unidimensional model fit better than the four orthogonal factor model, Δχ2 (3) = 141, p < .01. The second-order four-factor model fit better than either the unidimensional model, Δχ2 (7) = 144, p < .01, or the four orthogonal factor model, Δχ2 (5) = 285, p < .01. Comparing the BIC values of these models also supported the conclusion that the second-order four-factor model fit best, with the BIC for this model being more than 9.2 smaller than the BIC for either of the other models (ΔBIC=105.5)

Table 1.

Model Fit Indexes for the Unidimensional, First-Order Three Orthogonal Factor, Second-Order Three-Factor, and Other Modified Models.

Model Number
of items		 χ2 (df) χ2/df BIC GFI CFI RMSEA
Unidimensional 12 299.1(65) 4.60 373.57 .840 .775 .113
First-Order Four
Orthogonal Factor		 12 440.1(62) 7.10 530.49 .774 .637 .147
Second-Order Four-Factor 12 155.1(58) 2.67 268.07 .917 .907 .077
Modified Second-Order
Four-Factor		 12 79.5(47) 1.69 254.64 .958 .969 .049
Open in a new tab

Note. CFA = confirmatory factor analysis; BIC = Bayesian information criteria; GFI = graduated fit index, CFI = comparative fit index; RMSEA = root mean square error of approximation; 3 pairs of observed variables’ unique variances were allowed to correlate in the modified second-order four-factor model: (a) Items 15 and 16, (b) items 8 and 13, and (c) items 13 and 12.

Having determined that the second-order model fit best, we reran the CFA, freeing the item loadings in order to estimate the magnitude and significance of the factor loadings. In order to obtain accurate estimates of each factor loading and its standard error, we used modification indexes to guide us in allowing correlated pairs of error terms. Error terms were only correlated across subscales and only if there were similarities in item content or wording that would explain why they would produce covariances. MacCallum and Austin (2000) describe the problems inherent in the model-generation approach in which data are fit to a model and then model fit is improved solely on the basis of modification indices. Because such methods can lead different researchers with the same model to produce different final models, we were careful to correlate errors only after model selection was completed based on the tau-equivalent comparison and in cases of related item content. Following review of the modification indices and item content we chose to correlate the errors for 3 item pairs: 10 and 11, 8 and 4, and 8 and 7. Items 10 and 11 both discuss giving. We would expect a sense of having a lot to give (item 11) varying similarly across the sample with a belief in giving without expecting anything in return (item 10). Items 7 and 8 both refer to an awareness of the impact that one’s actions have on another person, and are further related due to their close physical proximity on the scale. Schwartz (1999) explains that people use surrounding items to interpret how to respond to any given items; hence proximity can have significant impact on how people respond to an item. Items 8 and 4 are related through an inverse logic with each other, with item 4 emphasizing how another’s emotional state impacts the self and item 8 emphasizing how the self’s actions on another can have a reciprocal impact on the self.

As can be seen in Table 1, the fit of the modified second-order model was acceptable, χ2 (47) = 79.5, p < .01, χ2/df = 1.69, GFI = .96, CFI = .97, and RMSEA = .049. The first-order and second-order standardized factor loadings of this model, displayed in Figure 2, were all greater than 0.40, and were significant, except in the cases where unstandardized loadings were initially constrained to 1 to scale the latent variables. Awareness-Community had the highest loading on the general factor (.89), indicating that this factor is the most associated with the underlying construct, awareness of connectedness. The loadings for the Awareness-Individual, Awareness-Family, and Awareness-Natural Environment subscales were .82, .78, and .73 respectively. These results support the notion that the ACS is comprised of four factors subsumed under one higher order factor.

Figure 2.

Figure 2

Open in a new tab

Modified second-order four-factor confirmatory factor analysis model. Three pairs of observed variables’ unique variances were allowed to correlate in this model.

