Table 1.
Description of primary and secondary outcome measures
| Outcome | Measure Description | Assessment Time Point | Data collection method and source |
|---|---|---|---|
| Mentor Attitude | |||
| aPrimary outcomes | |||
| Cultural humility | Mean score of Empathic Feeling and Expression Subscale of the Ethnocultural Empathy Scale (α = .90; 15 items) [30]. Items are rated on a 6-point Likert-type scale from 1 = strongly disagree to 6 = strongly agree. Subscale has been previously used in youth mentoring [31] | Baseline & 4.5 months | Mentor survey |
| Self-efficacy beliefs for providing ethnic/racial support | Mean score of self-efficacy beliefs for providing ethnic/racial support scale (α = .89 -.94; 6 items) [25]. Responses are on a 4-point Likert-type scale from 1 = not at all confident to 4 = very confident. Subscale demonstrated sensitivity to change in evaluations of mentor race/equity training [23, 25] | Baseline & 4.5 months | Mentor survey |
| Self-efficacy beliefs for engaging in conversations about racial injustice | Created for current study. Mean score of self-efficacy beliefs for engaging in conversations about racial injustice (4 items) [32]. Responses are on a 4-point Likert-type scale from 1 = not at all confident to 4 = very confident | Baseline, 4.5 months & 12 months | Mentor survey |
| Perceived knowledge | Mean score of adapted measure of Multicultural Awareness, Knowledge and Skills scale (α = .75; 13 items) [33]. Participants are asked to assess each skill on a scale from 1 = very limited to 4 = very good | Baseline, 4.5 months & 12 months | Mentor survey |
| Colorblind racial attitudes | Mean score of Unawareness of Racial Privilege subscale of the Color-Blind Racial Attitudes Scale ((CoBRAS) 7 items) indicated sensitivity to intervention in initial validation (α = .71-.83) [34]. Responses are on a 6-point Likert-type scale from 1 = strongly disagree to 6 = strongly agree. Higher scores reflect more unawareness. Subscale previously used in evaluation of mentor race/equity training [25] | Baseline & 4.5 months | Mentor survey |
| Secondary outcomes | |||
| Bias awareness | Mean score of adapted Bias Awareness Scale (4 items). Responses are on a 6-point Likert-type scale from 1 = strongly disagree to 6 = strongly agree. Initial scale validation with a sample of white, undergraduate students demonstrated good internal consistency (α = .81) and discriminant validity with measures of implicit and explicit prejudice [35] | Baseline & 4.5 months | Mentor survey |
| Social justice interest | Mean score of adapted Social Justice Interest subscale of the Social Issues Questionnaire (α = .81; 6 items) [36]. Responses range from 1 = not at all important to 6 = very important. Subscale previously used in evaluation of mentor race/equity training [25] | Baseline & 4.5 months | Mentor survey |
| Mentor-Youth Interactions and Relationship Quality | |||
| Primary outcomes | |||
| Mentor attunement | Mean score of Mentor Attunement Scale [37]. Mentor and youth scales have 7 and 15 items respectively. Response scales range from 1 = never to 5 = always. Validation of mentor scale had good internal consistency (α = .91) | 12 months | Mentor and youth surveys |
| Mentoring activities on race and social justice | Created for current study. Mean score on a scale about the frequency with which mentors and youth engage in activities related to race and social justice (9 items) [32]. Response scale ranges from 1 = never to 5 = all the time | 12 months | Mentor survey |
| Cultural clashes | Two items assessing the extent to which they experience challenges in bridging racial/ethnic differences with their mentee and with the mentee’s family [6]. Response scale is from 1 = not at all challenging to 4 = very challenging | 12 months | Mentor survey |
| Ethnic/racial socialization in mentoring | Created for current study. Mean scores of Ethnic/Racial Socialization in Mentoring scale (6 items) [32]. Response scale ranges from 1 = never to 5 = all the time | 12 months | Mentor survey |
| Mentor support for ethnic/racial identity | Mean score of the adapted Mentor Support for Ethnic/Racial Identity Scale (7 items) [13]. Scale used with adolescent girls of color (Mage = 11.75) demonstrated fair to good internal consistency (α = . 77 to .87). [13] | 12 months | Youth survey |
| Mentor support for coping with discrimination | Created for current study. Mean score of the Mentor Support for Coping with Discrimination Scale (5 items) [32]. Response scale ranges from 1 = not at all true to 4 = very true | 12 months | Youth survey |
