Instrumental Mentoring for Young Adults: A Multi-Method Study

Closeness between mentor and mentee is previously defined as an important indicator of relationship quality in youth mentoring, but whether this is the case in instrumental mentoring for young adults remains unclear. This is an exploratory study examining how instrumental mentoring serves young adults in their instrumental needs and how relational closeness develops. We applied a mixed-methods design, using quantitative data from a study of an instrumental mentoring program in Rotterdam, The Netherlands (N = 53), and qualitative data from a subsample of participants (N = 10). Two statistically distinctive clusters of closeness were found; 49% of the mentees reported high levels of closeness, and 51% reported low levels of closeness in their mentor relationship. Multivariate analyses of variance (MANOVAs) showed that the cluster with high levels of closeness was correlated with instrumental compatibility, satisfaction, and perceived attitude similarities. Semi-structured interviews were used to illustrate the role and development of closeness for mentees in both clusters, and three cases were presented. Experiencing closeness seemed a result of receiving instrumental support, not a precondition. Mentees’ previous experiences might in some cases explain the lower levels of closeness, but this did not always hinder mentees to profit from their mentors’ support.


Introduction
Positive relations with supportive adults are considered essential in youth development (Laursen & Birmingham, 2003;Theokas & Lerner, 2006). However, youths who experience both individual and environmental difficulties, in combination with insufficient protective factors, are considered at risk of various negative outcomes, such as school dropout, unemployment, or mental health problems (Jenson & Fraser, 2015). These risks may hinder youths to reach their full potential and effectively participate in society. Moreover, these youth's networks are often overburdened, or not able to provide specific forms of support (Schenk et al., 2018). Pairing youth with a caring, nonparental adult who puts the youth's need central, and meeting regularly, is assumed to be a preventive way of supporting youths. Mentoring programs are mainly focused on children and youths up to 18 years old. Young adults (age 18-28), however, may profit from mentoring too. Increasing calls upon self-sufficiency may be extra hard for this age group with multiple problems and a limited social support network. A mismatch between young adults' needs to become self-sufficient and the necessary contextual resources to do so, may be bridged by the support of a mentor.
Most mentoring research is based on mentoring programs that use a developmental approach which emphasizes a close, long-lasting relationship as the primary mechanism of mentees' development (Rhodes, 2005). Mentees' selfesteem, for example, is believed to increase through the presence and affirmation of a mentor (Rhodes, 2005). A close relation between mentor and mentee, therefore, is essential to mentoring with a developmental mentoring approach. For this close bond to arise, spending time together and having fun are the primary ingredients of developmental mentoring. Empirical research, indeed, shows that in this type of mentoring, close mentor relationships are associated with better youth outcomes (Cavell & Elledge, 2014;Kanchewa et al., 2018;Karcher et al., 2002). Although a close bond seems necessary to profit from mentoring, it is suggested that a relationship-based approach alone may not adequately address certain youth's needs (e.g., Bowers, 2019;Rhodes, 2019). In addition, outcomes of a recent meta-analysis showed that relationship-based programs yield smaller effect sizes than more targeted approaches (Christensen et al., 2020). The instrumental mentoring approach, unlike the developmental approach, facilitates space for these insights. The focus in this approach is on setting, pursuing, and achieving goals, such as improving mentees' competencies or school grades (Eby et al., 2007;Karcher & Nakkula, 2010). The mentor's behavior then is aimed at helping a mentee reach those goals. In this approach, goal-focused mentoring activities are perceived as equally important as the development of a close bond. It is very well possible that this is especially advantageous for young adults in mentoring programs. It may be that mentors who guide young adults through the obstacles that may accompany their transition to adulthood fit their needs better than mentor relations aimed at emotional development exclusively.
To date, research on instrumental mentoring is limited, and in particular for young adults (Balcazar & Keys, 2014). The central purpose of this study is to examine how goal-focused activities and relational closeness affect relationship quality in instrumental mentoring for young adults. We start by describing the features of instrumental mentoring and how this approach may be better suited for young adults, and thereafter will focus on the role that closeness may play in this type of mentoring.

