Subjects and lessons to explore.

Law

Developmental Psychology: Lifespan Development · Development

Research Methods and Ethics

6 min read
12 key termsReport an error
Want it in plain words first? Jump to Eli explains — the same idea, no jargon.
On this page 7 sections
  1. In 30 seconds
  2. Why this matters
  3. The college version
  4. Eli explains
  5. Worked example
  6. Key takeaway
  7. Study tools

In 30 seconds

Developmental researchers study change using descriptive methods (observation, surveys/interviews, ), correlational methods, and — each answering a different question. Because true experiments are often impossible or unethical with children, researchers rely on longitudinal, cross-sectional, and sequential designs to track change over time. Every study is governed by , for minors, , and . The field's most important distinction is : relatedness of two variables never, by itself, proves one caused the other.

Why this matters

Clinicians and educators constantly interpret developmental research, so method quality is a patient-safety skill. When a study claims "breastfeeding raises IQ," a nurse should recognize a correlational finding (confounded by income and parental education) before citing it to families. When describing a child's behavior, use descriptive and correlational language ("is associated with," "on average") rather than causal overclaims — and never use group-level research findings to diagnose or label an individual child. Obtaining a child's assent before a procedure, in language the child understands, mirrors research assent and respects the child's autonomy, while guardians provide consent.

The college version

1. Descriptive Methods

  • — systematically watching and recording behavior in natural or lab settings. Strength: captures real behavior. Limits: observer bias, reactivity, no causal inference.
  • — asking people to self-report attitudes, beliefs, or behaviors. Strength: efficient for large samples. Limits: self-report bias, memory error, social-desirability responding.
  • Case studies — in-depth study of one person or a small group. Strength: rich detail, hypothesis generation. Limits: may not generalize.

2. Correlation vs Causation and Experiments

  • Correlation measures whether two variables change together but cannot show that one causes the other — a third variable or reverse causation may explain the link.
  • Experiments manipulate an independent variable and randomly assign participants to conditions, allowing causal inference (treatment vs control).

3. Developmental Designs

  • — follows the same individuals over time. Strength: reveals true individual change. Limits: slow, costly, subject to attrition and practice effects.
  • — compares different age groups at one point in time. Strength: fast and inexpensive. Limits: cannot reveal individual change; vulnerable to cohort effects.
  • — follows multiple cohorts over time, separating age from cohort effects. Strength: most rigorous. Limits: complex and costly.

How it works

  1. Choose the right method: description (observation/survey/case study), association (correlation), or cause (experiment).
  2. For questions about change, select a design — longitudinal, cross-sectional, or sequential — matching it to the confounds you can tolerate.
  3. Obtain informed consent from adults (or guardians of minors), explaining purpose, procedures, risks, and the right to withdraw.
  4. Obtain assent from children in age-appropriate language, even when a guardian consents.
  5. Protect confidentiality by de-identifying data and storing it securely.
  6. Add safeguards for vulnerable populations, ensuring no coercion and minimal harm.

Common confusions

Do not confuseWithDifference
CorrelationCausationAssociation vs proof one variable changes the other
Longitudinal designCross-sectional designSame people over time vs different ages at one time
Cohort effectAge effectBirth-era differences vs changes from growing older
AssentConsentA child's agreement vs an adult/guardian's permission
Observational researchExperimentWatching without manipulating vs manipulating with random assignment

Memory aids

Designs — "Long, Cross, and Seq": Longitudinal = same people over time; Cross-sectional = a snapshot cut across ages; Sequential = several cohorts over time. Ethics — "C-C-A-C": Consent, Confidentiality, Assent, protection of Children (and other vulnerable populations). The big rule: "Correlation is not causation."

Quick review

Topic Recap

Developmental research uses descriptive methods (observation, surveys/interviews, case studies) for description, correlational methods for association, and experiments for causation. Because experiments are often impossible or unethical with children, longitudinal, cross-sectional, and sequential designs track change — each with distinct strengths and confounds. Ethical practice requires informed consent, assent from minors, confidentiality, and protection of vulnerable populations. Above all, correlation is not causation.

Knowledge Check

  1. A researcher finds children who own more books have larger vocabularies. Can we conclude owning books causes a larger vocabulary? Why or why not?
  2. What is the key difference between a longitudinal and a cross-sectional design?
  3. Why might a cross-sectional study overstate age-related cognitive decline?
  4. What is the difference between informed consent and assent?
  5. Which single design feature allows an experiment to support a causal claim?