Objective 2: Evaluate Item Functioning with Item Response Theory

In the next step of analysis we used IRT modeling to examine individual item functioning. We adopted Samejima’s (1996) graded response model, which is often used when items consist of more than two ordered response categories, and conducted the analyses using the ltm package through R (Rizopoulos, 2006). Table 2 displays the descriptive statistics using the 20- and 5-category response formats for the 12-item set. Item parameters for the 12 items are shown in Table 3, which lists discrimination or slope parameters (as) and location parameters (bs). Items 4 (Family’s happiness), 6 (Community’s happiness), and 9 (Give and receive abundantly) display the highest slopes, indicating that they are best at discriminating between participants with varying levels of awareness of connectedness. Items 10 (Give without expecting) and 12 (Come from and return to land) displayed the lowest slopes, indicating that they are least discriminating.

Table 2.

Means and Standard Deviations of the ACS items

20-point analog scale 5-Item Calibration
Item M SD M SD
1 12.06 5.72 3.36 1.30
2 11.38 6.04 3.19 1.36
3 13.96 4.91 3.79 1.11
4 16.71 4.43 4.37 .96
5 12.04 5.67 3.36 1.29
6 10.82 5.56 3.06 1.26
7 11.81 6.01 3.27 1.38
8 10.89 5.16 3.08 1.18
9 13.05 4.91 3.58 1.10
10 14.07 4.97 3.79 1.16
11 12.48 5.47 3.44 1.24
12 13.20 5.52 3.60 1.25
Open in a new tab

Note: ACS = Awareness of Connectedness Scale

Table 3.

Item Parameters for 12-Item Awareness of Connectedness Scale by Subscale

Item Abbreviated Content a b1 b2 b3 b4
Awareness-Individual
9 Give and receive abundantly 2.686 −2.064 −1.293 −0.026 0.766
10 Give without expecting 1.092 −3.023 −2.072 −0.600 0.769
Awareness-Family
2 Family hurts with me 1.340 −1.604 −.818 0.221 1.241
4 Family's happiness 4.218 −2.332 −1.774 −.980 −0.321
Awareness-Community
5 Community believes
important		 2.111 −1.599 −.841 0.137 .878
6 Community's happiness 2.901 −1.183 −0.534 0.376 1.180
8 Energy into community 2.169 −1.428 −0.614 0.444 1.428
11 A lot to give community 1.438 −2.173 −1.136 0.139 1.008
Awareness-Natural Environment
1 Feel connected to nature 1.442 −1.895 −1.051 0.064 1.065
3 Respect nature like family 2.404 −2.200 −1.398 −0.435 0.583
7 Mistreating nature 2.018 −1.380 −.722 0.133 0.855
12 Come from and return to
land		 1.060 −2.664 −1.852 −0.102 0.834
Open in a new tab

The location parameters (bs), suggest that although the items represent varying levels of the attribute, most represent lower levels of the latent trait. Values for location parameters from b2 through b4 for items 2 (Family hurts with me), 6 (Community’s happiness), and 8 (Energy into community) are relatively higher than the other items, indicating that these three items are more informative in distinguishing people with a higher awareness levels. While these three items are somewhat less likely to receive endorsement of the highest anchors, like most of the items on the scale they appear to represent much of the range of the latent variable. This is to say that none of the items on the scale would be endorsed only by people with high levels of the attribute. On the other hand, items 3 (Respect nature like family), 4 (Family’s happiness), and 10 (Give without expecting) have lower location parameters, indicating that they are better at distinguishing among people with low levels of awareness of connectedness.

Another means of assessing item functioning is through item information curves. These curves plot the amount of information each item provides at each level of the latent trait (low to high Awareness of Connectedness), allowing an examination of the precision of each item (Marshall, Orlando, Jaycox, Foy, & Belzberg, 2002). Information in IRT analysis is roughly equivalent to the inverse of standard error, as standard error can be understood to represent what we don’t know, but item information has the benefit of being additive such that the information provided by a test is equal to the sum of the information of each item (Edwards & Edelen, 2009).