Instrumental Mentoring for Young Adults
Whereas increased competencies and skills may be a result of the growing interpersonal mentor-mentee relationship in developmental mentoring, in instrumental mentoring increasing competencies and skills is the primary goal. In instrumental mentoring, the mentor supports the mentee to accomplish particular goals (e.g., increasing academic skills or building career knowledge) by providing advice, guidance, explanations, or suggestions (Karcher et al., 2006). Instrumental mentoring is often related to domains that are key to increase self-sufficiency, such as education, work, and mental health (Bannink et al., 2015). Youth in formal mentoring programs often experience individual and environmental difficulties, which hinders them to become self-sufficient (Herrera et al., 2013). Aiming for an effective match with the labor market, reducing debts, or addressing health problems are examples of important issues for young adults in mentoring programs. In instrumental mentoring, important goals are made explicit which allows for greater intentionality and definable structure of the mentoring, instead of spending time together to form a close bond. Advice and support from mentors can result in more knowledge, access to social resources, and self-confidence and eventually increased self-sufficiency.
Instrumental mentoring may not only be more suitable for mentees with personal and environmental challenges, but it also seems more appropriate to the developmental stage young adults are in. Setting clear, common goals in mentoring is emphasized as a promising starting point for young adults (Darling, 2005;Noam et al., 2014). Young adulthood (also referred to as emerging adulthood) is characterized by more transitory and inconsistent states and requires youth to become active agents to construct their future lives (Arnett, 2004;Shulman & Nurmi, 2010). Young adults, thus, are more likely to benefit from working on concrete, common goals in their transition to independence, compared to younger mentees (Darling, 2005;Hamilton & Hamilton, 2005;Musick, 1999;Noam et al., 2014). It is more effective for young adults to develop a relationship around these shared goals than to have relationship development as a separate, primary starting point (Hamilton & Hamilton, 2005). For young adults experiencing multiple obstacles in life, thus, instrumental mentoring seems to better address their needs that come with their developmental stage and challenges than developmental mentoring.

The Role of a Close Bond in Instrumental Mentoring for Young Adults
Closeness between mentor and mentee is often how relationship quality in mentoring is specified (e.g., De Wit et al., 2019;Lyons et al., 2019;Raposa et al., 2016;Rhodes, 2005). Closeness, in this sense, refers to mentees' feeling of a close bond with their mentor, and of being able to share (negative) experiences and concerns. Rather than relying on developing closeness alone, instrumental mentoring aims to combine goal-directed and relational activities to establish relationship quality. Research suggests that it is often the combination of goal-orientated activities and mentor-mentee closeness that is most effective in instrumental mentoring. To illustrate, Nakkula and Harris (2010) found the combination of goal-focusing and sharing thoughts and emotions to correspond to the greatest degree of mentees' satisfaction with their mentor in a Big Brothers Big Sisters program. In fact, the focus on goals solely, without the sharing aspect of the relationship, compromised the relationship quality.
The assumption that involving activities to develop a close and lasting relation in instrumental mentoring is most effective was tested in a schoolbased mentoring program (Lyons et al., 2019;McQuillin & Lyons, 2016). Results showed that a combination of instrumental approaches along with the development of a close relationship had the largest effects on mentees' outcomes. Another study showed that how mentors support their mentees is strongly related to the close bond mentees report (Lyons & Perrewé, 2014).
The authors conclude that even though a close bond and mentoring support behavior are two distinct constructs, they are hard to separate. This so-called sweet-spot of instrumental activities and the development of a close bond (Lyons et al., 2019), and the finding that perceived support and affective perceptions coincide (Lyons & Perrewé, 2014), provides new evidence for using hybrid models of mentoring. However, studies on this hybrid model are scarce and based on school-based mentoring programs (Lyons et al., 2019;McQuillin & Lyons, 2016) or postgraduate students (Lyons & Perrewé, 2014). Other relevant research on closeness in mentoring has been done by Liao and Sánchez (2019) and Hurd and Zimmerman (2014). Both studies showed how closeness is associated with various outcomes. The presence of relational closeness was a precondition for young adults to perceive benefits on psychological outcomes (Hurd & Zimmerman, 2014). Close and growthoriented relationship profiles were associated with various outcomes such as motivation, aspirations, and grades (Liao & Sánchez, 2019). However, both studies are based on natural (informal) mentoring relations, where youths identified someone other than their parents who provides additional support. By definition, these are people they already know. Second, natural mentoring often refrains from setting goals. The establishment and content of a relationship, thus, is different from formal mentoring programs that use an instrumental approach.

