# WhatIstheShapeofDevelopmentalChange.pdf

What Is the Shape of Developmental Change?

Karen E. Adolph1, Scott R. Robinson2, Jesse W. Young3, and Felix Gill-Alvarez1

1New York University

2University of Iowa

3Stony Brook University

AbstractDevelopmental trajectories provide the empirical foundation for theories about change processesduring development. However, the ability to distinguish among alternative trajectories depends onhow frequently observations are sampled. This study used real behavioral data, with real patterns ofvariability, to examine the effects of sampling at different intervals on characterization of theunderlying trajectory. Data were derived from a set of 32 infant motor skills indexed daily duringthe first 18 months. Larger sampling intervals (2-31 days) were simulated by systematically removingobservations from the daily data and interpolating over the gaps. Infrequent sampling causeddecreasing sensitivity to fluctuations in the daily data: Variable trajectories erroneously appeared asstep-functions and estimates of onset ages were increasingly off target. Sensitivity to variationdecreased as an inverse power function of sampling interval, resulting in severe degradation of thetrajectory with intervals longer than 7 days. These findings suggest that sampling rates typically usedby developmental researchers may be inadequate to accurately depict patterns of variability and theshape of developmental change. Inadequate sampling regimes therefore may seriously compromisetheories of development.

Developmental TrajectoriesUnderstanding developmental change is a central goal for developmental science. However,despite numerous treatises by prominent developmental theorists in a variety of areas urgingresearchers to focus on change processes (e.g., Elman, 2003; Flavell, 1971; Siegler, 1996;Thelen & Smith, 1994), developmental psychologists have made surprisingly little progresstoward understanding the process of developmental change. Part of the problem is historical.Much of the work in developmental psychology has concentrated on descriptions of children’sbehavior at various ages or on the earliest manifestations of particular abilities. Decades ofreliance on cross-sectional designs, demonstration proofs, and broad-sweeping longitudinalapproaches have left researchers with a gallery of before and after snapshots, studio portraitsof newborns, and fossilized milestones, but little understanding of the process of developmentitself. What we need are accurate, fine-grained depictions of developmental trajectories forcognitive, language, perceptual, motor, and social skills.

The staggering variety of developmental trajectories has also contributed to the lack of progressin understanding change processes. The shape of developmental change might assume anynumber of patterns (Figure 1). For instance, a trajectory might show smooth and monotonicimprovements with age, proceeding at a steady pace as in children’s use of retrieval strategiesin addition (Siegler, 1996), or with accelerating or decelerating rates of change, as in infants’

Correspondence should be addressed to: Karen E. Adolph, Department of Psychology, New York University, 4 Washington Place, Rm410, New York, NY 10003, Tel: (212) 998-9058, Email: Karen.Adolph@nyu.edu.

NIH Public AccessAuthor ManuscriptPsychol Rev. Author manuscript; available in PMC 2009 March 11.

Published in final edited form as:Psychol Rev. 2008 July ; 115(3): 527–543. doi:10.1037/0033-295X.115.3.527.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

acquisition of new words (McMurray, 2007) and improvements in toddlers’ walking skill(Adolph, Vereijken, & Shrout, 2003), respectively. The path of change may showdiscontinuities such as abrupt, stage-like shifts in performance between periods of relativestability, as in children’s stage-like success on many Piagetian tasks (Schultz, 1998), theirabrupt shift from ignoring to marking the past tense of verbs (Marcus, et al., 1992), and thesudden transition to grasping while reaching (Wimmers, Savelsbergh, Beek, & Hopkins,1998). Variability may increase during the period of acquisition, with a series of reversalsvacillating between less and more mature expressions of the skill, as in children’s conservationof volume (van der Maas & Molenaar, 1992). Or a variable acquisition period may entail useof multiple, unsystematic use of strategies between incorrect and correct endpoints, as in(Church & Goldin-Meadow, 1986) and their acquisition of a theory of mind (Flynn, 2006).Discontinuities can take on other shapes, such as episodic changes, where developmentadvances like climbing a staircase, with sudden improvements in children’s conceptualunderstanding separated by long periods in a single stage (Case & Okamoto, 1996) or smallfits and starts of physical growth separated by periods of stasis (Lampl, Veldhuis, & Johnson,1992). Discontinuities can involve reversible patterns of change, as in the U-shaped course ofchildren’s success on math equivalence problems (McNeil, 2007), infants’ alternating steppingmovements (Thelen, 1984), and the classic description of over-regularizations in past tenseverb forms (Marcus, 1992), or the inverted-U-shaped trajectory of cognition over the life span(Craik & Bialystok, 2006), and infants’ zigzag-shaped error rate in detecting threats to balanceas they learn to sit, crawl, cruise, and walk (Adolph, 2005).

Such descriptions of developmental trajectories play an instrumental role in formulating andtesting theories of development (Gottlieb, 1976; Siegler, 2006; Smotherman & Robinson,1995; Wohlwill, 1973). For example, a contentious theoretical debate was spurred bydescriptions of a sudden, stage-like increase in children’s rate of word learning, the so-called“vocabulary spurt,” or “naming explosion” (Bloom 2004; Ganger & Brent, 2004). Accordingto the classic description, at about 18 months of age, when children have acquiredapproximately 50 words, they display a sharp transition from an initial stage of slow vocabularygrowth to a later stage of faster growth. Several influential theories were advanced to explainthe putative shift, invoking major cognitive or linguistic changes that coincided with the spurt(e.g., Gopnik & Meltzoff, 1987; Reznick & Goldfield, 1992). However, recent work showsthat for most children the increase in the rate of word learning is best fit by a quadratic ratherthan a logistic function (Ganger & Brent, 2004). Without a stage-like spurt in the trajectory,theories positing a sudden, fundamental change in cognitive or linguistic abilities becomesuperfluous.

As illustrated by this example, regardless of whether the theoretical perspective is one ofdiscontinuity or continuity, spurts or quadratics, theoretical accounts of how change occurs arebuilt upon the foundation of an accurate portrayal of the pattern of developmental change(Wohlwill, 1970, 1973). And, as we demonstrate in this paper, an accurate characterization ofthe developmental trajectory depends on the rate at which observations are sampled.

The Problem of Sampling RateMore than 75 years ago, Vygotsky (1978) criticized researchers’ reliance on sampling methodsthat merely characterize the stable endpoints in cognitive development. As a remedy, heproposed a “microgenetic method” of sampling at small time intervals to observe developmentin progress. More recent researchers also have cautioned against over-reliance on cross-sectional and long-term longitudinal designs (Wohlwill, 1970, 1973), and have espoused themicrogenetic method for capturing the process of developmental change (e.g., Granott &Parziale, 2002; Kuhn, 1995; Siegler, 2006; Thelen & Ulrich, 1991).

Adolph et al. Page 2

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

However, apart from the general criticism that researchers’ typical sampling intervals are toolarge, the microgenetic method does not quantify the potential consequences of various ratesof data collection for detecting and characterizing different patterns of development.Proponents of the microgenetic method have offered general suggestions that researchersshould collect observations spanning the entire period of change from one stable state toanother, and that the frequency of observations should be high relative to the rate of change ofthe phenomenon (Siegler, 2006). But these proponents have not addressed the problem of howto decide whether a sampling interval is small enough to detect the shape of the underlyingtrajectory. That is, when does one stable state end and another begin? Is the development step-like or is there an intervening period of variability, partial or intermittent expression, ordisruption of performance? Similarly, critics of developmental methodology have recognizedthat overly large sampling intervals in longitudinal research can cause important patterns ofchange to go undetected, and have suggested that developmental researchers sample at smallerintervals (Burchinal & Appelbaum, 1991; Collins, 2006; Hertzog & Nesselroade, 2003;McArdle & Epstein, 1987). But how small is small enough?

