The Gold Standard of History of Developmental Psychology

Chapter index

A developmental question concerns a pathway through time. Asking whether two children differ in vocabulary describes a difference. Asking how each child’s vocabulary changes, which experiences precede that change, and whether the same relationship appears in other settings begins a developmental investigation. The unit of interest may be an individual, a relationship, a skill or a pattern of adaptation. These units should not be silently exchanged during an explanation.

Three distinctions organise this history. Description identifies what happened. Explanation proposes why it happened. Evaluation judges whether the change is desirable for a specified purpose. A child who answers more quickly has changed on a speed measure; that observation does not establish improved understanding. An adult who slows down while checking a difficult instruction may be performing more effectively. A direction on a scale becomes progress only after the goal and costs have been stated.

Age is a coordinate, not a complete mechanism

Imagine that twelve-year-olds complete more planning tasks than eight-year-olds. Chronological age records elapsed time, but it does not specify which processes produced the difference. The groups might differ in prior instruction, reading demands, familiarity with the materials or opportunities to organise activities. A biological explanation is possible, but it requires evidence beyond the age labels. An environmental explanation requires evidence too. Naming a plausible cause is the beginning of an investigation.

The same care applies to stability. If an individual’s score remains constant, the underlying work may have changed. A harder environment could offset improved skill; additional support could compensate for a new difficulty. Record the demands and resources alongside performance. Otherwise a flat line invites an interpretation that the measurement cannot support.

A useful reading method is to keep four columns in your notes: question, observation, proposed mechanism and unresolved alternative. For an infant diary, these might be the emergence of a response, a dated behaviour, a proposed association and an unmeasured sensory cue. For a longitudinal study, they might be change in reasoning, repeated scores, learning opportunities and selective participant loss. This method lets different historical approaches be compared without pretending that they asked identical questions.

Contents · Next chapter

Charles Darwin’s 1877 A Biographical Sketch of an Infant drew on notes about one of his children made decades earlier. He described dated observations of responses, expression and emerging abilities, and explicitly expected variation in developmental timing between infants. The paper places detailed observation beside interpretation; at several points Darwin acknowledged uncertainty about the cues guiding a response. Its value as a historical document is the visibility of the observer’s questions, not a claim that one child’s chronology establishes a universal sequence.

Original case: recording an event and recording its meaning

Suppose a caregiver writes, “At ten months, Mira deliberately shared a toy.” The visible event might instead be recorded as: Mira extended a toy towards another person, released it when that person touched it, and reached towards it again. Sharing is one possible interpretation. Requesting help, starting a familiar game or simply releasing an object are alternatives. A stronger record preserves the action, the preceding event and what happened afterwards.

Now imagine twenty opportunities during a week. The behaviour appears on six occasions, all with one familiar person. Reporting six successes out of twenty opportunities gives 30%, but the percentage is not a universal probability of sharing. The opportunities were not necessarily equivalent or independently sampled. The record nevertheless improves on recalling only the most memorable episode, because its denominator makes the selection visible.

Repair the observation by defining an opportunity before collecting more notes. Keep the activity ordinary and comfortable; record comparable occasions, including non-occurrences. Note who was present and whether anyone prompted the action. If the behaviour appears with another familiar person in a different game, the interpretation becomes broader. If it disappears without a specific cue, that cue belongs in the explanation.

This exercise concerns everyday observation, not experimentation on a child or developmental screening. Its historical lesson is methodological: careful records can generate questions that a striking anecdote cannot settle. A diary offers depth and continuity. It needs complementary comparisons before its writer can move from “this happened here” to “this is how children develop.” The strengths and limits arise from the same feature: unusually close access to a particular life.

Previous chapter · Contents · Next chapter

Jean Piaget’s 1964 Development and Learning distinguished the development of cognitive structures from learning tied to particular situations. The paper discussed maturation, experience, social transmission and equilibration as factors in development. The Piaget Foundation bibliography documents its conference and journal publication contexts. A productive historical question is what kind of organisation in thinking a task was meant to reveal, rather than which age label a reader can attach to a learner.

Original case: the same score, different reasoning

Consider two learners who answer eight of ten classification questions correctly. One sorts objects by a stated rule and explains exceptions. The other memorises the arrangement used in practice. Their scores are identical, but the explanations predict different responses when the materials change. Give both a new set in which colour and shape no longer coincide. Ask them to explain the sorting rule without supplying it. Performance on this changed task can distinguish some of the competing accounts.

