The Concept of Self

The self can be defined as "the totality of a complex and dynamic system of learned beliefs that an individual holds to be true about his or her personal existence and that gives consistency to his or her personality" (Purkey and Schmidt 1987, 32). At first glance, the concept of self may appear ill suited for a central place in the study of social problems, for, as Rosenberg writes, "in its essence, nothing is more quintessentially psychological; an unequivocally subjective phenomenon, its home is located in the inner world of thought and experience" (Rosenberg 1981, 593).

But the relevance of self-concept to the study of social problems is not difficult to grasp. The self develops out of the individual's social experiences and interactions in the different social contexts of the life

course—the family, the school, the workplace, and the community. Although the individual's perception of self is experienced internally, that perception is a product of social interaction.

The concept of self is by no means a simple one. One must attempt to sort out those features of the self that are stable and enduring and those that are "situated" (i.e., those parts of the self that are adjusted to social interaction). As James noted, on the one hand, "there is a certain average tone of self-feeling which each one of us carries about with him, and which is independent of the objective reasons we may have for satisfaction or discontent." But, on the other hand, a person also "has as many different social selves as there are distinct groups of persons about whose opinion he cares. He generally shows a different side of himself to each of these different groups" (James 1890, 294; emphasis in original). Thus, we may be restrained with our parents and teachers, and silly and adolescent with our friends. We do not reveal ourselves to clients and employers in the same way as we do to our colleagues and intimates.

There is an enormous amount of literature and a good deal of empirical research on the ways in which self is affected by social context. Much research has focused on the relation between social esteem and self-esteem. For example, it is commonly believed that minority children develop low self-esteem because they compare themselves unfavorably with the white majority in terms of social class and family structure (traditional nuclear families versus single-parent families). Research offers little support for this conclusion, however. Wylie's review of the research finds that minority group members do not have significantly lower self-esteem (1979, 57–116). Rosenberg's research suggests that all children and adolescents tend to compare themselves with those in their immediate interpersonal environments; when minority children compare themselves with their minority peers, there is a normal distribution of self-esteem. It is only when social circumstances place minority children in situations in which whites are the majority that their self-esteem may suffer (Rosenberg 1981, 605). This research should remind us that situations must be viewed from the perspectives of those who experience them and not just from the perspective of an observer. Findings of this sort are extremely useful in assessing such questions as the effects of school busing, the causes of poor academic achievement, and other problems related to race and ethnicity.

The concept of self has been a central theme of social psychology for the past thirty years. As Stryker (1977) points out, there are two major

traditions in social psychology—the psychological and the sociological . The psychological tradition (Rogers 1965; Epstein 1973; Bandura 1977) focuses on the consequences of the self-concept for individual functioning, that is, how feelings about one's self affect one's behavior and social interactions. In this tradition, Epstein defines the self-concept from an attribution perspective, as a self-theory that a person constructs as an experiencing, functioning individual in interaction with the world. He also points out that the self-theory is a mechanism that serves to "optimize the pleasure/pain balance of the individual over the course of a lifetime" (1973, 407). Thus, one of the important functions of self-theory is to help maintain self-esteem and to organize experience in a manner that enables one to cope with it effectively.

The sociological tradition in social psychology (Rosenberg 1979; Burke 1980) emphasizes how social-structural and contextual factors influence individuals' perceptions of self. The sociological perspective seems to see the structuring of self as developing throughout life, whereas the psychological tradition tends to see the self becoming structured in the earlier years of life. Rosenberg's work is primarily sociological, concentrating on the development of self-evaluative behavior in terms of how social milieu affects behavior (Rosenberg 1965, 1979).

The perspective from which the concept of self is seen obviously influences the ways in which we attempt to deal with social problems. Using the psychological perspective, we will tend to intervene at an individual level; that is, we will favor interventions that alter a person's perception of self in order to change his or her behavior and social interactions. Using the sociological perspective, we will be more likely to intervene at the social level; that is, we will favor interventions that alter social arrangements in order to change the individual's perception of self and his or her behavior and social interactions.

Dimensions of Self-Esteem

Most research on the self-concept focuses on self-esteem, and the self-concept is frequently equated with self-esteem. But self-esteem appears to occupy only a small part of our thoughts about ourselves. McGuire and Padawer-Singer (1976) report the interesting finding that when people are allowed a high degree of freedom in describing themselves, fewer than 10 percent of their responses deal with self-evaluation. Among sixth-graders who were asked, "Tell us about yourself," almost a quarter of all responses were devoted to habitual activities such as rec-

reation and daily routines. The second most frequently mentioned category involved significant others—mostly parents, siblings, and friends. The other categories mentioned were one's attitudes (17 percent), school (15 percent), and demographic characteristics (12 percent). Selfevaluation accounted for only 7 percent of all responses.

A distinction can be made between self-conception (identity), and self-evaluation (self-esteem). Self-conception usually refers to the concept that individuals hold of themselves as physical, social, and spiritual or moral beings. In Rosenberg's terms, the self-conception is broadly defined as the "totality of an individual's thoughts and feelings having reference to himself as an object" (1979, 7). Self-esteem, however, has been referred to as an individual's overall self-evaluation, the self as "an object of knowledge." James (1890) viewed self-esteem as the ratio of our actualities to our supposed potentialities, that is, the ratio of success to pretensions.

