Youth mentoring: the motivation of volunteering and the validity of the volunteer function inventory


Teye, ACPeaslee, L. (2020). Why mentor? A validation study of the volunteer functions inventory for use in youth mentoring. Journal of Community Psychology, 119.

Summarized by Ariel Ervin

Notes of Interest: 

  • Examines the validity of the Volunteer Functions Inventory, as well as the functional motivations of youth mentoring program volunteers 
  • 473 participants/mentors were matched into a one-on-one mentoring program
  • Mixed methods study (factor analyses and semi-structured interviews)
  • Results indicate the usefulness of a five-factor structure model
  • Qualitative analyses indicate that there are some new areas to explore for the modern volunteer population

Introduction (Reprinted from the Abstract)

This study explores the functional motivations of volunteers in youth mentoring programs and tests the validity of the Volunteer Functions Inventory (VFI) for this population. Participants included 473 volunteer mentors matched within a one‐to‐one mentoring program. The study utilizes a mixed‐methods approach to validate the VFI, which was administered before the match. Exploratory and confirmatory factor analyses were employed to test for content validity and determine fit. These were complemented by an emergent theme analysis from a semi‐structured interview, which included questions assessing mentor motivations. Findings suggest a five‐factor structure model best captures functional motivations. In addition, results from the qualitative analysis suggest there may be previously unexplored domains to consider for contemporary volunteer populations. These include community/civic responsibility, organizational structure and reputation, and self‐concept. Future research should seek to refine scales that are uniquely predictive of mentor motivations to expand practical applications for use.


Implications (Reprinted from the Discussion)

Previous research findings suggest that the VFI may be a useful tool for the youth mentoring field as a means of understanding mentor motivations for volunteering. Consistent with previous research across a range of volunteer fields, mentors in this study were predominantly motivated by Values (Chacón et al., 2017). This confirmed our intuition and other noted research findings that volunteers in youth development programs have greater motivations toward altruism and are more inclined to seek opportunities for learning and targeting skill development (though social desirability bias likely inflated responses in this category). However, work here revealed that the VFI Values domain lacked internal consistency and possibly depth in capturing a range of humanitarian and prosocial motives. These are explored further in our discussion on qualitative findings. Other results presented here suggest that the traditional six‐factor VFI structure proposed by Clary et al. (1998) and adopted by many others does not best describe the functional motivations of youth mentors. Findings may also raise questions about the applicability or the use of the VFI, as currently scaled, among the emerging adult volunteers more generally, as also noted by previous research (e.g., Francis, 2011; Hyde & Knowles, 2013). Still, some caution should be exercised in generalizing findings presented here to the broader mentoring population, given that the demographic composition of mentors in this study was predominately White, female, college students.

While initial reliability estimates were satisfactory, the six‐factor model did not withstand validity analysis. Moreover, key results from construct validity tests were in sync with much of the literature, suggesting alternative approaches may be a better fit. First, Enhancement and Protective scales should be reconceptualized as a single scale, excluding items that do not conceptually fit. Second, a few items consistently noted in the literature were confirmed here as poorly fitting with their corresponding scales (i.e., Item 15 and Item 29). Other items removed suggest there may be specific trends in the youth mentoring field. However, we would suggest further research to confirm these findings. Third, previously suggested factorial structures should be ruled out as predictive among the volunteer mentoring group. Specifically, the General Factor model and the Two‐Factor model failed to meet CFA diagnostic criteria. Given previous literature also failed to confirm these factorial structures, researchers might consider moving on from the consideration that all items on the VFI fall on a single construct or that they can be perceived dichotomously as altruistic or egotistic in nature. Rather, quantitative findings confirmed a corrected five‐factor model, dropping five items from the original VFI, as most explanatory for the study population. This structure may also work well for broader populations given similar findings elsewhere. Moreover, results from the MGCFA confirm configural invariance by gender, college‐attending status, and ethnicity. Future research in this area may look to apply more stringent invariance testing (e.g., metric) that constrains factor loadings. However, challenges with a low number of observations among men and community members limited the ability to perform these tests here. Research broadly confirms the notion that χ2 values in CFA are highly sensitive to sample size (Milfont & Fischer, 2010). Future research should focus on using this approach of understanding trends among these and other subgroups.

Qualitative findings presented in this paper suggest there may be more domains to consider for contemporary youth mentoring populations. These include community/civic responsibility, organizational structure and reputation, and self‐concept. The inclusion of these indicators may be essential for understanding volunteer functional motivations, which have been directly linked to volunteer engagement and satisfaction, as well as to framing effective volunteer marketing and recruitment efforts. In addition, inconsistencies between qualitative and quantitative findings may point another important conclusion. The VFI may be capturing domains that are not central among mentoring volunteers. Although, when prompted, mentors rated Understanding items highly, reflecting a more socially desirable tone, they were unlikely to offer responses connected with the domain during the semi‐structured interview. Thus, we might conclude that learning new things or gaining new perspectives were not highly salient functional motivations. Although not fully explored here, we believe that social desirable may overstate the degree to which mentors are motivated by Values and understate motivation driven by career, or Enhancement and Protective factors. Future research should be focused on further refining scales that are more reflective of a vast range of mentor motivations. This may be key to developing effective mentor recruitment messages. In addition, an important next step in mentor motivation research will be to expand our understanding of how distinct motivational domains are associated with mentor match preferences, match length, and relationship quality.


To access this article, click here.