Desktop version

Home arrow Mathematics

  • Increase font
  • Decrease font


<<   CONTENTS   >>

Methods

Table of Contents:

Sample

Data are from the baseline interview of the multi-site randomized control trial Bii-Zin- Da-De-Dah (Ьё-zen-da-de-da; Listening to One Another; BZDDD) program. BZDDD is a family-based culturally adapted evidence-based substance use prevention program for American Indian pre-adolescents aged 8-10 delivered in four communities that share a common language, culture, and history (for detailed discussion of the adaptation process, see Ivanich, et al. 2020).

Youth and their families were recruited through a school-based and community outreach recruitment strategy. Working with community partners, we created a web-based interest form for local recruiters to use at events, school take-home flyers, and other community outreach activities, resulting in 497 online interest forms across four community partner locations. These recruiting activities are commonly used in CBPR and with research in American Indian communities (Whitesell, et al. 2019). Local recruiters attempted visits with 440 families for an official recruitment home visit and to screen for eligible youth living in the home. All families received wild rice for their time learning about the program and each youth received a Bii-Zin-Da-De-Dah drawstring bag. Of the 440 attempted visits, 363 families (85.5%) were successfully visited, eligible, and were initially interested in participating in the program.

A total of 679 interviews were conducted with 303 (68.9%) families that met eligibility criteria (i.e., youth living in the home of the target 8-10 age range, and a caregiver willing to attend the 14-week program) and did not refuse enrollment, move, or become unreachable after the initial visit. Baseline data collection began in May 2017 and was completed in April 2018. Data were collected using in-person interviews with a trained interviewer from the local community in which the respondent lived. Each caregiver and child who completed the baseline survey received a $20.00 Visa gift card. All materials, processes, and surveys were reviewed and approved by our local research advisory boards called Prevention Research Councils for community and culture appropriateness prior to university institutional review board approval (IRB #20140214158FB). All participants were informed of policies, risks, benefits, and compensation before signing consent/assent agreement forms for enrollment and data collection.

Measures

The BZDDD program focuses on substance use, mental health, and cultural/traditional engagement. Mental health was measured using the Achenbach System of Empirically Based Assessment (ASEBA) youth self-reports (Achenbach 1991). A standardized scale of 102 questions, ASEBA captures information about psychopathology of youth, specifically internalizing and externalizing behavior. Due to the young age of participants (8—10 years old), permission was granted by ASEBA to remove a total of six questions: two focused on suicidal thoughts and attempts (1 and 2), two focused on sex and sexual thoughts (3 and 4), and two related to hearing things that others do not hear and seeing things that others do not see (5 and 6). The latter two (5 and 6) were removed upon advice from community members serving as field interviewers and the local research advisory boards. In this American Indian community, there is a history of being shamed, and institutionalized for similar cultural practices, and such questions were therefore not appropriate for this study.

Substance use is a dichotomous measure if youth said yes (1) to any of the following questions. If they have ever "had a drink of any beverage that contains alcohol, such as beer, wine, hard liquor, or mixed drinks," "smoked a cigarette," "used marijuana (pot/ weed)," "used any other type of drug or substance to get high," and "used prescription pills to get high." Youth that indicated that they had not tried any of these were assigned a value of (0).

Cultural participation is measured with a sum of dichotomous measures. Participants were asked if they had offered tobacco, participated in ceremonial songs, smudged or saged, participated in a ceremonial dance, gone to a traditional healer, sought advice from a spiritual advisor, or participated or sung in a drum group in the past 12 months. A yes is counted as a (1) and a no as a (0). The sum of these responses provides a measure of how much each youth participated in cultural practices.

Cultural discrimination is measured with a series of questions about being a tribal member. Individuals were asked how often you have been treated as if you were not smart because you are [redacted to maintain confidentiality of the communities], you have been treated as if you were not as good as others because you are [redacted], you have been treated as if you were not an honest person because you are [redacted], you have been treated with less respect than others because you are [redacted], you have been called names or insulted because you are [redacted], you have been treated negatively because you are [redacted], your family talks about incidents of discrimination, and you feel confident in discussing discrimination with your family. Youth could respond, never (0), rarely (1), sometimes (2), and often (3). We summed the numeric responses of all eight measures to create a single cultural discrimination measure. The Cronbach's alpha for this measure is 0.77.

The interviewer was asked a few questions once the interview was complete and they were alone. We use three questions about privacy in this post-interview questionnaire. We consider interviewers to be known (1) by the participant if they answered "very well" or "somewhat" to the question: "how well did you know the person you interviewed before you interviewed them?" and unknown (0) if they said, "not at all." The presence of an engaged third party was assessed by asking: "Was anyone else present during any portion of the interview? By present, we mean not just walking through the area where the interview was taking place, but listening to or taking part in the interview process," and coded yes (1) or no (0). Interviewers also noted if the interview took place in the participant's home (1) or somewhere else (0).

Analytic Strategy

We first present univariate descriptive statistics of all variables used in this study in Table 10.1. Then we present a series of multilevel linear random-effects models for four of the dependent variables and a logistic random-effects model for the substance use outcome. In these models, the participants of the study are treated as level 1 and the interviewers are level 2. The only information we have about interviewers is their unique ID, all other information used in the analysis is about the participant. Results for internalizing and externalizing measures are presented in Table 10.2. Results for the remaining dependent variables are presented in Table 10.3. All analytic models include covariates for the participant's gender and age. For every outcome, each privacy measure is tested on their

TABLE 10.1

Descriptive Statistics of Participants (N = 365)

Statistic

Mean/%

Std. Dev.

