The Nature of Adolescents’ Relationships With Their “Very Important” Nonparental Adults 1

As part of a larger program of research on the nature of adolescents’ relationships with very important nonparental adults (hereafter referred to as “VIPs”), a community sample of 243 eleventh graders (mean age D 16.6 years) was surveyed, and a subgroup of 55 adolescents and their VIPs were interviewed about the nature and quality of their relationships. Results showed that (a) adolescent–VIP relationships were a normative component of adolescent development, not a result of problems in adolescents’ lives; (b) adolescent–VIP relationships were generally of high quality (e.g., high support, low conﬂict, and high mutuality); (c) there were signiﬁcant differences between kin and nonkin VIPs in terms of the duration of relationships and frequency of contact, but not in the quality of relationships; and (d) VIPs whom adolescents designated as extremely important were distinguished from other VIPs in terms of providing a higher level of social support and a higher frequency of contact. Theoretical and practical implications of these ﬁndings are discussed.


INTRODUCTION
Parents are arguably the most important adults in the lives of most children. Parents have significant effects on their children's development of beliefs, goals, attitudes, and behaviors, as well as on other aspects of the individual. As children make the transition into adolescence, however, they come into contact with a broader array of adults and develop increasingly important relationships with peers. Although an abundance of research has focused on various aspects of adolescents' relationships with their parents (Dornbusch et al., 1985;Paikoff & Brooks-Gunn, 1991;Steinberg, 1990) and their peers (Berndt, 1996;Berndt & Hoyle, 1985;Morgan & Grube, 1991), relatively few studies have evaluated relationships between adolescents and their extended adult networks (i.e., community members, teachers, and extended family members).
Studies of young people's ties to nonparental adults-sometimes kin, sometimes others-have shown, however, that nonparental adults may exert a strong influence on psychosocial adjustment. For example, research on resiliency has demonstrated that nonparental adults may have positive effects among adolescents at high risk as a result of poverty conditions and parental mental illness (Cowen & Work, 1988;Garmezy, 1987;Luthar & Zigler, 1991;Werner & Smith, 1982). Similarly, in studies of pregnant and parenting African American teens, Rhodes and colleagues (Klaw & Rhodes, 1995;Rhodes, Ebert, & Fischer, 1992) found that the presence of a person the teenage mothers considered a mentor was associated with lower levels of depressive symptomatology, more positive attitudes toward career attainment, and greater life optimism. Taylor, Casten, and Flickenger (1993) found that the presence of kinship support (measured by proximity of kin and frequency of contact) was negatively related to African American adolescents' involvement in problem behavior in single-parent families-generally considered a high-risk context regardless of adolescents' ethnicity. Finally, Burton's programmatic research (Burton, 1996) has depicted the role that grandparents (especially, grandmothers) play in the lives of African American at-risk youths.
Positive effects of nonparental adults also have been documented for relatively low-risk samples of adolescents. For example, drawing from a lower middle-class population of 11th graders, Greenberger, Chen, and Beam (1998) surveyed students about their "very important" nonparental adult (VIP)-someone at least 21 years old who has had a significant influence on the adolescent or whom the adolescent could count on in times of need and who engaged in good "role-model" behavior. They found that adolescents with a VIP were significantly less likely to be involved in misconduct, regardless of the behavior of close friends and family members. Moreover, adolescents were significantly less likely to be involved in misconduct when VIPs showed disapproval of the adolescents' participation in such behaviors, even after controlling for the attitudes of peers and parents. Additionally, positive effects of nonparental adults have been found in studies of "assigned" mentors. A cross-regional study of the Big Brothers/Big Sisters program, for example, showed that program participants, when compared to a control group, were involved in fewer antisocial activities, had better academic performance, attitudes, and behaviors, and had better relationships with family and friends after 18 months (Tierney, Grossman, & Resch, 1995).