In fields of inquiry such as physiology, psychobiology, health psychology, and neuroscience,principles are available to guide the selection of an appropriate sampling rate to ensure recoveryof the underlying pattern. For instance, the Nyquist-Shannon sampling theorem (Nyquist,1928/2002; Shannon, 1949/1998) provides an algorithm for calculating the minimum samplingrate to fully characterize complex waveforms. The sampling theorem stipulates that for awaveform composed of one or more frequencies, with a maximum relevant bandwidth (B), theminimum sampling frequency (fs) necessary to reconstruct the original waveform must be atleast twice the bandwidth (fs > 2B). In other words, sampling frequency must be at least twiceas frequent as the highest frequency component. For example, recording sounds at 20 kHz, theupper limit for human auditory perception, would require sampling the waveform at a minimumof 40 kHz (which is one reason why mp3 digital sound files have such poor quality for higherfrequency sounds). Assumptions about the maximum relevant bandwidth are dictated by thenature of the research question. A study of human color discrimination would not require lightwavelengths to be sampled beyond the blue end of the visible spectrum.

Ironically, the same developmental psychologists who scrupulously use principles such as theNyquist-Shannon theorem to select sampling rates to estimate functions for physiological andpsychophysical variables rely on intuition, convenience, and tradition to select samplingintervals to characterize developmental change in said functions. For example, to describe age-related changes in the ERP associated with face and object recognition, Webb, Long, andNelson (2005) sampled the EEG at 100 Hz to ensure that they could characterize specificcomponents of the EEG response distributed during the first 1500 ms after presentation of thestimulus. But, they relied on arbitrary two-month intervals to chart the developmental trajectoryof the ERP signals. To describe the development of stereoacuity in infants, Held, Birch, andGwiazda (1980) estimated the psychophysical functions by ensuring a sufficiently highsampling rate to distribute intervals of visual angle along the inflection of the curve. Yet, theyrelied on an arbitrary, one-month sampling interval to estimate infants’ developmentaltrajectories and onset ages. Similarly, Adolph (1997) described developmental changes ininfants’ perception of affordances for crawling and walking by sampling at sufficiently smallintervals of difficulty to ensure robust estimates of the psychophysical functions, while relyingon an arbitrary three-week sampling interval to estimate the developmental trajectories.

A recommended remedy for researchers’ sampling dilemma is to design the spacing ofobservations based on a formal theoretical model about the shape of the underlyingdevelopmental function (Boker & Nesselroade, 2002; Burchinal & Appelbaum, 1991). Sucha model would dictate the minimum number of data points and their optimal spacing in time(e.g., a linear function requires only two observations at each end of the acquisition period).

Adolph et al. Page 3

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

However, formal rules such as the Nyquist theorem are applicable only when the data consistof complex waveforms and the maximum frequency of interest is known in advance. If thetemporal scale of developmental changes also were known in advance, then applying a formalrule like the Nyquist might be possible (e.g., sample at twice the frequency of the smallestsignificant change). Unfortunately, most developmental data are not periodic and are notgenerated by simple mathematical functions, where the relevant scale of temporal change canbe obtained by deduction. Thus, developmental researchers must determine patterns ofdevelopmental change empirically and discover, rather than deduce, the temporal scale ofevents that make a difference in the process of change.

The problem is compounded because, as Collins and Graham (2002) point out, empiricallyderived sampling intervals lead to a “chicken and egg” situation: Without prior knowledgeabout the shape of the underlying trajectory to inform a statistical function, researchers cannotknow how frequently to space their observations. And, the underlying function that determinesthe shape of the developmental trajectory cannot be discovered empirically without making adecision about sampling interval. Often, researchers do not even have prior information aboutthe approximate ages that span the period of developmental change.

Implications for the Shape of ChangeFew examples of developmental research have systematically assessed the empirical costs andbenefits of large and small sampling intervals on descriptions of developmental change. Anotable exception is Lampl and colleagues’ research on patterns of physical growth (Johnson,Veldhuis, & Lampl, 1996; Lampl, Johnson, & Frongillo, 2001; Lampl, Veldhuis, & Johnson,1992). Traditionally, children’s growth is characterized as a continuous function from birth toadulthood, with more rapid growth rates during infancy and adolescence. However, whenchildren’s height is measured every day, growth appears to be episodic. Infants’ height, forexample, can increase 1.65 cm in the course of a single day, separated by long periods of daysor weeks during which no growth occurs. Sampling at weekly intervals results in developmentaltrajectories that preserve the episodic nature of children’s growth but reduce the observednumber of growth spurts, increase the amplitude of the spurts, and prolong the periods of stasis.And sampling at quarterly or yearly intervals, as in traditional studies of growth, results in thesmooth, continuous growth curves on standard growth charts.

Even within a 24-hour period, growth is not continuous. In a tour de force of micro-measurement, Lampl and colleagues (Noonan et al., 2004) demonstrated episodic growth ontwo time scales: brief periods of substantial growth on a scale of minutes and days, flanked bylong periods of no growth on an hourly and weekly scale. Leg growth in freely moving lambswas measured with a microtransducer surgically implanted across the tibial growth plate. Bonelength was sampled at 167-sec intervals over a period of 3 weeks, synchronized with videorecordings of the lambs’ activity. Periods of bone growth revealed by the microtransducercoincided with periods of recumbency revealed by the video recordings, and periods whenbones did not grow coincided with periods of loading the limbs in stance or locomotion. Theauthors calculated that 90% of bone growth occurs while lying down, even though lambs spendjust over 50% of their time in a recumbent position, and little or no growth occurs while standingor walking. Clearly, tradition, intuition, and convenience that informed traditional studies ofphysical growth have been inadequate for capturing the richness of the actual trajectory.

The case of physical growth shows how increased sampling resolution from years to days tominutes can provide novel insights into developmental process. The episodic growth patternfrom minute to minute indicates that bones lengthen only when compressive forces on the legare absent. Paradoxically, other research has demonstrated that the presence of physical forcesapplied to bone promote growth by stimulating the expression of genes that regulate cartilageand bone formation (Muller, 2003). Together, these research findings imply that cellular

Adolph et al. Page 4

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

processes involved in regulating physical growth must be coordinated and synchronized on atemporal scale previously unsuspected.

As exemplified by the research on physical growth, overly large sampling intervals will causeinterval data to appear smooth and continuous, regardless of whether the underlying trajectoryis episodic or U-shaped. Similarly, overly large sampling intervals will distort the shape ofchange for binary data (skills that are indexed as absent or present). Figure 2 shows the potentialimpact of sampling at monthly intervals on characterizing patterns of development using actualdata from daily observations of two infants’ progress in balancing upright. The top panel (A)shows a step function, where the infant exhibited a single transition from not-standing tostanding, from one day to the next. The bottom panel (B) shows a variable developmentalfunction, where standing was expressed intermittently (21 times) over a protracted transitionperiod of several weeks. For skills with variable trajectories and reversals, interpolating overthe existing data points—which is what all developmental researchers do when measurementsare collected at weekly, monthly, and yearly intervals—can distort the shape of thedevelopmental trajectory. Infrequent observations will cause binary data to appear as a stepfunction, with a single abrupt transition, regardless of whether the underlying trajectory isvariable, with a series of reversals. As illustrated by the gray curves in the figure, the variabledata in (B) will appear to follow the same developmental path as the stage-like data in (A).

Implications for the Timing of ChangeOverly large sampling intervals are also likely to produce errors in estimating onset ages—theearliest age at which children consistently and reliably express a behavior, skill, orphysiological milestone. Identification of onset ages plays a prominent role in normative andclinical studies of human development, screening for developmental delay, and experimentalmanipulations of development in animals. The onset ages of cognitive and motor milestonesare commonly used to document developmental delays in clinical populations, such as thedelay in autistic and deaf children’s acquisition of theory of mind (Peterson & Siegal, 1999).Age at onset is used to compare the development of different skills such as languagecomprehension and production (Clark & Hecht, 1983), or to compare the development of thesame skill expressed in different contexts, such as the age of attaining conservation of quantitiesin different cultures (Dasen, 1984), or the age of reaching for objects in the light and in thedark (Clifton, Muir, Ashmead, & Clarkson, 1993). Researchers use age at onset to assess effectsof prior experiences on the development of a target skill, such as interactions with siblings onacquiring a theory of mind (Perner, Ruffman, Leekam, 1994), experience with pottery makingon the onset of conservation (Price-Williams, Gordon, & Ramirez, 1969), or the effect ofsleeping prone versus supine on the subsequent development of crawling (Majnemer & Barr,2005). Measures of experience in human infants typically are calculated as the number of daysbetween onset and test dates, for assessing effects of crawling experience, for example, onimprovements in perceptual, cognitive, and social tasks (Campos et al., 2000).

