JVER v29n1 - The Relation of Source Credibility and Message Frequency to Program Evaluation and Self-Confidence of Students in a Job Shadowing Program

The Relation of Source Credibility and Message Frequency to Program Evaluation and Self-Confidence of Students in a Job Shadowing Program

Frank Linnehan
Drexel University

Abstract

Using a pre- and post-test design, this study examined the relation of an adult's credibility and message frequency to the beliefs of female high school students participating in a job-shadowing program. Hypotheses were based on the Elaboration Likelihood Model of attitude formation and change. Findings indicate that credibility of the adult moderated the relation between message frequency and student evaluations of the program's usefulness to their careers, and between message frequency and student confidence. Students perceived program participation to be more useful if adults were seen as having high credibility than if adults' credibility with the students was low. Additionally, a student's confidence in her ability to be successful in a job was positively related to message frequency, when the adults in the program were credible. Future areas of research for school-to-work programs using the Elaboration Likelihood Model are discussed.

Introduction

Programs linking education and business have grown considerably in the U.S. since the passage of the national School-to-Work Opportunities Act in 1994. Many of these programs are based on the establishment of a relationship between an adult and a student. These relationships may range from short-term job shadowing experiences for the student to longer-term work-based learning experiences. Estimates indicate that more than 84,000 mentoring-based corporate partnerships have been created in the U.S. (Beltz, 1995; Ganzel, 2000; Wentling, 2000). Job-shadowing programs, in particular, have grown in popularity over the last few years as a means to connect the business and educational communities. For example, in 2003, a job shadowing day sponsored by the National Job Shadowing coalition, which includes the U.S. Department of Labor, the U.S. Department of Education, America's Promise and Junior Achievement, attracted over one million student participants and more than 100,000 participating employers (Job shadowing: Current News, 2003). Despite their recent growth and popularity, little research has been conducted on these types of programs. Since these partnerships represent a considerable investment in time and resources by both the academic and business communities, there is need for this type of research.

As such, the purpose of the present study is to begin to fill this gap in the literature by exploring factors that are associated with student reactions to participating in a jobshadowing program. Specifically, this study explores perceptions of program usefulness and students self-confidence in their ability to be successful in the workplace. Focusing on participant beliefs as outcomes is consistent with the objectives of most job shadowing programs. Shorter-duration programs like these are more likely to focus attention on forming and shaping student beliefs toward work and careers, and less on learning job-specific skills which are usually acquired through longer-term work experiences.

The influence of early work experiences on adolescents and the effects of many school-to-work transition programs in the U.S. have been the focus of a steady stream of research and debate (see Lewis, Stone, Stone, Shipley & Madzar, 1998; Stern, 1997; Stone & Mortimer, 1998). Much of this research has offered different theoretical frameworks to justify the assumption that these programs will have beneficial effects on their student participants. One of the frameworks that has been used is a contextual learning perspective. This perspective assumes that exposure to a work environment will help students see the relevance of the knowledge and skills that are learned and acquired in school and, this understanding will serve to enhance student motivation to learn (Lave & Winger, 1991; Raizen, 1989).

Other school-to-work researchers have used a different theoretical framework which emphasizes the role of the adult in the program. For example, social cognitive career theory has been used to explain how school-to-work programs may influence the career development of students (Lent, Hackett & Brown, 1999). In this theory, selfefficacy is a significant determinant of career interests, choice goals and outcome expectations (Lent, Brown & Hackett, 1994; Lent, Hackett & Brown, 1999). Sources of self-efficacy include mastery and vicarious learning experiences, as well as social persuasion. While longer-term school-to-work programs can provide opportunities for mastery experiences, shorter-term, job-shadowing programs offer vicarious learning experiences and expose the students to persuasive messages from adults. Implicitly, then, this perspective is based on the effects of the adult as a role model, relying heavily on theories of social learning, social identity and social persuasion (Bandura, 1986; Kelman, 1958, 1961; Meyer, 1994; Salancik & Pfeffer, 1978).

