Continuity in life-span career development: career exploration as a precursor to career establishment. (Articles).

The authors examined continuity in career development from adolescence to middle adulthood by testing the proposition that early developmental task-coping activity predicts later task-coping activity. One hundred forty-six rural high school graduates repotted career exploratory activity in 9th grade and 12th grade, occupational choice clarity in 12th grade, and occupational establishment activity 25 years later. Controlling for gender, school grades, and verbosity, occupational choice clarity predicted midcareer establishment activity. The prediction models were somewhat different by gender.

**********

The life-span developmental perspective on careers emphasizes continuity, that is, the progressive dynamic processes of maturation and adaptation. Life-span, life-space developmental theorists (Super, Savickas, & Super, 1996) adopted the constructs of (a) life stage (see Buehler, 1933) to depict predominant states in the unfolding process of career development and (b) developmental tasks (see Havighurst, 1953) to capture the predictable, socially impinged adaptive challenges within each stage. Developmental tasks are distinguished from unpredictable events, such as economic conditions and natural disasters, which also require adaptive processes. Life stages are segments of the life-span career, and developmental tasks are the socially normative, age-graded influences on behavior during each stage. The stage names signal the principal tasks to be mastered at each of five life career stages: growth (usually experienced between ages 4 and 13 years), exploration (ages 14-24 years), establishment (ages 25-44 years ), maintenance (ages 45-65 years), and disengagement (age 65 years and over).

Life-span theorists propose that developmental continuity is manifested through the progressive mastery of age-graded tasks (Super et al., 1996). Success in mastering tasks at one stage results in both effective functioning within that stage and preparation to address the tasks of subsequent stages. This principle is consistent with Havighurst's (1953) developmental proposition that "good performance on a task at one age will be followed by good performance on this or similar tasks at later ages" (p. 320). The details of good performance, or career development task mastery, are elaborated in Super's (1963) list of attitudes and behaviors that are associated with vocational developmental tasks. In addition, Savickas (1984) deconstructed task mastery into coping responses, that is, "behaviors instrumental to satisfactory and satisfying responses to the tasks" (p. 222). Identifying specific coping behaviors as good performance on stage-specific developmental tasks permits counselors and researchers to assess int erindividual differences within a given life stage and allows for examination of intraindividual developmental changes across the life course. In summary, the major constructs in career development theory are organized hierarchically, with life-span career divided into life stages, each of which contains developmental tasks and a range of coping behaviors that constitute task mastery.

The sequence of career life stages and tasks is not fixed and inevitable, as are stages in Piaget's developmental theory, for example. Although the career life stages and tasks are linear and predictable, they do not occur in invariant order. The ages and sequence at which people encounter the tasks of each life stage may vary, depending on individual biosocial development and life situations. For example, some people will cope successfully with exploratory tasks and enter the establishment stage in their early 20s, some will continue to explore indefinitely, and others will, in their 30s, return to exploratory tasks in order to enter a different type of work. According to developmental theory, one set of behaviors will be dominant during a life stage; for example, exploratory behavior will generally override other behaviors during adolescence. Life-span theorists argue for a form of continuity and are clear about the cumulative effect of task mastery on career development.

Success in adapting to each developmental task results in effective functioning as a student, worker or retiree and lays the groundwork for mastering the next tasks along the developmental continuum.... Skipping a stage can result in difficulties at a later stage. (Super et al., 1996, p. 131)

Super (1982) proposed that mastering tasks in an early stage is a precursor to but not a prerequisite of mastering tasks in succeeding stages. For example, he responded to evidence in Herr, Good, McCloskey, and Weitz's (1982) longitudinal study and cited anecdotes from his own career pattern study to affirm that "exploration is desirable but not that exploration must precede establishment" (Super, 1982, p. 255). Therefore, across many people, coping with developmental tasks follows a predominant, but not fixed, sequence. According to theory, exploratory coping behaviors are desirable precursors to career establishment coping behavior.

Contemporary life-span theorists also recognize the importance of continuity across developmental processes of career development. Savickas (2001) recommended a "new general theory of ontogenetic development from growth to disengagement" (p. 304) that emphasizes developmental processes rather than structures. These are issues "that can only be resolved through developmental research that examines connections between early developmental processes and later developmental processes and outcomes" (Savickas, 2001, p. 299).

The purpose of our study was to examine career development continuity by describing the connections between early (exploratory) career developmental task-coping behaviors and later (establishment) task-coping behaviors in the lives of the same people. The study had three objectives that were descriptive and predictive. The first was to describe developmental task-coping responses during the high school years and 25 years later. Second, we tested the hypothesis, derived from the progressive mastery proposition, that adolescent exploratory behaviors (exploratory activity and choice clarity) will predict adult establishment behaviors. Third, we examined the relationship between exploratory and establishment behaviors separately for male and female participants.

We focused on gender by incorporating it into the predictions. Gender must be considered in any study of careers across adolescence and adulthood during the late twentieth century. In a recent summary of women's career development, Fitzgerald and Harmon (2001) concluded that "life is, in many ways, less gendered that it was in the 1970s"; however, "there are still ways in which life and its opportunities are partially controlled by gender stereotypes" (p.226). Several writers have argued that "women's work lives are more complex than men's" (Fitzgerald, Fassinger, & Betz, 1995, p. 76). For example, women may encounter additional developmental tasks that men do not, or society may impose a different timetable for them than it does for men.