Answers and Rationales

  1. No — it is correlational. Families with more books may also read more or have higher income. Third variables and reverse causation are possible; only random assignment could test cause.
  2. Longitudinal follows the same individuals over time; cross-sectional compares different age groups at one point in time. Longitudinal reveals individual change; cross-sectional is faster but confounds age with cohort.
  3. Because of cohort effects. Older participants differ from younger ones in education, nutrition, and life experience — not just age — so the comparison is not a pure measure of aging.
  4. Consent is legal, informed permission from an adult or guardian; assent is the minor's own agreement, given in age-appropriate language.
  5. Random assignment to conditions. It distributes pre-existing differences across groups, isolating the effect of the manipulated variable.
Eli, the EliExplains learning guide

Eli explains

The same idea, in plain words

Explain it like I’m 10

Imagine you want to know whether screen time makes toddlers fussy. You could watch families and take notes (observational research), ask parents to fill out questionnaires (surveys/interviews), or study one child closely for months (a case study). Each gives a description but not proof of cause.

Now suppose children who use screens more are also fussier. That is a correlation — two things go together, like how people who carry umbrellas are more likely to get rained on. The umbrella did not cause the rain; both trace to a third thing (stormy weather). "Where it stops being exact": the umbrella metaphor captures the idea of a third variable but is not a rigorous test — proving cause requires a controlled experiment, which may be impossible or unethical with children.

Simple Example

A researcher randomly assigns one group of infants to a structured reading program and another to a control group. Because assignment is random, any later vocabulary difference is more plausibly caused by the program. Random assignment turns a correlation into a test of causation.

Worked example

  1. Suppose a cross-sectional study finds 60-year-olds score lower on a memory test than 20-year-olds. A naive reading is "memory declines with age," but this confuses cohort with age — the groups differ in education, nutrition, and life experience, not just age.
  2. A longitudinal study of the same people tracks genuine change but risks attrition and practice effects.
  3. A sequential design tests several cohorts repeatedly, teasing apart aging, cohort, and time-of-measurement effects — the most rigorous but most expensive approach.
  4. Throughout, correlation is not causation: even a well-measured association is not proof of cause unless a true experiment with random assignment is run — and many developmental questions cannot be studied experimentally for practical or ethical reasons.

Key takeaways

  • High yield: Correlation does not imply causation; only experiments with random assignment establish cause.
  • High yield: Cross-sectional studies confound age with cohort; longitudinal studies track true change but suffer attrition.
  • High yield: Sequential designs combine multiple cohorts and time points to separate age from cohort effects.
  • High yield: Children provide assent while guardians provide consent; both are ethically required.
  • Observational research, surveys, and case studies describe but cannot prove cause; confidentiality and protection of vulnerable populations are non-negotiable safeguards.

Keep learning

Ready to build on this? Continue to the next lesson.

Practice Developmental Psychology: Lifespan Development

This lesson has no separate scored set. Practice draws from the subject’s question bank.

Study tools & related lessonsYou’ll learn to · Key vocabulary · Related

You’ll learn to

  • Describe the descriptive methods — observational research, surveys and interviews, and case studies — and their strengths and limits.
  • Distinguish correlation from causation and explain why correlational findings cannot establish cause.
  • Compare longitudinal, cross-sectional, and sequential designs, including each design's advantages and confounds.
  • Identify the core ethical principles — informed consent, assent, confidentiality, and protection of vulnerable populations.

Key vocabulary

Observational research
Watching and recording behavior systematically
Surveys and interviews
Self-report questionnaires or conversations
Case studies
In-depth study of one person or small group
Correlation vs causation
Association vs one variable causing the other
Experiments
Manipulate a variable with random assignment
Longitudinal design
Same individuals studied over time
Cross-sectional design
Different ages compared at one time
Sequential design
Multiple cohorts followed over time
Informed consent
Voluntary, informed agreement to participate
Assent
A minor's agreement, with guardian consent
Confidentiality
Keeping participants' data private
Protection of vulnerable populations
Extra safeguards for those who cannot fully consent

Educational content only. It is not medical, legal or professional advice. Found an error? Tell us.