It is easy to imagine how sampling at longer intervals will result in reduced accuracy inestimating the onset age of skills that exhibit abrupt, step-like transitions (e.g., monthlysampling risks 1-month delays in estimates of onset ages; see Figure 2A). But it is less intuitivehow the choice of sampling interval affects the accuracy of estimating onset ages in skills withvariable developmental trajectories. As shown in Figure 2B, infrequent sampling is likely tomiss the period of variability, and thereby provide a later estimate of the onset age.Occasionally, the observations will fall on a day when the skill is present, but not yet stable,and thus distort the estimate of onset by providing a prematurely early estimate.

As we have argued in the foregoing account, the rate at which behavior is sampled is likely tohave a significant impact on our ability to discern the shape and timing of developmentalchange. Sampling at inappropriately large intervals can yield an erroneous picture of the

Adolph et al. Page 5

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

underlying developmental trajectory, which in turn may provide misleading inferences fordevelopmental theory. But the real cost is not just in misrepresenting the pattern of change. Itis the loss of the ability to distinguish among alternative trajectories, such as the ones depictedin Figure 1. An important principle of empirical science is that theories and hypotheses mustbe falsifiable. It should be possible in principle to obtain some set of measurements that wouldnot accord with the theory (Popper, 1959). Inferences about particular developmentaltrajectories are not falsifiable unless the data could have revealed alternative patterns of change.Confidence in the shape of a developmental trajectory depends on whether the data weresampled at appropriate intervals to permit the possible detection of alternative paths. Andbecause there are no generally accepted rules or theorems to guide selection of a samplinginterval in a particular developmental context, appropriate sampling intervals must bedetermined empirically.

Current StudyIn the present study, we aimed to meet the challenge of the microgenetic method by establishingempirically whether the different sampling rates typically used by developmental psychologistsin microgenetic and longitudinal research—days, weeks, and months—are sufficient toaccurately characterize the pattern of developmental change. Our aims were four-fold. First,we sought to demonstrate that real data with real patterns of variability could yield dramaticallydifferent trajectories when sampled at rates commonly used in developmental research. Second,we aimed to quantify how quickly researchers lose the picture of developmental change whensampling at increasingly large intervals. It is a mathematical certainty that coarser samplingwill be less sensitive to fluctuations in the data, but it is not clear at what rate researchers willincur the cost of misrepresenting the underlying trajectory. Third, we assessed the consequenceof different sampling intervals for estimating onset ages—the earliest manifestation of stableexpression of skills and abilities. And fourth, we tested whether the effects of sampling intervalgeneralize across children, the first 18 months of life, and a range of different skills.

Specifically, this study measured the impact of collecting developmental data at intervals ofvarying length on loss of sensitivity to detect the underlying trajectory. To ensure that naturalpatterns of variability would be included in the data, we compiled a real data set of daily changesin 32 infant motor skills (sitting, crawling, standing, walking, etc.) obtained from parentchecklist diaries, rather than an artificial data set of experimenter-generated data. We focusedon motor skills because motor performance is overt and amenable to objective, reliablemeasurement, new motor skills are highly salient to parents, and motor development has a longhistory of longitudinal and microgenetic research. However, in principle, the data set couldhave been constructed from any skills appearing at any point in the lifespan, indexed in termsof competence rather than performance, and obtained in the laboratory or during home visitsrather than by parents’ reports.

Following in the long tradition of language studies (e.g., Darwin, 1877/1974; Dromi, 1987),parents served as informants by noting the presence or absence of each skill at the end of theday in a checklist diary. Although readers’ first inclination may be skepticism regardingparental reports, home observations integrated over the course of the day may be the best wayto determine if a skill is in children’s repertoire because parents are with their children in manydifferent situations, including contexts that are likely to elicit and support the emergence ofnew skills (Bodnarchuk & Eaton, 2004). For language skills, laboratory tests and experimenterhome visits grossly underestimate children’s early abilities, necessitating parental reports toavoid false negatives (Bates, 1993). For motor skills, parent checklist diaries of basic motorskills are concordant with experimenter home visits (Bodnarchuk & Eaton, 2004).

Adolph et al. Page 6

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

As is customary in the literature, we treated the appearance and disappearance of motor skillsas binary, categorical data (present or absent). Like researchers in other developmental domainsthat treat skills categorically (e.g., object permanence, conservation, and theory of mind), weestablished operational definitions for the performance of each skill. For several skills, weincluded multiple criteria for successful performance (e.g., walking < 3 m and walking > 3 m)to determine whether more stringent criteria would affect the trajectory.

From the daily assessments, we constructed developmental trajectories for each skill at thefinest available grain of temporal resolution. Then we systematically removed observations tosimulate the effects of sampling at intervals ranging from daily to monthly, and reconstructedthe developmental trajectories based on the reduced number of observations. Key features ofthe resulting trajectories were compared to the original data to determine the loss of sensitivityfor detecting various patterns of developmental change that result from different samplingschedules. In addition, we formulated a method based on a neurally-inspired activation functionfor objectively estimating onset ages for each skill. We compared the estimated onset agesderived from the original daily observations with those derived from the simulations of largersampling intervals to determine the magnitude of error that could be attributed to samplingfrequency.

MethodChecklist Diary

We compiled a database of daily diary data from eleven families (5 boys, 6 girls). Nine infantswere Caucasian and two were Asian. All parents were middle class and highly educated. Eightinfants had parents who were doctoral students or professors in psychology or anthropology,including the daughter of the first author, and thus most respondents were experienced inmethods of behavioral data collection. Parents began keeping diary records before their infantscould perform any of the target skills, and ended participation several weeks after their infantscould walk independently. One family stopped participation abruptly when the infant was 9months old because of a medical emergency. For the other 10 infants, length of participationranged from 10.94 to 17.00 months (M = 12.59 months). One additional family ceasedparticipation after only 3 months because the parents found it to be too grueling; data from thisinfant were not included in the database.

Parents were trained to make daily entries into a 3-page, paper-and-pencil, checklist diarycontaining 32 gross motor skills involving balance and locomotion, all of which could beperformed in a minimally structured environment (i.e., with a floor and furniture). Instructionmanuals accompanied parents’ diaries with detailed descriptions of the criteria for each skill(see Appendix 1), and a reminder for how to fill out the diary. The diaries were similar to thoseused by Bodnarchuk and Eaton (2004), who showed that parents’ reports were concordant withhome visit observations. Data were collected for 22 additional stair climbing and sliding skills,but these were not included in the current study because they required access to specialequipment not readily available on a daily basis.

Parents noted whether they had observed infants perform each skill at any point over the courseof the day. The diaries provided space for additional written comments about observed skillsthat did not quite match criteria. Such comments about the first two participants—the firstauthor’s daughter and the son of another psychology professor—provided useful informationfor revising skill definitions and criteria. Only skills with uniform definitions and criteria wereincluded in the final data set. Parents entered a question mark for days when they could notremember whether they had witnessed the skill or if they had forgotten to fill in their diaries.Parents also noted days when infants did not have normal access to the floor or to furniture

Adolph et al. Page 7

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

(due to long car trips, camping trips, infants’ illness, etc.) and thus were precluded fromperforming various skills due to situational factors.

Diaries were distributed to parents each month and were organized to minimize errors inparents’ reports. Skills were grouped roughly by postural systems and order of appearance(sitting, prone/crawling, standing/cruising/walking). The first page of the diary containedsitting and early prone skills, the second page contained crawling and upright skills, and thethird page contained stair climbing and sliding skills. More stringent criteria for specific skills(e.g., “walking > 3 m”) followed more lenient criteria (“walking < 3 m”). Some of the skillsin our dataset were ordered hierarchically, where demonstrated facility for a stricter criterionnecessarily assumed facility under a more lenient criterion (noted by asterisks in Appendix 1).For example, once a child can consistently walk 3 meters or more it is not necessary to alsorecord walking less than 3 meters. Therefore, after infants demonstrated facility for at least 30consecutive days with the stricter criterion, the entries for the lenient criterion were assumedto be present.