The item information curves for each item by subscale are plotted in figure 3. Items 1 (Feel connected to nature), 2 (Family hurts with me), 10 (Give without expecting), 11 (A lot to give community), and 12 (Come from and return to land) have relatively flat information curves at lower levels of information than other items on their respective subscales. The relative levels of information displayed by these items and the flat nature of their curves indicates that they provide low to moderate levels of information across all levels of awareness. This relatively lower level of information indicates that the items may be individually less informative, but that they provide recognizable levels of information at all levels indicates that they remain useful in describing the construct and contribute meaningfully to the overall information provided by the scale. In contrast, items 3 (Respect nature like family), 5 (Community believes important), 7 (Mistreating nature), and 8 (Energy into community) offer relatively greater information at moderate levels of awareness, with the least information at the lowest and highest levels of awareness, indicating that they are best suited for characterizing people with moderate levels of awareness. Item 6 (Community’s happiness) provides moderate to high information at low to moderate levels of awareness, whereas item 4 (Family’s happiness) provides high information at low levels of the latent trait, indicating that both items are especially informative for people with low levels of awareness. The majority of items discriminate best at low levels of awareness while providing relatively low information at higher levels.

Figure 3.

Figure 3

Figure 3

Open in a new tab

Item information functions by Subscale

Objective 3: Convergent and Discriminant Validity

We assessed convergent and discriminant validity of the final 12-item ACS using Alaska Native Cultural Identification (ANCI), Reasons For Life (RFL), and the Multicultural Mastery Scale (MMS). All measures were range-standardized for comparability of data. As can be seen in Table 4, convergent validity was evidenced through moderate to strong positive correlations with the RFL, and the ANCI, Mastery-Family and Mastery-Friends subscales. While there was a statistically significant inverse relationship between the ACS and the White American Identification (WAI) subscale, the magnitude of correlation is small (r = −.12, p < .05), suggesting limited relationship between the two variables. This limited relationship, along with limited relationships between the ACS and Mastery-Self (r = .13, p < .05) and age (r = .05, ns) were consistent with expectations related to discriminant validity.

Table 4.

Correlation Between the ACS and Reasons for Life, Alaska Native Identification, White American Identification, Communal Mastery, Mastery-Self and Age.

Awareness of
Connectedness
Reasons for Life .62**
Alaska Native Identification .34**
White American Identification −.12*
Communal Mastery (Family & Friends) .44**
Mastery-Family .46**
Mastery-Friends .25**
Mastery-Self .13*
Age .05
Open in a new tab
*

p < .05.

**

p < .01

Cronbach’s alpha for the final 12-item ACS was an acceptable .85. Table 5 displays the internal consistency coefficients (Cronbach’s alpha) for the full scale and subscale scores. Alpha coefficients for the four-item subscales were in the conventionally acceptable range (Nunnally & Bernstein, 1994, p. 264), but alphas for the two-item subscales were lower (.54 and .61).

Table 5.

Internal Consistency (Cronbach’s alpha) for the full scale and the sub-scales

n 5-point
Calibration
M(SD) α
Awareness of Connectedness Scale 12 41.88(9.01) .85
   Awareness-Individual 2 7.37(1.86) .54
   Awareness-Family 2 7.56(1.99) .61
   Awareness-Community 4 12.93(3.94) .80
   Awareness-Natural Environment 4 14.03(3.74) .72
Open in a new tab

Discussion

This research suggests that a concept of awareness of connectedness as described in many Native American epistemologies can be measured as a multi-dimensional construct, and that this construct is related to hypothesized suicide and substance abuse protective factors such as reasons for living and communal mastery among AN youth. Awareness of connectedness as measured by the ACS includes items that assess the individual’s sense of connectedness in relation to their self, family, community, and the natural environment. The ACS assesses connectedness as evidenced by the degree to which a person endorses the view that they exist in a web of reciprocal relationships. Final scale analyses identified 12 items from the 18-item pool that fit the model and function best. These results confirm awareness of connectedness as a measurable construct and corroborate other researchers’ work regarding awareness of connectedness as a measurable construct that is positively related to wellbeing (Hill, 2006 & 2009), recovery processes (Hobfoll, Jackson, Young, Pierce & Hobfoll, 2002; Bowen, 2005; Gone, 2009; Mohatt et al., 2007), and a sober and healthy life (Hazel & Mohatt, 2001; Mohatt et al., 2004b; Allen, et al. 2006) among American Indian/Alaska Native cultures. Future research should continue to investigate these relationships in order to elucidate the complex ways in which this positive cultural value contributes to community and individual wellbeing.