Assume the first learner answers seven of eight new questions correctly and the second answers three. The descriptive transfer rates are 87.5% and 37.5%. With such a small invented set, these figures are not precise estimates of enduring abilities. They show why a single original score was insufficient. The next question is whether the difference survives changes in wording, examples and testing occasion.

An incorrect answer also needs interpretation. A learner may understand the relevant relation but misread an instruction. Another may repeat the correct phrase without understanding the relation. Combine actions, explanations and carefully varied tasks. Do not require a fluent verbal explanation as the sole proof of understanding when expressive language is itself a substantial demand.

Historical theories are most useful when they suggest discriminating observations. They become less useful when used as permissions to stop looking: “not ready” can conceal a missing prerequisite, an inaccessible instruction or an unfamiliar response format. Conversely, successful coaching on one item does not prove that an entire conceptual system has changed. The task is to specify the breadth, independence and persistence of the observed performance. These are questions to investigate, not automatic conclusions from either success or failure.

Previous chapter · Contents · Next chapter

David Wood, Jerome Bruner and Gail Ross’s 1976 The Role of Tutoring in Problem Solving examined an adult helping children aged three, four and five construct a three-dimensional structure. The paper used scaffolding to describe assistance that controls parts of a task beyond the learner’s unaided capacity. Its evidence concerned the interaction between helper, learner and task. It should not be reduced to a general promise that more assistance always produces more development.

Original case: assistance hides inside the success rate

Suppose a learner completes nine of ten planning tasks with an adult present. On four tasks the adult identifies the first step; on three others the adult corrects a missed condition. Only two are completed without either kind of intervention. “Nine out of ten completed” describes supported task completion. “Two out of ten completed independently” describes a different outcome. Both can be accurate, but they answer different questions.

In a second session, the learner completes eight tasks: six independently and two after a general reminder. Completion has fallen from 90% to 80%, while independent completion has risen from 20% to 60%. If independence is the goal, the lower overall completion rate may accompany meaningful improvement. We still need comparable task difficulty, a record of help and another observation before treating the change as dependable.

A practical recording scheme separates the task outcome from the support level. Record no help, a general prompt, a specific hint or direct demonstration. These categories need clear definitions, but they need not be converted into a misleading numerical scale. The distance from a hint to a demonstration is not automatically equivalent to the distance from no help to a prompt.

The repair is to fade assistance where it is no longer needed and preserve it where removing it would prevent participation. Then test a changed task with the agreed level of support. This makes the target explicit: learning to request suitable help can itself be a useful competence. Independence does not mean refusing all tools or other people. It means carrying the responsibilities appropriate to the activity rather than mistaking someone else’s hidden work for one’s own mastery.

For the educational connection, History of Educational Psychology examines how theories of learning became proposals about teaching and assessment.

Previous chapter · Contents · Next chapter

K. Warner Schaie published A General Model for the Study of Developmental Problems in 1965; the Seattle Longitudinal Study bibliography also records his subsequent work on cross-sequential designs and time and cohort differences. These titles mark a central methodological concern: people are measured at an age, in a historical period, after belonging to a particular birth cohort. Research design must confront their relationship.

Original calculation: the age difference is not a growth rate

Imagine testing a group aged twenty and a group aged forty in the same year. Their average scores on an unfamiliar digital navigation task are 80 and 60. The difference is twenty points. Dividing it by twenty years gives one point per year, but calling that an annual rate of decline would be an unsupported inference. No participant has been observed changing across those twenty years.

The older group might have had different exposure to the interface, education or recruitment procedures. These possibilities do not prove that age is irrelevant. They show that the design has not isolated it. A longitudinal study follows the same people, helping to address the within-person question, but then introduces other concerns: repeated testing, historical events and loss of participants.

For an idealised calendar calculation, age equals observation year minus birth year. A person born in 2000 is approximately twenty in 2020 and thirty in 2030, with exact age depending on the date of birth. Because these quantities are linked, unrestricted age, period and cohort effects cannot all be uniquely identified from that arithmetic alone. More rows of data do not remove an exact relationship. An analysis needs additional assumptions, restrictions or substantively justified comparisons.

Following several cohorts over overlapping ages can expose patterns that a single cross-section cannot. It does not make interpretation automatic. Write down which contrast supports each claim. Comparing different cohorts at the same age still compares different historical periods. Comparing the same cohort ten years later still combines ageing with everything else that occurred during those years.

