Postformal reasoning during adolescence and young adulthood: the influence of problem relevancy.

Sebby, Rickard A. ; Papini, Dennis R.

In an effort to more adequately conceptualize the nature of cognitive development following childhood and adolescence, several theorists (e.g., Arlin, 1984; Baltes, Dittmann-Kohli, & Dixon, 1985; Basseches, 1984; Labouvie-Vief, 1982; Kramer, 1989a; Perry, 1970; Sinnott, 1989) have proposed various stages or processes of thought that extend beyond Piaget's (1980) final cognitive developmental stage (formal operations). In general, these theorists posit a progression in thinking from dualistic or absolutist thought (truth vs. falsehood) to more subjectively determined modes of thinking (Kramer, 1983; Kramer, Kahlbaugh, & Goldsten, 1992; Perry, 1970) in which the relativistic and/or dialectical nature of knowledge is more thoroughly understood.

Various researchers who have tested this theoretical proposal have obtained results consistent with this view. Adams, Labouvie-Vief, Hakim-Larson, DeVoe, and Hayden (1989), for instance, report that adults demonstrated higher level reasoning (viz., they were more aware of multiple interpretations of a problem) when presented with problems that were embedded within situations typically faced by adults. By contrast, adolescent subjects performed at a lower reasoning level when presented with these same adult-relevant problems (viz., they gave solutions that were logical but exhibited little or no self-awareness of pragmatic or personal constraints).

Using a similar scoring sheme, Blanchard-Fields (1986) presented problems to subjects ranging in age from 14 to 46 years. These problems varied in emotional saliency (i.e., the degree of personal involvement subjects reported feeling for problems). Three problems (viz., war between fictional countries, a conflicted visit to grandparents by an adolescent and his parents, and the dilemma faced by a man and woman following an unintended pregnancy) were presented to each subject. Blanchard-Fields found that adolescents, relative to older individuals, were less able to engage in relativistic reasoning on problems found to be emotionally salient by the entire sample.

Following theoretical positions established by Labouvie-Vief(1982), Blanchard-Fields indicated that "an emotionally salient context may be more disruptive for younger than for older thinkers" due to younger people's inability to effectively integrate logical (cognitive) and emotional domains. For adolescents, the inability to integrate multiple domains of experience would seem to be derived from a dualistic (Perry, 1970) or an absolutist (Kramer, 1983) style of thinking in which they see people or situations as being fixed or unchanging. Other research (Kramer, 1989a, 1989b) indicates that adolescents' lack of integration may be due to their tendency to approach intimate reletionships with a sense of idealism or absolutism (e.g., this is the person for me, for always).

Two important observations are warranted at this point regarding the research conducted by Blanchard-Fields (1986). First, she presented problems to subjects that were "assumed" to be emotionally salient, and she only empirically verified the emotional saliency of the problems after they had been presented. Second, when emotional saliency was examined (post-hoc), no age-related effects or interactions were observed. That is, neither adolescents nor any other particular age group considered specific problems to be especially emotionally salient. Thus, one might argue that a more stringent test of the disruptive effect of emotional content on the logical interpretation of problems would be to utilize problems that had been determined a priori to be relevant to particular age groups. If emotional content is disruptive, then its maximal impact should be evidenced when subjects' reasoning is examined on age-specific relevant problems.

Recent research by Sebby and Papini (1991) would seem to underscore the importance of these considerations. These investigators utilized experimental problems that were developed specifically for young adults (M = 19 years), middle-aged adults (M = 31 years), and older adults (M = 68 years); the appropriateness of the problems for particular age groups was empirically established by subjects in a prior pilot study. Similar to the procedure used by Blanchard-Fields (1986), Sebby and Papini also asked their subjects (post-hoc) to rate the relevancy of each problem to their own personal life. For all age groups, problems with greater perceived relevancy were found to yield higher level reasoning, whereas problems with lower perceived relevancy evidenced lower level reasoning. Thus, Sebby and Papini provide evidence that the perceived relevancy of problems produces a direct and positive impact on reasoning performance. Whether a similar facilitative effect of problem relevancy would have resulted for adolescents is unknown, given that adolescents were not included in Sebby and Papini's study.