The development and validation of the ACS provides a measurement tool that may be useful for future examination of hypotheses regarding possible relationships between Native American cultural identification and positive behavioral health outcomes. Bates et al. (1997) suggest that internalization of cultural norms likely provides a more direct link to substance use outcomes than general cultural identification. They further suggest that cultural identification may be linked to behavioral outcomes via these specific cultural norms—greater identification with traditional culture leads to internalization of the culture-specific value system. Depending on whether the cultural norms that are internalized are protective or risk factors, this internalization could lead to either higher or lower prevalence of substance use. The moderate correlation between the ACS and ANCI support the theory that the ACS measures a culturally-based value. The moderate to strong correlations of ACS with RFL and Communal Mastery, two identified protective factors, suggests the ACS also taps a protective variable that could be useful in explaining the variance in substance use and other behavioral health outcomes among Native youth. Additionally, the relation with communal forms of mastery over self-based mastery is again suggestive of the cultural overlay in the construct tapped by the ACS, indicating its association with a preference to alternatives to individualistically focused coping and mastery.

The results of this research should not be interpreted to suggest that the ACS taps all levels of awareness of connectedness. IRT results indicate that the scale provides more information and discriminates participants best at low to moderate levels of the latent trait. This makes the scale relatively more useful for research seeking to identify and help people with low levels of awareness of connectedness, such as in suicide and substance abuse interventions and prevention programs. Further scale development could result in a more discriminating scale across the full range of the latent trait through identification of more challenging questions. However, in cases where it is most important for the measure to be sensitive to low to moderate levels of awareness of connectedness IRT results suggest the current 12-item ACS will be a valuable tool.

Gone (2009) recommends that the distance between evidence-based practices and culture-based treatment programs be bridged through the development of locally identified, culturally based outcomes. The ACS is one example of this type of development effort. IRT results suggest that the scale could easily be pared down to a brief measure of change suitable for repeated measurements in intervention outcomes analyses. Many of the items on the scale provide information across the same (low to moderate) levels of awareness of connectedness. Removal of the least informative and least discriminating items could result in a 4–5 item scale. These results indicate that the ACS could be a valuable quantitative assessment tool for intervention and prevention efforts that seek to increase cultural connectedness by assessing the degree to which these program enhance people’s endorsement of this key cultural value. While the ACS may prove useful in bridging the gap between culture-based interventions and evidence-based practice, it should not be construed to measure improvement in health or pathology. The ACS is instead conceptualized as a culturally-based protective factor contributing to resiliency and well-being. It is a targeted measure of one potential culture-based outcome independent from measures of pathology.

Future research into connectedness among Native Americans should investigate the different path relationships between awareness of connectedness and protective factors, risk factors, recovery, and resiliency. Of particular interest is the question of what levels of awareness of connectedness are considered protective and to what degree is there protective function. Also important is development of research into the complex relationships between behavioral health, cultural identity and enculturation, and the mediating role of specific cultural values such as awareness of connectedness. Development and validation of the ACS is a first step for these new lines of inquiry.

In future research, it will be important to explore the generalizability of the ACS with other Alaska Native and American Indian cultural groups. While interpretations of ACS scores were validated predominantly with one Native American cultural group, the Yup’ik people of southwestern Alaska, literature on the worldview of other indigenous cultures echoes similar traditional ways of thinking about the world. The reciprocal wellbeing of the self in relation to the broader life-world is similar across Alaska Native cultures (Mohatt et al., 2007) and the Lakota people (Bowen, 2005), as well as evidenced by related research with other tribes (Hill, 2006; Gone, 2009). Indigenous concepts such as ellam-iinga8, yuuyaraq9, and mitakuye oyas’in10 all indicate that the notion of connectedness is based on understanding how all spirit-beings, including family members, other people, and the natural environment, are related and equally demanding of respect. The land may be different than a human or an animal, but as a spirit that is part of the whole, any harm we cause the land may be visited back on us in the same way harm to a brother or sister can harm oneself. Conversely, life choices that positively impact the people and world around also positively impact the self. In this worldview, the self is defined broadly to encompass the relationship with all spirit-beings.