The useful discipline is to label the estimate accurately: an age-group difference, a within-person change, or a model-dependent estimate. Each can be informative. Confusion begins when one is reported as another, especially when a descriptive graph is presented as a causal biography of a person who was never followed.

Previous chapter · Contents · Next chapter

Urie Bronfenbrenner’s 1977 Toward an Experimental Ecology of Human Development proposed studying the changing relation between people and the environments in which they live. The paper included immediate settings and the larger contexts containing them, and argued for combining relevance with rigorous experimental work. It did not offer context as an excuse to abandon causal questions. Its challenge was to make the environment part of the research problem rather than treating the testing room as the whole developmental world.

Original case: a change outside the classroom

A school observes more incomplete assignments after a transport timetable changes. One interpretation is declining student motivation. Another is that a longer journey reduces the time available for homework. A third is that the timetable change coincides with a more difficult unit. An ecological description identifies the connected settings; a causal investigation must still distinguish the explanations.

Suppose the affected group moves from 80% to 65% assignment completion, a fall of fifteen percentage points. An unaffected comparison group moves from 82% to 77%, a fall of five points. The difference between changes is minus ten points. Under an assumption that the groups would otherwise have followed comparable trends, this contrast is consistent with an additional effect associated with the transport change. That assumption is substantial, not a fact supplied by subtraction.

Check whether both groups faced the same assignment changes, whether their previous trends were similar, and whether group membership changed. Ask whether the proposed pathway is observed: did travel time increase, did available work time shrink, and did completion change most where that disruption occurred? A coherent mechanism strengthens interpretation; it does not compensate for an unsuitable comparison group.

This example illustrates why adding the word context to an explanation is insufficient. Specify the setting, the connection, the exposure and the expected effect. A family schedule, peer activity or school rule can be relevant through a concrete pathway. Listing every possible influence produces a diagram without an explanation.

The developmental question is also reciprocal. A learner may reorganise routines, seek help or choose a different activity in response to the changed environment. Record these responses rather than assuming a passive person receiving a fixed treatment. The same external change can lead to different pathways because the available responses and constraints differ.

Previous chapter · Contents · Next chapter

Paul Baltes’s 1987 Theoretical Propositions of Life-Span Developmental Psychology set out a framework emphasising multidirectionality, plasticity, historical context and the interplay of gains and losses. It addressed constancy and change throughout life. These were theoretical propositions and research orientations, not a universal scoring formula. In particular, judging a gain or loss requires specifying the demands against which a person is adapting.

Original calculation: a composite can conceal a trade-off

Imagine a worker who completes an unfamiliar administrative task in eight minutes with two consequential mistakes. After practice and a new checking routine, completion takes ten minutes with no mistakes. If the only outcome is time, performance appears worse. If the cost is defined for this exercise as minutes plus five additional cost units for each mistake, the first result costs 8 + 5 × 2 = 18 units; the second costs 10 + 5 × 0 = 10.

The second performance is better under that rule. It is not better under every conceivable rule. If each mistake costs only half a unit, the first costs nine and the second ten. The assumed value of an error changes the ranking. Real decisions require appropriate units and a reasoned account of consequences, not arbitrary weights chosen after seeing the preferred result.

This is also why a single total score should not replace the underlying profile. Preserve speed, accuracy, effort and the kind of help used. A composite can serve a decision, but it cannot explain how the decision was reached unless its components and weights remain visible.

Plasticity can be investigated by changing resources or practice conditions and observing the response. Improvement under one arrangement establishes something about that arrangement and outcome. It does not establish unlimited change, equal opportunity or identical responses among individuals. Failure under one arrangement is equally limited evidence about what all other arrangements could achieve.

For readers comparing development with stable individual differences, History of Personality Psychology provides a complementary route through traits, situations and life stories. A person can remain relatively high compared with others while changing substantially in absolute performance. Relative position and individual change are separate questions.

Previous chapter · Contents · Next chapter

The ManyBabies Consortium’s 2020 Quantifying Sources of Variability in Infancy Research Using the Infant-Directed-Speech Preference reported a study across 67 laboratories. It found an overall preference for infant-directed over adult-directed speech, with a meta-analytic effect size of 0.35 and a 95% confidence interval from 0.29 to 0.42. The reported magnitude varied with age, language and dialect match, and testing procedure. The historical importance here is methodological: a shared investigation estimated both a phenomenon and some conditions associated with its variation.