The present investigation, then, was designed to examine the effects of perceived relevancy on the reasoning performance associated with problems that were developed specifically to reflect adolescent issues and concerns. If adolescents are unable to effectively integrate emotional content (Blanchard-Fields, 1986; Labouvie-Vief, 1982), then problems that reflect the particular concerns and issues of adolescents should produce poorer reasoning performance than do problems developed for other age domains. Two specific hypotheses were examined. First, it was predicted that particular age groups (adolescents, young adults) would perceive particular problems as being more relevant to their own lives in accord with the issues or situations the problems were developed to reflect. Second, adolescents' reasoning performance was expected to be poorer, relative to older individuals, particularly on those problems that were perceived by adolescents to have greater relevancy to their own lives. Specific predictions concerning the performance of the two adolescent age groups, relative to the older age group, were not specified.

METHOD

Subjects

One hundred and five subjects (55 males and 50 females) between the ages of 13 and 29 years agreed to participate in this study. Thirty-seven early/middle adolescents (13 to 16 years of age), 36 older adolescents (17 to 19 years of age) and 32 young adults (20 to 29 years of age) from a small midwestern city were included in the sample with all subjects attending either high school or college.

Tasks

Problems similar to those used by Sebby and Papini (1991) and by Adams et al., (1989) were developed in a pilot study. The problems consisted of syllogism-like dilemmas which three hundred individuals (ranging in age from 16 to 62) determined were particularly relevant for adolescents, adults, or older adults. These syllogisms were constructed using "if-then" statements, with the only difference between types of problems being the personal relevancy of the context in which they were embedded. Formal syllogisms were designed to minimize the potential relevance of the problem's embedding context. Conversely, contextual syllogisms sought to maximize the potential relevance of the problem's embedding context. Three different types of contextual syllogisms were presented: adult-relevant (family/marital relations), adolescent-relevant (parent-adolescent relations), and older adult-relevant (retirement, relocation). Only syllogisms rated as age-relevant by 75% of the pilot subjects were subsequently included in the study.

Pilot data were used to identify four problems within each of the four problem types (adolescent-relevant, adult-relevant, older adult-relevant, and formal), resulting in a total of 16 problems.

Examples of Adolescent-Relevant, Adult-Relevant, Older Adult-Relevant, and Formal Problem Types

Adolescent-relevant problem type. Jean does not keep her room clean, and it is especially messy by Friday. Sara, Jean's mother, warns her that if her room is not clean by the time she gets home from work, she will not be allowed to go to the basketball game at school that night. Jean does not clean her room. Sara arrives home from work.

Does Jean get to attend the basketball game? How certain are you of your answer?

Adult-relevant problem type. John is known to be a heavy drinker, especially when he goes to parties. Mary, John's wife, warns him that if he comes home drunk one more time, she will leave him and take the children. Tonight John is out late at an office party. John comes home drunk.

Does Mary leave John? How certain are you of your answer?

Older Adult-relevant problem type. Ellen, who has lived in the same house for the last 25 years, has almost fallen down the stairs on at least a half-dozen occasions. After her last near-fall, Ellen promised that if it ever happened again she would sell her house and move into another that had only one floor. Yesterday, Ellen had another near-fall but avoided it by catching herself on the handrailing.

Does Ellen move into a new house? How certain are you of your answer?

Formal problem type. Nancy shows Tom a stack of cards with numbers printed on both sides of each card. Nancy tells Tom that if there is a number 2 printed on one side of a card, then the number 9 and the number 4 are printed on the other side of the same card. Nancy hands Tom a card with the number 2 printed on one side.

Is there a 9 and a 4 on the other side of the card? How certain are you of your answer?

Procedure

Subjects were asked to engage in two types of activities as part of this investigation. First, through the use of a "thinking aloud" strategy, they were asked to solve 16 problems that were embedded in contexts designed to differ with respect to their personal relevancy. Second, subjects were asked to rate the relevancy of each problem to their own life using a five-point scale. The scale anchors ranged from "very relevant to my life" to "not at all relevant to my life." Following these activities, subjects were administered a demographic questionnaire.