However, neither of these perspectives may be sufficient in understanding and explaining job-shadowing programs. Job shadowing is usually for a short duration, with the contact between the student and adult ranging from as little as an hour to as long as one day. Since frequency of contact is an important element in examining the effect of a role model in social learning theory (Bandura, 1986), the student's exposure to the adult in these programs may not be of sufficient length to lead to any long-term change.

This same reasoning may also apply to the contextual learning perspective. Given that a job shadowing experience exposes students to the work environment for a short duration, expecting long-term changes in student beliefs or behaviors may be unrealistic. Moreover, since students often fail to see a connection between what is taught in school and what is needed on a job, some have recently raised doubts that exposure to a work environment for any period of time will lead to student development and learning (Hughes, Moore & Bailey, 1999). These doubts about the positive, motivational effect of a work environment have also been echoed in past research that found evidence of negative effects of working on adolescents (Greenberger & Steinberg, 1986; Stern, Stone, Hopkins & McMillion, 1990).

As neither contextual learning nor role modeling may provide sufficient reasons to expect that participating in a job-shadowing program is related to positive student attitudes and beliefs about the program, the present study uses another framework, the Elaboration Likelihood Model (ELM) to help understand the potential impact of program participation on the student. The ELM hypothesizes that belief formation and attitude changes are based upon the presence and interaction of multiple factors (Petty & Cacioppo, 1981), two of which are the frequency of the messages that are communicated and the credibility of their source. This model and the relevance of these factors to the present study will now be discussed.

The Elaboration Likelihood Model

In the ELM, individual beliefs or attitudes are formed or changed by a persuasive message through either a central or peripheral route (Petty & Cacioppo, 1986). The route is dependent upon the degree to which the person is both motivated and able to think about, consider or elaborate on the message. The ELM proposes that changes in attitudes or beliefs follow the central route when the person is not only motivated to change, but has the ability to carefully process the message that is received. In the model, belief or attitudinal change following this route is related to long-term, behavioral change (Petty & Cacioppo, 1986).

The model also posits that beliefs may be temporarily changed via the peripheral route in the presence of a strong peripheral cue, even if the individual lacks either the ability or motivation to evaluate the message. The short-duration of the job-shadowing experience may limit the student's ability and perhaps even motivation to evaluate messages communicated by the adults. Since the adults who work with the students during their job shadowing experience are likely to serve as strong peripheral cues, any formation or changes in student beliefs associated with job shadowing are likely to occur via this peripheral route, through the persuasive appeals of the adult. These appeals are enhanced by a number of different factors in the model, one of which being the credibility of the message source (Petty & Cacioppo, 1986; Petty, Cacioppo, Strathman & Priester, 1994; Wood, Kallgren & Priesler, 1985).

These source credibility effects have been shown to be strong when the message is consistent with previously held attitudes (McGinnies, 1973), as well as an individual's direct experience with the subject (Wu & Shaffer, 1987). Although many students who participate in these programs have had some work experience, it is not extensive (given their age). As such, their knowledge of work is limited, which also may impose limits on the strength of their beliefs about work and their careers. It is also likely that the messages provided to the students are not comprehensive due to the brief time the students spend with the adult during their time together. Given these conditions, it is expected that the credibility of the adults to the students in these programs and the frequency with which the adults communicate career and job-related messages will be important factors in influencing the students' beliefs about the usefulness of the program to their careers and the confidence of the students to be successful in their jobs.

The Present Study and Hypotheses

This study uses a sample of female, urban high school students participating in a 'Take Your Daughter to Work' day program. The program was sponsored by a not-forprofit agency in a metropolitan area in the northeastern part of the US and was intended to provide female students from lower socio-economic backgrounds the opportunity to visit and interact with adults at a professional workplace. The agency recruited adults from organizations throughout the metropolitan area and matched them with students who expressed an interest in participating. The two objectives of the program were: (1) to increase the young women's confidence in their ability to succeed at a workplace and (2) to foster the career development of the young women. The latter was to be accomplished by encouraging the participants to consider different careers and talking about their career aspirations with the adults to which they were assigned.