Research Review

The relationship between exploration and establishment behavior within a life span has been studied as part of several prospective longitudinal research projects. The best known study of this kind is the Career Pattern Study (CPS), which was conducted by Super and his colleagues (see Super, 1985, for a complete summary). In 1951, they interviewed 100 ninth-grade boys from families of predominantly blue-collar workers in Middletown, New York. Follow-up surveys and interviews were conducted when the participants were in high school and again when they were 21, 25, and 36 years old. The longitudinal design allowed them to examine relationships between adolescent and young adult career behaviors. Ninth-grade measures of vocational maturity--the extent to which the boys mastered exploration stage tasks--were characterized as "weak, even though useful" (Super, 1985. p. 409) predictors of age 25 career establishment scales. Twelfth-grade vocational maturity measures were about as powerful predictors of career establ ishment as conventional measures (i.e., socioeconomic status, high school grade point average [GPA], and self-concept; Super, 1985). On the other hand, anomalies were discovered among the extensive list of coping behaviors used as individual predictors of career establishment at age 25. For example, the direct relationship between 9th-grade occupational information and later career establishment is negative and countertheoretical (Super, 1985). Kleinberg (1976) studied an expanded sample of 148 CPS men, including boys from the succeeding class, and found a significant negative correlation between high school occupational information and adult occupational stability.

In accordance with developmental stage theory, exploratory career decisions (i.e., deciding to seek more information) are expected to emerge and predominate before terminal career decisions (i.e., deciding to stabilize). Phillips (1982a) analyzed the CPS data and found that exploratory career decisions were exhibited in high school and also at ages 21, 25, and 36 years, although with decreasing frequency, and that stabilizing behavior increased from age 21 to age 25 and to age 36. In a second study of 95 GPS participants, Phillips (1982b) tested the hypothesis that individuals who follow the prescribed pattern in which exploration preceded commitment would achieve more desirable career outcomes than would those who did not. No significant differences among exploration patterns were found on participants' career establishment scores at age 25, but, at age 36, participants with exploration patterns of increasing commitment scored significantly higher on the career progress factor than participants who had patte rns of consistently low commitment. In summary, Phillips's results supported the sequence of stage-relevant behaviors. Adolescent predictors were not as strong for early (age 25) as they were for later (age 36) establishment behaviors. Super (1985) reported on two unpublished dissertations that used CPS exploration and establishment data. Walvoord (1979) found that exploration from ages 15 to 25 years was positively related to several outcomes at age 36 years, thus supporting the prediction of adult career status from adolescent career processes. Exploration before age 18 was correlated with occupational satisfaction and attained status. Using a different approach, Lin (1981) found that patterns of adolescent exploration predicted patterns of early adult establishment.

The CPS findings analyzed by several investigators and methods generally supported the progressive mastery proposition that careers develop through a sequential progression of life stages and that this progression is contingent upon mastery of the developmental tasks of earlier life stages. The findings must be considered in light of GPS design limitations: The sample was small, all male, and from one community during a unique period in history. The most notable limitation is the failure to consider gender and gender socialization as important shapers of career development.

In the 1960s, Gribbons and Lohnes (1982) began a well-designed 20-year study in suburban Boston of the careers of 110 male and female participants from adolescence through their transition to early adulthood. Although they did not directly test a progressive mastery proposition, Gribbons and Lohnes did find that readiness for vocational planning (i.e., exploration task mastery) was a weak predictor of global clinical transitional coping behavior ratings for 2 to 4 years after high school. Over several Markov chain analyses, they found "an innate orderliness" (p. 59), suggesting that predictions of late (rather than very early) transition behavior may yield more substantial relationships.

Two longitudinal studies conducted on a large, mixed gender population tested hypotheses from Super's developmental theory (Super et al., 1996), including the progressive mastery principle. Herr et al. (1982) examined the effects of high school curricula (academic vs. vocational) and gender on completion of adolescent career exploration tasks and young adult career establishment tasks. More than 1,000 young adults were surveyed 6 or 8 years after they graduated from one of three Pennsylvania high schools. Curriculum, but not gender, contributed significantly to career developmental task completion; however, the authors noted that only 3% of task mastery variability was predicted. Both curriculum and gender predicted certainty about occupational plans but not career satisfaction. Herr et al. (1982) suggested that curriculum incumbency is associated with exploratory opportunities, time sequences, and concepts of choice. Using a randomly drawn sample from the same data set, Niles and Herr (1989) questioned wheth er successful coping with the developmental task of crystallizing a vocational preference while in secondary school leads to success in developmental tasks in early adulthood, ages 24-27 years. The aggregate adolescent measures of vocational attitude maturity and knowledge of interests and abilities were significant predictors of the establishment of coping behaviors. None of the individual variables were a significant predictor of career establishment coping behaviors. Niles and Herr interpreted their findings as identifying significant predictors but "not of sufficient power to warrant individual prediction" (p. 351).

An example of a social psychological perspective on life-span career is Clausen's (1991) archival study of life course development. In a complex analysis of a 50-year longitudinal data set, Clausen found that planful competence in adolescence (ages 15-18 years) correlated .38 and .45 with career orderliness (progressively advancing positions) and career satisfaction, respectively, for 64 men at ages 53-62 years. Clausen interpreted planful competence as a trait rather than as coping responses to developmental tasks. He argued that individuals who are more competent in adolescence tend to assess the options available to them and make considered choices and, thus, are better prepared than less competent planners to work through the problems of adult career adaptation.