The subjects were individually interviewed by a trained experimenter in a quiet laboratory room. They were asked to verbally report thought processes as they solved each of the 16 problems (Giambra & Arenberg, 1980). A review of the thinking aloud strategy (Ericsson & Simon, 1980) suggests that it is an appropriate means of maximizing the report of mental processes involved in problem solving. The experimenters were trained to probe the subject's statements in order to ascertain the level of reasoning involved in the solution of problems. Specific attention was given during the training of the experimenters to methods by which subjects' thoughts could be elicited without biasing them toward particular answers. For example, experimenters were trained to encourage the thinking aloud process by using open-ended statements (e.g., "tell me more about what you're thinking").

Each participant's verbalizations were tape recorded and later independently scored by one male and one female rater. The raters were unaware of specific characteristics of the subjects (e.g., their age or education) when the ratings were performed. The average level of agreement reached by the two raters was 86%. When conflicts between the two raters arose, a third rater evaluated the problem and a discussion of the difference in the ratings resolved the conflict.

Coding Scheme

Using a scheme developed by Adams et al., (1989) and similar to that used by Blanchard-Fields (1986), each of the 16 problems was scored. In brief, the scoring scheme allowed the identification of the following three levels of reasoning specified by Labouvie-Viefs (1982) theory: intrasystemic, intersystemic, and integrated.

Level 1--Intrasystemic Reasoning. Problem-solving at the Intrasystemic level is based on the more or less absolute certainty of logical truth, with the implication that there is only one truth to be constructed during problem solving. Intrasystemic reasoning is characterized by the individual's inability to recognize that the premises upon which truth statements are founded are subjectively interpreted, and that the development of cognitive structures are embedded in socialization processes. The self-system is not integrated with formal logic during decision-making, and this results in solutions to problems that are logical but idealistic.

Level 3--Intersystemic Reasoning. Intersystemic reasoning emerges as the individual begins to establish mastery over formal operational logic. Problem-solving with intersystemic reasoning is evidenced by a transition from logical absolutism toward logical relativism. Adolescents and young adults recognize that logical truth may depend upon the embedding context of the problem. As this breakdown in logical certainty occurs, the individual begins to recognize that the self can integrate logical operations with pragmatic concerns. Thus, the relativism of truth contrasts with the many practical solutions to problems that the individual must choose from, with the knowledge that the specific solution selected is not more truthful than another.

Level 5--Integrated Reasoning. Integrated reasoning reflects the highest form of problem-solving in adulthood. At this level, structural reorganization integrates the regulatory systems of self and logic. The individual becomes consciously aware of the sells role in interpreting the premises of a problem and evaluating each potential solution in terms of pragmatic, social, cultural, and personal truth.

For each problem, raters assigned scores of 1, 3, or 5 corresponding to the intrasystemic, intersystemic, and integrated levels of reasoning. Scores of 2 and 4 were assigned to transitional level responses.

RESULTS

The prediction that reasoning would be affected by problem relevancy was examined in two steps. First, subjects' perceptions of problem relevancy were examined in order to determine whether particular problem types (i.e., adolescent-relevant, adult-relevant, older adult-relevant problems) were perceived differently by each of the three age groups. Second, given the results of the first step, age differences in reasoning for particular problem types were examined. For example, if subjects in one of the adolescent age groups perceived adolescent-relevant problems to be significantly more relevant to their lives than the other problem types, reasoning on these types of problems should be poorer (following Blanchard-Fields, 1986), relative to other age groups. Analysis of variance was used to test differences between the means. Tukey HSD tests were conducted (p [is less than] .05) for all of the follow-up comparisons between means reported.

Problem Relevancy

Subjects' ratings of the relevancy of each of the 16 problems were summed across the four problems within each problem type. The resulting sums were examined in a 3 x 4 (age group by problem type) analysis of variance with the last factor being repeated across subjects. The analysis indicated that problem type was significant, F(3,306) = 55.40, p [is less than] .001. Adolescent-relevant problems (M = 3.48) were perceived as being significantly more relevant than either adult-relevant (M = 2.70) or older adult-relevant problems (M = 2.54), with each of these problem types significantly differing from formal problems (M = 2.00).

