Chapter 2

Models and Theories

Models reflect basic assumptions about the nature of reality, and although too general to test, are useful because they generate theories. Theories can be verified or disconfirmed. Theories that derive from the same model bear a family resemblance reflecting the assumptions characteristic of the model that generated them.
Key Terms: model, theory

The Environmental Model: The environmental model traces behavior to external, environmental antecedents.
Key Terms: environmental model

The Organismic Model: The organismic model focuses on processes within the individual, assuming that individuals are active rather than passive.
Key Terms: organismic model, habituation

Environmental Theories

B. F. Skinner: Skinner assumed that most behavior is learned, coming under the control of the events that follow it (reinforcers), rather than being controlled by preceding motives or intentions.
Key Terms: reinforcement

Albert Bandura: Bandura’s social-cognitive theory emphasizes the importance of inner processes such as attention and memory for learning. Bandura assumes that most human learning occurs through observing others rather than through direct conditioning.
Key Terms: reciprocal determinism

Organismic Theories

Jean Piaget: Piaget viewed intelligence as biologically based. He assumed that knowledge, rather than being a simple copy of reality, is an active construction of what we know of the world.

Robert Kegan: Kegan assumes that the most central human activity is “meaning making,” or constructing a reality that corresponds to our sense of self in relation to events and other people. Development is the cumulative process of differentiating our sense of “me” from “not me.”

Sigmund Freud: Freud formulated his theory of personality development around sexually-based tensions that develop when children are attracted to the opposite-sex parent.
Key Terms: Oedipus complex, Electra complex, castration anxiety, superego

Erik Erikson: Erikson assumed that society challenges us with new demands as we age and that we experience these as psychosocial crises. Each crisis takes a slightly different form and gives each stage its unique characteristics. Achievement of a personal identity is the central crisis of adolescence.
Key Terms: identity

Nancy Chodorow: Chodorow attributes gender differences to the social fact that for almost all children the first intimate relationship is with a female—their mother. Girls can continue to define themselves within the context of this relationship, but boys must separate themselves in order to develop as males. As a consequence, girls’ development is characterized by attachment, and boys’ development by separation and individuation.

Carol Gilligan: Gilligan notes differences in the ways males and females define themselves, and finds these differences extend to the ways they resolve issues involving others. Males tend to see themselves as separate from others, whereas females are more likely to describe themselves in terms of their relationships with others.

A Scientific Approach to the Study of Adolescence

Research Strategies: Research strategies are the procedures scientists follow to safeguard against making faulty observations. Each strategy carries its own advantages and problems. Archival research, for instance, uses data that already exist, such as census data and public records. Its advantages include ease of access to data, large data sets (e.g., the U.S. Census), and use of unobtrusive measures. Disadvantages over time include loss of data and changes in quality of record keeping.
Key Terms: scientific method, archival research, naturalistic observation, quasi-experimental designs, confounding, maturation, testing effect, history effect, statistical regression, experiment, independent variable, dependent variable, random assignment, correlational research, classification variable, age changes, cohort group, cohort differences, time of measurement differences, cross-sectional design, longitudinal design, sequential design

Research Issues: Research quality can be evaluated by the answers it provides. Are answers clear and unambiguous? Or are they open to interpretation? A variety of factors affect research validity. For instance, internal validity is provided by research that gives unambiguous answers to the questions it was designed to address. External validity exists when the findings of a study can be generalized to other populations and contexts. A test of significance is a statistical procedure for determining whether group differences are due to random error or can be attributed to the variable being studied.
Key Terms: internal validity, confounding, external validity, population, sample, bias, double-blind controls, test of significance, degrees of freedom

Archival research uses data that already exist, such as census data and public records. Advantages to this type of research are ease of access to data, large data sets (e.g., U.S. Census), and use of unobtrusive measures. Disadvantages over time include loss of data and changes in quality of record keeping.
Key Terms: archival research

Naturalistic observation involves observation of behavior as it occurs in a natural setting. This research gives richly detailed descriptions of behavior, but does not allow researchers to rule out competing explanations for why behaviors occur.
Key Terms: naturalistic observation

Quasi-experimental designs work with existing, or intact, groups, introducing a treatment and observing any resulting changes. A disadvantage is that one can’t be sure the differences observed actually reflect the treatment. Potential confounds include maturation, testing, history effects, and statistical regression.
Key Terms: quasi-experimental research, confounding, maturation, testing effect, history effect, statistical regression

Experiments offer the greatest control over possible confounds. In an experiment, the experimenter randomly assigns participants to groups, which are then exposed to different treatments. Random assignment ensures the groups are initially equivalent and allows the investigator to attribute observed differences to the way the groups are treated.
Key Terms: experiment, independent variable, dependent variable, random assignment

Correlational research works with variables in which individuals are classified according to existing characteristics—such as age, sex, and ethnicity. Age changes, the biological and experiential changes that accompany aging, can be confounded with cohort differences, differences among people born at different periods of time, and with time of measurement differences, which are due to conditions prevailing when measurements are taken.
Key Terms: correlational research, classification variable, age changes, cohort group, cohort differences, time of measurement differences

In cross-sectional designs, several age cohorts are tested at a single time of measurement. This method takes less time to complete but may miss developmental patterns and potentially confounds age changes with cohort differences.
Key Terms: cross-sectional design

In longitudinal designs, a single cohort group is followed over time and tested at several times of measurement.
Key Terms: longitudinal design, subject mortality

Sequential designs, in which several cohorts are tested at several times of measurement, allow investigators to estimate time of measurement and cohort effects and to isolate these from age changes.
Key Terms: sequential design

Research Issues: Internal validity is provided by research that gives unambiguous answers to the questions it was designed to address. External validity exists when the findings of a study can be generalized to other populations and contexts.
Key Terms: internal validity, external validity

A population is the entire group of adolescents in which one is interested. A sample is a subgroup drawn from this population. If a sample is drawn at random from a population, to the extent it is large enough, one can be confident that it is representative of that population.
Key Terms: population, sample, error, bias, double-blind controls

A test of significance is a statistical procedure for determining whether group differences are due to random error or can be attributed to the variable being studied.
Key Terms: test of significance, degrees of freedom

Go