The Current Study
Although existing research has clearly documented the important contributions that nonparental adults (i.e., mentors or VIPs 3 ) make in adolescents' lives, little is known about the nature of adolescent-VIP relationships, for example, their initiation, stability and change over time; the frequency, contents, and quality of the interactions; and the linkages between adolescent-VIP relationships and adolescents' other social contexts such as family and peers. In this exploratory study, we examine some of these important aspects of adolescent-VIP relationships with both quantitative and qualitative data from surveys and interviews with adolescents and their VIPs. Specifically, we investigate four research questions: (a) is the development of a close relationship with a nonparental adult a normative part of adolescent development, or is it specific to adolescents who do not have a positive relationship with their parents?; (b) what is the overall quality of adolescent-VIP relationships, and do they represent a unique social context that serves different functions from relationships with parents and peers?; (c) are there significant differences in adolescents' relationships to kin and nonkin VIPs?; and (d) which factors distinguish VIPs whom adolescents consider extremely important from other VIPs?
Much of the literature on the role of nonparental adults in adolescent development has focused on at-risk samples. The theoretical discussions surrounding such relationships often have focused on their compensatory role. Mentors or VIPs are typically conceptualized as individuals who provide the at-risk adolescents with support or resources that are not adequately provided by parents. On the other hand, it is possible that adolescent-VIP relationships develop naturally in adolescents' various social contexts 3 Different definitions and terminology regarding nonparental adults have been used in the literature. For example, Rhodes and colleagues have preferred the term "natural mentors." Natural mentors have been identified by asking participants, "other than your parents or whoever raised you, is there an older person in your life (a mentor, or positive role model) who you go to for support and guidance?" (Rhodes et al., 1992). We use the term "VIP" because it covers important nonparental adults who play a wider range of roles in adolescents' lives (see our definition of "VIP" in the Measures section).
as part of normative development. From an early age, children have the experience of interacting with nonparental adults (e.g., relatives, doctors, babysitters, teachers). It seems plausible that at some point in time, youths may develop a close relationship with one or more of these nonparental adults.
To test the normative hypothesis versus compensatory hypothesis, we examine the prevalence of adolescent-VIP relationships, adolescents' life situations at the onset of the relationships with a VIP, and VIPs' reasons for involvement. Additionally, we compare adolescent-VIP relationships with parent-adolescent relationships. The compensatory hypothesis would predict a low prevalence of VIPs among adolescents from a community sample, a high number of negative life events at the time of the relationship onset, situation-specific reasons rather than person-specific reasons for the formation of the relationship, and negative correlations between parentadolescent and adolescent-VIP relationships. The normative hypothesis, on the other hand, would predict a high rate of prevalence, few special life situations at the onset of relationships, person-specific ("he's a really neat person") rather than situation-specific ("his parents were getting divorced") reasons for relationship formation, and a lack of an association between parent-adolescent relationships and adolescent-VIP relationships.
To examine the overall quality of adolescent-VIP relationships, we investigate the following dimensions: degree of importance of the VIP to adolescents, perceived support (type and frequency), frequency of contact and conflicts, and mutuality of the relationship. Furthermore, we explore whether adolescent-VIP relationships represent a unique type of relationship. VIPs may provide a type of support that cannot be provided by peers because of peers' lack of social experience, or by parents because of adolescents' (and parents') discomfort in discussing certain issues.
Previous research (Greenberger et al., 1998;Rhodes et al., 1992) has shown that about half of young people's important nonparental adults or natural mentors are relatives such as aunts or uncles and grandparents, and the other half are nonfamilial adults such as teachers, church representatives, friends' parents, and coaches, among others. The differential roles of kin versus nonkin in adults' social networks have been studied extensively by sociologists (e.g., Litwak & Szelenyi, 1969;Wellman & Wortley, 1990). Most research has shown signicant differences between individuals' kin and nonkin social networks. For example, relationships with kin-group members tend to be relatively permanent, more affective (i.e., more emotionally involving), and more diffuse (i.e., providing a wide array of social and instrumental support), whereas relationships with nonkin tend to be more temporary, instrumental, and specific. In this study, we explored whether the nature and quality of adolescent-VIP relationships (i.e., length of relationships, perceived support, mutuality in relationships, frequency of contact) differed among youth with kin and nonkin VIPs.
Finally, because VIPs differ in their degree of importance to adolescents, we examine which factors distinguish between VIPs whom adolescents consider extremely important figures in their lives and VIPs of lesser importance. It is plausible that VIPs who are extremely important to adolescents might differ in the type of support they provide and in the quality of their relationships with youths.