Despite promising results, this research has certain limitations. This study was conducted with youth from predominantly a single cultural group, yet the theory behind the scale suggests that the construct could be useful with adults, and across other American Indian and Alaska Native cultures. Future research should seek validation with other cultural and age groups.

Another important limitation is the low reliability of the subscales. In particular the Individual and Family subscales, each composed of only two items, had low alphas in this sample. For this reason, we recommend that if the ACS is to be used in clinical applications, only the full scale be used. We believe that the subscales can still be appropriate for research, where measures with modest reliabilities can be useful for research, particularly in the early stages of construct validation (Nunnally & Bernstein, 1994). Future research should seek to develop other items and to elaborate the constructs underlying these two subscales in order to provide more precision at the subscale level. Finally, the IRT analysis suggested that few ACS items discriminate among people at higher levels of awareness of connectedness. This may not be a limitation for identifying people who are at risk, but for conducting broader research it will be useful to develop additional more “difficult” items.

As we advance research into these indigenous ways of thinking about our place in the world, we must retain the same respect for cultural difference that led us to develop the ACS. This research does not yet answer more nuanced questions of how the people understand and situate the measured construct. For example, although the Yup’ik and Lakota people express a similar epistemology of connectedness, they use different linguistic constructions to express the concept. Future research should also explore cultural difference in understandings of the notion of connectedness. By better understanding potential cultural variation with regard to the construct, we will be better equipped to interpret research findings and identify how the value of connectedness contributes to individual, community, and cultural wellbeing from one culture to the next. While the ACS was developed in collaboration with Alaska Native people, the construct holds significant promise for research into how different cultures, Native and otherwise, view and relate to a sense of connectedness with the entire life-world.

The ACS measures a multidimensional construct of awareness of connectedness based on a Native American worldview. The scale is intended for use with youth, and is related to other hypothesized protective factors for Native youth. Native people across the United States have emphasized that a notion of a holistic connectedness with the larger spiritual universe is an important cultural value underpinning a healthy Native lifestyle. This work represents a first step in taking this Native way of thinking as a basis for empirical research.

Acknowledgments

We thank all of the People Awakening Team including participants, community co-researchers, Coordinating Council and Center for Alaska Native Health Research project staff for their assistance in completing this research. We also acknowledge the invaluable contributions of James A. Walsh. This research was funded by the National Institute of Alcohol Abuse and Alcoholism (NIAAA) [R21AA0016098] and National Institute for Minority Health Disparities (NIMHD) [R24MD001626], PI: James Allen; NIAAA R21AA015541; NIAAA & NCMHD [R01AA11446] and National Center for Research Resources [P20RR061430], PI: Gerald V. Mohatt, and a University of Alaska International Polar Year Postdoctoral Fellowship award to the second author.

Appendix: Awareness of Connectedness Scale

Awareness-Individual

  • 9.

    I believe that if I give I will receive abundantly.

  • 10.

    I give without expecting anything in return.

Awareness-Family

  • 2.

    When I am hurting, my family hurts with me.

  • 4.

    My family’s happiness is part of my happiness.

Awareness-Community

  • 5.

    My community believes I am important.

  • 6.

    My community’s happiness is part of my happiness.

  • 8.

    The energy I put into the community comes back to me.

  • 11.

    I believe I have a lot to give my community.

Awareness-Natural Environment

  • 1.

    I feel connected to nature.

  • 3.

    I treat nature with respect like a family member.

  • 7.

    Mistreating nature is the same as mistreating myself.

  • 12.

    We come from the land and will return to the land.

Response Scale: The response scale used for this measure was a continuous analog scale presented on a computer with semantic anchors of “Not at all,” “Somewhat,” and “Alot.”

Footnotes

Publisher's Disclaimer: The following manuscript is the final accepted manuscript. It has not been subjected to the final copyediting, fact-checking, and proofreading required for formal publication. It is not the definitive, publisher-authenticated version. The American Psychological Association and its Council of Editors disclaim any responsibility or liabilities for errors or omissions of this manuscript version, any version derived from this manuscript by NIH, or other third parties. The published version is available at www.apa.org/pubs/journals/cdp.