In summary, longitudinal studies have produced consistent but not strong relationships that support the proposition of progressive mastery of stage-specific career developmental tasks. Exploratory coping behaviors emerge early in adolescence, dominate, and then decline at theoretically predicted times and comparative rates. Evidence generally supports the proposition that early career coping behaviors predict later coping behaviors, although prediction seems to be stronger for criteria observed after about age 30 years than before that age. The few tests that have investigated gender differences have produced mixed results.

Method

The design of this empirical test of life-span career development continuity across stages was strengthened by at least two design conditions: (a) a longitudinal data set with appropriately timed assessments and (b) comparable measures of task-coping behaviors at adjacent stages.

Longitudinal Data

We used portions of a 29-year longitudinal data archive for this study. (See Jepsen, 1972, 1975; Jepsen & Choudhuri, 2001, for more details.) Participants were 146 members of the 1973 graduating classes from three contiguous rural Wisconsin high schools who completed questionnaires in 9th grade (November 1969) and 12th grade (March 1973) and a 25-year follow-up questionnaire (summer 1998). Two hundred forty-nine students graduated from the three schools, and high school data were available for all but a few who were absent on the data collection days. All participants were White and, with very few exceptions, were from families with two parents. Approximately one third were from farm families, 60% were from working-class families, and the remaining participants were from business and professional families. Additional descriptions of social and historical context are reported in Jepsen and Choudhuri (2001).

Information provided by class reunion committees allowed us to follow up a large portion of the graduates. We obtained addresses for 228 graduates in 1998) but 6 were deceased and accurate addresses were not available for 15 graduates. The return rate from living graduates with available addresses was 75%. Nonresponse was higher for men than for women and for those who left their home communities than for those who stayed. When we compared our response rates with those for one class's 20th reunion, we found that, generally, the same people did not respond to us or to their classmates. The sample is remarkable among follow-up studies for retaining farming and working-class members.

Measures of Coping Behaviors

Exploratory activity. Measures of coping behaviors were derived from the explication of developmental tasks for two consecutive life career stages: exploration and establishment. Exploration is the name for both a stage and an activity. Exploratory activity, the coping responses expected during this life career stage, has been construed and measured from two major viewpoints (Jepsen, 1984). The first is Jordaan's (1963) multidimensional definition, widely accepted among developmental theorists, that refers to exploratory behavior as

activities, mental or physical, undertaken with thc marc or less conscious purpose or hope of eliciting information about oneself or one's environment, or of verifying or arriving at a basis for a conclusion or hypothesis which will aid one in choosing, preparing for, entering, adjusting to, or progressing in an occupation. (p. 59)

Exploratory behavior involves experimentation, investigation, trial, searching/questing, seeking out, trying out, and/or hypothesis testing. Hence, the GPS surveyed a wide range of activities--for example, curricular, cocurricular, employment, and reading--as indicators of exploratory activity (see Jordaan & Heyde, 1979). The second viewpoint on career exploration led to a series of important counseling outcome studies. Working from information-processing assumptions, Krumboltz and associates (e.g., Krumboltz & Thoresen, 1964) argued that instrumental exploratory behavior, called information-seeking behaviors, is necessary to reduce uncertainty about choosing a career option. Therefore, we defined instrumental career exploratory coping behavior as a sampling of public, observable activity that occurs in distinct incidents. The frequency of these incidents is an index of career exploration mastery, which we named exploratory activity.

The cognitive aspect of career exploration behavior involves crystallizing and specifying a choice (Super, 1963). Developmental theorists construe choice as a subjective sense of purpose or goals rather than as an objective indicator of personality traits. Successful developmental taskcoping in adolescence is manifested in the verbal behavior of occupational choice expression. During adolescence, occupational choice expressions are expected to become more specific, clear, and certain. Super (1963) identified three prominent cognitive process manifestations of crystallizing and specifying a choice: (a) consistency across choices, (b) specification of choices, and (c) confidence in choices. We reasoned that variables representing consistency across three choices, occupational choice clarity, and confidence in occupational choices provide reasonable estimates of cognitive exploratory behavior.

Establishment activity. As the social expectations of early adulthood occur, individuals between 25 and 44 years old seek to establish their niche in society and especially within the occupational world (Buehler, 1933; Super, 1963). During the establishment stage, people are expected to master three molar tasks. First, they must become stabilized in their occupation (i.e., adapt to the work environment and perform satisfactorily so that they would settle down in an occupation). Once stabilized, people must consolidate and maintain effective productivity records and relations with supervisors and coworkers to secure their position within an organization. Finally, some people advance (i.e., move ahead into new levels of responsibility; Super, 1963).

Coherent patterns of coping responses pertinent to each developmental task during the establishment stage were created by Dix and Savickas (1995), who clustered reports from 50 successful midcareer workers. For example, persons coping successfully with the task of stabilization reported that they functioned as a team player, sought to increase their job expertise, established high standards for themselves, and set measurable goals. Persons coping with the tasks of consolidation and advancement reported that they organized their time, worked hard, promoted open communications, networked with colleagues, and kept current in their field. The resulting taxonomy allows for the comparison of individuals within the establishment stage. Many of these behaviors also appear on an extensive taxonomy of competencies that are necessary for women's successful pursuit of academic careers (Hackett, Betz, & Doty, 1985). Therefore, we adopted an operational definition for establishment activity as the frequency of reporting in strumental responses to six of Dix and Savickas's coping behavior clusters.