What a pooled result does and does not say

A positive average does not mean that every participant exhibits the preference on every trial. A confidence interval for an average effect is not an interval containing 95% of individual infants’ responses. Nor does a preference measure directly establish later language attainment. Each additional claim needs a suitable outcome and design. The published result is narrower, and therefore more informative, than a slogan about what all infants need.

Consider an original pooling example. Laboratory A records an average difference of two seconds in ten participants. Laboratory B records a difference of one second in ninety participants. An unweighted mean of the two laboratory averages is 1.5 seconds. Weighting by participant count gives (10 × 2 + 90 × 1) ÷ 100 = 1.1 seconds. The calculations answer different aggregation questions.

Neither calculation is a substitute for a proper meta-analysis, which must consider uncertainty and dependence among observations. The example simply shows why “the average finding” requires a stated averaging rule. A large laboratory need not be more representative of every population, and equal laboratory weights need not accurately reflect precision. Sample size, precision and representativeness are different properties.

When reading a collaborative study, look for common materials, planned exclusions, age ranges, language exposure, recruitment differences and the relationship between exploratory and planned analyses. A shared procedure helps make comparisons interpretable; it does not erase local differences. Variation can be a clue about a mechanism, a measurement problem or both.

The move from a close diary to coordinated laboratories need not be described as replacing one useful method with another. Rich observation helps identify phenomena and alternative interpretations. Shared studies examine how well specified findings travel. The question determines which combination of depth, comparison and replication is needed.

Previous chapter · Contents · Next chapter

Following the same people sounds like a direct solution to developmental questions. It is a major improvement when the question concerns within-person change, but the people who return can differ from those who do not. A follow-up average therefore needs a description of who contributed to it. Attrition is not merely a smaller sample size; it can change which pathway the remaining data represent.

Original calculation: apparent improvement without individual improvement

Begin with one hundred participants. Fifty score 80 and fifty score 40, giving an average of (50 × 80 + 50 × 40) ÷ 100 = 60. At follow-up, all fifty higher-scoring participants return, but only ten lower-scoring participants return. Assume, for this demonstration, that every individual’s score is completely unchanged.

The follow-up average is (50 × 80 + 10 × 40) ÷ 60 = 73.33. Comparing 73.33 with the original average of 60 suggests a gain of 13.33 points. Yet nobody improved. The apparent change comes entirely from the composition of the observed group. Among the sixty returners, the baseline average was already 73.33, so their average within-person change is zero.

This does not imply that matching returners solves every problem. Their observed change may be estimated correctly while failing to describe the forty absent people. If missing outcomes are related to unobserved difficulties or successes, statistical adjustment depends on assumptions. Report the missingness, compare available baseline characteristics and explain what the data cannot recover.

Practice is another pathway to a higher score

Suppose repeat participants gain six points on a puzzle while comparable first-time participants at the later occasion score only one point above the original baseline group. The five-point contrast is suggestive of a retesting contribution under comparability assumptions. It is not an exact decomposition of development and practice simply because the arithmetic works. The groups could differ in recruitment, exposure or other characteristics.

A good report distinguishes observed change from its interpretation. It states the testing interval, whether tasks were repeated, who was retained and what changed outside the study. It also reports individual variation when available. Two people gaining twelve and zero points produce the same average gain as two people gaining six each, but the developmental stories are different. A mean trajectory should not quietly become the trajectory of every participant.

Previous chapter · Contents · Next chapter

A developmental comparison depends on what a score means at each occasion. If the task changes its demands, a numerical difference may combine change in the target ability with change in something else. Keeping identical wording is not sufficient: the same words or materials can have different familiarity, relevance or difficulty across age groups and settings.

Original case: language conceals reasoning

Imagine a task intended to assess whether a learner can identify a rule. The instructions contain unfamiliar vocabulary. A learner who misunderstands the instruction produces an incorrect response, but the response cannot distinguish a reasoning difficulty from an instruction difficulty. A translated instruction might help, yet translation itself may alter clues or demand. The repair is to examine comprehension and task equivalence rather than assuming either version is a transparent window into the intended ability.

Use several routes to the same relation where appropriate: an action, a visual arrangement, a brief verbal explanation and a new example. Agreement across routes strengthens some interpretations. Disagreement is useful evidence about where the task demands differ. It should not be averaged away before the discrepancy is understood.