During monthly lab visits, researchers collected parents’ completed diaries from the previousmonth, interviewed parents about diary entries (confirmed infants’ expression of new skills,cessation of old skills, and question-mark and no-access days), and distributed a new diary forthe current month. The interviewer reminded parents about the criteria for the various skillsusing verbal descriptions, physical demonstrations of the behaviors, and by directing them tothe relevant definitions in the instruction manual.

Missing DataBecause our aim was to assess effects of sampling interval on characterization of the underlyingdevelopmental trajectories, it was especially important to maintain high confidence in theintegrity of the time series. Days that parents noted with question marks and days in whichinfants had no access to the floor constituted missing data. Given that the aim of the study wasto detect variability, we adopted a conservative strategy for interpolating over missing data.For each skill, a software program written in our laboratory searched for the first instance ofexisting data prior to the day for which data were missing and replaced the missing data entrywith that notation. The assumption underlying the interpolation rule was that infants were likelyto have continued doing what they last did until otherwise noted. At most, two consecutivedays of missing data were reconstructed in this way. If a skill contained more than twoconsecutive days of missing data or if missing data constituted 5% or more of all entries, thetime series was not used for further analyses.

Overall, each infant contributed 4-30 skills (M = 23.73 skills) for a total of 99,971 usable diaryentries in the final data set across infants and skills. Several factors caused the large range inthe number of skills that each infant contributed. For the first two infants in the sample, werevised the definitions and criteria for several skills, and thus eliminated several time seriescollected under earlier definitions and criteria. For other infants, some of the time seriesincluded more than 5% missing data due to days noted with question marks, days when infantsdid not have access to the floor, and in the case of one infant, a lost month of entries. Acrossthe sample, some infants never performed certain skills (e.g., never crawled > 3 m). Finally,several time series were either cut short or were not performed by the infant who withdrewfrom the study because of a medical emergency.

Manipulation of Sampling FrequencyThe critical tests involved varying sampling frequency, then interpolating over the interveningpoints. The actual daily data entries provided the smallest sampling interval. We wrote softwareto simulate the effect of sampling at longer intervals by systematically selecting observation

Adolph et al. Page 8

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

points at 2- to 31-day intervals for each skill reported by each parent in the data set. For example,to simulate a 2-day sampling interval, the program selected every second data point; to simulatea 3-day sampling interval, the program selected every third data point. After resampling,removed days were replaced with interpolated values. The process continued for each of theremaining sampling intervals until the least frequent sample at 31-day intervals. Therefore,every simulated time series had the same number of days as the original.

When observation points are distributed in time, the specific day that each sample is collectedcan vary (e.g., sampling at a 2-day interval could be initiated on the first available day thatmeasurements were collected and on all odd numbered days thereafter, or the sample couldbegin on the second day of data collection and proceed on all even numbered days). Failure totake phase into account would allow for random sampling effects to influence the overalltrajectory, particularly if performance of the skill is variable. For example, the singularoccurrence of a skill on Day 31, but not on surrounding days, would appear as a stage-liketransition a month earlier if sampling at 30-day intervals beginning on Day 1 (i.e., 1, 31, 61,91…) compared to the same rate of sampling beginning on Day 2 (2, 32, 62, 92…). To allowfor random variation due to the phase of sampling, the final data set was exhaustive and includedall possible phases at each sampling interval (e.g., 30 phase sets were created for each skillwhen sampling was conducted at 30-day intervals).

The resulting final data set included the original data collected daily, and data sets resultingfrom sampling at 2-31 day intervals at all possible phases. After sampling each simulated seriesof observations, the software program interpolated over missing values by filling in daily valuesbased on the last available observation point. Although it would have been possible to use analternative rule, such as retroactively filling in missing data according to the next availabledata point, we adopted a conservative assumption that a binary function continues on the sametrajectory until a demonstrated instance of a change. Because each of the original time seriesresulted in 495 additional sampled time series, the original data set of 261 (infant × skill) timeseries yielded a final data set of 129,456 unique time series.

ResultsEffects of Sampling Interval on Observed Trajectories

We assessed the effect of variations in sampling interval on the shape of the observed trajectoryby counting the number of transitions between absent and present for each time series. A singletransition would represent an abrupt step-like trajectory from absent to present, as exemplifiedby infant 11 who began standing on one day and stood every day thereafter (see Figure 2A,which is also depicted as the data point nearest the origin in Figure 3A). Alternatively, multipletransitions would represent a variable trajectory between absent and present, as exemplifiedby infant 7 who vacillated 21 times between standing and not-standing (see Figure 2B and topdata point in left-most panel of Figure 3A).

Of the 261 time series in the data set, only 15.7% showed single abrupt transitions (either onsetsalone or a single onset and offset) at a one-day sampling interval. For the remaining 84.3% oftime series, the daily diary data showed variable trajectories, ranging from 3 to 72 transitionsduring the acquisition period (M = 13.37 for those time series showing variable trajectories).Inspection of all time series revealed that variable trajectories were characteristic of all infantsand skills. Between 65% and 100% of the time series for each infant showed multipletransitions, regardless of sex. Similarly, between 67% and 100% of the time series for eachskill (expressed by at least two infants) showed multiple transitions, regardless of the kind ofskill, the strictness of the criterion for judging skill occurrence, or the average age at which theskill was expressed.

Adolph et al. Page 9

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

The consequence of larger sampling intervals was to obscure the true shape of thedevelopmental trajectory. Across the 32 motor skills in the data set, sampling at the simulatedrate of once per week caused 51.2% of the 220 variable time series to show a single transitionfrom absent to present. At the simulated rate of once per month, 91.4% of the variable timeseries appeared to involve a single, abrupt transition. Overall, with monthly sampling, 242(92.7%) of the 261 time series appeared to follow step-like trajectories (compared to the 15.7%based on daily samples), yielding a very different picture of developmental change from thatof the daily data.

How quickly did we lose the picture of developmental change? Given fewer observations atlarger sampling intervals, one would expect a general loss of sensitivity to detect variability.In fact, sampling at progressively larger intervals carried a tremendous cost: Sensitivity todetect variability in the time series declined dramatically each time we widened the samplinginterval by one day. Figure 3A illustrates this precipitous drop-off in sensitivity for each childfor one skill, standing (also represented in Figure 2). Of the 8 children depicted in the graph,only one (infant 11) exhibited a step-like transition from absence to presence when standingwas indexed daily. By the time sampling approached 14-day intervals, however, the child with21 transitions (infant 7) was indistinguishable from the child with a single transition.

The black curves in Figure 3B show that the dramatic decrease in sensitivity was evident forall of the 32 skills in the data set. The gray curve in the figure shows the group average acrosssampling intervals. Although infants averaged 11.74 transitions (SD = 10.48) in their actualdaily diaries across all 32 skills, sampling once per week yielded only 2.51 transitions (SD =2.10), on average, and sampling once per month yielded only 1.20 transitions (SD = 0.83). Thedrop-off in sensitivity is more evident in Figure 3C, which depicts these same data expressedas a percentage of the number of transitions observed with daily sampling. As shown by theconcentration of trajectories in the lower left of the figure, for most time series, fewer than 1in 4 transitions (25%) were detected when sampling at larger than one-week intervals.Moreover, time series with frequent transitions were disproportionately mischaracterized. Theonly trajectories that were depicted accurately at larger sampling intervals were the 41 timeseries (15.7% of all time series) with only 1 abrupt transition from absent to present, shownby the superimposed horizontal lines at 100%.

Each day that the sampling interval widened resulted in fewer transitions detected. To quantifyhow quickly sensitivity to variability was lost, we fit a variety of mathematical functions to thedata shown in Figure 3B. The loss of sensitivity to detect transitions was best described by aninverse power function, meaning that the rate of loss of sensitivity was greatest at the smallestsampling intervals and declined as intervals grew larger. As shown in Figure 3D, most of theR2 values exceeded 0.8 for power functions fit through the data for each of the 240 time serieswith multiple transitions at each of the 31 possible phases.