Procedure

When the participants were in the 9th and 12th grades, questionnaires were administered under classroom conditions. (For more complete descriptions of participants and procedures, see Jepsen, 1972, 1975.) At each grade level, participants were asked to construct answers to questions on the Occupational Information Inventory (OII), a questionnaire comprising 13 open questions that are derived primarily from the Readiness for Vocational Planning Interview schedule (Gribbons & Lohnes, 1968). The two Oils were identical except for the addition ofa question about choice confidence in the 12th-grade form. The 9th-grade OII responses were categorized by trained raters for career exploration activity and measures of cognitive career exploration--choice clarity and choice consistency. The 12th-grade Oil responses were categorized for exploration activity and also for measures of cognitive career exploration: choice clarity, choice consistency, and choice confidence.

The 1998 Follow-Up Survey (FUS 98) is a five-page questionnaire that contains more than 40 open-ended questions or checklists in sections tided Present Job, Family, Leisure Pursuits and Community Service, Job History, Training and Education, Future Career Plans, and Twenty-Five Year Perspective. (A copy of the survey is available from the first author upon request.) The last section includes six questions about career establishment activities. The FUS 98 was mailed to all potential participants; individuals who did not respond to two letters were called and invited to complete the survey by telephone. Respondents were compensated $10, which they could use for themselves or contribute to their schools.

Career exploration activity is the number of separate information-seeking behaviors reported by participants on the OII in both the 9th and 12th grade. At each grade level, OII questions elicited participants' reports about reading materials, talking with people, and other investigative activities. Thus, we obtained two samples of self-reported career exploratory behaviors--one from the 2nd month into high school and the other from the 3rd month prior to graduation. The broad behavioral samples are assumed to be representative of career exploration during the adolescent years because they inquire at two intervals when choosing an occupational goal is particularly salient, orientation into high school and anticipation of the transition out of high school.

Reliability of career exploration activity was established by interrater comparisons. Raters scored the responses to three OII questions by counting the people with whom participants discussed occupational plans, the pieces of material they read, and the number of other investigative activities they completed. They were instructed to count plurals as two activities (e.g., "parents," several books," and "several visits" were counted as two activities). Several estimates of interrater reliability have been obtained over the years. A median correlation of .98 for was obtained for three raters on 9th- and 12th-grade OII responses (Jepsen, 1975), and later a score generalizability estimate of .92 was calculated across six raters and a sample of both 9th-and 12th-grade responses (Jepsen & Grove, 1986).

The validity of self-reported information-seeking behavior was checked in early studies (e.g., Krumboltz & Schroeder, 1965; Krumboltz & Thoreson, 1964; Thoreson & Hamilton, 1972) by asking persons who may have witnessed the episodes. All behaviors reported were either confirmed or were not confirmable. Despite the laudatory precaution of confirming self-observation, there remains the possibility of error due to respondents' ability to recall and compose thoughts in writing. In an attempt to reduce these extraneous influences, portions of the data archive were used as statistical controls. We reasoned that thinking and writing ability are reflected by academic achievement. Therefore, we adjusted career exploration activity scores by controlling for students' self-reported classroom grades (GPA) for the first semester of high school and averaged on a 4-point scale.

Choice clarity was assessed by having raters categorize the 12th-grade OII preferred occupational choice as specific, general, or vague. Later, the general and vague categories were collapsed to form a two-category scale. Generalizability of choice clarity scores on the three-category scale was estimated as .47 across six raters and two occasions (Jepsen & Grove, 1986), suggesting modest reliabilities. Additional analyses revealed that most of the error variance was due to fluctuation over time rather than due to raters.

Scales were constructed to assess choice consistency, choice clarity, and choice confidence as separate components of cognitive exploratory behavior. Unfortunately, the performance of these scales in earlier studies and in preliminary analyses for this study did not support construct validity. The three scales were not highly correlated (all rs < .18) nor did they demonstrate high stability over participants' high school years (Jepsen, 1975). The exception was choice clarity, which had moderate stability (9th to 12th correlation, .40 for girls) and relatively high scorer reliability and generalizability (Jepsen, 1972, 1975; Jepsen & Grove, 1986). Preliminary analysis also revealed that clarity, in contrast to consistency and confidence, was not correlated significantly with high school grades. This finding was similar to those of Jordaan and Heyde (1979) who found that choice consistency and specificity (i.e., clarity) measures loaded on different oblique factors for both 9th and 12th grade CPS boys and that both were independent of exploratory behavior measures, a factor they named "extensiveness and quality of information sources." Thus, based on empirical grounds, we decided to retain choice clarity as the sole variable representing cognitive career exploration.

Our operational definition of career establishment activity was the frequency of establishment activities reported on the 25-year follow-up survey. Six activities, identified by Dix and Savickas (1995) as representing general career establishment coping behaviors, were assessed using participants' responses to open-ended questions. The activities were (a) earning degrees and certifications, (b) receiving support from others, (c) achieving recognition for job performance, (d) getting ahead within an organization, (e) getting along with organizational objectives, and (f) fitting in with coworkers. The scale used to assess establishment activity consisted of the six coping activity categories for which the person reported behavioral episodes. The internal consistency of the six-item scale was measured by Cronbach alpha and yielded a coefficient of .71 for the total sample, .70 for male participants, and .73 for female participants. Adult establishment activity is theoretically associated with career satisfaction , that is, a person's general positive feeling about his or her overall career. Again using archival data, we found that career establishment activity was correlated .42 with a one-item measure of 25-year career satisfaction, thus providing support for the construct validity of the scale used to measure career establishment activity.