Adolescents
Participants in the study were recruited from a greater Los Angeles high school with achievement indicators that were average for 11th graders in California schools, and thus, below national norms. On the basis of standardized achievement tests, school versus statewide scores were 39th versus 36th percentile, respectively, for reading; 49th versus 48th percentile for mathematics; and 46th versus 45th percentile for language. The percentage of families receiving AFDC (Aid to Families with Dependent Children) was lower for the participating school than for the state as a whole (10 vs. 18%; California Department of Education, 1999).
The sample consisted of 11th graders who are part of a longitudinal study that examines adolescents' transition into early adulthood. Of the 300 enrolled 11th graders, 243 (81%) participated in the survey study (M age = 16.6 years, 57% female). The sample was representative of the ethnic diversity of this metropolitan area: 54% European American, 16% Latino, 12% Asian American/Pacific islander, 8% African American, and 10% mixed ethnicity. Modal education for mothers was high school completion, and for fathers, completion of vocational or technical school. Twenty-three percent of mothers and 28% of fathers had earned a college degree or higher. Approximately half of the sample (52%) came from intact families, 20% had parents who were remarried, 20% had single parents, and 8% were classified as "other" (e.g., living with relatives, joint custody).
Additional interview data were collected from a subsample of 61 adolescents. After excluding six cases with incomplete data, the final interview sample of adolescents consisted of 55 individuals. The interview sample did not differ significantly from the larger survey sample on demographic characteristics (see Table I).

VIPs
We surveyed and interviewed the 58 adults who were the VIPs of the 61 adolescents we interviewed. (Four adolescents identified the same person-a teacher-as their VIP. 4 ) We obtained 61 completed surveys and 56 completed interviews from VIPs. VIPs' age ranged from 21 to 82 years with a mean of 36.5 years (SD = 12.8). Slightly more than half of the VIPs (53%) were married, 22% were single, and 14% were divorced. About one third of the VIPs (31%) had a 4-year college education or higher, 43% had some college background, and 26% had a high school education or less. A majority of VIPs (65%) were working full time, and another 12% were working part time. Their occupations were diverse, including police officer, heavy equipment operator, teacher, chiropractor's aide, and author.

Adolescents
Active consent of adolescents and their parents was obtained prior to survey administration. Students who wanted to participate in the study were given letters and consent forms to bring to their parents, with a request that parents allow their youngsters to participate. We administered surveys only to adolescents who returned both their own and their parent's signed forms.
Student survey data were obtained by means of a questionnaire administered by the researchers during a 50-minute class period at school. The questionnaire covered a wide variety of topics, including family demographics and measures of parental warmth, adolescent disclosure to parents, and adolescent problem behaviors. We included an extensive series of questions about the relationship between the adolescent and his or her adult VIP. All participants in the survey were entered into a lottery for gift certificates to a record store.
Students who completed the surveys were asked to indicate whether they would be willing to participate in a subsequent paid interview. All individuals who indicated "yes" and provided contact information were called by the researchers 1 month after the survey administration and asked to continue their participation in the study. Adolescents who consented to the interview and to our inviting the important nonparental adult in their life (VIP) to be interviewed were included in the interview sample. (Parental consent was again obtained.) Interviews with adolescents took place after the VIP interviews were completed (see below). Semistructured interviews that lasted approximately 1 hr were scheduled individually with each adolescent interviewee. Adolescent interviews were led by trained graduate and undergraduate interviewers and took place at a location selected by the adolescent (e.g., local coffee shop, home). Interviews were audiotaped with the adolescent's permission. Participants were given a check for $20 at the end of the interview.
Researchers telephoned eligible VIPs and scheduled semistructured interviews that lasted from 1 to 2 hr. Trained interviewers, accompanied by a backup notetaker, met the VIP at a convenient and appropriate location selected by the VIP (in a quiet coffee shop, or in some cases, the VIP's home). For the eight VIPs who lived outside the Los Angeles area, telephone interviews were conducted. Interviews were audiotaped with the VIP's permission. Following the interview, VIPs were given or sent a short survey to complete at home and return by mail. Upon completion of the interview and survey, VIPs were sent a check for $40.
Interview questions were coded by two undergraduate or graduate research assistants, all of whom received several hours of training and met the standard of 80% correct on a test-set of interviews. Interrater agreement was calculated for all coded responses from surveys and interviews. In the case of ordinal scales (e.g., judgment of importance of relationships), we used the conventional criterion of agreement within a 1-point difference in ratings. The average interrater agreement was 91%, ranging from 74 to 100% (see Measures section for details). Disagreements were resolved through discussion by the coders.