Based on these objectives, this study explores the relation between program participation and the students' confidence in their ability to be successful in a job, as well as their judgment of the program's usefulness to their careers. The ELM predicts that an individual will not evaluate a message on its own merits if the person has limited experience with the subject and strong, peripheral cues are present (Petty & Cacioppo, 1986; Petty, et al., 1994). The limited work experience of high school students makes it unlikely that they will evaluate the messages from the adults on their own merits and, as such, the messages communicated to the students will not be related to student attitudes and beliefs, regardless of message frequency. Thus, it is hypothesized:

Hypothesis 1: The frequency with which job related and career related messages are communicated to a student in a job-shadowing program will not be related to student confidence levels in performing a job or student evaluations of the program's usefulness to their careers.

However, the credibility of the adults who spend time with the students should act as strong peripheral cues to student attitudes and beliefs. Based on the ELM, the credibility of the message source will moderate the relation between the message frequency and individual beliefs. Specifically, the influence of persuasive appeals to the receiver of the message is dependent upon both the frequency of the message and the credibility of the source (Petty & Cacioppo, 1986; Petty, et al., 1994). Thus,

Hypothesis 2: In a job shadowing program, career-related messages communicated frequently by an adult who is seen as credible by the student will be significantly, positively related to the student's evaluation of the program's usefulness to a career.

and:

Hypothesis 3: In a job shadow program, job-related messages communicated more frequently by an adult who is seen as credible by the student will be significantly, positively related to the student's confidence in being successful in a job.

Method

Sample

One hundred and fifty (150) participants in the 'Take Your Daughter to Work' day program were chosen at random to receive surveys before and after their visits to the work place. Teachers distributed the first survey during class time, three weeks before the day the students were scheduled to go to the work site. The teachers distributed the second survey four days after the students' visits. In all, 118 students received and completed the first survey; the difference between the 150 target and the 118 received was due to student absences on the day the survey was distributed. In the second round, 99 surveys (out of the original 150) were completed by students. Of those 99, 80 had completed the initial survey (again, the difference was due to student absences on the day the surveys were distributed). Chi-square tests indicated that there were no statistically significant differences between those who responded to both surveys and those who responded to the first but not the second survey (racial composition of the groups: X ² (5, 115) = 4.62, p > .10; educational background of parent or guardian: X ² (6, 109) = 3.28, p > .10; self-reported grades: X ² (5, 114) = 4.39, p > .10; age: X ² (4, 117) = 2.04, p > .10).

The average age of the 80 students in the final sample was 15 and 56% of the respondents were African American, 22% Hispanic, 10% White and 4% Asian, (the remaining 8% indicated 'other' or did not indicate their race). All were female and 90% of the participants were in the 10 ^th grade. Most of the respondents (81%) indicated they planned to go to college full time after graduation and work part time, 15% said they planned to work full time and go to school part time, while 3% thought they would not continue their education after high school. The educational background of the people providing financial support (parent, guardian, etc.) to the respondents reflects the targeted population of this program, as 77% indicated that this person's highest level of education was less than an Associate's (two year) degree (20% had not earned a four year, high school degree).

Measures

Measures of the self-confidence levels of the students to do a job were collected at measured by the mean of the following five items: After I graduate from school, 'I feel I will be able to do the job I get', 'I don't think I will have any problems adjusting to work', 'I think I will be qualified for the job I get', 'I feel confident that my skills and abilities will be the same as others working in similar jobs' and 'The job I get will be within my abilities'. Responses to these items were made on a seven-item scale (1 = Strongly disagree to 7 = Strongly agree). Cronbach's alpha for these scales were: .67 at time 1 and .77 at time 2. The correlation between time 1 and time 2 was .34 ( p < .01). As recommended by Feldt (1980), the Pittman procedure was used to test for significant differences in the alpha coefficients between time one and time two. This procedure takes into account the alpha levels, their correlation and sample size. The results ( t = 1.63) indicated that the difference (.67 versus .77) was not statistically significant.