Subsequent analyses conducted to examine whether particular age groups perceived specific problem types to be more relevant indicated an age-related difference in perceived relevancy. A one-way analysis of variance, F(3, 144) = 23.80, p [is less than] .001, indicated that subjects in the early/middle adolescent age group perceived adolescent-relevant problems (M = 3.72) to be the most relevant to their lives, while all of the other problem types were perceived as being less relevant (adult-relevant, M = 2.80; older adult-relevant, M = 2.47; formal, M = 1.97). Similar analyses conducted within the older adolescent (F(3, 140) = 14.85, p [is less than] .001) and the young adult age groups (F(3, 124) = 10.26, p [is less than] .001) indicated that these two age groups also perceived that particular problem types were more relevant than others. Older adolescents, like younger adolescents, perceived adolescent-relevant problems to be more relevant (M = 3.48) than any of the other problem types (Ms = 2.54, 2.45, and 2.08 for older adult-relevant, adult-relevant, and formal problem types, respectively). Young adults did not distinguish among adolescent-relevant (M = 3.2), adult-relevant (M = 2.85), or older adult-relevant (M = 2.62) problem types but did perceive them to be more relevant than formal problems (M = 1.94).

Reasoning Performance

The results of the analyses examining problem relevancy indicated that all age groups perceived specific problems types to be more relevant to their lives than did others. Following the theoretical formulations advanced by Blanchard-Fields (1986), the second step in the analysis procedure was to examine the effect of particular problem types on the reasoning performance of the three age groups. Scores (1 to 5) assigned by raters on each of the 16 problems were averaged and computed for each problem type. The resulting four average scores (one for each problem type) served as dependent variables in the following analyses.

Because the previous analysis of problem relevancy revealed that the two groups of adolescents in this study (early/middle and older adolescents) attributed greater relevancy to adolescent-relevant problems, the results of the analysis for this particular type of problem is reported first. A one-way analysis of variance examining differences among the three age groups on adolescent-relevant problems found a significant main effect of age group, F(2, 102) = 12.28, p [is less than] .001. Early/middle adolescents exhibited significantly poorer performance on these problems (M = 1.92), relative to the other two age groups who did not differ significantly from one another (older adolescent, M = 2.40; young adult, M = 2.89).

An examination of reasoning scores on adult-relevant, F(2, 102) = 16.15, p [is less than] .001, and formal problems, F(2, 102) = 13.49, p [is less than] .001, revealed a significant pattern of results, similar to that reported for adult-relevant problems. Subjects in the early/middle adolescent age group had significantly poorer performance on adult-relevant (M = 1.84) and formal problems (M = 1.52) than did either older adolescents (Ms = 2.44 and 2.00; adult-relevant and formal problems, respectively) or young adults Ms = 2.92 and 2.34; adult-relevant and formal problems, respectively). For the final problem type (i.e., older adult-relevant) a significant main effect of age group was found, F(2, 102) = 11.45, p [is less than] .001, with early/middle adolescents (M = 2.14) and older adolescents (M = 2.60) having poorer performance than young adults (M = 3.15).

DISCUSSION

The two hypotheses examined in this study concerned (a) the perception of problem relevancy by subjects in the three age groups and (b) whether differences in perceived problem relevancy would have more disruptive effects upon adolescents' reasoning than upon young adults. Adolescent subjects (both early/middle and older age groups) were found to attribute greater relevancy to problems tailored to reflect issues affecting their age group. More significantly, perceived relevancy was found to exert a more disruptive effect upon the reasoning ability of the youngest group of adolescents (i.e., the early/middle adolescent age group), relative to the other two age groups. On problems perceived by older adolescents and young adults to be relevant to their own lives (adolescent-relevant, adult-relevant, and older adult-relevant), reasoning performance was always higher for these two age groups than for the early/middle adolescent age group. In fact, the early/middle adolescent age group had significantly lower scores on all four problem types, relative to the other two age groups. These results are similar to those found by Blanchard-Fields (1986) in that adolescents in her study (aged 14 to 16.5 years) also performed at a lower level on the three tasks she included in her study. However, Blanchard-Fields reported that adolescents' and young adults' reasoning responses did not differ when performance on the Livia task (lowest in emotional saliency) was examined. In the two other tasks (i.e., pregnancy and visit to grandparents) that were higher in emotional saliency, adolescents' reasoning responses were found to be at a lower level than those of young or middle-aged adults. In the present study, by contrast, early/middle adolescents had lower level reasoning regardless of the problem type.

