Theories and Methods of Development

This is Unit One, part three, on the last learning theories, the systems around a person, and the methods of developmental science. Part two closed on classical conditioning, learning by association. Part three opens with learning from consequences, then turns from the theories to the methods used to test them. Carry one question through this part: what can a given study actually show? A correlation, a case, an observation, a comparison of ages: each shows something, and each has a limit. At the entry on experiments, the part names the one method that can establish cause and effect. By the end you will be able to tell reinforcement from punishment, and describe learning by observation. To name Bronfenbrenner's five systems. To explain why a correlation cannot show cause, and name the method that can. To tell reliability from validity. And to compare the three designs that bring in time.

Three entries: learning from consequences, learning by watching others, and the theory Bandura built from it.

Learning from what follows a behaviour.

HGD 0043 Operant Conditioning

Operant conditioning, developed by B. F. Skinner from Thorndike's law of effect, holds that consequences change how often a behaviour occurs. Reinforcement makes a behaviour more frequent; punishment makes it less frequent. Shaping reinforces successive approximations of the goal, as when praise comes only as a toddler's attempt at the word apple gets closer to it.

Learning by watching others.

HGD 0044 Observational Learning

Observational learning, described by Albert Bandura, is learning by watching others and the consequences they meet, rather than by one's own reinforcement. Children who watched an adult hit a Bobo doll hit it themselves, and they also invented new ways to attack it. Models can be live, or symbolic, as on television.

The theory Bandura built from it.

HGD 0045 Social Cognitive Theory

Bandura's social learning theory, later called social cognitive theory, holds that people and their environments influence each other, a process he named reciprocal determinism. Parents shape a child's world while the child reshapes how the parents behave. People are not only products of their surroundings; the way they interpret events shapes how those events influence them.

Three entries on the settings that surround a developing person, from the family to history itself.

A model of nested settings.

HGD 0046 Bronfenbrenner's Ecological Model

Urie Bronfenbrenner's ecological systems model places the person at the centre of nested systems, circles within circles, whose influence becomes less direct moving outward: family and school closest, culture and history farthest out. Bronfenbrenner had co-founded Head Start, an early-education programme, the decade before. Later revisions made the person an active shaper of each system.

The five systems, one by one.

HGD 0047 The Five Systems

Bronfenbrenner's model names five systems. The microsystem holds the settings a person directly experiences: family, school, peers, health care. The mesosystem is the interaction between microsystems, such as a parent–teacher conference. The exosystem affects the person indirectly, through a parent's workplace policy or the local school board. The macrosystem is the culture's beliefs, laws and economic system. The chronosystem is all of it moving through historical time. Influence runs both ways, and the person is active within every system. On the exam, a parent's job affecting a child points to the exosystem.

The system that brings in history.

HGD 0048 Cohorts, Not Generations

A cohort effect is the influence of a shared experience on people of about the same age at the same historical time. Gen X, born about nineteen sixty-five to nineteen eighty, are digital immigrants beside later cohorts of digital natives. Cohort is more accurate than generation, which invites age stereotyping. A cohort is a shared history, not a shared personality.

Two entries on who research describes: the categories it uses, and the people it has mostly studied.

Three categories researchers must define carefully.

HGD 0049 Race, Ethnicity and Culture

Race is socially constructed, with no genetic basis; ethnicity is shared cultural or national heritage; culture is meaning shared and passed between generations. The misconception is that race is a biological category. Researchers now say ethnoracial when their findings centre on ethnicity yet acknowledge that race is constructed.

Who psychology has mostly studied.

HGD 0050 WEIRD Samples

Much psychology rests on WEIRD samples, people from Western, educated, industrialised, rich and democratic societies, and that limits how far findings generalise: one country's university students do not stand for the whole world. A representative sample reflects the population a conclusion is about. Babies across cultures, though, do share a preference for infant-directed speech.

Four entries on how developmental science builds knowledge, and on two common errors in reading a finding.

Explanations, and the predictions drawn from them.

HGD 0051 Theories and Hypotheses

A theory is a well-developed explanation; a hypothesis is a testable prediction drawn from it, often in if-then form. If attachment shapes later relationships, then secure infants should form closer friendships later. Research supports, modifies or rejects theories; it never proves them. Science cycles between deductive testing and inductive observation.

Two tests every measure must pass.

HGD 0052 Reliability and Validity

Reliability is the consistency of a measure; validity is whether it measures what it is intended to measure, and a measure can have one without the other. Picture a kitchen scale that reads two hundred grams heavy, every single time. It is reliable and invalid. The misconception is that consistent measurement is accurate measurement.

The most common error of inference.

HGD 0053 Correlation and Causation

A correlation shows that two variables move together, but it cannot show that one causes the other. A third variable can drive both: people who eat breakfast cereal may differ in other habits that explain their heart health. People also see correlations that are not there. The belief that a full moon changes behaviour is an illusory correlation; a review of nearly forty studies found no relationship, yet confirmation bias keeps noticing the cases that fit. Illusory correlations do real harm, helping to sustain prejudice about groups of people. Only an experiment can show cause and effect.