Effects of Sampling Interval on Estimated Onset AgesDevelopmental researchers rely on onset age—the earliest date at which children canconsistently and reliably express a particular motor or cognitive skill—as a primary index ofdevelopmental progress. As the foregoing discussion of sampling intervals suggests, measuringdevelopmental change at long intervals is likely to result in greater error in identifying the onsetof skill performance than measuring at shorter intervals. We sought to quantify the expectedmagnitude of error in estimating the onset ages by calculating the deviation between the datedetermined by a particular sampling frequency and the date determined by daily sampling.

When sampling at 31-day intervals, each unique phase set provided a separate estimate of theonset age, and thus a distribution of 31 different estimates of the error of measuring onset agerelative to daily samples. In addition to phase differences, if onset is determined by a criterion

Adolph et al. Page 10

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

other than first day of expression, the specific pattern of days in which a skill occurs within avariable trajectory can influence the age identified as the onset of stable performance. Becausemost time series were variable, we sought to vary the particular sequence of days in which theskill was expressed to obtain a larger and more robust set of time series. By creating variantsthrough a constrained randomization procedure, we provided more time series for analysis atall sampling intervals, including short intervals that provided fewer estimates of onset age (e.g.,sampling on alternate days provides only two estimates, one for each phase). In applying arandomization procedure, however, it was important to constrain the procedure to localsequences within the time series, thereby leaving the overall arc of the trajectory the same.

We used a Monte Carlo randomization procedure to introduce slight variations in the dates atwhich skills were expressed. In a typical randomization procedure, the sequence of events inthe entire time series is shuffled, producing a random reordering of the data set (e.g., Johnsonet al., 1996; Kleven, Lane, & Robinson, 2004). Clearly, if the sequence of events in the originaltime series were completely randomized, no developmental pattern could be discerned. Topreserve the overall developmental pattern while creating random variations in the daily events,randomization was constrained to restrict the temporal range within which shuffling occurred.A similar procedure was applied by Loreau (1989), who constrained randomization on aseasonal basis to maintain biological realism in a model of annual activity cycles and ecologicalcompetition.

To implement our randomization procedure, each binary time series, after simulated samplingand interpolation, was parsed into a sequence of bins comprising 14 consecutive days. The sizeof this bin (14 days) was selected after exploring alternative bin widths, and was chosen toprovide a diversity of permutations while introducing minimal error in the overalldevelopmental profiles. Within each bin, daily events were randomly resampled withoutreplacement, creating a sequential permutation of the original bin (Crowley, 1992). Althoughthe specific dates of occurrence were reordered within bins throughout the time series, thesequence of 14-day bins was not modified. Thus, for a time series of daily samples spanninga year, there would be 26 14-day bins and therefore (14!)26 possible permutations. We selected25 randomly generated time series from this set of possible permutations for each unique phaseset for further analysis. This approach resulted in the creation of many alternative time seriesthat differed in the specific dates that skills were expressed, but which preserved the samegeneral developmental trajectory.

The 129,456 density x phase combinations and 25 randomization procedures applied at eachsimulated sampling interval resulted in a total of 3,236,400 time series of skill performance.For each of these time series, we applied an objective algorithm to identify the onset age basedon the earliest age at which the skill was consistently and reliably performed. Determinationof the onset age is straightforward when the underlying developmental trajectory is a step-function because skill performance exhibits a single transition from absence to presence in theinfant’s repertoire (see Figure 2A). However, objectively defining skill onset is moreproblematic when the skill is performed on one day and not on the next (see Figure 2B).

In determining onset age, one might simply report the first day on which the skill was observed.In some developmental research, however, the first date of observation is not used as thecriterion for onset because a singular performance followed by weeks of no expression maybe interpreted as anomalous or unrepresentative of a stable ability. Other criteria (e.g., skillmust be expressed on three consecutive days) are also arbitrary and seem to lead to exceptionsand additional criteria requiring qualitative inspection of each time series. In lieu of theseoptions, and to provide an automated method of determining the onset of stable performanceof each skill that could be applied to three million time series, we applied an objective algorithm

Adolph et al. Page 11

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

to summate over variable periods of skill expression until a criterion level of skill performancewas reached.

The algorithm we used to objectively determine the onset age consisted of an activationfunction that summated across consecutive days of variable skill performance. Althoughinspired by the rules of summation that generate action potentials in a neuron, this function didnot involve processing of the data with an artificial neural network, but acted more generallyas a smoothing function over periods of variable expression. The critical parameters of theactivation function were a decay rate (d), a criterion onset threshold (Ton), and a criterioninactivity threshold (Toff). Activity accumulated or decayed in the function over successivedays following equation 1, where At is the accumulated activity at time t, E is the value of theevent at time t (1 = skill present, 0 = skill absent), and d is decay rate, which specifies theamount of activity that carries over from one day to the next.

(1)

With this simple smoothing rule, each day that a skill was performed added activity to thefunction (much like a small synaptic potential contributes to the net depolarization of a neuron),but activity decayed from one day to the next. When the skill was performed over consecutivedays, the function approximated a logarithmic function and activity summated toward anasymptote that represented consistent and reliable performance. Over a span of days when theskill was not performed, activity decayed toward zero as a negative exponential function. Whenthe cumulative activity, as determined by the particular pattern of skill expression oversuccessive days, exceeded the criterion onset threshold (Ton), the skill was considered stableand the onset age was determined by tracing the rising slope of activity back to the precedingminimum below the inactivity threshold (Toff, see Figure 4A). In practice, this algorithmidentified the first day a skill was expressed in cases where there was a single step-like transitionfrom one day to the next, and it consistently identified a date between the first day a skill wasexpressed and the asymptote in trajectories with periods of variable expression.

We systematically explored the effects of varying different parameters in this function with asubset of the data to maximize the number of time series for which an objective onset age couldbe determined. To confirm the validity of this function, all four authors visually examinedrepresentative graphs of the time series to identify an age by consensus for the onset of stableand consistent performance. The subset of time series included skills that exhibited suddenonset from one day to the next, and skills that showed protracted periods of intermittentexpression before skills were consistently expressed. Parameters of the activation function thenwere adjusted to identify the same ages in the exemplar trajectories. For the results reportedbelow, we used a decay rate of 0.8, an upper onset threshold of 75% of asymptote, and a lowerinactivity threshold of 10% of asymptote as optimal for identifying onset ages across all typesof trajectories. With these settings, we identified onset ages for 3,045,764 time series (94.1%).In most instances, failure to identify an onset age by these objective criteria was due to theinfrequent expression of the skill (on five or fewer days) in the time series (and thus, insufficientactivity accumulated to exceed the onset threshold).

For each child and each skill, we used the activation function to identify an onset age from theoriginal daily diary data. Then, we compared the original onset ages with estimated onset agesfor all of the other time series generated by the randomization procedure at each of the simulatedsampling intervals and phases. Figure 4B shows a series of histograms charting the distributionsof error estimates for one representative skill, standing, in all 8 of the infants for whom we haduseable data. As revealed by reading down the column of histograms, the magnitude of errorincreased systematically with larger sampling intervals. As sampling interval increased, the

Adolph et al. Page 12

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

distributions progressively shifted to the right, reflecting delays in identifying onset. At themost extreme, the estimated onset age was delayed by 55 days.

The pattern of increasing error exemplified by standing was characteristic across the entiredata set (Figure 4C). With daily sampling, the magnitude of error introduced by the MonteCarlo randomization procedure averaged 4.31 days (SD = 1.97) across infants and skills. Inother words, constraining our randomization of skill sequences within 14-day bins resulted inrelatively small variations in onset age. However, with progressively longer sampling intervals,the average magnitude and range of error in estimating onset ages increased sharply. Forexample, sampling at weekly intervals resulted in a mean absolute error of 6.31 days (SD =4.43), and absolute errors >14 days occurred in 7.7% of estimated onset ages. (Errors largerthan about 14 days can seriously compromise theorizing about motor skills.) Sampling at 20day intervals resulted in a mean absolute error of 11.06 days (SD = 7.74), and absolute errors> 14 days in 21.4% of estimates. At a 30-day sampling interval, the mean absolute errorcompared to daily samples was 15.06 days (SD = 9.86), and absolute errors > 14 daysconstituted 37.5% of estimates. At the most extreme, the estimate of onset age differed fromthe actual onset age calculated from daily sampling by 109 days.