High frequency of adult self-reported activities, such as career establishment activity, may be simply explained as verbosity or the number of words and phrases an individual uses to compose questionnaire responses. For example, studies of supply response questions in the 1930s found that respondents who gave lengthy answers on one question tended to do likewise on another question (Wagner, 1939). Therefore, we adjusted career establishment activity scores statistically by controlling for verbosity and assumed that the resulting index was less contaminated and, hence, a more valid estimate of career establishment activity. Verbosity was estimated by summing the number of concepts and phrases written in response to five FUS 98 questions that elicited responses about topics other than establishment behaviors: expressed job satisfaction and dissatisfaction, leisure and recreational activities, community service activities, and self-descriptive adjectives. One additional piece of evidence from the data suggested that many of the FUS 98 participants accurately recalled their earlier thinking about careers. Sixty percent of the respondents named their high school occupational choice as one of three they had actually expressed when they were in the 12th grade.

In summary, career development task mastery for two life stages was measured as self-reported public coping behaviors. The participants reconstructed behavioral incidents from memory and reported them on questionnaires. The resulting molar scales, career exploration activity, choice clarity, and career establishment activity, condensed large samples of behavior to scales with limited range but that were distributed approximately normally and were interpretable with a small sample. We argue that scores can be compared across the life span because the scales (a) have similar demand characteristics and (b) sample theoretically stage-appropriate activity.

Data Analysis

The progressive mastery proposition was tested by both a simple bivariate correlation analysis and a complex hierarchical multiple regression analysis. The correlation between adolescent coping behavior and midcareer coping behavior provides an index of the direct relationship between the two molar indexes of-self-reported activity separated by 25 years. In other words, the correlation represents the relative ordering of people on stage-relevant activity at two stages in their careers. In fact, this was the design used by Havighurst (1953) in early studies of developmental tasks.

The regression procedure estimates the accuracy of the predicted relationship and the degree of relationship between a linear combination of predictors and the criterion (Licht, 1995). In addition, theoretically competing explanatory factors--gender, grades, and verbosity--are controlled statistically. In our study, we analyzed the aggregate and independent contributions of developmentally earlier events (career exploration and choice clarity) in predicting later events (career establishment) within the same life-span career. The dependent variable was career establishment activity, with controls for verbosity; the independent variables were 9th-grade and 12th-grade career exploration activity and 12th-grade choice clarity, with controls for gender and grades. A three-step regression analysis was used with the controls--gender, grades, and, verbosity--entered in the first step, both career exploration activity variables in the second step, and choice clarity entered in the third step. Semipartial correlations were examined to determine the correlation between individual predictors and the criterion when all other predictors were controlled. Because we conceived of the study as exploratory, a .05 alpha level was set for all statistical tests.

The independent contribution of gender role to predicting career establishment activity was examined using two procedures. First, gender was entered as a predictor in the first step of the multiple regression analysis. Second, separate three-step hierarchical multiple regression equations were constructed for men (n = 56) and women (n = 90), with controls only for GPA and verbosity. The validity of these analyses were threatened by low power, thus increasing the likelihood of not accepting a true difference.

Results

The frequency, distribution, and intercorrelations for the four coping behaviors are presented in Table 1. These data summarize career development task mastery over 29 years and reveal trends that are consistent with life-span theory. As expected, exploratory activity increased between the 9th and 12th grades. (See Jepsen, 1975, for a more detailed analysis.) Jordaan and Heyde (1979) also found that exploratory behavior increased from the 9th to 12th grade. Two thirds of our participants expressed a specific occupational choice in the 12th grade, but clarity of expression was not associated with exploratory activity as information processing theory would suggest (e.g., Krumboltz & Thoresen, 1964).

When we compared male and female participants, using a dependent t test, there were no differences on 9th-grade career exploration activity (p < .09), 12th-grade choice clarity (p < .44), and midcareer career establishment activity (p < .43). There were large and significant differences on 9th-grade GPA (p < .0001), 12th-grade career exploration activity (p < .0001), and midcareer verbosity (p .006). The mean GPA for male participants was 2.06 (SD = 0.50), whereas the mean GPA for female participants was 2.54 (SD = 0.69); male participants' 12th-grade mean career exploration activity was 3.63 (SD = 2.65), whereas 12th-grade mean career exploration activity for female participants was 5.89 (SD = 2.86). Such large differences in achievement and instrumental exploration seemed to represent different social norms among male and female adolescents.

The simple bivariate correlation between molar scales of exploratory activity and career establishment activity was .216 and .207 for 9th-grade and 12th-grade career exploration activity, respectively (both p < .01), demonstrating that these activities do co-occur across the career life span. Adolescent clarity of choice expression was correlated .213 with adult career establishment activity, which was also significant beyond chance (p < .01). All separate gender bivariate correlations between exploratory and establishment activities were statistically significant, with the exception of 9th-grade career exploration and establishment activities for male participants.