Constructs That Make Up Relationship Quality
Although relational closeness is likely to remain an important indicator relationship quality, there are additional aspects of relationship quality that need consideration in instrumental mentoring, such as mentees' satisfaction with the mentors' effectiveness in supporting their mentee in goal attainment. A construct that is often associated with a close mentor-mentee bond is perceived similarities between mentor and mentee. Shared characteristics have been linked to relationship quality, in general (Byrne, 1971), and in mentoring relations, in particular (Allen & Eby, 2003;Raposa, Ben-Eliyahu, Olsho, & Rhodes, 2019). To illustrate, in a developmental mentoring program similar racial and ethnic backgrounds of mentor and mentee were predictive of longer lasting mentoring relations (Raposa et al., 2019). In practice, mentoring programs frequently match higher-class position mentors with lowincome youth from minority backgrounds (Tierney & Grossman, 1995). With few background similarities between mentor and mentees, bridging these differences might hinder the development of a close bond. Within the field of instrumental mentoring, however, research on perceived similarities is scarce. With shifting point of views on the role of a close bond in instrumental mentoring, the question arises how important perceived similarities are. In other words, working on set goals in the instrumental approach might not be hindered by differences between mentor and mentee.
In contrast, mentees' perception of the supportive role of the mentor might be more important than similarities and levels of closeness for their satisfaction with the relationship . Whereas open discussions and problem solving skills seem important for younger youths, the need for structured and meaningful activities grows increasingly important in late adolescence and young adulthood (Larose et al., 2010). Mentors need to be able to provide these activities. In addition, a study in an academic context showed that perceived effectiveness of the mentors' support leads to increased levels of satisfaction with the relationship (Lyons & Perrewé, 2014). Thus, with increasing age, the way a mentor can contribute to and support the mentee's goals may lead to satisfaction in the relationship. Satisfaction and compatibility in this way might also be better indicators of relationship quality in instrumental mentoring for young adults. Another feature of instrumental mentoring may be that a close bond does not precede effective support of a mentor, but conversely, by working on goals, a close bond between mentor and mentee can arise (Hamilton & Hamilton, 2005). Satisfaction and the perception of how mentors contribute to the set goals in this way might be a better indicator of the mentor relation, than the presence of a close bond.
Young adults who have been in contact with many social workers and who find it hard to trust new people may have negative expectations or may be resistant to develop a close bond (Barnhoorn et al., 2013;DiGiuseppe et al., 1996;Lenkens et al., 2020). Indeed, Raposa et al. (2016) found that youths' multiple stress factors at the individual and environmental level affect their ability to form a close and lasting bond with their mentor. Also, risk factors such as behavioral problems or drug use predict lower levels of satisfaction in the mentor relationship and early match closure (Kupersmidt et al., 2017;Raposa et al., 2016). For older youth with direct needs regarding self-sufficiency, moreover, meaningful activities seem to be a more fulfilling and natural way of interacting with a mentor instead of spending time together in funfocused activities (Larose et al., 2010). Taking into account young adults' characteristics may offer new insights in how young adults perceive their mentors' effectiveness. Furthermore, more knowledge on the role of a close bond as a requirement of mentoring, or the result of effective mentoring, is needed.

The Present Study
Young adults' needs and developmental characteristics should be taken into account when studying mentoring. This exploratory study examines how instrumental mentoring serves young adults and how relational closeness develops together with goal-focused activities. To do so, we apply a multimethod approach. First, we study the levels of experienced closeness and how these levels of closeness are related to relationship indicators such as satisfaction, perceived compatibility, and similarities. In addition, we provide three case studies to illustrate the development of closeness and how youth experience various levels of this closeness. As the instrumental approach of mentoring yields better results than developmental mentoring practices (Christensen et al., 2020), it is relevant to examine the role and development of relational closeness in instrumental mentoring, as it seems understudied in former research. In addition, the developmental status and direct needs of youths may influence their perceptions of closeness with their mentor. Gaining more insight in what the sweet-spot of relational and instrumental activities might entail for young adults may have implications for mentoring programs with this target population.

Design
We adopt a mixed-methods design, using both quantitative and qualitative data to study instrumental mentoring for young adults. Adding qualitative interviews to quantitative data allows for participants' perspectives on relationship quality that the deductive methods do not take into account (Johnson & Onwuegbuzie, 2004). More specifically, by using cluster analysis, we investigate whether we can differentiate different subtypes of mentoring relations. We then illustrate the validated clusters by presenting case studies, based on interviews with mentees. We have selected three participants whose interviews were rich enough to vividly demonstrate their perception of the development of their mentoring relation.

Participants
Mentees were recruited from a local mentoring program in Rotterdam, The Netherlands. The program is aimed at young adults (age 18-28), who are mainly referred to the mentoring program at the municipal agency for young adults (in Dutch: Jongerenloket). At this site, young adults can get legal support when they, for instance, want to go back to school or apply for social welfare. Young adults' self-sufficiency in the most important life domains such as income, daily activities, addiction, justice, social support, housing, mental health, and employment is assessed. When there are multiple problems in one or more of these domains and no protective factors in their immediate environment, young adults are considered at risk of a variety of outcomes that hinder them to participate in society or reach their full potential. These young adults are introduced to the mentoring program and can choose to sign up to enter the program voluntarily and for free. Many participants have received professional assistance in their current and past lives, but the mentoring program is distinct in that it is based on young adults' voluntary involvement, their own formulated needs, and on volunteers (cf., professionals). The mentoring program adopts an instrumental approach for young adults with a specific request for support in self-sufficiency, such as reducing debts or finding employment. Support is provided through one-on-one mentoring whereby a mentor is linked to a young adult to meet and support the mentee on a regular basis. The mentoring program recruits mentors who are highly educated and are highly active in working life. Mentors are initially screened by the program staff and then matched with mentees based on personality and shared interests. Mentors receive a neuro-linguistic programming training and information session on practical subjects such as debt restructuring. At the start of the relation, mentor and mentee set goals to work toward together. The program supports mentors and mentees through digital contact along with face-to-face interactions.
Interviews were conducted with a subsample of participants in the quantitative study and included 10 participants. At the time of the interview, participants' mean age was 25.3 years. Five participants identified as men, six as Dutch, two as Antillean, one as Moroccan, and one as Surinamese. Scores on the relationship quality measures did not differ between the qualitative subsample and main sample. Characteristics of both samples are presented in Table 1.