Moreover, errors were not distributed symmetrically around the daily estimates of skill onset;most errors were greater than 0, indicating a delayed estimate of the onset age. Sampling atlonger intervals resulted in estimates that were increasingly delayed. When sampled at 2-dayintervals, 19.5% of estimates were delayed relative to the actual onset age, compared with20.1% occurring earlier and 60.4% on the correct date. Sampling at weekly intervals resultedin 34.3% of all estimates occurring later than the actual onset age. At 30-day sampling intervals,delay errors increased to 59.0% of all estimates. For all skills, acceleration errors did not changeacross sampling intervals. But delay errors increased with longer sampling intervals: The rateof increase followed a power function, R2 = 0.96.

DiscussionA fundamental goal of developmental science is to understand change processes. To achievethis goal, researchers need accurate pictures of the shape of change, and such pictures requirerepeated observations. Most developmental researchers, however, do not conduct longitudinaland microgenetic studies because repeated observations are difficult and expensive to collect.The problem is compounded because overly large sampling intervals distort depictions ofdevelopmental change by obscuring important fluctuations in the data: Trajectories chartedwith binary data will appear more abrupt than they really are, and trajectories charted withinterval or ratio data will smooth over important irregularities such as regressions and suddenchanges in the rate of change.

The present study addressed the problem of selecting sampling intervals for developmentaldata by assessing the empirical costs of sampling at progressively larger intervals. The aimwas not merely to confirm the loss of detail with coarser sampling, but to determine how quicklydepictions of development may be altered by sampling data at the rates typically used bydevelopmental researchers. We compiled an illustrative dataset of 32 infant motor skills, andsampled daily to provide a fine-grained depiction of developmental change. We used real,rather than hypothetical data to ensure that our sampling regimes incorporated actual patternsof variability into depictions of the shape of developmental change. Most skills showed a periodof variability (vacillating between occurrence and absence) before acquiring a stable period ofdaily expression. When we simulated sampling at longer intervals (2-31 days), the picture ofa variable acquisition period was quickly lost, so that skills with variable trajectories showeda single, step-like transition. Other critical aspects of the trajectories were also distorted: Mostskills showed large delays in estimating onset ages.

Adolph et al. Page 13

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Daily Changes in Infant Motor Skill AcquisitionA surprising finding that emerged from analyses of the original, daily time series was the largenumber of transitions preceding stable performance. The widespread practice of using point-onset dates for motor skills (e.g., Adolph, et al., 2003; Campos et al., 2000; Frankenburg &Dodds, 1967) presupposes that most skills appear suddenly and are consistently expressedthereafter. However, in the current study, a variable acquisition period characterized most skillsfor every infant across the entire age range. For example, infants averaged 14.57 transitions(SD = 4.96) for standing, as illustrated in Figure 2, and 13.37 transitions across all skills (SD= 10.35). Is it really possible that infants vacillate between occurrence and absence of a skillon a day-to-day basis? Perhaps the variability is just noise and is not developmentallysignificant.

Several factors lend assurance that the diary reports were reliable indicators of dailyperformance. First, the parents were a select group of observers. Eight infants had parents whowere professors or doctoral students, and who conducted behavioral research of their own. Inaddition, parents were carefully trained on the criteria for each skill, and understood theimportance of noting question-mark days when they had insufficient data to mark a skill presentor absent. Most parents spontaneously annotated their diaries when infants’ performance didnot meet criterion (e.g., number of crawling steps, seconds of independent sitting), suggestingthat they took the criteria seriously, and were eagerly waiting for performance to reachthreshold. A second factor that inspires confidence in the daily data is a reliability study: Aless select group of 95 parents provided reliable reports of sitting, crawling, standing, andwalking skills using a daily checklist diary designed after the one used here (Bodnarchuk &Eaton, 2004). Home visits by experimenters blinded to the diary entries yielded concordantdata for 11 of 12 measures. A third factor concerns the directional bias of parents’ errors. Ifparents did err, the most likely errors were false positives. That is, observing infants passcriterion on one day may have biased parents to produce “present” responses on the followingdays. False positives, however, would produce fewer transitions in the time series for any skill,suggesting that the number of transitions reported here are, if anything, an underestimate ofthe true day-to-day variability.

Why then might infants have failed to express sitting, crawling, walking or other basic motorpatterns after demonstrating the ability to do so? Variable acquisition periods cannot beattributed to a lack of opportunity. We only analyzed skills that could be performed in a normalhome environment (with floor, furniture, etc.), and that did not require special equipment orresources (e.g., stairs). Moreover, we eliminated days when the family situation precludedaccess to the floor (traveling, illness, etc.). Variable acquisition periods also cannot beexplained as an artifact of low base-rate levels of performance. Nearly all (94%) of the 261time series eventually reached a stable pattern of daily performance, suggesting that infantswere highly motivated to perform the indexed skills.

A remaining possibility is that variable acquisition periods reflect a biological reality: Asinfants acquire new motor skills, they perform close to the limits of their abilities, much likeathletes struggling to meet their personal best during competition. In early periods of skillacquisition, infants’ peak skill level is far below the criterion level, and on a binary scale, theskill is considered absent. At later periods, as infants’ abilities hover around the criterionthreshold, their top level of performance exceeds criterion on some days, but not others,resulting in variable trajectories. Eventually, infants’ peak skill level comfortably surpassesthreshold, and skills are expressed on a consistent, daily basis. To achieve a more stringentcriterion for the same skill (e.g., walks > 3 m versus walks < 3 m), infants must acquire a stillhigher level of peak performance.

Adolph et al. Page 14

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

This “peak performance” interpretation implies that at least for gross motor skills, over the firstyear and a half of life, infants continually push the envelope of possibility by attempting actionsthat they haven’t quite mastered. Like Vygotsky’s (1978) concept of a “zone of proximaldevelopment,” day-to-day variability in motor skill performance may reflect periods ofdevelopment when infants are operating close to their limits; they are most disrupted byperturbations, and can benefit most from external support. This account also accords withprevious proposals that motor skills are more unstable and sensitive to context when they firstappear in infants’ repertoires (Thelen, Fisher, & Ridley-Johnson, 1984; Robinson &Smotherman, 1992; Garciaguirre, Adolph, & Shrout, 2007). As infants’ peak abilities expand,performance improves, and skills are expressed for longer durations and under more variableand challenging circumstances.

A question that arises about daily variation in infants’ motor skills is whether even smallersampling intervals would have revealed something additional. As in the example of physicalgrowth, where episodic growth across days encompassed episodic growth across minutes, likea set of nested Russian dolls, smaller, meaningful units of motor action are nested within dailysamples. For example, nested within the stuttering day-to-day trajectory of performance incrawling and walking, infants also show a variable trajectory in their expression of locomotion.On the scale of minutes and seconds, infants vacillate between short bouts of locomotion andlonger periods of rest (Adolph, Badaly, Garciaguirre, & Sosky, 2008; Badaly & Adolph,2008; Chan, Lu, Marin, & Adolph, 1999). Variable expression from step to step produces atemporally distributed and spatially variable practice regimen that is most effective inpromoting motor learning (Adolph & Berger, 2006). The intervening rest periods provide timeto consolidate effects of practice and to renew infants’ motivation. Thus, intermittent restperiods may be especially important when infants must operate at peak performance simply toexecute crawling or walking steps.

These theoretical speculations about variable acquisition periods, however, depend on thecharacterization of the developmental trajectory. Without evidence for variable acquisitionperiods, the foregoing discussion of theoretical implications for motor skill acquisition wouldbe moot. And without sampling at a rate that renders the same picture as the daily data, therewould be no evidence for variable acquisition periods. Instead, we would be constructing anaccount to explain step-like transitions in the development of motor skills. A similar dilemmais posed for sampling development in other domains.