In order to more fully explore the consistently lower reasoning scores found for the youngest age group, a post-hoc analysis of differences in reasoning level associated with each problem type was conducted for this group. The analysis indicated that only formal and older adult-relevant reasoning differed significantly, F(3, 144) = 3.15, p [is less than] .05, with reasoning responses on older adult-relevant problems (M = 2.14) being higher than that on formal problems (M = 1.52). While reasoning responses on these two problem types were found to differ, relevancy ratings of these two problem types by early/middle adolescents did not differ significantly. Relevancy ratings of adolescent-relevant problems by early/middle adolescents were significantly higher than any of the other problem types but reasoning responses on adolescent-relevant problems (M = 1.92) did not significantly differ from older adult-relevant or formal problems. This finding indicates that when differences in reasoning among problem types are examined for a group of adolescents (early/middle adolescent age group) comparable to that of Blanchard-Fields (1986), a pattern similar to that found by Blanchard-Fields can be observed. Reasoning scores were highest on problems perceived to be less relevant (older adult-relevant), relative to problems having the highest perceived relevancy (adolescent-relevant). Thus, perceived relevancy appeared to disrupt reasoning performance of the youngest age group.

A similar post-hoc analysis was performed for subjects in each of the other two age groups to determine whether problem relevancy exerted disruptive effects similar to those found for the youngest age group. Although older adolescents also perceived adolescent-relevant problems to be the most relevant, relative to other problem types, the reasoning responses of the individuals in this age group were not found to differ significantly across the four problem types, F(3, 140) = 1.43, p [is greater than] .05. An effect of problem relevancy was noted only when reasoning performance was examined for young adults on the four problem types, F(3, 124) = 5.07, p [is less than] .05. Young adults' reasoning performance was lowest on formal problems (M = 2.34) and differed significantly from reasoning on the other three problem types (adolescent-relevant, M = 2.89; adult-relevant, M = 2.92; older adult-relevant, M = 3.15). In contrast to the findings obtained for the other two age groups (i.e., the early/middle adolescents and the older adolescents), problem relevancy can be seen to have a facilitative effect on the reasoning performance of young adults by noting that these subjects perceived formal problems to be the least relevant, with the other problem types all perceived to be of greater relevancy.

In summary, this investigation revealed that persons of different ages do exhibit different types of cognitive reasoning in response to problems that vary in their relevancy. These age-specific perceptions of relevancy were found to facilitate performance only among the oldest group of subjects. By comparison, the present findings indicate that problem relevancy exerted a disruptive effect on the reasoning performance of early and middle adolescents. A relatively neutral effect of problem relevancy was found when the performance of older adolescents was examined. Thus, a sequential progression in the ability to incorporate aspects of a problem context becomes evident when the performance of the three age groups are compared. Relevancy moves from having a disruptive to a neutral to a facilitative effect upon reasoning. These findings are in accord with those of Blanchard-Fields (1986) and provide support for the theoretical positions advanced by Labeuvie-Vief (1982). Adolescents, especially those in early to mid-adolescence, appear to have difficulty integrating relevant (or emotionally salient) information with cognitive or logical task demands.

While these age-specific responses are quite apparent, the mechanism that explains these findings is less apparent. Whether adolescents have difficulty integrating emotional and logical demands when they tackle problems in a real-world context remains to be determined. Such real-world contexts obviously present adolescents with emotionally charged situations. The disruptive effect of such emotions on an adolescent's attempt to understand and adapt to such experiences as parental divorce, stress associated with increasing intimacy among heterosexual partners, sexual exploration and its attendant complications can easily produce reactions that have life-long implications.