Min

Max

Female

52%

0.50

0

1

Age

9.10

0.91

7

11

Know interviewer

26%

0.44

0

1

Present third party

31%

0.46

0

1

Interviewed in home

73%

0.45

0

1

Internalizing - total

16.77

8.99

0

53

Anxiety/depressed

5.53

3.96

0

20

Withdrawn

4.67

2.92

0

15

Somatic complaints

6.57

3.89

0

19

Externalizing - total

7.95

6.92

0

35

Rule breaking

2.15

2.50

0

15

Aggressive behavior

5.80

4.91

0

24

Substance use - yes

11%

0.31

0

1

Cultural participation

2.41

2.24

0

7

Cultural discrimination

13.59

5.44

0

31

own (with demographic covariates) in their own model, and then all three privacy measures are placed in a combined model. All analyses were conducted in R (R. Core Team 2014) using the glmmTMB package.

TABLE 10.2

Linear Mixed-Effects Models Predicting Internalizing and Externalizing Behaviors

Internalizing

Externalizing

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

Female

  • 0.874
  • (0.918)
  • 0.808
  • (0.933)
  • 0.848
  • (0.934)
  • 0.952
  • (0.923)
  • -1.358*
  • (0.724)
  • -1.348*
  • (0.724)
  • -1.418*
  • (0.725)

О 00 § £! 7 S

Age

  • -1.573”
  • (0.506)
  • -1.594"
  • (0.514)
  • -1.557”
  • (0.514)
  • -1.555”
  • (0.508)
  • -0.253
  • (0.399)
  • -0.260
  • (0.399)
  • -0.277
  • (0.399)
  • -0.279
  • (0.400)

Know

interviewer

  • -3.539”
  • (1.042)
  • -3.459”
  • (1.049)
  • -0.201
  • (0.822)
  • -0.200
  • (0.827)

Present third party

  • 0.684
  • (1.045)
  • 0.530
  • (1.052)
  • 0.443
  • (0.811)
  • 0.289
  • (0.829)

Interviewed in home

  • -0.942
  • (1.013)
  • -0.924
  • (1.016)
  • 0.812
  • (0.786)
  • 0.770
  • (0.801)

Constant

  • 31.556”
  • (4.669)
  • 30.352”
  • (4.777)
  • 30.780”
  • (4.732)
  • 31.228”
  • (4.723)
  • 11.011“
  • (3.682)
  • 10.696”
  • (3.709)
  • 10.956"
  • (3.672)
  • 10.824”
  • (3.723)

N

365

365

365

365

365

365

365

365

AIC

2,628.065

2,639.116

2,638.674

2,631.101

2,454.627

2,454.386

2,453.611

2,457.409

*p < 0.1; ‘p < 0.05; **p < 0.01; ***p < 0.001.

TABLE 10.3

Logistic and Linear Mixed-Effects Models Predicting Substance Use, Cultural Participation, and Cultural Discrimination

Substance Usea

Cultural Participation

Cultural Discrimination

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

Female

1.19

1.17

1.18

1.17

  • 0.162
  • (0.22)
  • 0.156
  • (0.22)
  • 0.158
  • (0.22)
  • 0.153
  • (0.22)
  • 0.274
  • (0.57)
  • 0.267
  • (0.57)
  • 0.290
  • (0.57)
  • 0.251
  • (0.57)

Age

1.04

1.03

1.03

1.03

  • -0.034
  • (0.12)
  • -0.032
  • (0.12)
  • -0.037
  • (0.12)
  • -0.035
  • (0.12)
  • -0.804'
  • (0.31)
  • -0.788'
  • (0.31)
  • -0.800'
  • (0.31)
  • -0.794'
  • (0.31)

Know interviewer

0.57

0.57

  • 0.112
  • (0.27)
  • 0.111
  • (0.27)
  • 0.597
  • (0.65)
  • 0.538
  • (0.66)

Present third party

1.19

1.22

  • -0.156
  • (0.26)
  • -0.166
  • (0.26)
  • -0.684
  • (0.64)
  • -0.648
  • (0.65)

Interviewed in home

0.97

0.92

  • 0.067
  • (0.26)
  • 0.093
  • (0.26)
  • -0.017
  • (0.62)
  • 0.075
  • (0.63)

Constant

0.10

0.09

0.09

0.10

  • 2.524'
  • (1.15)
  • 2.655'
  • (1.16)
  • 2.555'
  • (1.15)
  • 2.625'
  • (1.16)
  • 20.605'"
  • (2.88)
  • 21.110'"
  • (2.91)
  • 20.718'"
  • (2.88)
  • 20.982'"
  • (2.92)

N

365

365

365

365

365

365

365

365

365

365

365

365

AIC

195

197

197

199

1,599

1,599

1,599

1,603

2,275

2,275

2,276

2,278

Odds ratios presented.

•p <0.1; * p < 0.05; **p < 0.01; *** p < 0.001.

 
<<   CONTENTS   >>

Related topics