Procedure
All mentees enrolled in the program were contacted by a researcher. The researchers provided information about the study and invited mentees to participate in the quantitative study. Mentees who did not respond within 2 weeks received a reminder email, and a second after another week. After informed consent, participants could fill in the online questionnaire on their mobile phone, computer, or tablet, on their own preferred time and place. Participants were compensated with 15 euro after completion of the questionnaire.
Qualitative interviews were conducted with a small sample of mentees who filled out the online survey. All participants of the quantitative study were eligible for participating in our qualitative study. They were contacted several times via email and invited for an interview for another 15 euro compensation. Ten mentees responded and were interviewed. Interviews took place in a separate room of the mentoring programs' office, a central place that most of the mentees were familiar with, or in the central library. An interview protocol was designed with guidelines for the structure of the interview and information for the interviewees. Written informed consent for the interview and use for scientific purposes was obtained and the interviews were audio-recorded after verbal consent. The purpose of the interview was described, as was the confidential character of the interview. Participants were told that they could (temporarily) stop the interview when necessary, or that they could leave a question unanswered. Semi-structured interviews were conducted by a research assistant and the first author and lasted between 45 and 60 minutes each. Participants were asked to come up with a pseudonym for themselves and their mentor under which the interviews were transcribed and presented in the current study. Approval of the design of the study was obtained from the institutional ethics board.

Instruments
Closeness, Instrumental compatibility, and Satisfaction were adapted from the Match Characteristics Questionnaire for college mentees (MCQ; Harris & Nakkula, 2018) and translated to Dutch (see the appendix). All questions in these scales were rated using a 6-point Likert-type scale ranging from I totally disagree to I totally agree.
Closeness. The level of closeness that mentees perceive was measured with five items from the scale "Closeness" and "Personal Support" from the MCQ for college mentees (Harris & Nakkula, 2018). Items included, for example, "My mentor and I have a close bond" and "We talk about negative or stressful things that were happening in my life." A mean score was created based on these items, with higher scores indicating higher levels of closeness. Cronbach's alpha of the scale was .74.
Instrumental compatibility. How mentees perceived their mentors to be compatible with their needs was measured with the scale "Instrumental compatibility" from the MCQ (Harris & Nakkula, 2018). This scale consists of three items such as "My mentor is well-suited to help me with the most important challenges in my life." A mean score was created based on these items, with higher scores indicating more instrumental compatibility. Cronbach's alpha was .63.
Satisfaction. Mentees' satisfaction with their mentor is measured by the "Satisfaction" scale of the MCQ (Harris & Nakkula, 2018). It includes four items, for example, "This year would have been much harder for me if I had not had my mentor." A mean score was created based on these items, with higher scores indicating greater satisfaction. Cronbach's alpha was .59.
Perceived similarities. How mentees perceived similarities between them and their mentor was measured using the Homophily Scale (McCroskey et al., 2006). The questionnaire consists of two scales. The background homophily scale (α = .71) consists of six items questioning the similarities in background (economic and social status), such as "My mentor has a different background than me." The attitude homophily scale (α = .81) consists of 15 items, such as "My mentor and I share the same values." Questions were rated using a 7-point Likert-type scale ranging from Totally disagree to Totally agree. A mean score was created based on these items, with higher scores indicating more perceived similarities.
Interviews. The first author and a research assistant conducted the interviews. Both were trained in doing interviews, and the first author was experienced in conducting interviews with this target group in particular. We had no prior relationships with participants before the interviews. We used semi-structured protocols with questions designed to elicit perceptions of relationship quality. Topics addressed in the interviews were how mentees experienced the beginning of their mentor relationship, the development of the relation over time, reasons for having a mentor, goals, goal attainment, and similarities and differences between them and their mentor. Interviewers asked openended questions, followed by follow-up questions. Participants were as well provided with the opportunity to talk about negative aspects of their relationship in a hypothetical way, to avoid participants' tendency to mainly talk about positive aspects of their relationship. For example, "Suppose you are the director of this mentoring program, what would be your main concern in matching mentors to mentees?" and "Describe what you in general think is a good mentor."

