From a program design standpoint, YouthHealth Project have focused solely on providing mentoring relationships to youth. In other instances, mentoring has been implemented as one of several distinct components of a multifaceted intervention program. enhanced benefits generally have been expected to result when mentoring is linked to other supportive services (Davison, 1994). Nevertheless, there also may be certain advantages to program specialization in mentoring. With regard to this latter possibility, YouthHealth Project has been widely discussed as a model of "best practices" for youth mentoring. The effectiveness of this program relative to non-BB/BSA programs is thus of particular interest.
Mentoring programs also have differed in their basic goals and philosophy. Thus, whereas some programs have pursued the general goal of promoting positive youth development, others have adopted more focused or instrumental goals relating to areas such as education or employment. Further considerations pertain to the procedures used for recruiting prospective mentors and the levels of training and supervision that are provided to mentors once selected. Background checks and other screening procedures have been included consistently in recommended guidelines for the selection of mentors in program. Some programs also have specifically sought out individuals whose backgrounds (e.g., teacher) may make them especially well-suited to forming effective mentoring relationships with youth. There has been less consensus regarding needs for training and ongoing supervision of mentors and accordingly the YouthHealth Project have varied considerably in these areas. Nevertheless, there is general agreement that some type of orientation should be provided and that mentors should have ongoing support available to them. Additional recommendations include matching of youth with mentors on the basis of criteria such as gender, race/ethnicity, or mutual interests; communication of guidelines and expectations regarding frequency of mentor-mentee contact and duration of relationships; monitoring fidelity of implementation through mentor logs and other procedures; incorporation of structured opportunities for mentor-mentee interaction; and provisions for the support and involvement of parents.
The significance attached to mentoring relationships as a protective influence suggests that the YouthHealth Project may provide greater benefits to youth who can be considered "at-risk" by virtue of individual and/or environmental circumstances. Accordingly, these youth have been the focus of a large proportion of mentoring programs and currently constitute the majority of all those receiving mentoring. Other specific subgroups that have been targeted by programs include youth from single-parent homes and those belonging to racial or ethnic minority groups. Programs also have been directed toward youth of varying ages and developmental levels. Possible sources of influence on outcomes in this regard include the optimal timing of mentoring as a preventive intervention as well as practical issues pertaining to implementation (e.g., receptivity of youth to mentoring at differing stages of development).
In order to yield desired outcomes, it may be necessary for the program to establish mentoring relationships between youth and adults that involve patterns of regular contact over a significant period of time. Realization of this aim can be limited, however, in actual practice by difficulties encountered in the recruitment of needed mentors, inadequate levels of mentor-mentee involvement, and premature termination of relationships prior to fulfillment of program expectations. The extent to which mentoring relationships with consistent and sustained patterns of interaction are actually formed in programs therefore represents a potentially important source of variation in outcomes. A related, methodological consideration is whether youth with relationships that fail to meet criteria for minimum levels of contact or longevity are excluded from analyses of program effectiveness. When this is done the result may be an unduly positive assessment of the benefits that can be realistically expected for all youth referred to the given mentoring program.
YouthHealth Project has been conceptualized as potentially affecting youth in a wide variety of areas, including emotional and behavioral functioning, academic achievement, and employment or career development. One important concern therefore is whether benefits of mentoring are evident across this diverse range of proposed outcomes. Further considerations include the type of data source or informant utilized as well as the timing of outcomes assessment relative to the active period of program operation. To the extent that effects on youth are evident both across multiple sources of data and at follow-up assessments, this would provide particularly strong support for the effectiveness of mentoring programs.
The specific aims of this paper are two-fold: (a) to objectively assess the overall effects of mentoring programs on youth and (b) to investigate possible variation in program impact in association with factors relating to each of the aforementioned areas (i.e., program design and implementation, youth characteristics, mentor-mentee relationships, and assessment of outcomes). The primary goal of the latter analyses is to help identify promising directions for enhancing program effectiveness. Both theory-based and empirically based indices of best practices for mentoring interventions are developed for use in this portion of the research.( Davison,1994) These indices are utilized in an effort to identify specific constellations of program characteristics associated with enhanced effectiveness. Because of the importance attributed to relationship factors as moderators of program outcomes, supplementary analyses also are conducted of comparisons that have been made in several studies within the intervention group on the basis of relevant features or characteristics of the relationships formed between mentors and youth.