| Youth-centered relationships | Mean score of the Youth Centeredness scale (5 items) [38]. Response scale ranges from 1 = not true at all to 4 = very true. Scale used with a racially/ethnically diverse sample of youth (M age = 11.15) demonstrated acceptable internal consistency (α = .70) [39] | 12 months | Youth survey |
| Secondary outcomes | |||
| Mentoring relationship duration | Duration of the youth-mentor relationship is calculated based on date of the first mentor-youth meeting and the date of the match’s closing or (for ongoing matches) the date the information was retrieved from BBBS records | 12 months | Agency case records |
| Emotional closeness—Youth | 1 item asking how close does youth feel to their mentor. Response options range from 1 = not close at all to 10 = extremely close | 12 months | Youth survey |
| Emotional closeness—Mentor | Mean score of the closeness subscale of the Match Characteristics Questionnaire (α = .83; 4 items) [40]. Responses range from 1 = never to 5 = always. Scale used in prior research with volunteer mentors [40] | 12 months | Mentor survey |
| Relationship satisfaction | Mean score of the mentor’s satisfaction with the mentoring relationship subscale of the Investment Model Scale (5 items) [41]. Likert-type response scale ranging from 1 = strongly disagree to 4 = strongly agree. Research has used this scale with volunteer mentors (α = .85). [42] | 12 months | Mentor survey |
| Authentic self | Mean score of the modified Level of Adolescent Voice Scale (8 items) [43]. Response scale ranges from 1 = really true for me to 4 = really not true for me | 12 months | Youth survey |
| Advocacy | Mean score of the Advocacy subscale of the Mentoring Processes Scale (MPS): Mentor Version [44]. Likert-type response scale ranging from 1 = strongly disagree to 4 = strongly agree | 12 months | Mentor survey |
| Support seeking | Mean score of the Support Seeking scale (α = .88; 4 items) [45]. Likert-type response scale ranging from 1 = strongly disagree to 4 = strongly agree | 12 months | Mentor survey |
| Strength of relationship | Mean score of the youth and mentors’ satisfaction in the relationship. Youth measure has 10 items (α = .79) and mentor measure has 14 items (α = .85). [46]. Likert-type response scale ranging from 1 = strongly disagree/not at all true to 4 = strongly agree/completely true | 12 months | Agency case records, SOR |
| Youth Outcomes | |||
| Primary outcomes | |||
| Openness to adult support | Mean score of the modified Interpersonal Trust Scale (5 items) [47]. Likert-type response scale ranging from 1 = strongly disagree to 4 = strongly agree. Scale has been used with Latinx adolescents (α = .73) and has been found to be correlated with mentoring [48] | Baseline and 12 months | Youth survey |
| Community connectedness | Mean score of the modified Community Connectedness Scale (5 items) [49]. Likert-type response scale ranging from 1 = strongly disagree to 4 = strongly agree. The scale has been used with adolescents (α = .72) [49] | Baseline and 12 months | Youth survey |
| Mattering | Mean score of the adapted Sense of Mattering Scale (5 items) [50]. Likert-type response scale ranging from 1 = strongly disagree to 4 = strongly agree. The scale had good internal consistency in samples of adolescents and emerging adults (α > .80) [51] | Baseline and 12 months | Youth survey |
| Ethnic/racial identity (ERI) | Mean scores of each of two subscales adapted from the Multidimensional Inventory of Black Identity -Teen (MIBI-T). The subscales are the private regard subscale (α = .76; 3 items) and public regard subscale (α = .77; 3 items [52]. Likert-type response scale ranging from 1 = strongly disagree to 4 = strongly agree. Scale was initially developed to examine ERI with an ethnically/racially diverse sample of sixth graders (Mage = 11.5) | Baseline and 12 months | Youth survey |
| Ethnic identity (exploration & affirmation) | Mean scores of each of two subscales of the Ethnic Identity Scale (EIS) [53]. The subscales assess exploration (α = .79-.89; 3 items) and affirmation (α = .70-.75; 3 items). Response scale ranges from 1 = does not describe me at all to 4 = describes me very well. Scale has been previously used in longitudinal research with Latino youth | Baseline and 12 months | Youth survey |