Data Analyses
A concurrent mixed-method design was used to answer our research questions (Leech & Onwuegbuzie, 2009). To find clusters of observations with similar values on the close bond items (see the appendix), a cluster analysis was carried out using IBM Statistical Package for Social Scientist (SPSS 24). This way, clusters are created such that the within-cluster differences are as small as possible, and differences between clusters are maximized (Pastor et al., 2007). We followed a two-step cluster analyses (Gordon, 1999). In Step 1, we used agglomerative hierarchical techniques for small sample sizes and Ward's method for combining clusters (Rapkin & Luke, 1993;Ward, 1963). We determined the cluster solution based on the number of cases within clusters, stability of solutions, interpretability, and distinctiveness of the clusters (Rapkin & Luke, 1993). In Step 2, we validated the clusters found in Step 1, using nonhierarchical k-means clustering. Here we enter the cluster centers as determined in Step 1 and used Euclidean distance as similarity measure. In addition, to see whether the clusters were significantly different from each other, we conducted an analysis of variance (ANOVA). After the determination of the number of clusters and their distinctiveness, we conducted chisquare tests and multivariate analyses of variance (MANOVAs) to see how participants in the clusters differed from each other in gender, socioeconomic status (SES), educational level, instrumental compatibility of the mentor, satisfaction, and perceived similarities.
To explore mentees' views and experiences of the mentor relation in instrumental mentoring, interviews were transcribed verbatim and analyzed using ATLAS.ti. Using sensitizing concepts (Bowen, 2006), the first author and a research assistant independently identified themes present in mentees' descriptions of relations with their mentor, and how they valued their mentors' characteristics in the context of their practical and emotional needs. The first two transcribed interviews were open-coded independently by the two researchers, and similarities and differences in coding were analyzed. This resulted in a coding scheme used to code the remaining interviews. Minor adjustments were made in the coding scheme based on the following two interviews, but was fully applicable to the final five interviews. Axial coding was applied, making connections between categories, split, and merge codes. This was followed by selective coding, to identify relations between the themes (Strauss & Corbin, 1998). During the process of analyzing, memos were made to make expectations and assumptions of the researchers explicit and these memos were discussed to see whether the researchers were not led by implicit assumptions not reflected in the data. From these analyzed interviews, we purposively selected three cases that represent the identified main themes in the clusters. Each case study stresses the role of closeness and the related constructs.

Cluster Analyses
To identify different subtypes in mentoring relations regarding levels of closeness, we performed a cluster analysis. We used mentoring relationship quality items indicating a close bond in the mentoring relation to create the cluster groups (see the appendix). First, we performed a hierarchical cluster analysis, to study how many relationship quality profiles could be identified based on items indicating a close bond. A two-cluster solution appeared to be the best fit. Second, we used nonhierarchical k-means to determine whether the clusters represent meaningful subtypes. The first cluster (n = 26, 49%) is characterized by high scores on closeness. The second cluster (n = 27, 51%) differed from the first cluster in lower levels of closeness. We used the label "High closeness" for Cluster 1, and label "Low closeness" for Cluster 2. We conducted an ANOVA to compare and validate the two relationship profiles. Mentees in the High closeness cluster had a significantly closer bond (M = 4.43, SD = .68) than mentees in the Low closeness cluster (M = 2.42, SD = .77). The profiles significantly differed on levels of closeness, F(1, 51) = 102.36, p <.001, showing the distinctiveness of the two clusters.

How Do Clusters Differ Based on Mentees' Characteristics?
To see how mentees' characteristics such as gender and mentoring experiences were associated with the two relationship profiles, we conducted several analyses. Participants in both profiles differed significantly from each other in their scores on instrumental compatibility, F(1, 51) = 25.77, p = <.001; satisfaction, F(1, 51) = 48.43, p = <.001; and perceived attitude similarities with their mentor, F(1, 51) = 7.08, p = .01 (see Table 2). Mentees experiencing higher levels of closeness (High closeness cluster) reported higher levels of satisfaction, experienced their mentor to be more instrumentally compatible, and experienced more similarities with their mentor in attitude than mentees with lower levels of closeness in their relation (Low closeness cluster). Using chi-square test of independence, we found no associations between gender, SES, educational level, and the relationship profiles.

Case Studies
The cluster analyses revealed two meaningful subtypes of closeness in mentoring relationships. To illustrate the background, thoughts, and feelings of mentees in both clusters, we present three case studies. Results from interviews with mentees from the High closeness cluster were relatively uniform in how mentees experienced closeness. Therefore, we present one case that illustrates the role of a close bond the best. Maira's case is illustrative for most of the mentees in this cluster; perceiving instrumental support makes mentees feel they are not alone and this results in the feeling of a close connection. In the Low closeness cluster, there was variability in how mentees perceived the lower levels of closeness. We, therefore, present two cases from this cluster. Daniel and Laura both indicated to have a less close bond. From the interview it became apparent that for Daniel this is exactly the way he likes his relationship with his mentor to be. Undertaking activities together Note. MANOVA = multivariate analysis of variance. *p < .05. **p < .01.
without talking too much about private issues was a catalyst for him to bring change in other life domains. For Laura, however, the lack of closeness seems problematic, since she does indicate the need for more closeness. Characteristics of these three case studies are presented in Table 3.