Adolescents' Descriptions of VIPs
The social identity and functions of the VIPs were assessed through a series of questions. Participants first were asked to consider whether they had an "important adult" in their lives other than a parent-"someone at least 21 years old who has had a significant influence on you or whom you can count on in times of need." To stimulate participants' thinking on this topic, we provided examples of possible VIPs, such as an aunt, teacher, or friend's parent.
For screening purposes, adolescents also rated the importance of the person in their lives on a 5-point scale marked not really all that important (0.8% of the total sample), somewhat important (5.3%), important (20.2%), very important (34.2%), or a truly key person (27.2%). Respondents who rated their VIP as "not all that important" or only "somewhat important" were dropped from further analyses. To examine whether "truly key" VIPs were distinguishable from VIPs who were "important" or "very important" in terms of their roles and functions in adolescents' lives, we recoded VIPs into two categories: "truly key" VIPs and "not truly key" VIPs.
VIPs' social identity was coded into the following categories: grandparent, aunt/uncle, cousin, sibling, parent's significant other, friend's parent, neighbor, older friend, parent or other relative of boyfriend/girlfriend, teacher, coach, academic or psychological counselor, church representative, and other nonkin. Intercoder agreement was 100%. For most analyses we report, these responses were recoded into kin and nonkin VIPs.

Onset of the VIP-Adolescent Relationship
Information about the onset of the VIP-adolescent relationship was obtained from both surveys and interviews. In the adolescent survey, we asked open-ended questions pertaining to the age at which the VIP became important to the adolescent and events that were occurring at the time the VIP became important. Significant events were coded by research assistants into several categories such as "family problems" and "adolescent's peer problems." Intercoder agreement was 92%. Qualitative data concerning the reasons for VIP's involvement with adolescents were obtained through interviews.

Quality of the Relationship
Quality of the relationship was assessed through both survey and interview questions. Four indices were used. First, as indicated earlier, adolescents rated on the survey the importance of the VIP in their lives. Second, on a checklist of 10 items developed on the basis of open-ended data obtained in earlier research (Greenberger et al., 1998), adolescents indicated ways in which their VIP was important to them. Scores on the 10 items were summarized to create a scale of ways of importance (Cronbach α = .75). Third, adolescents reported frequency of received support-that is, the frequency with which the VIP had done such things in the past 6 months as "expressed interest and concern in my well-being," "let me know that I did something well," "provided me with money or something else I needed," and "provided me with a place to stay." (1 = never, 2 = once or twice, 3 = three or four times, 4 = more often). Adapted from Barrera, Sandler, Ramsey's Inventory of Socially Supportive Behaviors (Barrera, Sandler, & Ramsey, 1981), this 16-item scale had a Cronbach alpha of .91. Fourth, we obtained information from adolescent and VIP interviews on the frequency of conflicts between them. Responses were grouped into three categories: "never/very seldom," "occasionally," or "often." The intercoder agreement was 89%.

Frequency of Contact
On their surveys, adolescents indicated the frequency of talking to their VIP on a 4-point scale ranging from "a few times a year" to "nearly every day."

Importance of Relationship to VIPs
To assess the mutuality of the adolescent-VIP relationships, VIPs were asked during the interviews how much they value their relationship with the adolescent. They responded on a 4-point scale ranging from "would not be too upset/sad if the relationship ended" to "would be very upset/sad if the relationship ended." In addition, on the basis of the entire interview protocol, coders also rated the overall importance of the adolescent to the VIP on a 5-point response scale from "not really all that important" to "truly key." Intercoder agreement for these ratings was 95%.

Adolescent-VIP Relationships as Compared to Other Social Relationships
In the interviews, adolescents reported how they viewed their relationship with their VIP in comparison to their relationships with parents and peers. Two specific aspects of these relationships were compared: adolescents' perception of others as judgmental and punitive (e.g., "my parents criticize me, but he [the VIP] just listens") and adolescents' comfort with disclosure (e.g., "I can talk to her [the VIP] about things I can't talk to my friends about"). Interrater agreement was 81% for the comparison between VIP and parents and 74% for the comparison between VIP and friends.

Qualities of Adolescent Relationships With Parents and Peers
To examine the association between adolescent-VIP relationships and family and peer contexts, we measured perceived parental warmth, perceived conflict with parents, and perceived peer warmth through survey questions. Adolescents rated parental warmth on the 11-item Parental Warmth and Acceptance scale (Greenberger et al., 1998). Sample items, answered on a 6-point continuum of disagreement/agreement, were, "I know that my parents will 'be there' for me if I need them" and "I find it hard to please my parents" (reverse coded). Cronbach alpha for the parental warmth scale was .89. Adolescent-parent conflict was assessed by means of a 12-item scale (4-point response scale, from "never" to "all the time"). Topics included school-related issues, chores, friends, money, personal habits, and family relations, among others. Cronbach alpha for this scale was .85. Peer warmth was measured with a scale analogous to the Parental Warmth scale. This 11-item scale had a Cronbach alpha of .82.