For a related applied explanation, eduKateSingapore’s How Children Process English separates input, recognition, meaning, integration and response demands. That distinction helps explain why a language-heavy task may measure more than the concept named in its heading.

Ceiling effects and changing opportunities

Suppose a ten-item task produces scores of nine and ten at successive occasions. The observed gain is one point, but the scale provides almost no room to detect further improvement. A harder task might reveal differences that the original task cannot register. It might also measure additional skills, so the scores cannot simply be joined into one growth line without checking comparability.

Similarly, a child cannot demonstrate independent travel planning if no opportunity to plan a journey is offered. Absence of an observed behaviour can reflect absence of opportunity. Record what the person was asked or allowed to do, not only what they did. Developmental evidence becomes more interpretable when competence, performance conditions and opportunity are documented separately. None can be read reliably from an age label or an isolated score.

Previous chapter · Contents · Next chapter

Developmental history offers a practical habit: inspect the explanation before acting on it. A description such as “immature planning” may compress several different problems into one label. It might refer to unclear instructions, limited experience, difficulty monitoring progress or an unsuitable task. The label does not identify the repair.

A worked diagnostic conversation

An adult says that a learner cannot work independently because the learner asks for help repeatedly. First define the behaviour. Are the questions requests to understand the goal, checks on correctness, requests for materials or attempts to have someone perform the task? Next inspect the task. Does it state the outcome clearly? Are the necessary resources available? Has a comparable task ever been completed without help?

Suppose observation shows that most questions concern where to find materials. Making the materials accessible reduces those questions, while requests about reasoning remain. The intervention has repaired an environmental obstacle; it has not demonstrated a sudden global change in maturity. The remaining questions can now be investigated more precisely. This is progress because the explanation has become narrower and more testable.

If performance improves only while an adult supplies the first step, record that dependency. If it persists when the first step must be chosen independently, the evidence supports a broader change. If it fails in another setting, examine which demand changed before assuming that the learner forgot everything.

Repair rules for common mistaken conclusions

When a claim rests on a single striking event, collect comparable observations and counterexamples. When it rests on an age-group difference, ask for within-person evidence or a qualified description. When it rests on a supported success, record the support and test the intended degree of independence. When it rests on a follow-up average, inspect retention and baseline composition. When it rests on one language-heavy task, check whether the instruction is understood.

The site’s How a Learner Changes from Primary 1 to Primary 6 offers a practical route into changing responsibilities, judgement and support. Read such educational descriptions as prompts for observation, not as fixed rules for every child. The appropriate next step comes from the particular task and evidence, not from forcing a person to match a narrative.

Previous chapter · Contents · Next chapter

Here is an original exercise. A community programme claims that participants became better planners over one year. At the beginning, one hundred participants complete an activity and average 50 points. Sixty return at the end and average 68. The organisers describe an eighteen-point developmental gain. The final task allows a checklist and a helper; the first task allowed neither.

Before reading the answer, identify at least three reasons why the headline exceeds the evidence. Then propose a better comparison. Finally, decide what useful result could still be reported without claiming a general developmental gain. The exercise tests whether the distinctions in this article survive a new setting.

Answer: separate people, conditions and outcomes

First, the initial and final averages describe different observed samples. Calculate the returners’ own baseline average before estimating their change. Second, the performance conditions differ: the checklist and helper may account for some improvement. Third, experience during the year, familiarity with the activity and other events remain possible contributors. Ageing over the interval is not a unique explanation.

Now add information. The sixty returners averaged 60 at baseline. At follow-up they average 68 with support and 63 on a comparable unsupported task. The observed supported increase among returners is eight points, not eighteen. Their unsupported increase is three points. These are descriptive comparisons, conditional on task comparability; they do not isolate the programme’s causal effect or represent outcomes for absent participants.

The difference between 68 and 63 is five points in these observed conditions. It must not automatically be called the pure effect of assistance. Task order, fatigue and differences between the supposedly comparable forms could contribute. A stronger design would plan task assignment and order, specify support, and include appropriate comparisons before data collection.

Transfer beyond the original activity

Ask participants to plan a different low-stakes activity with new constraints. Record whether they identify the goal, sequence steps, anticipate a missing resource and revise after a change. Keep the target responsibility stable even though the surface story changes. A checklist may remain an appropriate real-world tool; report tool-supported and unsupported outcomes separately if both matter.