Maira (High closeness): "My mentor is helping me 100% and she doesn't even know me": How instrumental support leads to a close connection.
We speak Maira about 9 months after she entered the mentoring program. Maira is 24 years old. She was referred to the mentoring program at the municipal agency for social welfare. Maira dropped out of school, experienced mental health issues, and had no income. She also had debts at five or six agencies. One week after sending an email, Maira was invited to the mentoring program's office to meet the program staff and had an interview with a potential mentor, Vanessa. They set up a meeting right away and during that meeting they decided to look for the right study for Maira, while working on her financial debts and looking for a job. Maira felt there was an instant connection after the first meeting. Since Vanessa has experience in guiding college students and also has a lot of contacts at the municipal agency, Maira feels that Vanessa knows exactly how to support her. They meet regularly and then create a to-do list. When Maira finishes the to-do list, she contacts Vanessa to set up another meeting. Vanessa would never tell Maira what to do, but offers suggestions, or helps putting Maira's needs and wishes into words. By making to-do lists, Vanessa helps Maira putting things in perspective and get her going. Regarding similarities between her and Vanessa, Maira thinks that they are on the same page because they both want to get things done. Maira, however, is sometimes stuck in Even though all of the support that Maira describes to receive from her mentor is instrumental, she experiences a close bond with Vanessa. ". . . in that period of time I was clueless and I felt like me against the world. But with her help it became more easy for me to confront what was happening in my life at that time." It is the feeling that Vanessa understood her situation so well was able and available to provide support, that means a lot to Maira.
Her support, the way she came to me, she was like a 100% interested in me and she didn't even know me. 100% helping me. So actually by her actions she was telling "hey you're not alone, we're going to get this done." This indicates that perceiving instrumental support stimulates the formation of a close bond. While still struggling with her mental health, Maira found a job as a house cleaner, paid off her debts at four agencies, and enrolled in a study program.

Daniel (Low closeness): "We are too down-to-earth for that": How undertaking activities together appears sufficient for growth in multiple life domains.
Daniel is a 26-year-old young adult, living in Rotterdam. At the time of the interview, he has had a mentor for almost 7 months. His mentor is Jord, a retired entrepreneur. Daniel was involved in multiple reintegration programs, but after 3 months or so, this ended and would leave Daniel sitting at home again with no job or daily structure. This, together with changing contact persons from the involved agencies, frustrated him. According to Daniel, he lived in social isolation and was addicted to drugs at the time of entering the mentoring program. His wish was to be able to take better care of himself by cooking, having a job, and daily structure.
When Daniel and Jord met, they hit it off right away, according to Daniel: For us it was actually there right from the start, because, yeah, you have common grounds, you like cycling, you like other sports, he also went to [the same sort of] school, he told me right away when we met. So you immediately have things to talk about.
Jord and Daniel share the experience of attending a certain school and using drugs. For Daniel these similarities made it easy to connect with Jord when they first met. Jord being retired might indicate a generation gap, but it also leaves him with a lot of spare time to invest in Daniel. Jord is able to see Daniel very regularly, and sometimes also joins Daniel last-minute during important appointments with institutions.
At the start of their relationship, Jord and Daniel mostly spend time together on their racing bikes, at least once, and sometimes twice a week. For Daniel this was much better than the emphasis on goal setting that he saw in other mentoring couples. He felt annoyed when he was asked to set up goals and felt like others were telling him what to do. Daniel has an aversion of talking about "emotional stuff" and thinks he and Jord are too "down-to-earth" for that. Also, his experiences have led him to prefer a certain distance: Look, I have seen one hundred care providers come and go so to speak, well, I do think it's one hundred. Well, and it doesn't immediately incite you to think "I am going to explain my whole story and express my emotions etcetera." For this reason he would rather not talk about too many private issues with his mentor. However, as his social isolation was one of his reasons to sign up for the mentoring program, spending time with Jord through cycling was making him feel better already. During these rides they do not talk a lot according to Daniel. Only after a while Jord would ask Daniel "Come on, what are you waiting for?" referring to Daniel's growing insight in the need to stop using drugs in order to get that daily structure and a job.
Daniel is very satisfied with the way that Jord does not push him too much, but makes him realize that change is necessary. After a couple of months Jord asked Daniel explicitly what Daniel expected him to do in their relationship. Daniel then told him to take a step back regarding some issues, but concerning withdrawal from drugs, Daniel asked Jord to chase him more about the registration at a rehabilitation program. At that time, Daniel already had the insight he needed and wanted to stop using drugs, but it was hard for him to actually take action. Jord would then call him and ask if he already made "the" phone call. According to Daniel, it is this regular activity with Jord, and slowly gaining insight in his own situation that made him decide to sign up for a rehabilitation program for his drug addiction, and he is applying for jobs now too.
Laura (Low closeness): "Just a text would do": How the need for closeness is hard to express when experiencing rejection. Laura is 19 years old, and as most of the mentees in the program, living in Rotterdam. Approximately 1 year ago she was permanently expelled from school. She then went to the municipal agency to ask for support and apply for social welfare, and at this site she was introduced to the mentoring program. At the time of the interview she met her mentor almost 1 year ago but it has been a long time since they last met. At the beginning of the interview, Laura needed to be reminded which mentor the interview was about, since she had multiple mentors and coaches.
Laura and her mentor met for the first time at the mentoring program's office. Laura was struggling with the lack of daily activities and with financial problems. After their first meeting, Laura and her mentor started messaging each other and met multiple times. They then talked about what Laura needed since she was expelled from school. According to Laura, no concrete objectives were formulated, and her needs at the time remained vague: "I just needed help with my life situation" . . . [support] in a nice way. Support includes what I want to do in my life." Laura feels the mentor could not really support her, neither emotional nor instrumental.
During the interview Laura indicates the need for some closeness, only if it is just a text saying "Hey, how are you?" She would like to talk about issues and receive positive feedback from her mentor. At the same time Laura tells about the negative experiences she has with teachers and social workers, and how she feels that they are never really on her side. Laura thinks that this also led to the fact that she rather does things on her own. Even though she has the need for support, she does not think that her mentor can really support her. She rather handles private issues on her own, because she does not like to ask for help. She has experiences of rejection after asking for help, so she does not do that any longer. In her own words, she is used to doing things on her own now.