Why findings must be repeated.

HGD 0054 Replication and the Mozart Effect

Replication means repeating a study to see whether its finding holds, and findings that fail to replicate lose support. The Mozart effect, the claim that listening to Mozart boosts reasoning, found little to no support in later studies. The vaccine–autism claim rested on falsified findings, since retracted.

Four entries on the descriptive methods, and on the number that says how strongly two things are related.

One person, studied in depth.

HGD 0055 Case Studies

A case study examines one person or a few people in great detail, often revealing insights other methods cannot reach, such as a neurologist's account of one patient who could no longer recognise objects by sight. The limit is generalisability: what holds for one case may apply only to that person. Case studies lean on inductive reasoning.

Watching behaviour where it happens.

HGD 0056 Naturalistic Observation

Naturalistic observation records behaviour in real-life settings without control or interaction, such as children at play on a school playground. That gives it high ecological validity, meaning realism. Observer bias is the risk; clear coding criteria and inter-rater reliability, agreement between observers, guard against it.

Observation with the same set tasks.

HGD 0057 Structured Observation

Structured observation gives every participant the same set tasks and records the responses, so observations can be compared. Ainsworth's Strange Situation is the example: a stranger enters, the caregiver leaves, and later the caregiver returns. The procedure measures infant attachment, and the moment of reunion with the caregiver is the most telling.

How strongly two variables are related.

HGD 0058 The Correlation Coefficient

The correlation coefficient runs from minus one to plus one: its size gives the strength of a relationship, and its sign gives the direction. Height and weight rise together; more sleep goes with less tiredness. A negative correlation is still a relationship. Zero means none, and point nine is far stronger than point three.

Four entries: the one method that shows cause, and the three designs that bring age and time into a study.

The payoff: the one method that shows cause.

HGD 0059 Experiments

An experiment is the only research method that can establish cause and effect, and it answers the question this part opened with. The researcher manipulates the independent variable, measures the dependent variable, and uses random assignment, giving every participant an equal chance of each condition, so participant differences spread evenly. In one classic study, observers watched the same snowball fight labelled with different gender pairings, and rated the fight between two boys as least aggressive; the label was the independent variable. Experiments can be artificial, limiting external validity, and some questions cannot be manipulated ethically: no one can assign a family to divorce. There, a quasi-experimental design compares groups that arose naturally, giving up some control.

A design that follows the same people.

HGD 0060 Longitudinal Designs

A longitudinal design follows the same people over time, reassessing them to see how each one changes with age, such as fifty sixth-graders reassessed every year for five or six years. It is slow, and it suffers attrition as people drop out, practice effects from repeated testing, and, most seriously, cohort effects.

The faster alternative.

HGD 0061 Cross-Sectional Designs

A cross-sectional design compares different age groups at a single point in time, such as twelve-, fourteen-, sixteen- and eighteen-year-olds tested in the same week. It is fast and free of attrition and practice effects, but it cannot track individuals, and cohort effects remain.

All three time designs, compared.

HGD 0062 Three Designs Compared

Developmental research brings in time in three ways. A longitudinal design follows the same people for years and shows individual change, at the cost of time, attrition and practice effects. A cross-sectional design compares different ages at once, quickly, but cannot follow anyone and mixes age with cohort. A cross-sequential design, also called sequential, follows several age groups for a shorter time, combining both, as when five-, seven- and nine-year-olds were each retested a year later. Choose by the question. On the exam, a study that tests different ages once is cross-sectional, and cohort effects are its classic weakness.

One entry to close the unit: the study that reshaped the ethics of research on people.

The case that reshaped research ethics.

HGD 0063 Tuskegee

From nineteen thirty-two the United States Public Health Service studied six hundred Black men in Tuskegee, never telling those with syphilis. Penicillin was curing syphilis by nineteen forty-seven, yet it was withheld until the study was exposed in nineteen seventy-two. The outcry produced the National Research Act of nineteen seventy-four. Today, review boards and informed consent, with a child's assent, protect participants.

Look at what you can now do. You can tell reinforcement, which makes a behaviour more frequent, from punishment, which makes it less frequent, and describe how Bandura's children learned by watching a model hit a Bobo doll. You can name Bronfenbrenner's five systems, micro, meso, exo, macro and chrono, and say that a parent's job affecting a child points to the exosystem. You can explain that a correlation cannot show cause because a third variable can drive both, and that only an experiment, with random assignment, can. You can tell reliability, consistency, from validity, measuring what is intended, like the scale that is consistently wrong. And you can compare the designs: longitudinal follows the same people, cross-sectional compares ages at once, and cross-sequential combines both. Unit Two turns to heredity, prenatal development and birth.