1

The YRCC is a 10-member advisory council that acts as the CBPR partner voice for the regions engaged in research and ensures cultural validity of the projects.

2

Cuqyun, means “measurement tool” in the Yup’ik language.

3

Cultural identification is the degree to which a person self-identifies with a culture (Oetting & Beauvais, 1990; Bates et al., 1997).

4

enculturation is the degree to which a person is steeped in the cultural traditions and values (Whitbeck et al., 2002)

5

Salmon are a central cultural icon of the region and were used based on input from local advisors.

6

Nathaniel Mohatt (UAF), Rebekah Burket (UAF), and Tonie Quaintaince University of Alaska Anchorage

7

“I put myself first when making decisions,” “my spirit affects the people around me,” “my actions have an effect my family,” “I show respect to my family members in different ways depending upon who they are,” “life in my community has little effect on me,” and “the land, water and air keep us alive” were removed based on PCA results.

8

Yup’ik for “Eye of Awareness”

9

Yup’ik for “Way of the Human Being”

10

Lakota Sioux for “All of My Relatives”

Contributor Information

Nathaniel V. Mohatt, The Consultation Center and Division of Prevention and Community Research, Department of Psychiatry, Yale University School of Medicine

Carlotta Ching Ting Fok, Center for Alaska Native Health Research (CANHR), University of Alaska Fairbanks (UAF).

Rebekah Burket, CANHR and Department of Psychology, UAF.

David Henry, Institute for Health Research and Policy, University of Illinois at Chicago.

James Allen, CANHR and Department of Psychology, UAF.