Are Adolescent-VIP Relationships a Normative Part of Adolescent Development?
A large majority (N = 199, 82%) of the 243 adolescents in the study reported having a nonparental adult who played an "important," "very important," or "truly key" role in their lives (12% did not have a VIP, and 6% reported that their VIP was "not very important" or "somewhat important"). Of the 199 adolescents who had a VIP, only a minority (23%) indicated that a significant event was occurring at the time the VIP became important. Most frequently cited events were family problems (46%), followed by personal problems such as emotional, physical, or behavioral problems (27%); changes in the environment or other transition (14%); peer problems (9%); and school-related problems (5%).
Interviews with VIPs provided further indication that adolescent-VIP relationships develop naturally and gradually, rather than suddenly as a result of special circumstances. As an illustration, one VIP, who was a teacher, described the development of his relationship with the adolescent in the following way: She tested me for a while. She would do whatever a student would do to fail my class. She really pushed me. But because of the amount that she read, I knew she had a mind, and that she thought and cared-that there was something else going on somewhere. So I was like, "I'm not letting you get me to hate you." And eventually she gave up and decided to be my friend.
At least three things are obvious from this quote about the onset of this particular adolescent-VIP relationship. First, the relationship between the teacher and his student did not form because the adolescent was having family or peer problems. Second, the teacher took an interest in the adolescent because of her personal qualities-"she had a mind, and she thought and cared." Indeed, when asked about reasons for becoming involved in the adolescents' lives, 92% of the VIPs we interviewed checked the reason "she/he is fun to be with." Third, the process of forming an adolescent-VIP relationship appears to be gradual-as indicated by the teacher's choice of the word "eventually." The gradual process of developing an adolescent-VIP relationship was explicitly emphasized by another VIP, an older sister of the adolescent. I think I've always been important-as far as becoming connected, I would say as he started to get to know himself, I started to become more important because I made it really clear to him about how much I loved him and how much I needed him. It's been gradual, we just started being there for each other as we got older.
Finally, consistent with the normative hypothesis, correlational analyses revealed that the quality of adolescent-VIP relationships was generally not associated with that of adolescents' relationships with parents or peers. Out of 18 correlations between the five indices of adolescents' relationships with their VIP (importance of VIP, ways in which they were important, and frequency of received support, conflicts, and contact) and three measures of adolescents' relationships with parents and peers (parental warmth, conflict with parents, and peer warmth), only one correlation was significant: Frequency of support received from VIPs was modestly correlated with the frequency of parent-adolescent conflict, r(192) = .21, p < .01. All other correlations ranged from −.05 to .10, ns.

Quality of Adolescent-VIP Relationships
Table II presents the means and standard deviations of the variables used to measure the quality of adolescent-VIP relationships and related constructs. The results clearly showed a high-quality relationship between adolescents and their VIP.
The screening criterion used for identifying VIPs (see Measures section) ensured that all VIPs were at least "important" to adolescents. Indeed, a large majority (72%) were identified by adolescents as either a "very important" or "truly key" person in their lives. The average rating of importance was 4.08 (4 = very important).
Most adolescents (more than 80%) reported that the VIPs were important to them in the following ways: showing respect, providing emotional support, serving as someone to talk to, and supporting various activities adolescents were engaged in. For the other six types of support (e.g., helping with school, relationships, personal issues, and financial issues, serving as a companion and as a role model), at least half of the adolescents indicated that their VIP was important in these ways. On average, adolescents checked 7.5 of the 10 items reflecting "ways of importance." Adolescents reported a high frequency of overall support provided by VIPs (see Table II). The mean frequency for all 16 types of support provided by VIPs during the previous 6 months was 2.67 (between 2 = once or twice and 3 = three or four times). In other words, in a 6-month period, the VIPs exhibited, on average, one or two supportive behaviors toward adolescents per week. Consistent with such results, adolescent participants reported talking to or seeing their VIP an average of "once or twice a week" (M = 3.11).
Relationships between adolescents and their VIPs also can be characterized as low in conflict. In fact, a large majority of VIPs and adolescents we interviewed (79% of VIPs, 68% of adolescents) indicated that they never or very seldom had arguments with each other.
Adolescent-VIP relationships were important not only to the adolescents (M = 4.08; 4 = very important), but to their VIPs (M = 3.87). Furthermore, VIPs indicated that they would be quite sad or upset if the relationship were to end (M = 3.20, where 3 = pretty upset/sad, 4 = very upset/sad ).