An honest programme report could say that returning participants performed better under the observed conditions, describe the amount of support, report missing follow-up data and state the limits of attribution. That is a useful result. The stronger claim that planning developed generally requires broader evidence across tasks, occasions and conditions. The repair is not to discard the data, but to match the claim to what was actually observed.

Previous chapter · Contents · Next chapter

Is developmental psychology only about children?

No. A developmental question can concern continuity and change across a life, including adulthood. The history considered here includes infant and child research alongside lifespan perspectives. Do not assume that the aim is always to reach one final adult state; specify the particular function, setting and period being investigated.

Does an age-related pattern prescribe what an individual should do?

No. A group pattern describes the sample and measure used. It does not by itself determine an individual’s capacities, opportunities or needs. A responsible application begins with the person’s actual performance and context. These examples explain research reasoning; they are not diagnostic milestones or screening instruments.

Can stages and gradual change both be useful ideas?

They propose different descriptions that need operational definitions. A sudden shift in an observed score could reflect a threshold in the task rather than a sudden change in the underlying process. Conversely, an average smooth curve can conceal abrupt changes in different individuals at different times. Inspect individual observations and the measurement rule before choosing a description.

Does improvement after practice prove development?

It proves improvement on the observed outcome under the recorded conditions. Whether the change is durable, general or best explained by a proposed developmental mechanism requires more evidence. Practice is not an illegitimate explanation. It becomes a problem when a study intends to estimate another process and fails to distinguish them.

How should a student study this history?

Build a short timeline with a question beside each source: Darwin and detailed observation; Piaget and the organisation of knowledge; Schaie and developmental comparisons; Wood, Bruner and Ross and assisted performance; Bronfenbrenner and connected environments; Baltes and lifespan adaptation; ManyBabies and collaborative replication. The timeline is a reading route, not a claim that one contribution replaced every earlier approach.

For each source, write a claim it can support and one it cannot. Then solve the attrition and transfer examples without looking at the answers. Finally, find an ordinary claim about age or learning and rewrite it so that the population, task, support and observation period are explicit. This practice connects historical reading to everyday judgement.

Readers interested in self-monitoring can continue to The Development of Metacognition in Children. Use it to formulate questions about how learners check their own work, while preserving the distinctions between a proposed mechanism, an observed behaviour and evidence of change.

Previous chapter · Contents · Next chapter

Darwin, C. R. (1877), A Biographical Sketch of an Infant, Mind, 2, 285–294. Read it as a dated observational document. Separate the recorded behaviour from the author’s interpretation, and avoid converting its individual chronology into a population norm.

Piaget, J. (1964), Cognitive Development in Children: Development and Learning, Journal of Research in Science Teaching, 2, 176–186. The foundation’s bibliography documents the relationship to the conference text. Historical publication information and claims about the theory are different kinds of evidence.

Schaie, K. W. (1965), A General Model for the Study of Developmental Problems, Psychological Bulletin, 64, 92–107. The University of Washington research bibliography verifies this publication and related later design papers. The calendar and comparison examples in this guide are original illustrations, not results taken from those studies.

Wood, D., Bruner, J. S., and Ross, G. (1976), The Role of Tutoring in Problem Solving, Journal of Child Psychology and Psychiatry, 17, 89–100. Read the actual task and interaction evidence before transferring the scaffolding idea to a different activity.

Bronfenbrenner, U. (1977), Toward an Experimental Ecology of Human Development, American Psychologist, 32, 513–531. Distinguish a proposal about the scope of inquiry from evidence for a particular environmental cause.

Baltes, P. B. (1987), Theoretical Propositions of Life-Span Developmental Psychology: On the Dynamics Between Growth and Decline, Developmental Psychology, 23, 611–626. Notice the theoretical status of the propositions and the need to specify criteria when judging gains and losses.

The ManyBabies Consortium (2020), Quantifying Sources of Variability in Infancy Research Using the Infant-Directed-Speech Preference, Advances in Methods and Practices in Psychological Science, 3, 24–52. Examine the outcome, procedures and populations before extending its result to another claim.

Across these readings, keep the evidence attached to its conditions. A dated observation, an experimental comparison, a theoretical framework and a pooled estimate contribute in different ways. Understanding their differences is what turns a history of developmental psychology into a usable method for thinking about change.

Previous chapter · Contents · Gold Standard master