Discussion
The present study explored the role of closeness in instrumental mentoring for young adults. Young adults with practical needs in mentoring programs require guidance, support, and advocacy, which makes instrumental mentoring better suited for young adults than developmental mentoring (Bowers, 2019;Cavell & Elledge, 2014;Rhodes, 2019). As research on the role of closeness in instrumental mentoring is limited (see Lyons et al., 2019), and especially on how instrumental mentoring supports young adults, the present study set out to explore the role of closeness in instrumental mentoring for this specific group. A two-cluster solution was validated based on the levels of closeness mentees indicated to experience with their mentor. The first group reflected mentees experiencing high levels of closeness from their mentor. The second group reflected mentees who experienced lower levels of closeness with their mentor. Compared to mentees with low levels of closeness, mentees with high levels of closeness perceived their mentor to be more compatible with their instrumental needs, were more satisfied with their mentor relation, and perceived more similarities in attitude between them and their mentor. Case studies illustrated the way closeness developed, mainly as a result from receiving instrumental support. In the group of low levels of closeness, there was more variation in how mentees experience this lack of closeness. For some mentees this was problematic, for others this was their preference as the result of their experiences with social services.
Although mentees in the High closeness cluster were more satisfied with the relation and perceived their mentor as more compatible with their needs, the mentees in the Low closeness cluster were not unanimously dissatisfied with their mentoring relation. For some, the emotional distance between them and their mentor was how they liked their relation to be, and still led to the achievement of some very important goals. For others, the lack of emotional support seemed more problematic. Here, the lack of setting goals seemed to hinder the development of the relation. Previous research has indicated the importance of concrete goal setting in instrumental mentoring (Keller, 2005). With no close bond and no concrete goals to work on, the contact remained superficial and vague and may lead to early closure of the match. For young adults this experience on top of their previous experiences with social services is rather problematic (Spencer, 2007).
Mentees with higher levels of closeness were characterized by having more perceived similarities in attitude with their mentor, but not with more perceived similarities in background. Interviews with mentees showed that indeed, mentees did not see their mentors' background as dissimilar to theirs, but they focused on details that would underscore their similarity. For example, mentor and mentee that both spend their younger years in the same type of school, or sharing same interests in sports, or having the same mindset. Mentees would also seize these similarities as indicators of an instant connection with their mentor. They mentioned that because of this connection they had the idea this match was going to be a good one. Although the present study was not set up to identify the minimal basis of trust and empathy, our results suggest that that even for mentees where closeness developed as a result of instrumental support, some basic levels of trust between mentor and mentee is necessary. This trust is most likely related to levels of similarities. The types of similarities mentioned as important for a connection varied broadly but were not focused on background similarities such as social class or ethnicity. This finding complies with previous mentoring research that finds surface similarities (gender and ethnicity) are inconsistently linked to perceptions of mentoring, and that deep-level similarities (attitudes and believes) are related to more support (Eby et al., 2013). Similarities on the experiential level (educational background or job tenure), however, are believed to be associated with more instrumental support (Eby et al., 2013). In our study, mentors and mentees were often dissimilar in their educational background and jobs, but this did not seem to hinder the mentors' effectiveness in providing instrumental support. Mentors' ability to connect and navigate through networks seemed sufficient in the development of mentees' social capital (Bourdieu, 1986).
As Hamilton and Hamilton (2005) suggested, our interviewees indicated a close bond to arise from the mentor's supporting behavior. In our study, closeness seemed to be the result of the instrumental support of the mentor, instead of a precondition of working on set goals. Research suggests that behavior that is perceived to be performed voluntarily, rather than formally required, is an indicator of someone's trustworthiness (De Jong et al., 2007). Mentees in our study often talked about everything their mentors did for them with amazement. A mentor supporting a mentee without immediate selfinterest is signalizing a positive orientation toward the relationship, and repeated support over time, can lead to the formation of a close bond between mentor and mentee (McAllister, 1995).
Something many mentees in our study struggled with was the prescriptive attitude they were used to from previous encounters with social care providers. Several mentees indicated the need to make their own decisions and to see where things were going without an explicit focus on goals during the mentoring relation. This was in accordance with previous research on at-risk young adults' needs to do things on their own (Lenkens et al., 2020). Although a close bond in our study did not appear to be a precondition of effective instrumental mentoring, it could be the case that the presence of an emotional bond makes it easier to work toward goals. According to Karcher and Nakkula (2010), sharing thoughts and emotions with a mentor may prevent instrumentally focused interactions from feeling prescriptive. An emotional bond, thus, does not seem to be a precondition in instrumental mentoring, but it may make it easier to set and attain goals, and, in turn, to keep the relation going. At the beginning of the relationship, goals should, therefore, be primarily based on the mentees' needs. Optionally, new goals could be introduced later in the relation when some level of closeness has been established.
Satisfaction and mentors' compatibility were both associated with mentees experiencing more closeness (High closeness cluster). Satisfaction in the quantitative measure concerned a broad sense of being satisfied with having a mentor, both instrumental and relational. Mentors' compatibility regarded their skills and background with respect to supporting the mentee. From the interviews these two constructs were hard to separate. Mentees' satisfaction with the relation was often related to how they saw their mentor contribute to their goals. Indeed, mentees' dissatisfaction with the relationship was previously found to be associated with insufficient instrumental support (Nakkula & Harris, 2010). Most importantly, there were cases of mentees in our study, indicating to have low levels of closeness, but still were satisfied with having a mentor, and their mentors' effectiveness in supporting them. This finding might be specific for the age and needs of our sample and suggests that the sweet-spot of combining goal-directed and relational activities may differ per match (Lyons et al., 2019).