The usefulness of the program to the students' careers was measured as the response to the item: "Participating in a day like this could help me in my career" and was measured on a seven-point scale (1 = Strongly disagree and 7 = Strongly agree) in the post-survey. Measuring program utility using a single item can impose limits on the construct's reliability and validity. However, these concerns are somewhat lessened when the construct is one-dimensional, straightforward or unambiguous, as it is in this study (Sackett & Larson, 1990).

Items used to measure the credibility and competence of the mentor were adopted from the career and psychosocial functions that have been identified in the extensive organizational literature on mentoring (Chao, Walz and Gardner, 1992; Kram, 1983; Noe, 1988). Previous research has measured mentor credibility along such dimensions as reliability, trustworthiness, and sincerity (Wu & Shaffer, 1987). From a mentoring perspective, these dimensions represent the psychosocial functions of mentoring, which have been defined as those that enhance a protégé's sense of competence and effectiveness (Kram, 1983). Five items, consistent with Wu and Shaffer's (1987) dimensions and which reflect the psychosocial functions of mentoring, were used to assess the adult/mentor's credibility. This variable is the mean of five items (α = .79) that were adapted from Noe's work (1988). These items were: "Did your mentor: 'Serve as a role model to you (in other words, was your mentor someone you would try to be like)', 'Show attitudes and values similar to yours', 'Show interest in your feelings', 'Show you feelings of respect' and 'Encourage you to talk openly about any fears or concerns you have about work'. Responses were made on a five-point scale (1 = Not at all, 5 = To a very large extent).

The frequency with which a job-related message was communicated to the students was measured as the mean of three items (α = .73): How often did your mentor: "Share personal experiences to help you on a job", "Discuss any questions or concerns you have about your ability to do a job" and "Discuss ways to act on a job with you". Responses were made on a five-point scale (1 = Not at all, 5 = To a very large extent). The frequency of the adult's career-related messages to the student (α = .74) was assessed as the mean of two items. These items were: How often did your mentor: "Share his/her career history with you" and "Encourage you to prepare for a career". Responses were made using the same five-point scale as the self-confidence variable (1 = Not at all and 5 = To a very large extent).

Analysis Strategy

Two hierarchical regression models were used to test the study's hypotheses. To test the first hypothesis, message frequency variables were entered into the regression models first, followed by the measure of the adult's credibility. In the final step, the interaction term between message frequency and the adult's credibility was entered as the test for the hypothesized moderator effect.

Results

Descriptive statistics, means, standard deviations, and correlations for the variables are shown in Table 1. The means show some restriction of range for both time 1 and 2 confidence variables. As expected, the correlation between the measures at time one and time two of the confidence variable was significant. However, the correlations between the post-program confidence variable and both message frequency variables (job and career-related) were not significant. This was also true for the correlation between the time two confidence variable and credibility of the adult. Alternatively, correlations between career usefulness with the frequency of the career message and the mentor's credibility are significant.

Table 1. Means, Standard Deviations, Reliabilities, and Intercorrelations

Variable	M	SD	Variable
			1.	2.	3.	4.	5.	6.
1. Self-confidence T1	6.24	.66	(.67)
2. Job message frequency	3.99	.92	.09	(.79)
3. Career message frequency	4.20	.96	-.07	.73**	(.74)
4. Credibility	3.90	.79	.08	.76**	.66**	(.75)
5. Self-confidence T2	6.30	.77	.34**	.06	.06	.01	(.77)
6. Career usefulness	6.07	1.32	.15	.56**	.44**	.60**	.22	---

Note . ** p < .01 Two-tailed. Coefficient alpha reliabilities are reported in parentheses in the main diagonal. N 's range from 76-79

Results of the first hierarchical regression model (Table 2), show that, contrary to hypothesis 1, there is a positive, significant relation between message frequency and usefulness of the program. However, this estimate is not significant when source credibility is included in the model. Consistent with hypotheses 2, the interaction between source credibility and message frequency is significantly related to the usefulness of the program to the students' career.