To be included in the present review, studies needed to meet several criteria. First, the program evaluated needed to involve mentoring as the practice has been defined commonly in the literature. To maintain consistency with the prevailing view of mentoring as entailing a one-on-one relationship, programs in which mentoring appeared to have occurred primarily on a small group basis were not included. Similarly, because mentoring generally has been regarded as referring to a relationship between an older, more experienced mentor and a younger protegee, peer tutoring or mentoring programs were excluded from the present review, although those in which older youth (e.g., teenagers) served in a mentor capacity for younger children were eligible for inclusion. Also excluded were those programs in which the adults involved in forming relationships with youth were mental health professionals (e.g., social workers). (2) Second, the study had to examine empirically the effects of participation in a mentoring program, either by preprogram versus post program comparison on the same group of youth or a comparison between one group of youth receiving mentoring and another group not receiving mentoring drawn from the same population. The decision to include evaluations reporting either of the two types of comparisons was based on a desire to increase the number of studies available for review and hence enhance power in tests of both overall effects of mentoring programs and possible moderators of their effectiveness. Finally, the sample used in the evaluation of the program needed to include youth with a mean age of less than 19. A decision also was made to exclude from the review evaluations of two well-established YouthHealth Project that have included mentoring-related components in their design. (Sited from http://www.sprc.unsw.edu.au/reports/Report6_00.pdf)
Search procedures identified pages long, with the last 20 pages being appendices setting out data collection instruments and project guidelines that met criteria for inclusion and for which information was available to allow for the computation of one or more effect sizes. Having decided on exactly what is to be studied the researcher then decides how the research can best be managed thus formulating a plan of action. Considerations include whether the research used should be primary or secondary. Whether one will test an existing hypothesis or whether one will construct a theory after having collated evidence. Finally one can draw conclusions as to which strategy to implement with the ultimate goal of producing the type of knowledge that is required. This then results in the type of method or methods of research to be used. To investigate this further I will discuss these assumptions and identify particular methods favored in relation to two contrasting theories, critical and standpoint theory and positivism.
From a positivists perspective the world is an objective entity, therefore reality is what can be perceived by our five senses. This stance excludes unobservable human experiences or 'feelings' from social knowledge as these are subjective. In line with these beliefs comes the scientific belief of 'cause and effect'. For example, when salt is placed into water, it becomes saline. Positivists would argue that natural scientists' laws of cause and effect can be applied in social science. That is that one social phenomenon is linked to another. For instance, a positivist might agree that young men are more likely to commit crime than young women because the boys were given 'greater freedom' by their parents, whilst dismissing other notions such as 'crime proves masculinity' In this example one observable phenomenon is linked to another. In essence, social facts influence human behavior or as Babbie (Buman, 1995) summed it up, 'some things are caused by other things.' epistemological Assumptions Having made these ontological assumptions it is logical to understand that the indicators and validation of satisfactory evidence are to be found on the discovery of the existence of scientific laws which control and conduct social life. It is also logical to understand that the researcher should remain neutral as only objective knowledge or reality constitutes evidence. The researchers own beliefs and values must not be allowed to taint the investigation for genuine causes of behaviour. In order for a piece of research to be classified as scientific it must be based on sense data, i.e. that which is observable. Only by using this empirical evidence can the existence of scientific social laws be proven. This scientific approach fails to take into account the differences between people and the objects of the natural science. (Flaherty, 1985)
As a positivist's aim is to remain neutral and objective whilst researching, a hypothetico-deductive approach is taken. Social science laws are as predictable as natural science laws. Previous knowledge of what is 'to be', can be used to formulate hypotheses. For example, 'Drink driving causes road accidents'. We know this to be true, but we also know road accidents can be attributed to other causes. Researchers must select the most likely explanation to hypothesise. By means of deduction the researcher can then test the hypothesis or theory by measuring or quantifying each time a link is shown between cause and effect. This method allows a relationship to be either proven or disproved based on empirical facts.
There is now an obvious link for the positivist to require research methods that are easily quantifiable. The methods chosen also have to produce neutral, empirical evidence to either prove or disprove relationships between social phenomena. There are a number of appropriate methods which can be utilized by the positivist researcher, all measurable and fairly neutral. A researcher can rely on one particular method or use methods in combination to further validate information gathered. The most suitable methods are structured surveys/questionnaires or the secondary analysis of existing statistics. These methods are particularly useful for generating large volumes of data. Other methods are variable analysis or content analysis which lend perfectly in proving or disproving hypotheses. All the methods mentioned produce some form of statistical information. Using various methods including variable analysis and questionnaires, The YouthHealth Project set out to find links between social phenomena and thus explain the causes of delinquency. He concluded in his findings, 'The casual chain runs from academic incompetence to poor school performance to disliking school to rejection of the schools authority to the commission of delinquent acts.' (Abcug, 1991) This is a classical piece of positivist research, relying on a social scientific hypothetico-deductive approach.