| Coping efficacy with discrimination | Mean score of the modified Coping Efficacy with Discrimination Scale (α = .78-.91; 3 items) [19]. Response scale ranges from 0 = very badly to 3 = very well. Prior mentoring research has used this scale with low-income, Latinx adolescents (Mage = 15) [19] | Baseline and 12 months | Youth survey |
| Cultural mistrust | Mean score of the Cultural Mistrust Inventory for Adolescents, adapted from the cultural mistrust in interpersonal relationships subscale (11 items) [54]. Likert-type response scale ranging from 1 = strongly disagree to 4 = strongly agree. Scale demonstrated fair to excellent internal consistency in mentoring research with adolescent girls of color (α = . 78 to .91) [13] | Baseline and 12 months | Youth survey |
| Secondary outcome | |||
| Academic performance | Mean of youth self-reported grades for 4 academic subjects: math, reading, social studies, science. A through F scale | 12 months | Agency case records, YOS |
| Risky behaviors | 3 subscales regarding risky behaviors youth have engaged in during the last 3 months: substance use (4 items adapted from Brown) [55], status offenses (e.g., skipping school; 3 items created for survey), school discipline (4 items adapted from Vandell) [56]. Mean scores are calculated for each subscale. Responses are on a 3 point scale: 0 = 0 times, 1 = 1 or 2 times, 2 = 3 or more times | 12 months | Agency case records, YOS |
| Bullying | Mean score on the frequency with which youth has engaged in bullying during the past month (5 items; adapted from Bosworth [57] and Dahlberg [58]). Responses are on a 3-point scale: 0 = 0 times, 1 = 1 or 2 times, 2 = 3 or more times | 12 months | Agency case records, YOS |
| Police & juvenile justice contact | Sum score on youth contact with the police and juvenile justice system (3 items; 0 = No, 1 = Yes; created by BBBS) | 12 months | Agency case records, YOS |
| Emotional regulation | Mean score on 4-item scale; responses range from 0 = not at all true to 3 = totally true (adapted from the Middle School Survey [59] | 12 months | Agency case records, YOS |
| Depressive symptoms | Mean score on 10-item scale; responses range from 0 = Never to 4 = Almost Always [60] | 12 months | Agency case records, YOS |
| Loneliness | Mean score of the adapted Loneliness Questionnaire-Short Version (9 items) [61]. Likert-type response scale ranging from 1 = strongly disagree to 4 = strongly agree. The measure had excellent internal consistency in a sample of adolescents (α = . 92) [61] | Baseline and 12 months | Youth survey |
| Hopeful future expectations | Mean score of the modified abbreviated version of the Hopeful Future Expectations Scale (8 items) [62]. Response scale ranges from 1 = very low to 5 = very high. The original measure had acceptable internal consistency in samples of adolescents and young adults (α’s = .71 to 0.95) [63] | Baseline and 12 months | Youth survey |
| Critical reflection about societal inequality | Mean score of the Critical Reflection Subscale of the short Critical Consciousness Scale (3 items) [63]. Likert-type response scale ranging from 1 = strongly disagree to 4 = strongly agree. The measure had acceptable internal consistency in samples of adolescents (α’s > .70) [64] | Baseline and 12 months | Youth survey |
| Growth mindset | Mean score of the Revised Implicit Theories of Intelligence (Self-Theory Scale; 8 items) [64]. Likert-type response scale ranging from 1 = strongly disagree to 4 = strongly agree. The scale was developed to be used with children and teens, and has demonstrated acceptable internal consistency in research conducted with youth of color (α = . 75) [65] | Baseline and 12 months | Youth survey |
| Academic self-efficacy | Mean score of the Academic Self-Efficacy Subscale of the Patterns of Adaptive Learning Scale (5 items) [66]. Likert-type response scale ranging from 1 = strongly disagree to 4 = strongly agree. Internal consistency was acceptable (α = .78) in a sample of adolescents [66] | Baseline and 12 months | Youth survey |
Higher scores on each measure mean higher levels of that measure
YOS Youth Outcomes Survey. SOR = Strength of Relationships
aWe acknowledge that there are many primary outcomes which could affect study power. Of the various primary outcomes, we believe that the most primary are the mentor attitudinal outcomes given that mentors are the direct recipient of the intervention activities, and these are intended to have their most immediate impact on mentor attitudes