References

  1. Alcantara C, Gone J. Reviewing suicide in Native American communities: Situating risk and protective factors within a transactional-ecological framework. Death Studies. 2007;31:457–477. doi: 10.1080/07481180701244587. [DOI] [PubMed] [Google Scholar]
  2. Allen J, Mohatt GV, Fok CCT, Henry D, Burkett R. Predictors of Alaska Native Adolescent Sobriety and Reasons for Living: A Protective Factors Model for Substance Abuse and Suicide Prevention among Alaska Native Youth; Presentation at Native Children’s Research Exchange; Denver, CO.. 2008. [Google Scholar]
  3. People Awakening Team. Allen J, Mohatt G, Fok CCT, Henry D. Suicide prevention as a community development process: Understanding circumpolar youth suicide prevention through community level outcomes. International Journal of Circumpolar Health. 2009;68:274–291. doi: 10.3402/ijch.v68i3.18328. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. People Awakening Team. Allen J, Mohatt GV, Rasmus SM, Hazel KL, Thomas L, Lindley S. The tools to understand: Community as Co-Researcher on culture-specific protective factors for Alaska Natives. Journal of Prevention and Intervention in the Community. 2006;32:41–59. doi: 10.1300/J005v32n01_04. [DOI] [PubMed] [Google Scholar]
  5. Arbuckle JL. AMOS (Version 16.0) [Computer Program] Chicago: SPSS; 2006. [Google Scholar]
  6. Bates SC, Beauvais F, Trimble JE. American Indian adolescent alcohol involvement and ethnic identification. Substance Use and Misuse. 1997;32:2013–2031. doi: 10.3109/10826089709035617. [DOI] [PubMed] [Google Scholar]
  7. Bentler PM. Comparative fit indexes in structural models. Psychological Bulletin. 1990;107:238–246. doi: 10.1037/0033-2909.107.2.238. [DOI] [PubMed] [Google Scholar]
  8. Bentler PM, Bonett DG. Significance tests and goodness of fit in the Analysis of covariance structures. Psychological Bulletin. 1980;107:588–606. [Google Scholar]
  9. Bollen KA. Issues in the comparative measurement of political democracy. American Sociological Review. 1980;45:370–389. [Google Scholar]
  10. Bowen DE. Honoring the elders: Interviews with two Lakota men. Journal of Sociology and Social Welfare. 2005;32:125–134. [Google Scholar]
  11. Browne MW, Cudeck R. Alternative ways of assessing model fit. Sociological Methods & Research. 1993;21:230–239. [Google Scholar]
  12. Dana RH. The cultural self as locus for assessment and intervention with American Indians/Alaska Natives. Journal of Multicultural Counseling and Development. 2000;28:66–82. [Google Scholar]
  13. Duran E. Healing the soul wound: Counseling with American Indians and Alaska Native peoples. New York, London: Teachers College Press; 2006. [Google Scholar]
  14. Edwards MC, Edelen MO. Special topics in Item Response Theory. In: Millsap RE, Maydeu-Olivares A, editors. The SAGE handbook of quantitative methods in psychology. London: SAGE Publications Ltd.; 2009. [Google Scholar]
  15. Fok C, Allen J, Henry D. Multidimensional Assessment of Culturally Mediated Coping Strategies Using the Multicultural Mastery Scale for Youth; Poster presented at the Biennial Meeting of the Society for Research in Child Development; Montreal, Quebec, Canada. 2011. Apr, [Google Scholar]
  16. Gone J. A community-based treatment for Native American historical trauma: Prospects for evidence-based practice. Journal of Consulting and Clinical Psychology. 2009;77:751–762. doi: 10.1037/a0015390. [DOI] [PubMed] [Google Scholar]
  17. Guo X, Morales S, Schwartz SJ, Szapocznik J. Some evidence for multidimensional biculturalism: Confirmatory factor analysis and measurement invariance analysis on the bicultural involvement questionnaire short version. Psychological Assessment. 2009;21:22–31. doi: 10.1037/a0014495. [DOI] [PubMed] [Google Scholar]
  18. Hazel K, Mohatt GV. Cultural and spiritual coping in sobriety: Informing substance abuse prevention for Alaska Native communities. Journal of Community Psychology. 2001;29:541–562. [Google Scholar]
  19. Hatcher L. A step-by-step approach to using the SAS system for factor analysis and structural equation modeling. Cary, NC: SAS Institute; 1994. [Google Scholar]
  20. Hawkins EH, Cummins LH, Marlatt GA. Preventing substance abuse in American Indian and Alaska Native youth: Promising strategies for healthier communities. Psychological Bulletin. 2004;130:304–323. doi: 10.1037/0033-2909.130.2.304. [DOI] [PubMed] [Google Scholar]
  21. Hill DL. Sense of belonging as connectedness, Alerican Indian worldview, and mental health. Archives of Psychiatric Nursing. 2006;20:210–216. doi: 10.1016/j.apnu.2006.04.003. [DOI] [PubMed] [Google Scholar]
  22. Hill DL. Relationship between sense of belonging as connectedness and suicide in American Indians. Archives of Psychiatric Nursing. 2009;23:65–74. doi: 10.1016/j.apnu.2008.03.003. [DOI] [PubMed] [Google Scholar]