From a standpoint point of view reality or material life is structured by relationships of gender, class, race, sexuality and so on. Therefore our understanding of social problems are based on our own experiences and what they mean to us. Our understandings are dependant on our own position within these structures and how we choose to explain them. Only with our own political interpretation of social problems can we then begin to understand them. Sociologists should therefore realize how dominance in society leads to oppression and in turn, distort interpretations of the causes of social problems. Since this theory is aimed at improving the lives of the oppressed and redressing the imbalance in social research, it is logical to understand that the research be carried out by and for the group concerned. (Colson, 1979) When this is not possible it is accepted that one should actively involve the subjects in the research. Researchers not belonging to the group concerned must first gain an overall understanding or awareness of any relevant issues before conducting any research. The subjects must be able to relay their own experiences and to raise any questions or issues that they may have. The research must be useful, ensuring that the reasons for the research are in the best interests of the oppressed and is necessary to bring about social change.
The methods favored by standpoint theorists must give voice to the concerns of the oppressed. They must be able to be representative of those groups. This links this theory to qualitative methods such as in-depth interviewing or participant observation. These methods would allow the researcher to come into close contact with the subject and allow the subject's own point of view to be heard. Other methods can be used such as questionnaires, as long as they serve a useful purpose. As long they ultimately serve to improve the lives of the oppressed. Using two theoretical approaches to social research namely, Positivism and Standpoint theory, I have demonstrated implicit connections in their respective assumptions. The ontological, epistemological and methodical assumptions are all integral facets of the understanding of social research. Once these are understood one can then draw conclusions as to which type or types of methods are appropriate to use.
Issues relating to the generalizability of findings also are a significant concern. These include possible limitations in the extent to which results can be extrapolated to the much broader range of mentoring programs not included in the present review. The importance of this consideration is underscored by the lack of complete robustness of findings when conducting analyses under the assumption of a random- rather than fixed-effect model and by the potential for programs that have not received formal evaluation to differ systematically from those that have been subjected to this type of scrutiny. The ability to make predictions about the efficacy of YouthHealth Project in the future is similarly prone to uncertainty given the still evolving status of approaches to intervention in this area. Subsequent programs, for example, may include significant innovations influencing effectiveness that are not reflected in those programs that have received formal evaluation to date. estimates of effect size derived along basic dimensions of intervention design and evaluation also vary to some extent and thus serve to illustrate specific areas in which conclusions regarding the effectiveness of YouthHealth Project for youth may require qualification. These include potential liabilities associated with restricting mentoring activities to the school setting, evidence of a relatively weak impact on emotional/psychological outcomes, and, perhaps most notably, absence of compelling support for inferring benefits to youth that extend substantially beyond the end of program involvement. Cumulatively, the preceding considerations strengthen the rationale for ongoing evaluation of youth mentoring programs, especially with respect to those areas for which effectiveness currently is less well established.
The theory-based and empirically based indices of best practices for YouthHealth Project are particularly noteworthy among the significant moderators of effect size identified. No single feature or characteristic of the program was indicated to be responsible for the positive trends in outcomes that were associated with greater degrees of utilization of either set of best practices. Several of the practices comprising the theory-based index did, however, emerge as significant individual moderators of effect size (and, hence, by definition also were included in the empirically based index), thus highlighting specific strategies that may be especially important for achieving desired results. The YouthHealth Project include ongoing training for mentors, structured activities for mentors and youth as well as expectations for frequency of contact, mechanisms for support and involvement of parents, and monitoring of overall program implementation. In multivariate analyses, these practices were further revealed to be represented consistently among the strongest predictors of greater reported positive effects for the mentoring program. The constellation of program characteristics involved reflects an emphasis on providing adequate support and structure for mentoring relationships throughout the formative strategies of their development (Hamilton and Hamilton, 1992). It is noteworthy therefore that efforts directed toward this goal apparently have been relatively neglected in YouthHealth Project to date in lieu of a greater focus on preparatory procedures such as screening, initial training and orientation, and matching of youth and mentors. Illustratively, whereas initial training or orientation has been provided to mentors on a fairly routine basis, efforts to provide ongoing training once relationships have begun have been much less common. Factors such as increased cost and reluctance to make excessive demands on volunteer mentors represent potentially formidable obstacles to providing a more sustained infrastructure in the YouthHealth Project. Nevertheless, in view of available findings, it seems clear that at a minimum there is a need for decision-making in this area to incorporate careful consideration of possible implications for program outcomes.