Limitations and Suggestions for Future Research
The present study explored relationship quality in instrumental mentoring among young adults, a so far understudied sample and program type. Findings are based on a small sample and the use of case studies can only be seen as indicators of relevant issues in instrumental mentoring for this group. Although we were aware of mentees' possible restraint to talk about negative elements in their relation, it could be the case that they did not want to open up about this element in their relation, that they did not want to talk about differences between them and their mentor. In addition, mentees could voluntarily enter the study, and this might have resulted in a biased sample. It may be that mentees who had negative experiences were less motivated to enter the study, and most importantly, were reluctant to do an additional interview after completing the survey. We showed the variety in the Low closeness cluster by presenting two cases of mentees both experiencing lower levels of closeness but with various levels of satisfaction and instrumental compatibility. Daniel's scores on the correlates of closeness (satisfaction, instrumental compatibility, and similarities) were indeed higher than Laura's and this difference was reflected in their interviews. However, the satisfaction and instrumental compatibility scales were of low reliability. Future research, therefore, should be conducted with validated instruments that are able to differentiate between instrumental and relational elements in mentoring. We also suggest future research to take gender into account when studying the development of relational closeness in instrumental mentoring. Closeness in our study was largely constructed of items that considered talking about personal things and problems (see the appendix). For women, self-disclosure is considered a sign of closeness, whereas for men, engaging in activities is generally more important (Liang et al., 2014). In our sample there were more men than women, and this may explain our finding that mentees (64% men) preferred instrumental support over experiencing closeness. In addition, in order to examine the sequential order of the development of closeness and instrumental support, future research should use longitudinal data of mentoring relationships. Also, it is important to identify the minimal conditions that mentees need in instrumental mentoring, such as mentors' empathy and levels of trust. Finally, the present study only used mentees as informants, but mentors' perceptions should be taken into account as well. This could provide more insight in the dynamics between mentor and mentee. For example, if mentors perceive their mentees to avoid closeness, the question is whether they see this as a hurdle to provide instrumental support.

Practical Implications
Our study indicated the importance of providing youth with support that meets their instrumental needs. To formulate and monitor the progression of this need fulfillment, goal setting seems useful. Setting goals to work on may give youths a sense of control with regard to the problems they are dealing with in their life stage, but also seems necessary to start the mentoring relation without just spending time together to get to know each other. However, setting goals should also be handled with caution. As previous experiences with support influence how young adults perceive support and goal setting, mentees' preference in setting goals should be leading instead of prescriptive goal setting by the mentor or the program. For some mentees, the presence of clear goals seems to provide concrete agreements on how and when mentor and mentee will meet. To set appropriate expectations and effective communication, mentors need skills to do this (Nakkula & Harris, 2014).
Based on our findings that mentors' contribution to the relation is not only providing emotional support, but also advising, networking, and advocating, we would suggest matching mentors and mentees based on the mentor's compatibility to the mentee's (instrumental) needs. Mentees frequently indicated that their mentors had many useful connections and knew how to navigate the bureaucratic structure in order to support mentees in their obstacles. In addition, although future research is needed, matching based on shared interests seemed more important for experiencing closeness than shared backgrounds. Even one similarity could provide a mentee with the confidence that the match is going to be successful. Mentors can also be trained in self-disclosure, which is thought of as a stimulator to identify similarities enhancing the relationship (Dutton et al., 2019).

Conclusion
Questions have been raised about the role of closeness in instrumental mentoring of young adults. The findings of the present study suggest that it is worthwhile to further explore the role of closeness in instrumental mentoring, since cluster analyses and case studies showed variation in how mentees perceived closeness. For some young adults, closeness was a result of perceiving instrumental support, whereas for other young adults, the lack of closeness was problematic. The findings of our study suggest that for young adults in instrumental mentoring, findings on relationship quality in developmental mentoring (c.q. closeness) cannot be translated one on one to instrumental mentoring. The developmental stage of the mentees, and their history of social care, seemed to relate to their perceptions and preferences in mentor relations. For mentoring to serve as an intervention strategy for young adults, improvement in both research and practice is required.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.