Clearly there is a need for effective counseling methods that will enable adolescents to conceptualize emotional issues in a more mature way. It would seem reasonable that such methods would emphasize that emotions are generated and derived, in part, from within the individual (Kelly, 1955) and that the capacity to fully understand one's own emotional perspectives may require reorganization of existing cognitive structures in order to make more accurate predictions about the outcomes of emotional situations. Accurate predictions would also seem to depend upon the ability to perceive and incorporate others' perspectives as well. Recent research by Kramer and Melchior (1990) indicates that adolescents' reasoning may be related to the nature of the real-world choices they face. Adolescents who experienced multiple role demands (e.g., females) experienced more role-conflict but also developed more sophisticated modes of reasoning in order to respond to those demands. Other research by Labouvie-Vief, Hakim-Larson, and Hobart (1987) indicates that individuals with more mature reasoning styles employed more effective coping mechanisms in adulthood. Further research is necessary to explicate the experiential and/or maturational factors that may underlie the development of adaptive reasoning during adolescence.

REFERENCES

Adams, C., Labouvie-Vief, G., Hakim-Larson J., DeVoe, M., & Hayden, M. (1989). Modes of thinking and problem solving: Developmental transitions. Manuscript submitted for publication.

Arlin, P. (1984). Adolescent and adult thought: A structural interpretation. In M. L. Commons, F. A. Richards, & C. Armon (Eds.) Beyond formal operations. New York: Praeger.

Baltes, P. B., Dittmann-Kohli, F., & Dixon, R. A. (1985). New perspectives on the development of intelligence in adulthood: Toward a dual-process conception and a model of selective optimization with compensation. In P. B. Baltes, & O. G. Brim, Jr. (Eds.), Life-span development and behavior (Vol. 6, pp. 33-76). New York: Academic Press.

Basseches, M. (1984). Dialectical thinking and adult development. Norwood, NJ: Ablex Publishing Corporation.

Blanchard-Fields, F. (1986). Reasoning on social dilemmas varying in emotional saliency: An adult developmental perspective. Psychology and Aging 1, 325-333.

Ericsson, K. A., & Simon, H. A. (1980). Verbal reports as data. Psychological Review, 87, 215-251.

Giambra, L. M., & Arenberg, D. (1980). Problem solving, concept learning, and aging. In L. Poon (Ed.), Aging in the 1980s. Washington, DC: Gerontological Society.

Kelley, G. A. (1955). The psychology of personal constructs. New York: W. W. Norton.

Kramer, D. A. (1983). Postformal operations? A need for further conceptualization. Human Development, 26, 91-105.

Kramer, D. A. (1989a). A developmental framework for understanding conflict resolution processes. In J. D. Sinnott (Ed.), Everyday problem solving. New York: Praeger.

Kramer, D. A. (1989b). Change and stability in marital interaction patterns: A developmental model. In D. A. Kramer, & M. J. Bopp (Eds.), Transformation in clinical and developmental psychology. New York: Springer-Verlag.

Kramer, D. A., & Melchior, J. (1990). Gender, role conflict, and the development of relativistic and dialectical thinking. Sex Roles, 23, 553-575.

Kramer, D. A., Kahlbaugh, P. E., & Goldston, R. B. (1992). A measure of paradigm beliefs about the social world. Journal of Gerontology, 47, 180-189.

Labouvie-Vief, G. (1982). Dynamic development and mature autonomy. Human Development, 25, 161-191.

Labouvie-Vief, G., Hakim-Larson, J., & Hobart, C. (1987). Age, ego level, and the life-span development of coping and defense processes. Psychology and Aging, 2, 286-293.

Perry, W. G. (1970). Forms of intellectual and ethical development in the college years: A scheme. New York: Rinehart & Winston.

Piaget, J. (1980). Experiments in contradiction. Chicago: University of Chicago Press.

Sebby, R. A., & Papini, D. R. (1991). Perceived problem relevancy and its relationshp to reasoning on everyday problems. In J. D. Sinnott, & J. C. Cavanaugh (Eds.), Bridging paradigms: Positive development in adulthood and cognitive aging. New York: Praeger.

Sinnott, J. D. (1989). A model for solution of ill-structured problems: Implications for everyday and abstract problem solving. In J. D. Sinnott (Ed.), Everyday problem solving: Theory and applications. New York: Praeger. Dennis R. Papini, Ph.D., Western Illinois University