According to Gecas, "identity focuses on the meanings comprising the self as an object, gives structure and content to the self-concept, and anchors the self to social systems. Self-evaluation or self-esteem refers to the evaluative and affective aspects of the self-concept" (1982, 4). These aspects of the self-concept are closely interrelated: self-evaluation is typically based on substantive aspects of the self-concept, and identities typically have evaluative components.

Increasingly, aspects of self-esteem have been differentiated: inner and outer self-esteem (Franks and Marolla 1976), self-evaluation and self-worth (Brissett 1972), and sense of power and sense of worth (Gecas 1971), for example. Franks and Marolla use a two-dimensional approach, distinguishing inner self-esteem and outer self-esteem. The sense of inner self-esteem derives from feelings of one's own efficacy and competence—that is, the effects that one's actions have on the environment account for inner self-esteem. In contrast, outer self-esteem is bestowed by others; it deals with approval or acceptance by significant persons. Outer self-esteem refers to our desire to be connected affectively with others. One's sense of power and potency (inner self-esteem) can be quite independent of a sense of being accepted and liked by others (outer self-esteem). Social comparisons become very important in this context. It should be noted, however, that both the inner and outer dimensions of self-esteem can be seen as belonging in a sociological frame of reference (Franks and Marolla 1976). For example, minorities may worry about being accepted by white co-workers, but they

may not be concerned about acceptance from whites with whom they have no relationship.

As Franks and Marolla (1976) show, two dimensions of self-esteem often seem to parallel each other: those situations in which one evaluates oneself highly are often those situations in which one experiences a sense of mastery. The distinction between self-evaluation and self-worth, though conceptually important, tends to blur at the experiential level.

Knapp (1973) raises questions about the stability of self-esteem at different developmental stages in childhood, noting that it may seem to vary dramatically even within short periods of time. Little is known about whether self-esteem can, in fact, be changed and, if it can, the forms of intervention that are effective. The influences of culture, race, and sex-role identification on self-esteem are also relatively unknown.

Martinek and Zaichkowsky (1977), along with Knapp, suggest that the self-esteem of young children may be very unstable until about the age of seven. (Piers [1969] has confirmed this in her extensive investigations with elementary school and secondary school children.) Martinek and Zaichkowsky refer to Jersild's (1969) findings that certain emotional tendencies, feelings for others, and character traits are globally related to the self.

Battle (1981) reports that children's self-esteem has a weak, but statistically significant, positive relationship with intelligence. (Other researchers, such as Coopersmith [1967], have reported similar findings.) Teachers' ratings of students' self-esteem and students' self-reports correlate highly, but, as Piers (1969) and Coopersmith (1967) also found, teachers' ratings and students' school self-esteem (related to academic achievement) are not significantly related. Battle concludes that "one's self-esteem is not dependent on any particular factor (e.g., academic achievement), but on a combination of factors" (1981, 16). Battle also found, as did Coopersmith (1967), that among both adults and youths self-esteem is negatively correlated with depression. Finally, the self-esteem of academically successful students was significantly higher than that of less successful children with learning disabilities.

Sources of Self-Esteem

The difference between a sense of competence based on self-evaluation and a sense of value based on self-worth is important, because each arises

in a different process of self-concept formation. Briefly, the process of reflected appraisal contributes to the formation of self-worth, whereas competency-based self-evaluation is associated with self-attribution and social comparison.

There are four sources for the formation of self-esteem. The first is reflected appraisal, which is grounded in Cooley's concept of the "looking-glass self" (1902) and in Mead's idea of "role-taking" (1934). Cooley's concept of reflected appraisal is the notion that our sense of self is derived largely from our perceptions of how others regard us. In Cooley's view, self-esteem may be more strongly associated with the perceived appraisals of others than with actual appraisals. For example, Rosenberg (1979) found that the association between self-perception and reflected appraisals or social comparisons was stronger if the significant other was highly valued by the subject. Mead (1934) noted that self-image arises in social interaction as an outcome of the individual's concern about how others react to him or her. The "generalized other" serves as a source of internal regulation to guide and stabilize the individual's behavior in response to certain actions.

A second influential source of self-esteem is the social comparison process (Festinger 1954, 117–140), by which individuals assess their own abilities and virtues by comparing themselves to others. Social comparisons are most likely to occur in situations where information and standards are ambiguous or uncertain.

Self-attribution is a third source of self-esteem. It refers to how individuals explain their behavior (e.g., "I did poorly on the test because the teacher doesn't like me"). The notion of self-attribution suggests that self-esteem is tied not so much to an individual's behavior as to his or her interpretation of the behavior. As Gecas (1982) points out, attribution theory is, in general, more appropriate to consideration of self as a causal factor in social interaction than to explanations of the development of self.