  23. Hobfoll SE, Jackson A, Young S, Pierce CA, Hobfoll I. The impact of communal-mastery versus self-mastery on emotional outcomes during stressful conditions: A prospective study of Native American women. American Journal of Community Psychology. 2002;30:853–871. doi: 10.1023/A:1020209220214. [DOI] [PubMed] [Google Scholar]
  24. Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling. 1999;6:1–55. [Google Scholar]
  25. Jackson T, McKenzie J, Hobfoll SE. Communal aspects of self-regulation. In: Boekaerts M, Pintrich PR, Zeider M, editors. Handbook of self-regulation. San Diego, CA: Academic Press; 2000. [Google Scholar]
  26. LaFromboise TD, Hoyt DR, Oliver L, Whitbeck LB. Family, community, and school influences on resilience among American Indian adolescents in the upper midewest. Journal of Community Psychology. 2006;34:193–209. [Google Scholar]
  27. Lee G, Dunbar SB, Frisbie DA. The relative appropriateness of eight measurement models for analyzing scores from tests composed of testlets. Educational and Psychological Measurement. 2001;61:958–975. [Google Scholar]
  28. Lord FM. Applications of item response theory to practical testing problems. New York, NY: Erlbaum Associates; 1980. [Google Scholar]
  29. Marshall GN, Orlando M, Jaycox LH, Foy DW, Belzberg H. Development and validation of a modified version of the Peritraumatic Dissociative Experiences Questionnaire. Psychological Assessment. 2001;14:123–134. doi: 10.1037//1040-3590.14.2.123. [DOI] [PubMed] [Google Scholar]
  30. MacCallum RC, Austin JT. Applications of structural equation modeling in psychological research. Annual Review of Psychology. 2000;51:201–226. doi: 10.1146/annurev.psych.51.1.201. [DOI] [PubMed] [Google Scholar]
  31. Mohatt GV, Hazel KL, Allen J, Stachelrodt M, Hensel C, Fath R. Unheard Alaska: Participatory action research on sobriety with Alaska Natives. American Journal of Community Psychology. 2004a;33:263–273. doi: 10.1023/b:ajcp.0000027011.12346.70. [DOI] [PubMed] [Google Scholar]
  32. Mohatt GV, Rasmus SM, Thomas L, Allen J, Hazel K, Hensel C. “Tied together like a woven hat:” Protective pathways to Alaska Native sobriety. Harm Reduction Journal. 2004b;1 doi: 10.1186/1477-7517-1-10. Retrieved April 19, 2010 from http://www.harmreductionjournal.com/content/1/1/10. [DOI] [PMC free article] [PubMed]
  33. Mohatt GV, Rasmus SM, Thomas L, Allen J, Hazel K, Marlatt GA. Risk, resilience, and natural recovery: A model of recovery from alcohol abuse for Alaska Natives. Addicition. 2007;103:205–215. doi: 10.1111/j.1360-0443.2007.02057.x. [DOI] [PubMed] [Google Scholar]
  34. Napoleon H. Yuuyaraq: The way of the Human Being. Fairbanks, AK: Alaska Native Knowledge Network; 1996. [Google Scholar]
  35. Nunnally JC, Bernstein IH. Psychometric Theory. 3d ed. New York: McGraw-Hill; 1994. [Google Scholar]
  36. Oetting ER, Beauvais F. Orthogonal cultural identification theory: The cultural identification of minority adolescents. The international Journal of the Addictions. 1990;25:655–685. doi: 10.3109/10826089109077265. [DOI] [PubMed] [Google Scholar]
  37. Okazaki S, Sue S. Methodological issues in assessment research with ethnic minorities. Psychological Assessment. 1995;7:267–375. [Google Scholar]
  38. Osman A, Kopper BA, Barrios FX, Osman JR, Besett T, Linehan MM. The brief reasons for living inventory for adolescents (BRFL-A) Journal of Abnormal Child Psychology. 1996;24:433–443. doi: 10.1007/BF01441566. [DOI] [PubMed] [Google Scholar]
  39. Pearlin LI, Menaghan EG, Lieberman MA, Mullan JT. The stress process. Journal of Health & Social Behavior. 1981;22:337–356. [PubMed] [Google Scholar]
  40. Raferty AE. Bayesian model selection in structural equation models. In: Bollen KA, Long JS, editors. Testing structural equation models. Newbury Park, CA: Sage; 1993. pp. 63–180. [Google Scholar]
  41. Rizopoulos D. ltm: An R package for latent variable modeling and item response theory analyses. Journal of Statistical Software. 2006;17(5):1–25. [Google Scholar]
  42. Samejima F. Graded response model. In: van der Linden WJ, Hambleton PK, editors. Handbook of Modern Item response Theory. New York: Springer Publishing Company; 1996. pp. 85–100. [Google Scholar]
  43. Schwartz N. Self-reports: How the questions shape the answers. American Psychologist. 1999;54:93–105. [Google Scholar]
  44. Wolsko C, Mohatt GV, Lardon C, Burket R. Smoking, chewing, and cultural identity: Prevalence and correlates of tobacco use among the Yu’pik—The Center for Alaska Native Health Research (CANHR) study. Cultural Diversity and Ethnic Minority Psychology. 2009;15:165–172. doi: 10.1037/a0015323. [DOI] [PubMed] [Google Scholar]
  45. Whitbeck LB, McMorris BJ, Hoyt DR, Stubben JD, Laframboise TA. Perceived discrimination, traditional practices, and depressive symptoms among American Indians in the Upper Midwest. Journal of Health and Social Behavior. 2002;43:400–418. [PubMed] [Google Scholar]

RESOURCES