Adolescent-VIP Relationship as a Unique Social Context
To illustrate the unique role of adolescent-VIP relationships, we present excerpts from the interviews with two adolescents.

Case 1
This 16-year-old male adolescent was an only child who lived with both of his parents. The quality of his relationship with his parents was average for the adolescents in this study. He reported his parents as being moderately warm and having strong sanctions against his involvement in misconduct. He argued with them occasionally. He had several friends with whom he had positive relationships, and he told his friends just about everything.
His VIP, however, played a special role. In this case, the adolescent's parents were close friends of the VIP and his wife. The VIP was 32 years old and had no children of his own. The relationship between the adolescent and the VIP gradually developed as a result of the two families' doing things together. The VIP became important to the adolescent when he was about 11 or 12 years old.
The adolescent characterized his relationship with his VIP as being like a sibling relationship, "He's like an older brother to me." The VIP played a role that was unique from the role of his parents. "I can talk to him about things that I wouldn't talk to my parents about, like relationships or sex. If I did talk to my parents about some stuff, they might set new rules and regulations, and stuff, but if I talk to [VIP], nothing happens-I just get a lot of freebee advice," said the adolescent. Other adolescents also frequently mentioned topics that are notoriously difficult to discuss with parents-sex and drugs. This adolescent's sentiment about limiting disclosure to parents was echoed by many adolescents. The majority (60%) of adolescents we interviewed said they were able to disclose more to VIPs than to parents, whereas only 9% said they were able to disclose more to parents than to VIPs. The rest of the sample (31%) said they could disclose to their parents and their VIP equally.
This adolescent's VIP not only played a role that his parents could not fulfill, but also one that peers typically do not play: "I can tell him anything that I can tell my friends, but he gives me correct advice, as an adult-he wouldn't lead me the wrong way. He knows that I look up to him-he's like a role model." Although peers could have played some of the same roles, the social experiences and status of nonparental adults are likely to make them a better source of guidance and role models. In this particular sense, VIPs are more parent-like. In fact, about two thirds of the adolescents indicated that they considered their VIP to be "like a parent."