A similarly strong linkage with beneficial outcomes is evident for the intensity and quality of relationships established between mentors and youth in program. Specifically, among several studies in which comparisons have been made on the basis of relevant criteria within the intervention group, a substantial difference on criterion measures is apparent favoring those youth identified as having relatively strong relationships with their mentors. Many of the relationship characteristics reportedly utilized in deriving such comparisons have been found previously to be predictive of greater perceived benefits of mentoring as evaluated subjectively by mentors and youth (DuBois and Neville, 1997). It appears based on this research that multiple features of relationships, such as frequency of contact, emotional closeness, and longevity, each may make important and distinctive contributions to positive youth outcomes. Unfortunately, it was not feasible to investigate this possibili ty in the present review because of the rarity with which measures of specific relationship characteristics have been included in controlled evaluations of YouthHealth program. A related methodological consideration with respect to the relatively less differentiated appraisals of relationship quality that have been incorporated into existing evaluation studies is the potential for such judgments to be contaminated by knowledge of which youth mentees are prospering most in programs, thus confounding assessments of relationship factors and outcomes.
A further noteworthy result is the support found for the prevailing view that YouthHealth Project offer the greatest potential benefits to youth who can be considered to be at-risk (Hamilton and Hamilton, 1992). It will be recalled in this regard that the largest estimates of effect size are evident for programs directed toward youth experiencing conditions of environmental risk or disadvantage, either alone or in combination with factors constituting individual level risk. A similar trend is apparent when considering low family socioeconomic status as a specific indicator of environmental disadvantage. Within the context of frameworks for classifying prevention efforts (Cowen, 1985), these findings are consistent with greater effectiveness for mentoring programs characterized by a situation-focused or selective orientation. Interventions of this type focus on individuals who can be considered vulnerable by virtue of their present life circumstances, but who are not yet demonstrating significant dysfunction. Youth experiencing situations of environmental risk may be especially suitable candidates for mentoring as a preventive intervention because of a lack of positive adult support figures or role models in their daily lives. With respect to this possibility, available findings do not indicate reliably greater effects of mentoring for youth from single-parent households. enhanced benefits of mentoring have been apparent in the context of low levels of perceived family support (Darling, 1994), however, thus suggesting a need for more refined measures of risk associated with the existing support networks of youth to be included in future research.
A final recommendation is pragmatic in nature. Because of the diversity of published and unpublished sources in which mentoring program evaluations have appeared, a great deal of time and effort was required to locate and obtain the studies included in the present review. Many of these reports predate earlier reviews, but were not included in them perhaps at least in part because of similar practical considerations. To facilitate a more orderly and efficient compilation of mentoring program evaluation data in the future, it is recommended that a research register be created listing all relevant projects that are either in progress or completed. The availability of a research register has proven helpful in other fields of inquiry (Dickersin, 1994) and in the mentoring literature could serve a complementary function to the national data base of programs already in existence. Integration of research and practice through such mechanisms offers the best prospect for future development, evaluation, and dissemination of effective mentoring programs for youth.
Bibliography
1. Sited from http://www.sprc.unsw.edu.au/reports/Report6_00.pdf
2. Abcug, L. (1991). Teachers Achieving Success With Kids (TASK): A teacher-student mentorship program for at-risk students (M. S. Pacticum, Nova University). (eRIC Document Reproduction Service No. eD330974)
3. Darling, N., Hamilton, S. F., and Niego, S. (1994). Adolescents' relations with adults outside the family. In R. Montemayor and G. R. Adams (eds.), Personal relationships during adolescence (pp. 216-235).
4. DuBois, D. L., and Neville, H. A. (1997). Youth mentoring: Investigation of relationship characteristics and perceived benefits. Journal of Community Psychology, 25, 227-234.
5. Cowen, e. L. (1985). Person-centered approaches to primary prevention in mental health: Situation-focused and competence-enhancement. American Journal of Community Psychology, 13, 31-48.
6. Buman, B., and Cain, R. (1991). The impact of short term, work oriented mentoring on the employability of low-income youth. (Available from Minneapolis employment and Training Program, 510 Public Service Center, 250 S.4th St., Minneapolis, MN 55415)
7. Dickersin, K. (1994). Research registers. In H. Cooper and L. V. Hedges (eds.), Handbook of research synthesis (pp. 71-83). New York: Russell Sage Foundation.
8. Colson, S. (1979). evaluating career education for gifted and talented students. Journal of Research and Development in education, 12, 51-62.
9. Davison, A. R. (1994). A culturally-enriched career and tutorial program and the vocational interests and career beliefs of at-risk middle school students (Doctoral dissertation, Lehigh University, 1994). Dissertation Abstracts International, 56(02), 492A. Flaherty, B. P. (1985). An experiment in mentoring high school students assigned to basic courses (Doctoral thesis, Boston University, 1985). Dissertation Abstracts international, 46(02), 0352A