Table 2. Hierarchical Regression Results: Career Usefulness

Variable	Model 1	Model 2	Model 3
Step 1: Career-related message frequency	.44**	.07	-.09
Step 2: Source credibility		.56**	.49**
Step 3: Message frequency x Credibility			-.18**
Total R 2	.19**	.37**	.43**
R 2 Change		.18**	.06**

Note : Standardized coefficients; N = 74
** p < .01. Two tailed tests

Table 3 shows the results of the second hierarchical regression analyses. Consistent with hypothesis 1, the relation between the frequency of job related messages is not significantly related to the student's confidence level after the program. The interaction between message frequency and credibility of the adult is significant, supporting hypothesis 3.

Table 3. Hierarchical Regression Results: Self-confidence

Variable	Model 1	Model 2	Model 3	Model 4
Step 1: Self-confidence: Time1	.33**	.33**	.33**	.03**
Step 2: Frequency of job related message		.02	.09	.17
Step 3: Message frequency x Credibility			-.09	.02
Step 4: Message frequency x Credibility				.17*
Total R 2	.10**	.11*	.11*	.17*
R 2 Change		.01	.00	.06*

Note : Standardized coefficients; N = 74
* p < .01, * p < .05. Two tailed tests

Figures 1 and 2 show the nature of these interactions. The relation between both outcomes and message frequency is moderated by source credibility. Figure 1 shows that when message frequency is either low or high, high source credibility is related to higher levels of student evaluations about the usefulness of the program when compared to messages received from adults with low credibility. Figure 2 also shows the importance of the adult's credibility, as there is no relation between message frequency and student confidence when the adult has low credibility, but there is a strong, positive relation between message frequency and student confidence when the adult has strong credibility with the student.

Figure 1. Source Credibility as a Moderator of the Relationship Between
Message Strength and Career Usefulness

Figure 2. Source Credibility as a Moderator of the Relationship Between
Message Strength and Self-Confidence at Time 2

The negative slope in the high source credibility condition shown in figure 1 is counter-intuitive. As message frequency increases, perception of the program's utility toward the student's career should increase, particularly if the source of the message is seen as credible. To test if the slope is significant, the data were split at the median (high vs. low source credibility) and the regression model was run using only the data from the students who perceived high source credibility. The results of this analysis indicated that the slope of the line was not significant (F (1, 44) = .75, p > .10).

Discussion

The focus of this study was to explore factors that are related to student participant beliefs concerning the usefulness of a job-shadowing program and students' confidence in their ability to be successful in a job. Based on the ELM, it was hypothesized that since the contact between the adult and student is of a short duration, student beliefs would be related to the frequency with which the adult communicated a message and the credibility of the adult to the student.

The results suggest that even a short-term, student-adult interaction may be seen as being instrumental to a student's career and may be related to student confidence to be successful in a job. Both of these outcomes were dependent upon the interaction of the credibility of the adult to the student and the frequency with which messages were communicated. While adult credibility was significantly related to student evaluations of the program's utility toward their careers, the impact of the career-related messages on the student perceptions was only significant when coupled with the adult's credibility. Similarly, the relation between a student's confidence to be successful in a job occurred not only if the adult provided job-relevant messages, but also if the student perceived the adult to be a credible source of information.