Case 2
The adolescent was a 17-year-old female. She lived with both of her parents, two older sisters (ages 19 and 23), and two younger brothers. She had close relationships with all her family members, but a particularly strong and intimate relationship with her oldest sister whom she identified as her VIP. Her sister became important to her when she was 13 or 14 years old. Similar to earlier examples of gradual development of adolescent-VIP relationships, this adolescent said, She [the VIP] started taking me places-to concerts and stuff. That's when I started to hang out with her. Nothing [special] was going on in my life. That's just it. She taught me to get interested in stuff-like music. I really started loving music because of her.
She continued to point out the special role of her VIP in providing guidance and support, She's been there for anything important or exciting that has happened in my life. She's either been there or been the one to make it happen. I owe a lot of my outlook on life to her. She got me interested in groups and organizations that help people and do work around the world, and because of that, I want to dedicate my life to doing that kind of work.
When asked about the differences between her relationships with the VIP and those with her close friends, she said, "Even my closest friend-I don't tell her everything like I would [VIP]. I leave out parts of the story when I tell my friends." As in Case 1, this adolescent's VIP played a role that was not fulfilled by her relationships with her peers. In this case, the VIP served as a safe repository for deep feelings and personal secrets.
This adolescent also echoed the sentiment of Case 1 and other adolescents about disclosure to parents "I've shared experiences with her [the VIP] that I would never share with my parents. I think they [parents] would be more shocked or something-like 'Oh my God, I can't believe you're thinking that or doing that kind of stuff-shame on you!'" In contrast, "She [the VIP] doesn't give me that," the adolescent added. This difference between parents and VIPs was also evident for many adolescents. A substantial number of adolescents (31%) described their VIP as less judgmental and punitive than their parents, whereas only few (2%) indicated the reverse. The remaining 67% said that their parents and VIP were equally judgmental and punitive.
In sum, adolescent-VIP relationships appear to be unique and often have both parent-like and peer-like qualities. On the one hand, VIPs provide much support that peers usually cannot provide (guidance, role model, financial, etc.). On the other hand, most adolescents who were interviewed (85%) considered their VIPs and peers to be similar with respect to judgmental tendencies; a minority (11%) said their VIP was less judgmental than their friends or reported the reverse (4%). Similarly, 53% of adolescents said they were able to disclose equally to VIPs and peers, with the other half about equally divided between those who would disclose more to VIPs than to peers (27%) and those who would disclose more to peers than to VIPs (20%). These data are in clear contrast with adolescents' views about parents in comparison to VIPs, as noted earlier.
Relationships between kin and nonkin VIPs differed substantially in their onset and duration. Most of the nonkin VIPs became important during the past 4 years (78%), whereas the onset of kin VIP relationships was almost evenly distributed over the 17 years in the adolescents' lives. ANOVAs indicated significant differences in the length of adolescent-VIP relationships with kin (M = 9.56 years) versus nonkin VIPs (M = 2.74 years), F(1, 191) = 105.83, p < .001.
As shown in Table II, kin and nonkin VIPs differed significantly in frequency of contact, conflicts, and the mutual value of the relationship, but not in degree of VIP importance to adolescents, ways of importance, and frequency of support. Adolescents with nonkin VIPs had more frequent contact than did adolescents with kin VIPs. In fact, 85% of nonkin VIPs, but only 66% of kin VIPs had at least weekly contact with their adolescent, χ 2 (3, N = 197) = 11.26, p = .01. Kin VIPs had more conflict with adolescents than did nonkin VIPs, mainly because of a higher rate of conflict between adolescents and "truly key" kin VIPs (M = 2.00 or "occasional"; see the significant interaction between kin/nonkin and key/nonkey VIPs, Table II). All the other three groups (key or nonkey VIPs who were not relatives, and nonkey kin VIPs) had a low rate of conflict (M ≈ 1.20).
Kin VIPs rated the importance of the adolescents in their own life as higher than did nonkin VIPs (M = 4.25 and 3.50, respectively, where 3 = important, 4 = very important and 5 = a truly key person). Results also showed that kin VIPs placed significantly greater value on the relationship than did nonkin VIPs, as indicated by responses to the question "How would you feel if the relationship ended?" Mean ratings for this item on a 4-point scale (1 = would not be too upset to 4 = would be very upset/sad ) were 3.83 for kin VIPs and 2.75 for nonkin VIPs (see Table II).
Although there were no significant differences between kin and nonkin VIPs in the mean frequency of support provided, results showed that nonkin VIPs tended to provide more specific types of support, whereas kin VIPs were perceived to provide more diffuse support. Across the 16 different types of support listed, support from nonkin VIPs was concentrated on "expressed interest and concern for adolescent's well-being" (M = 3.23), "let adolescent know that he/she did something well" (M = 3.22), and "indicated that he/she will always be around if adolescent needs help" (M = 3.20). They rarely provided the following support: "gave or loaned money or something else that was needed" (M = 2.00) and "provided a place to stay" (M = 1.69). For kin VIPs, however, support was moderate to high for all 16 categories of support (min M = 2.20 and max M = 3.09).

"Truly Key" Versus Other VIPs
To explore which factors may distinguish "truly key" VIPs from other VIPs, we examined both demographic and relationship factors. Demographic factors included the age, gender, and kinship status of the VIPs and gender of the adolescents. Relationship factors included length, quality, quantity, and mutuality of the relationships.
In terms of relationship factors, there were significant differences in the frequency of support, ways in which VIPs were important, and frequency of contact. Truly key VIPs were reported to be important in more ways (M = 8.32) and engage in a higher frequency of supportive behavior (M = 3.04), even though they had less frequent contact with the adolescent (M = 3.02) than did other VIPs (Ms = 7.06, 2.48, and 3.30, respectively). As mentioned earlier, truly key kin VIPs also reported more conflict with the adolescent. There was no significant difference in the length of relationships and mutuality of the relationships between adolescents with key and nonkey VIPs.