The multiple regression results, as shown in Table 2, identify the variables that contributed to predicting midlife career establishment. The first group of predictors revealed that a large portion of the prediction of career establishment activity was due to the assessment of verbosity, which was correlated .49 with career establishment activity and shared 24% of the variance. Gender and 9th-grade GPA together accounted for only 1% of the variance in establishment activity. Apparently, people who generally supply expanded responses are also likely to list activities that achieve establishment tasks. Gender and academic achievement did not appreciably influence establishment activity.

When the two exploratory activity predictors were added in Step 2, the predictability did not increase significantly, [R.sup.2] only increased from .251 to .271 (p < .16). Apparently, high school exploratory activity, when controlled for verbosity, gender, and GPA, does not predict adult establishment activity. The bivariate correlations were reduced from .216 and .207 to partial correlations of .104 and .100 for 9th- and 12th-grade career exploration activity, respectively.

Choice expression clarity was added as a predictor in Step 3, and the prediction of career exploration activity increased significantly from .271 to .302 (p < .014). Choice clarity alone was correlated .207 with career establishment activity when all other variables were controlled, suggesting that it is a relatively independent predictor. Choice clarity predicted career establishment activity both independent of and in combination with career exploration activity. Career exploration, on the other hand, only predicted establishment when combined with clarity or when there were no controls for verbosity. Gender did not contribute directly to predicting career establishment.

Results of separate regression analyses conducted for male and female participants are reported in Table 3. For women, but not for men, clarity added significantly (p < .05) to the prediction of establishment activity when other variables were controlled. The control variables exhibiting significant effects on establishment activity were slightly different: For women, only verbosity predicted establishment activity, and for men, both verbosity and GPA predicted establishment activity. In fact, verbosity and GPA accounted for 35% of the variance in the establishment activity of men, whereas they only accounted for 22% of the variance in women's establishment activity. Curiously, lower grades were associated with higher establishment activity scores among men. Apparently, women have a slightly different set of precursors to career establishment than do men.

Diagnosis of several key regression assumptions did not reveal serious threats to the validity of the results. The low intercorrelations among predictor variables in Table 1 suggest that multicolinearity was not a problem. Visual scanning of plots did not reveal patterns of nonlinear relationships. The possibility of specification errors (i.e., not including plausible predictor variables) remains an open question. Future studies may reveal that other measures of exploration contribute to predictions of midcareer development. Perhaps underlying personality factors or features of the social structure will contribute to explanations of continuity.

Discussion

We conclude that mastery of cognitive career exploration coping behavior, especially choice clarity, seemed to be linked, albeit weakly, with self- reported mastery of occupational establishment coping behaviors among rural high school graduates who entered the labor force in the 1970s. This claim is subject to several limitations. The sample was culturally and geographically homogeneous, and about one fourth did not participate fully. Coping responses were measured by self-report and retrospective accounts and, as such, were a social representation of self. Selective memory or concern with social appropriateness may have influenced responses. Finally, the accounts were responses to specific questions that did not tap all relevant coping responses (e.g., part-time jobs as exploratory activity or prospectively planned advancement initiatives as establishment activity).

Despite these limitations, and the untold intervening unpredictable events influencing careers over the 25 years after high school, the findings corroborate evidence from earlier longitudinal studies that show small but statistically significant relationships between adolescent and adult career development. The data generally support the progressive mastery proposition in life-span theory; continuity was more clearly evident for women. Careers progress in a moderately predictable order, and early coping successes are precursors to later coping successes. Co-occurring factors within a life span are not sufficient evidence for causation, of course, but the observed relationships reveal the connections or linkages that were deduced from theory. When antecedents correlate with consequences as hypothesized, the combination of data and logic is persuasive.

Conclusions about continuity will be strengthened by replications under similar conditions (i.e., life-span conceptualizations of developmental tasks, comparable measures across stages, and longitudinal designs). Direct replication seems unlikely due to the extraordinary costs in time and resources required to conduct prospective longitudinal studies. Nevertheless, as longitudinal data sets become more accessible, archival analyses can be conducted. Investigators will need to construct theoretically derived measures that best fit the data set. The Clausen (1991) study that we reviewed earlier in this article is an excellent example. In addition, life-span career studies should be expanded to include adult life roles other than occupation. For example, family, community service, and leisure roles undoubtedly influenced the participants' occupations and vice versa.

The results have important implications for career guidance policy and practice, especially for high school counseling programs. Many states, including Wisconsin where the study participants attended high school, have developed policies for "comprehensive sequential developmental guidance programs in schools" (Schutt, Brittingham, & Perrone, 1997, p. ix). The Wisconsin model, an excellent example of policy that is based on human development theory, translates developmental concepts into specific competencies that, in turn, are blended into goals, content areas, and interventions. The age-graded structure of the competencies assumes the progressive mastery principle. For example, Schutt et al. stated that "new learning expands previous levels of knowledge and skills" (p. 6).

This study provides longitudinal evidence that is relevant to such a model. Among the many developmental concepts that life-span career theory offers, we found two that are viable for planning high school counseling programs: exploration activity and choice (or goal) clarity. We believe that the two concepts are strong candidates for building school career guidance programs, but our emphasis is slightly different from the usual policy statements. We observed that public behavior (instrumental activity and expressed choices) were critical ingredients in exploration. Thus, facilitating the overt, active processes of exploring and clarifying may be especially critical in successful high school guidance programs.