Empirical and Theoretical Costs of Sampling DecisionsGiven the long history of microgenetic research (Vygotsky, 1978), methodologists’exhortations to select sampling intervals for reasons other than convenience, tradition, orintuition (Wohlwill, 1970, 1973), and formal demonstrations that long sampling intervals cancompromise conclusions about development (Boker & Nesselroade, 2002; Collins, 2006), onemight expect that developmental research would reflect the same care in choice of samplingregime as in experimental design. Unfortunately, it does not. The general principle that wemust take sampling interval seriously in designing developmental studies is not reflected incurrent practice.

Possibly, general awareness about sampling on a developmental time scale has not yet filtereddown to the rank and file. As Collins and Graham (2002) commented, a similar situationprevailed 40 years ago for the use of power analyses to determine sample size: Originally,power was a concept that statisticians worried about, but it was not widely applied in actualresearch settings. Now researchers routinely use power analyses to design their experimentsas they balance the practical demands of minimizing sample size while avoiding the empiricaland theoretical pitfalls of type two errors.

Adolph et al. Page 15

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

How quickly we lose the picture of developmental change—How seriously mustdevelopmental researchers consider the problem of selecting a sampling interval? In previousstudies of infant motor skill acquisition that are touted in the literature as heroic examples ofmicrogenetic research, observations were collected at weekly or monthly intervals, two of thelarger intervals among our simulated sampling frequencies (Adolph, 1997; Corbetta &Bojczyk, 2002; Thelen et al., 1993; Thelen & Ulrich, 1991; Vereijken & Thelen, 1997). In thecurrent study, daily sampling revealed that 84% of time series exhibited a variable pattern ofemergence. When we simulated sampling infants’ daily motor performance at larger intervals,the picture of day-to-day variability was quickly lost. When sampled once per week, fewerthan half of these time series appeared variable, and when sampled monthly, only 9% appearedvariable. In other words, sampling motor skills once a month caused 75% of the developmentaltrajectories to erroneously look abrupt and step-like, thus characterizing 93% of the entire timeseries with step-like trajectories. It should come as no surprise then that researchers typicallyconsider the first appearance of motor skills to be the onset of a stable period of expression.

The shape of developmental change was not just distorted at the largest sampling intervals.Relatively small increases in interval length resulted in unexpectedly large decrements insensitivity to variability. In fact, an inverse power function accurately described the rate of lossof sensitivity in portraying actual developmental trajectories. These findings indicate that, inthe realm of motor development, the ability to detect variable developmental trajectories dropsoff extremely rapidly at sampling intervals longer than 2 to 3 days. It is the rapidity of thisdrop-off in sensitivity that is counter-intuitive, not the fact that infrequent sampling generallyreduces precision.

A second aspect of developmental profiles that was significantly affected by different samplingrates was estimates of onset ages. Increasingly large sampling intervals caused an increasedrate of errors in estimating the earliest age of stable expression for motor skills. With one-month sampling intervals, the average absolute error was 15 days, and 59% of errors werebiased toward delays. In areas such as infant motor development and language acquisitionwhere skills appear and disappear in relatively quick succession, errors of this magnitude arelikely to have serious consequences for both theory and application in studies of development.Erroneous onset ages carry concomitant costs for estimating durations of experience (e.g., howlong a child has been walking or talking), developmental sequences (e.g., the ordering of motorand linguistic events), and the duration of stable periods (e.g., telegraphic speech, over-regularization of verb tense).

Risks of over-sampling—Of course, frequent sampling also carries potential costs. Asothers have pointed out (Cohen, 1991; McCartney, Burchinal, & Bub, 2006), substantialpractical costs can be incurred by dense sampling. Collection of behavioral data, particularlyin experimental settings, often entails considerable time, effort, and expense that may presentlogistical difficulties. Frequent sampling may have adverse effects on subject recruitment andattrition because demands on participation can be considerable and onerous. Repeated testingcan alter participants’ responses to the experimental condition, although this problem can beaddressed explicitly by including a control group sampled less frequently. Dense samplingover a long period exacerbates problems of data management and methods for summarizingand analyzing data.

But does over-sampling carry the risk of misrepresenting developmental trajectories, therebycausing researchers to misinterpret the research findings? Many time-based phenomena areevident only when assessed on the appropriate time scale. For example, it might be difficult todiscern a 24-hr circadian rhythm while viewing an activity record plotted on a time scale ofseconds, or to detect the day-to-day variability in infants’ acquisition of crawling and walkingover a trajectory that included the bout-rest periods of locomotion on a time scale of seconds.

Adolph et al. Page 16

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

More generally, researchers might fail to detect a developmentally significant pattern on alarger time scale that is obscured by abundant low-level variability or noise in a denselysampled time series.

The interpretive problem, however, arises solely from failure to adequately summarize dataobtained from dense sampling. There are no intrinsic interpretational problems that arise fromsampling frequently, because any time series can be resampled at a reduced rate or smoothedto faithfully represent patterns at a lower grain of resolution. In fact, researchers routinely over-sample physiological and movement data and then apply various smoothing functions to reducenoise and to detect underlying patterns in the data. In other words, researchers can recover thedevelopmental pattern from over-sampled data, but the converse is not true: Researchers cannotrecover the developmental pattern from data sampled with overly large intervals.

Moreover, as illustrated by the findings in the present study, variable developmental trajectoriesare not an inevitable consequence of high sampling rates. Although we found that infant motordevelopment is most often characterized by variable trajectories, the data also demonstratedthat 15.7% of the daily time series showed a sudden, step-like transition, with the skill appearingfrom one day to the next. Contrary to the notion that high sampling rates might create the falseimpression of variability, only with sufficiently frequent sampling is it possible to refute thepossibility that a developmental trajectory is variable, and that a step-function is a moreaccurate depiction of the underlying pattern. This fact is well appreciated by evolutionaryscientists, who acknowledge the need for much finer resolution in the fossil record, on ageological time scale, to distinguish between competing theories of evolutionary change, suchas gradualism versus punctuated equilibrium (Gould & Eldredge, 1993; Gingerich, 2001).

Beyond binary data—Can conclusions regarding the effects of sampling interval generalizebeyond the specifics of the dataset reported here? Sitting, standing, walking, and so on werescored as binary data (either present or absent over the course of each day), and all skills reacheda level of stable, daily performance. Likewise, skills in other domains can be expected to attainstable, daily performance (e.g., correct production of words, learning the multiplication tables).Skills such as crawling, and cruising (and in other domains, weaning from breastfeeding, theability to distinguish speech sounds outside the native language, etc.) also attain stable offsetperiods, where children never produce them again. But what of skills scored as a binary processwith base rates between 0 and 1? Symbolic play, for example, might achieve a stable base rateduring the preschool years between 0.8 and 1, and professional hitting averages in baseballonly rise to the neighborhood of 0.2 to 0.3. Going in the other direction, crying begins at 1 fornewborns, but thankfully decreases to a base rate closer to 0. How does a base rate less than 1(and for offsets, a base rate greater than 0) affect the optimal selection of sampling interval?

A simple Markov switching model can help to clarify the issue of generalization to skills withintermediate base rates. Even a high base rate will result in some days when the skill is notexpressed. Figure 5 provides an illustration. The black curves represent data from threehypothetical time series; the gray curves represent a 15-day moving average that smoothesover the same data. Suppose that the developmental trajectory involves a step-like switch froman early period of absence (pE = 0) to a later period of probabilistic occurrence (pL < 1.0). Forinstance, as shown in Figure 5A, a sudden, step-like shift from absence to a 0.95 probabilityof daily expression would result in an average of five days when the skill was not expressed,and 11 concomitant transitions (between absence and presence, and vice versa) within a 100-day period that actually represents the stable base rate of the skill. Under these conditions,sampling on a daily basis would reveal occasional transitions in the stable base rate, whichmight be misidentified as a variable period of acquisition. In such a situation, it might seempreferable to sample less frequently, say, once a week or once a month, to reduce the chanceof erroneously attributing transitions to a variable acquisition period rather than to a stable,

Adolph et al. Page 17

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

more mature period with a base rate < 1. Obviously, if fewer samples are collected, fewer falsetransitions would be detected.