Finally, social identity is also closely tied to self-esteem. Social identity refers to socially recognized belonging, such as one's social class and status, race, religion, or organizational affiliations. Low prestige in terms of social stratification does not necessarily produce correspondingly low self-esteem, however. According to Rosenberg's notion of "psychological centrality," the impact of any given stimulus depends on its centrality in the individual's cognitive structure. Thus, individuals must first become aware of their lower status in respect to significant others in the environment, as in the example of minority children in

majority-dominated educational settings cited earlier (Rosenberg 1981; Rosenberg and Pearlin 1978).

Instruments for Measuring Self-Esteem

Instruments that measure self-esteem generally fall into one of four categories, each with certain limitations: behavioral trace reports, direct observations, projective techniques, and self-reports (Knapp 1973). Behavioral trace reports attempt to base judgments on concrete behaviors, such as grades and teachers' comments, thereby eliminating observer bias; problems with memory (e.g., the teachers') and validity (e.g., grades) do exist, however. Direct observations are used with very young children who are not yet able to communicate effectively on a verbal level; but it is possible that the values, feelings, and attitudes of the observer may bias the results. Projective techniques can reveal unconscious processes in children and adults, but scoring procedures are difficult and may be neither objective nor valid. Self-reports are practical and easily scored, but respondents may manipulate their self-reports to obtain desirable results (e.g., to elicit sympathy from the observer). Of course, techniques can be combined in various ways to achieve reliability, but this makes assessment more expensive.

Hughes (1984) reviewed the nineteen most-used instruments for evaluating self-esteem among children aged three to twelve, concluding that although therapists working with children and adults frequently cite changes in self-concept and self-esteem as goals, there is no widely adopted, coherent theory of self-esteem and the self-concept. She believes the Piers-Harris Children's Self-Concept Scale to be the best for clinical application with children aged nine to twelve, because of its high reliability and validity, but she found no measure equally adequate for use with younger children. The McDaniel-Piers Scale was recommended for children aged six to nine because of its reliability and popularity. Hughes also recommends the Behavioral Academic Self-Esteem Scale as the best "teacher report" measure. She notes a distinction made in the literature between self-concept, as "the descriptive perception of the self," and self-esteem, as "the evaluative assessment of those descriptions" (1984, 659).

The Martinek-Zaichkowsky Self-Concept Scale for Children (Martinek and Zaichkowsky 1977) is designed to measure the "global self-concept of children from first grade through eighth grade." Children are given a self-report instrument that utilizes pictures instead of words.

The authors claim that the instrument is "culture-free." Although they have not yet established validity and reliability measures for this scale, as Wylie (1979) indicates in a review of the literature, there is a critical need for a well-validated scale that measures the self-esteem of younger children and that does not require the ability to read or understand English.

The Self-Observation Scale (Katzen and Stenner 1975) also utilizes a self-report instrument in assessing the self-esteem of children at the primary, junior high, and senior high school levels. The authors have completed a fairly extensive validation study of their instrument, and they maintain that it emphasizes the healthy and positive aspects of self rather than the more negative and pathological ones.

Katzen and Stenner use a "practical decision-making orientation" instead of the more traditional orientation of theory and research. According to the authors, the emotional development of children has not received enough attention; the emphasis in research has traditionally been on cognitive development. Their instruments are designed to help educators and psychologists attend more effectively to emotional development in this age group. Their instruments use written questions that are organized into seven scales: self-acceptance, self-security, social confidence, self-assertion, peer affiliation, teacher affiliation, and school affiliation. The forms developed for the primary school level use pictures for student responses, whereas the other forms use a verbal format.

The Culture-Free Self-Esteem Inventory (Battle 1981) provides instruments for use with children from grades one through twelve, as well as with adults. These instruments are designed both to help identify children, youths, and adults who are in need of psychiatric help and to provide general information for the professional helper or researcher. The author claims to have developed inventories that are useful with clients of all cultures and races.

The Coopersmith Self-Esteem Inventory (1975) is designed for subjects aged nine years and older. This instrument is based on a widely known study of self-esteem and has demonstrated a degree of reliability and validity that has made it very popular with researchers. Respondents are asked to check various columns of responses to questions.

The Tennessee Self-Concept Scale (TSCS), developed by Fitts (1965), is one of the most widely utilized self-esteem scales. It is also one of the few well-developed measures of self-esteem for use with adults. Using self-reporting, it measures self-concept across many subareas, providing both an overall self-esteem score and a complex self-concept profile.

The ninety statements (evenly balanced for positivity-negativity) fall into one of five general categories: physical self, moral-ethical self, personal self, family self, and social self. Each category is divided into statements of self-identity, self-acceptance, and behavior. There are also ten items from the Minnesota Multiphasic Personality Inventory (MMPI) lie scale. Each question has five response categories, from completely true to completely false. The TSCS yields an overall self-esteem score, a total positive score, along with self-esteem subscale scores related to different dimensions of perceiving the self.

The Rosenberg Self-Esteem Scale (1965), originally developed for use with high school students, measures the self-acceptance aspect of self-esteem. The scale consists of ten Guttman-type items with four responses, from strongly agree to strongly disagree, which are, however, scored only as agreement or disagreement. The scale is designed with brevity and ease of administration in mind.