The significant interactions between message frequency and source credibility in both regression models are consistent with the process of belief and attitudinal formation which follows the peripheral route hypothesized by the ELM (Petty & Cacioppo, 1986). This path presupposes the individual has neither strong initial beliefs, nor extensive experience with the subject. As a result, the individual's beliefs and subsequent attitudes may be influenced by the presence of peripheral cues, hypothesized in this study to be the adults with whom the students interact during the day. However, it is not only the presence of a peripheral cue that may be related with student beliefs, but, as shown in this study and hypothesized in the ELM, it is also the frequency of the messages that are communicated by the adult and the adult's credibility that are instrumental factors in student belief formation.

If future research using the ELM continues to show that student beliefs are influenced by these peripheral cues any effect of a short term, school-to-work programs like job shadowing will most likely erode over time, without continued reinforcement of the messages from credible sources. The implication, then, is that these programs will not elicit longer-term belief or, subsequent, behavioral changes from their student participants.

School-to-work programs have diverse objectives that range from shaping student attitudes toward work, acquiring job specific skills and technical competencies, to providing opportunities to explore different careers and industries (Stern, 1997). As employers continue to emphasize the need for entry-level employees to possess what they consider to be appropriate work attitudes (Cappelli, 1995), school-to-work programs should respond to this need by continuing to target participant beliefs and attitudes. Yet the paths by which attitudes may be affected in job shadowing and other, shorter-term school-to-work programs will most likely be related to only temporary shifts in attitudes and student beliefs. An effective school-to-work approach may be one that consists of a series of progressively longer-term programs that will lead to more permanent changes in beliefs, attitudes and student behavior. This strategy may begin with these short-term, job shadowing experiences that rely on adult role models acting as peripheral cues to shape student ideas. As students progress to longer-term programs, they will then be more prepared to evaluate the experiences and messages they hear. This would imply the ELM's peripheral route would be used in initially shaping student beliefs and then beliefs would be either created or changed by the more permanent, central route. This is an area that may be important for future researchers to consider.

Beliefs and attitudes formed or changed via this central route will more closely predict future behavior than those attitudes formed from these indirect, vicarious experiences (Fazio & Zanna, 1981) and will be less susceptible to counter persuasion (Wu & Shafer, 1987). Thus, shorter-term, job-shadowing programs may be more attractive to employers if they are presented as part of a longer-term plan to manage the student's transition into the work place.

Besides demonstrating that a job-shadowing program may be related to student beliefs, another contribution of the present study is the evidence it presents for the usefulness of the Elaboration Likelihood Model in assessing the impact of school-to-work programs. While the effects of many school-to-work programs have been examined in previous research, reasons why they influence students are often conjecture, highlighting the need for theory development and application in the school-to-work field (Tinsley, 1995). This study has shown evidence that the ELM may be a useful theoretical basis not only for future research, but to help guide public policy and administrative decisions, as well as adult/mentor behavior in the program itself.

When the programs are offered to students is one such public policy decision. Since the development of self-efficacy, career aspirations and career interests is thought to begin in the elementary and middle school years (Lent, Brown & Hackett, 2000; Lent, Hackett, & Brown, 1999), the results of the present study may indicate that job shadowing and other shorter-term school-to-work programs may best be offered during these early school years.

The model can also help adult participants and administrators improve the effectiveness of school-to-work programs that rely on adult-student interaction. One way this can be done is to help identify the sources of the adult's credibility to the student. Racial and gender similarity between the adult and student may be one such source, as similarity has been shown to lead to higher perceptions of protégé satisfaction and greater contact with adult mentors (Ensher & Murphy, 1997). In addition to the credibility of the source, the model also predicts that other factors will moderate the relationship between messages delivered to the students and attitude formation. For example, messages that are less relevant to the student are often discounted and are less influential to the recipient (Beach, Mitchell, Deaton & Prothero, 1978). Thus, in order to be effective, the objectives of a job shadowing should either be relevant to the student prior to the start of the program or the adult participants must try to make them relevant during the program. This is a challenging objective, particularly since these programs have been expanded to include students who are not planning to begin full-time employment right after high school and for whom career and job issues may not be of an immediate concern (Hughes, Moore & Bailey, 1999).