DISCUSSION
On the basis of this study of a community sample, it seems clear that a large majority of adolescents had an important relationship with a nonparental adult, indicating that this is a normative component of adolescent development. As further evidence for the normative nature of adolescent-VIP relationships, fewer than one in four adolescents reported that significant life events triggered the onset of the relationship, and quality of adolescent-VIP relationships was not associated with quality of parentadolescent relationships. Instead, one factor that seems to contribute to the gradual establishment of adolescent-VIP relationships is the VIPs' interest in the adolescent as an individual. For example, VIPs cited "[the adolescent] is fun to be around" as one of the major reasons for becoming involved with him/her. Future research might explore the extent to which it is normative for adults to take a special interest in young people, as Erikson's concept of generativity (Erikson, 1956) might suggest.
Data from this study also suggest that VIPs tend to provide a combination of positive adult qualities (e.g., providing advice, serving as a role model) and "peer-like" relations (e.g., nonjudgmental, nonpunitive, fun). Thus, the special niche provided by VIPs appears to be qualitatively different from that provided by parents or peers. Through their relationships with VIPs, adolescents often have an experientially rich and interpersonally supportive environment for development. In short, it appears that VIPs comprise an additional, and important, component of adolescents' lives. This view is consistent with previous research showing that effects of VIPs on behavior and mood are independent of analogous family and peer factors (Greenberger et al., 1998).
The literature on adults' social networks portrays important differences between the function of individuals who are kin and those who are not kin (e.g., Litwak & Szelenyi, 1969;Wellman & Wortley, 1990). Consistent with findings from this line of research, we found that adolescents' relationships with their kin VIPs were more likely than those with nonkin VIPs to be long term and diffuse. More specifically, relationships with kin VIPs were of longer duration, and were characterized as providing a wider array of social and instrumental support. Thus, the nature of adolescents' relationships with their VIPs may mark the beginning of the structure and nature of adult social networks.
Demographic factors generally did not distinguish truly key VIPs from the other VIPs. Instead, VIPs were considered to be truly key because of the ways in which they had become important to adolescents and the frequency with which they provided support to them.

Implications of the Findings
Results of this study suggest that naturally occurring VIPs are an important but underrecognized context for adolescent development. Psychologists and service providers have very limited understanding of the intricate processes of socialization that take place in this context. The findings of this study, if replicated in other samples, have implications for practitioners and policy makers. Most programs in place today focus on high-risk populations and "assigned" mentors. Although many "assigned" mentoring programs such as Big Brothers/Big Sisters are effective, unfortunately, over half of the number of youths who are matched remain on a waiting list (Freedman, 1993). This large population of unmatched youths indicates that assigned mentoring programs do not have the capacity to fulfill the needs of all children. In addition, although assigned mentor relationships have been shown to have substantial positive impacts on certain youths, the mentor-child "match" is often less than ideal. Although it has great potential, assigned mentorship is not the cure-all for all youths in need of adult support.
Conversely, the more "naturally occurring" adolescent-VIP relationships we described in this paper may well be a normative component of adolescent development. Several key studies on naturally occurring relationships of this kind (i.e., Garmezy, 1985;Rutter, 1987;Werner & Smith, 1982) have demonstrated links between the presence of an important nonparental adult and better child outcomes. Practitioners and program planners should consider facilitating and strengthening the naturally occurring important relationships between adolescents and nonparental adults. On the one hand, many youths in need of additional adult investment languish on the waiting lists of assigned mentorship programs or may be ineligible for such programs. On the other hand, some youths do have existing relationships with adults in their environment who might play a more consequential role in their development. It might be especially worthwhile to develop programs that expand opportunities for kin VIPs to receive services, education or training (e.g., in ways of managing conflict with adolescents' parents), and other forms of support similar to those that typically are provided to assigned mentors. As shown in this study, kin VIPs are especially important for and very invested in their adolescents.

Limitations of This Study
Although this study included extensive surveys and in-depth interviews with both adolescents and their VIPs, several limitations of the study should be noted. First, we relied on retrospective report to obtain information about the onset of the relationships. It is possible that adolescents who have had long-standing relationships with their VIPs are likely to have provided less reliable data than adolescents whose VIP relationship onset was more recent. Another potential problem involves the restricted range of age, settings (suburban), and socioeconomic status (i.e., working and middle class of our sample). Also, the number of youths from various minority backgrounds was insufficient to permit ethnic subgroup analyses. Results of this exploratory research, therefore, need to be examined in other youth-populations. Readers should be cautious about generalizing the findings of the current study to adolescents of a different age and with different demographic attributes. Younger adolescents, for example, because of their more restricted range of social contacts, may be more likely to identify kinfolk than nonkin as VIPs. Similarly, youths from ethnic groups that emphasize strong familial ties also may be more likely to have VIPs who are kin.