Implications for high school career counseling practice are straightforward: Use interventions that are designed to stimulate public career exploration activity and clear expressions of occupational choices or goals. For example, evidence from many field experiments has shown that peer models and social reinforcement influence overt exploratory activity (e.g., Krumboltz & Schroeder, 1965; Krumboltz & Thoreson, 1964; Meyer, Strowig, & Hosford, 1970; Thoreson & Hamilton, 1972). We have also observed, in earlier field experiments with rural high school juniors, for example, that an "imaginary friend" simulation technique in small groups had a positive effect on choice-basis complexity (Warner & Jepsen, 1979) and that participating in problem-solving groups led to increased career exploratory behavior (Jepsen, Dustin, & Miars, 1982). Additional rested career counseling interventions have been identified by Baker and Taylor (1998). Increasing adolescents' exploratory activity and choice clarity may have important long-term payoffs, such as increased occupational establishment at midcareer.

TABLE 1

Means, Standard Deviations, and Intercorrelations for Measures of Coping
Behavior (N= 146)

Coping Behavior    Ages    M     SD   EXA9  EXA12   CLA    ESA

Exploration, 9th
 grade (EXA9)      13-15  3.23  2.46        .255 *  .061  .216 *
Exploration, 12th
 grade (EXA12)     17-19  5.02  2.98                .044  .207 *
Choice clarity,
 12th grade
 (CLA)             17-19  1.67  0.47                      .213 *
Establishment,
 midcareer
 (ESA)             42-44  3.53  1.83

* p < .01.

TABLE 2

Hierarchical Multiple Regression Analyses Predicting Midcareer
Establishment From High School Exploration and Choice Clarity (N = 146)

                                             [R.sup.2]
Step and Predictor    df     R    [R.sup.2]   Change      F      p

1 Gender            3, 142  .501    .251       .251     15.895  .000
  Verbosity
  9th-grade GPA
2 Exploration, 9th  2, 140  .520    .271       .019      1.855  .160
  grade
  Exploration,
  12th grade
3 Choice clarity,   1, 139  .550    .302       .031      6.243  .014 *
  12th grade

Note: GPA = grade point average.

* p < .05.

TABLE 3

Hierarchical Multiple Regression Analysis by Gender Predicting Midcareer
Establishment From High School Exploration and Choice Clarity

                                            [R.sup.2]
Step and Predictor   df     R    [R.sup.2]   Change      F     p

Women (n = 90)
1 Verbosity         2, 87  .466    .217       .217     12.073  .000
   9th-grade GPA
2 Exploration, 9th  2, 85  .489    .239       .022      1.220  .300
    grade
  Exploration,
   12th grade
3 Choice clarity,   1, 84  .523    .274       .034      3.985  .049 *
   12th grade

Men (n = 56)
1 Verbosity         2, 53  .589    .347       .347     14.082  .000
  9th-grade GPA
2 Exploration, 9th  2, 51  .607    .369       .022      0.890  .417
   grade
  Exploration,
   12th grade
3 Choice clarity,   1, 50  .627    .393       .024      1.992  .164
   12th grade

Note. GPA = grade point average.

* p < .05.

References

Baker, S. B., & Taylor, J. G. (1998). Effects of career education interventions: A meta-analysis. The Career Development Quarterly, 46, 376-385.

Buehler, C. (1933). Der mensehlich lebenslauf asl psychologisches problem [The human life course as a psychological subject]. Leipzig, Germany: Hirzel.

Clausen, J. S. (1991). Adolescent competence and the shaping of the life course. American Journal of Sociology, 96, 805-842.

Dix, J. E., & Savickas, M. L. (1995). Establishing a career: Developmental tasks and coping responses. Journal of Vocational Behavior, 47, 93-107.

Fitzgerald, L. F., Fassinger, R. E., & Betz, N. E. (1995). Theoretical advances in the study of women's career development. In W. B. Walsh & S. H. Osipow (Eds.), Handbook of vocational psychology (pp. 67-109). Hillsdale, NJ: Erlbaum.

Fitzgerald, L. F., & Harmon, L. W. (2001). Women's career development: A post-modern update. In F. T. L. Leong & A. Barak (Eds.), Contemporary models in vocational psychology (pp. 207-230). Mahwah, NJ: Erlbaum.

Gribbons, W. D., & Lohnes, P. R. (1968). Emerging careers. New York: Teachers College Press.

Gribbons, W. D., & Lohnes, P. R. (1982). Careers in theory and experience: A twenty-year longitudinal study. Albany: State University of New York Press.

Hackett, G., Betz, N. E., & Doty, M. S. (1985). The development of a taxonomy of career competencies for professional women. Sex Roles, 12, 393-409.

Havighurst, R. J. (1953). Human development and education. White Plains, NY: Longmans.

Herr, E. L., Good, R. H., III, McCloskey, G., & Weitz, A. D. (1982). Secondary school curriculum and career behavior in young adults. Journal of Vocational Behavior, 21, 243-253.

Jepsen, D. A. (1972). The impact of videotaped occupational field trips on occupational knowledge. The Vocational Guidance Quarterly, 21, 54-62.

Jepsen, D. A. (1975). Occupational decision development over the high school years. Journal of Vocational Behavior, 7, 225-237.

Jepsen, D. A. (1984). The developmental perspective on vocational behavior: A review of theory and research. In S. D. Brown & R. W. Lent (Eds.), Handbook of counseling psychology (pp. 178-215). New York: Wiley.