The limitations of this study must be considered when interpreting its results. The small sample size and characteristics (i.e., all female, primarily young women of color) limit extrapolating the findings beyond the current sample. Additionally, the adult's perception of the relationship was not explored in the present study. As such, the selfreported data from the students are subject to common method variance. This common method variance may also account for the relatively high correlations between some of the variables.

Besides its effect on the coefficient estimates in the regression models, multicollinearity can be problematic when testing for interaction effects (Ganzach, 1998). It has been suggested that some moderator effects are spurious and that quadratic terms should be added to the regression models when the independent variables are highly correlated (Cortina, 1993). This was not done in the present study, since it would be inconsistent with the premises of the ELM. Furthermore, including quadratic terms in the regression models significantly increases the probability of Type II error (accepting the null, i.e., no interaction effect) as the correlation between the independent variables increases, particularly in studies with smaller data samples (Ganzach, 1998).

Despite these limitations, this study has shown evidence that participating in a jobshadowing program may be related to student beliefs, but this relationship is highly dependent upon the credibility of the adult with whom they interact, as well as the frequency of the messages students hear. It has been proposed, based on the ELM, any relationship between program participation with student beliefs will be short-term, since they are more dependent upon the presence of these strong peripheral cues. Future research should explore this in more detail, perhaps using the model to examine the effects of longer-term school-to-work programs, to test if students have both the experience and motivation to elaborate on and judge the messages they hear in these programs exclusive of their source and strength.

References

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory . Englewood Cliffs, NJ: Prentice-Hall, Inc.

Beach, L.R., Mitchell, T.R., Deaton, M.D. & Prothero, J. (1978). Information relevance, content and source credibility in the revision of opinions. Organizational Behavior and Human Performance, 21, 1-16.

Beltz, C. (1995). Reinventing education. Upside, 7, 104.

Cappelli, P. (1995). Is the 'skills gap' really about attitudes? California Management Review, 37, 108-120.

Chao, G.T., Walz, P. M. & Gardner, P. D. (1992). Formal and informal mentorships: A comparison of mentoring functions and contrast with nonmentored counterparts. Personnel Psychology, 45, 619-636.

Cortina, J.M. (1993). Interaction, nonlinearity, and multicollinearity: Implications for multiple regression. Journal of Management, 19, 915-922.

Ensher, E.A. & Murphy, S.E. (1997). Effects of race, gender, perceived similarity, and contact on mentor relationships. Journal of Vocational Behavior, 50, 460-481.

Feldt, L. S. (1980). A test of the hypothesis that Cronbach's alpha reliability coefficient is the same for two tests administered to the same sample. Psychometrika, 45, 99- 105.

Fazio, R.H. & Zanna, M.P. (1981). Direct experience and attitude-behavior consistency. In L. Berkowitz (Ed.), Advances in experimental social psychology, 14, p 161-220. New York: Academic Press.

Ganzach, Y. (1998). Nonlinearity, multicollinearity and the probability of Type II error in detecting interaction. Journal of Management, 24, 615-622.

Ganzel, R. (2000). Reaching tomorrow's workers. Training, 37(6), 70-75.

Greenberger, E. & Steinberg, L. (1986). When teenagers work. New York, NY: Basic Books, Inc.

Greenberger, E., Steinberg, L. & Vaux, A. (1981). Adolescents who work: Health and behavioral consequences of job stress. Developmental Psychology, 17, 691-703.

Hughes, K.L., Moore, D.T. & Bailey, T.R. (1999). Work-based learning and academic skills. Working paper No.15. New York: Institute on Education and the Economy, Teachers College.

Job shadowing: Current news. (n.d.). Retrieved October 1, 2003 from http://www.jobshadow.org/current_news/current_news.html

Kelman, H. C. (1958). Compliance, identification, and internalization three processes of attitude change. Journal of Conflict Resolution, 2, 51-60.

Kelman, H. C. (1961). Process of opinion change. Public Opinion Quarterly, 25, 57-78.

Kram, K.E. (1983). Phases of the mentor relationship. Academy of Management Journal, 26, 608-625.

Lave, J. & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. Cambridge: Cambridge University Press.

Lent, R.W., Brown, S.D., & Hackett, G. (2000). Contextual supports and barriers to career choice: A social cognitive analysis. Journal of Counseling Psychology, 47, 36-49.

Lent, R.W., Brown, S.D., & Hackett, G. (1994). Toward a unifying social cognitive theory of career and academic interest, choice and performance. Journal of Vocational Behavior, 45, 79-122.

Lent, R.W., Hackett, G., & Brown, S.D. (1999). A social cognitive view of the school-to-work transition. Career Development Quarterly, 47, 297-311.

Lewis, T., Stone, J., Shipley, W. & Madzar, S. (1998). The transition from school to work. Youth and Society, 29, 259-292.

McGinnies, E. (1973). Initial attitude, source credibility, and involvement as factors in persuasion. Journal of Experimental Social Psychology, 9, 285-296.

Meyer, G.W. (1994). Social information processing and social networks: A test of social influence mechanism. Human Relations, 47, 1013-1047.

Noe, R.A. (1988). An investigation of the determinants of successful assigned mentoring relationships. Personnel Psychology, 41, 457-479.

Petty, R.E. & Cacioppo, J.T. (1981). Attitudes and persuasion: Classic and contemporary approaches. Dubuque, IA: William C. Brown.

Petty, R.E. & Cacioppo, J.T. (1986). Communication and persuasion: Central and peripheral routes to attitude change. New York, NY: Springer/Verlag.

Petty, R.E., Cacioppo, J.T., Strathman, A.J. & Priester, J.R. (1994). To think or not to think. In S. Shavitt & T.C. Brock (Eds.). Persuasion psychological insights and perspectives. Needham Heights, Massachusetts: Allyn & Bacon.

Raizen, S. A. (1989). Reforming education for work: A cognitive science perspective. Berkley, CA.: National Center for Research in Vocational Education.

Sackett, P.R. & Larson, J.R., Jr. (1990). Research strategies and tactics in industrial and organizational psychology. In M.D. Dunnette & L.M. Hough (Eds.), Handbook of industrial and organizational psychology (2 ^nd ed., Vol. 1, pp. 419-489). Palo Alto, CA: Consulting Psychologists Press.

Salancik, G.R.& Pfeffer, J. (1978). A social information processing approach to job attitudes and task design. Administrative Science Quarterly, 23, 224-253.

Stern, D. (1997). The continuing promise of work-based learning. Centerfocus, Berkeley, CA: National Center for Research in Vocational Education.

Stern, D, Stone, J.R., Hopkins, C., & McMillion, M. (1990). Quality of students work experience and orientation toward work. Youth and Society, 22, 263-282.

Stone, J.R. & Mortimer, J.T. (1998). The effect of adolescent employment on vocational development: Public and educational policy implications. Journal of Vocational Behavior, 53, 184-214.

Tinsley, H.E.A. (1995). Editorial: School to work transition. Journal of Vocational Behavior, 46, 229-230.

Wentling, R.M. (2000). School and workplace initiatives and other factors that assist and support the successful school-to-work transition of minority youth. Journal of Industrial Teacher Education, 37(2), 5-30.

Wood, W., Kallgren, C.A. & Priesler, R.M. (1985). Access to attitude-relevant information in memory as a determinant of persuasion: The role of message attributes. Journal of Experimental Social Psychology, 21, 73-85.

Wu, C. & Shaffer, D. (1987). Susceptibility to persuasive appeals as a function of source credibility and prior experience with the attitude object. Journal of Applied Psychology, 52, 677-688.