Jepsen, D. A., & Choudhuri, E. (2001). Stability and change in 25-year occupational career patterns. The Career Development Quarterly, 50, 3-19.

Jepsen, D. A., Dustin, R., & Miars, R. (1982). The effects of problem-solving training on adolescents' career exploration and career decision making. The Personnel and Guidance Journal, 61, 149-153.

Jepsen, D. A., & Grove, W. (1986). Generalizability of vocational decision-making scales. Measurement and Evaluation in Counseling and Development, 19, 67-76.

Jordaan, J. P. (1963). Exploratory behavior: The formation of self and occupational concepts. In D. E. Super, R. I. Starishevsky, N. Matlin, & J. P. Jordaan, Career development: Self-concept theory (pp. 42-78). New York: College Entrance Examination Board.

Jordaan, J. P., & Heyde, M. B. (1979). Vocational maturity during the high school years. New York: Teachers College Press.

Kleinberg, J. L. (1976). Adolescent correlates of occupational stability and change. Journal of Vocational Behavior, 9, 219-232.

Krumboltz, J. D., & Schroeder, W. W. (1965). Promoting career planning through reinforcement. The Personnel and Guidance Journal, 43, 19-26.

Krumboltz, J. D., & Thoreson, C. E. (1964). The effect of behavioral counseling in group and individual settings on information seeking behavior. Journal of Counseling Psychology, 11, 324-333.

Licht, M. H. (1995). Multiple regression and correlation. In L. G. Grimm & P. R. Yarnold (Eds.), Reading and understanding multivariate statistics (pp. 19-64). Washington, DC: American Psychological Association.

Lin, J. (1981). The career patterns of men in the exploration and establishment stages. Unpublished doctoral dissertation, Teachers College, Columbia University, New York.

Meyer, J. R., Strowig, R. W., & Hosford, R. E. (1970) Behavioral reinforcement counseling with rural high school youth. Journal of Counseling Psychology, 17, 127-132.

Niles, S., & Herr, E. L. (1989). Using secondary school behaviors to predict career behaviors in young adulthood: Does "success" breed "success?" The Career Development Quarterly, 37, 345-354.

Phillips, S. D. (1982a). Career exploration in adulthood. Journal of Vocational Behavior, 20, 129-140.

Phillips, S. D. (1982b). The development of career choices: The relationship between patterns of commitment and career outcomes in adulthood. Journal of Vocational Behavior, 20, 141-152.

Savickas, M. L. (1984). Career maturity: The construct and its measurement. The Vocational Guidance Quarterly, 32, 222-231.

Savickas, M. L. (2001). Toward a comprehensive theory of career development: Dispositions, concerns, and narratives. In F. T. L. Leong & A. Barak (Eds.), Contemporary models in vocational psychology: A volume in honor of Samuel H. Osipow (pp. 295-320). Mahwah, NJ: Erlbaum.

Schutt, D. A., Jr., Brittingham, K. V., & Perrone, P. A. (1997). The Wisconsin Developmental Guidance Model: A resource and planning guide for school-community teams. Madison, WI: Wisconsin Department of Public Instruction.

Super, D. E. (1963). Vocational development in adolescence and early adulthood: Tasks and behaviors. In D. E. Super, R. Starishevsky, N. Matlin, & J. P. Jordaan, Career development; Self-concept theory (pp. 79-95). New York: College Entrance Examination Board.

Super, D. E. (1982). Comments on Herr, Good, McCloskey & Weitz: "Career behavior." Journal of Vocational Behavior, 21, 254-256.

Super, D. E. (1985). Coming of age in Middletown: Careers in the making. American Psychologist, 40, 405-411.

Super, D. E., Savickas, M. L., & Super, C. M. (1996). The life-span, life-space approach to careers. In D. Brown, L. Brooks, & Associates. Career choice and development (3rd ed., pp. 121-178). San Francisco: Jossey-Bass

Thoreson, C. E., & Hamilton, J. A. (1972). Peer social modeling in promoting career behaviors. The Vocational Guidance Quarterly, 20, 210-216.

Wagner, I. F. (1939). Articulate and inarticulate replies to questionnaires. Journal of Applied Psychology, 23, 104-115.

Walvoord, J. E. (1979). Correlates and outcomes of exploratory behavior in adolescence and early adulthood. Unpublished doctoral dissertation, Teachers College, Columbia University, New York.

Warner, S. G., & Jepsen, D. A. (1979). Differential effects of conceptual level and group counseling format on adolescent career decision-making processes. Journal of Counseling Psychology, 26, 497-503.

David A. Jepsen, Division of Counseling, Rehabilitation and Student Development, The University of Iowa; Ginger L. Dickson, Lincoln University. This article is part of a project titled "Occupational Career Development From Rural Beginnings," which is supported by timely and generous grants to the first author from the University of Iowa Graduate College and the Spencer Foundation. The data and conclusions are solely the responsibility of the authors. Earlier versions of this article were presented at the National Career Development Association Conference, Portland, Oregon, July 1999, and the American Counseling Association World Conference, March 2001, San Antonio, Texas. Correspondence concerning this article should be addressed to David A. Jepsen, Division of Counseling, Rehabilitation and Student Development, The University of Iowa, N338B Lindquist Center, Iowa City, IA 52242 (e-mail: [e-mail]).