However, a reduced sampling rate would not provide a more accurate measure of thedevelopmental profile. Instead, it would fail to identify the correct onset age in a step-liketrajectory, and it would decrease the estimate of the number of transitions during the acquisitionperiod for time series with variable trajectories. In contrast, for step-like trajectories, densesampling would pinpoint the onset age. For variable trajectories, dense sampling would allowresearchers to distinguish a variable acquisition period from a post-acquisition period with astable base rate < 1, using the difference in the number of transitions (or some other measureof variability, as revealed by a smoothing function) as an index.

For example, Figure 5B presents the same two-state model (pE and pL) as in 5A, but nowseparated by a 60-day window (representing a variable acquisition period) in which theunderlying process randomly shifts between pE and pL. If pL is high (pL > .8), then the numberof transitions detected by daily sampling will be greater during the variable acquisition periodthan during the later period of stable expression. But, as shown in Figure 5C, if pL is low (pL< .5), then the number of transitions during the variable acquisition period will be less than thenumber observed after the onset of stable expression. In both cases, a simple smoothingfunction can reveal differences in the level of expression during the variable acquisition andstable periods. Thus, the difference in the number of transitions over the entire time seriesprovides a clue as to whether the change from absence to stable expression is step-like orvariable. Even though the absolute number of transitions can be inflated during acquisition forskills with base rates < 1, only dense sampling can reveal differences in the rate of expressionwhen expression of the skill is probabilistic.

Recording skills with greater precision does not alleviate the need for dense sampling tocharacterize the developmental trajectory. The findings regarding sampling intervals shouldalso generalize from the binary data presented here to more precise levels of measurement(ordinal, interval, and ratio scales). Ordinal data, for instance, present much the same problemas binary data for dealing with variable trajectories. On an ordinal scale, sampling lessfrequently would pose the risk of missing periods of vacillation between higher and lowerlevels of performance, periods of inconsistent fluctuation between several levels, periods ofconsistent expression at intermediate levels, or wholesale reversals in levels of performance.

Similarly, when skills are indexed with interval or ratio data, infrequent sampling overdevelopment can lead to decreased sensitivity to detect periods of stability and instability (insingle observations, means, and measures of variability such as the coefficient of variation).Interval and ratio data are conducive to interpolation and curve-fitting that smooth overvariations in the developmental trajectory, so that researchers may fail to detect brief episodesof improvement (as in the case of physical growth) or decrement, interruptions in the expressionof skills, spikes in activity, accelerations (e.g., the vocabulary explosion) and decelerations, orother changes in rate that are evident only when sampling on a finer time scale. Moreover,because sensitivity to random sampling error is greater with larger sampling intervals,estimation of onset ages based on achievement of a criterion level of performance, as measuredwith ordinal or interval data, also would be subject to the same types of errors that we havedescribed for binary data.

Theoretical consequences—Developmental trajectories provide more than empiricalsummaries of change over time: Historically, evidence that cognitive, perceptual, social, ormotor skills exhibit particular developmental trajectories (step-like, variable, linear, episodic,U-shaped, etc.) has stimulated some of the most important theories in developmentalpsychology. The concept of developmental stages illustrates the profound influence of

Adolph et al. Page 18

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

empirical claims about developmental trajectories on theoretical work about developmentalchange. Stage theories have enjoyed a long and influential history in developmental psychology(e.g., Baltes, Reese, & Nesselroade, 1977; Brainerd, 1978; Piaget, 1954). Although the conceptof developmental stages encompasses qualitative changes, hierarchical reorganizations,universal sequences, and so on (Fischer & Silvern, 1985), typically, a central feature of stagetheories involves the timing of development—extended, stable periods interrupted by shorterperiods of developmental change. Rapid, stage-like transitions from one stable pattern ofperformance to the next are characteristic of phenomena ranging from moment-to-momentfluctuations in sleep and waking states in rat pups (Blumberg, Seelke, Lowen, & Karlsson,2005) to patterns of change on an evolutionary time scale (Gould & Eldredge, 1993).

Considerable effort has been devoted to constructing formal models that can account for abrupttransitions between developmental stages. For example, theoretical accounts of stage-likecognitive development include simulations using connectionist models (McClelland, 1989)and rule-based approaches (Siegler, 1976), and mathematical models based on catastrophetheory (van der Maas & Molenaar, 1992), dynamic systems theory (Thelen & Smith, 1994),and other mathematical frameworks (van Rijn, van Someren, & van der Maas, 2003). Forinstance, according to catastrophe theory and dynamic systems theory, enhanced variability isa hallmark of transitions between stable attractors (Kelso, 1995; Thelen & Smith, 1994) suchas successive stages (Raijmakers & Molenaar, 2004). Thus, accurate assessment of the amountand timing of variability is critical for empirically evaluating such models of cognitivedevelopment.

The present study suggests that researchers’ ability to accurately characterize variable and step-like trajectories in development is profoundly affected by sampling rate, and either trajectorymay be inferred erroneously as an artifact of inadequate sampling. Models of developmentalchange become moot if the empirical evidence cannot distinguish among alternativetrajectories. That is, without an appropriate sampling interval, researchers would not be ableto detect a sufficient amount of variability to distinguish between punctate onset dates(Wimmers et al., 1998), instability around times of transitions (Kelso, 1995), expression ofpartial knowledge (Munakata, McClelland, Johnson, & Siegler, 1997), or other patterns of skillonset.

Sampling DevelopmentHow can developmental researchers avoid the pitfalls of under-sampling? Unfortunately,formal principles such as the Nyquist theorem are not applicable to developmental time seriesbecause developmental trajectories can assume many different shapes, few of which areperiodic or conform strictly to mathematical functions. Instead, sampling rates must bedetermined empirically based on the questions being addressed and developmental processesbeing studied. Building on previous work (Siegler, 2006; Thelen & Ulrich, 1991), the presentstudy suggests some precepts to guide the empirical enterprise of identifying optimal samplingrates to accurately capture the shape of developmental change.

(1). Determine the base rate—In most cases, skills of interest to developmentalpsychologists eventually reach a level of stable, consistent performance. Estimating the typicalrate at which the skill is expressed is important in planning how to sample the acquisition periodand/or the more mature, stable period. For skills with stable periods of occurrence (or stableperiods of absence in the case of skills that disappear from children’s repertoires), determiningthe base rate of occurrence will depend on the number of observations collected, not on therate of sampling. If the base rate is less than 1, applying a smoothing function may be usefulin determining an average rate of occurrence (see Figure 5).

Adolph et al. Page 19

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

(2). Find the acquisition period—For most kinds of skills, researchers are likely to initiatea study with some knowledge of the time frame encompassing significant development. Someaspects of children’s behavior emerge over a span of weeks; other aspects may require years.A preliminary investigation with economical sampling (at monthly intervals or longer) maybe useful to identify the approximate age range for the acquisition period, and thereby narrowthe span of time requiring more detailed examination. Note that the initial characterization ofthe developmental trajectory from the preliminary study is unlikely to reveal a detailed andaccurate picture of the shape of developmental change, but may be necessary in planning moredetailed sampling in future efforts.

(3). Sample as small as you can—If the objective is to accurately portray the shape of adevelopmental process, it is crucial to sample data at the minimum, practicable interval,especially over the ages spanning the acquisition period. Researchers should consider thedefault rate for most kinds of child behavior to be daily sampling: absence or presence overthe course of the day for skills indexed with binary data; highest level attained over the courseof the day for skills indexed with ordinal data; and a summary score such as the mean, sum,or coefficient of variation over the course of the day for skills indexed with interval or ratiodata. One reason to consider daily sampling a privileged sampling interval is that it reflects thenearly ubiquitous influence of 24-hour circadian rhythms on human psychological functions.Skills expressed each day are interrupted by sleep each night, during which the day’s activitiesand experiences may be absent, suppressed, forgotten, or consolidated (e.g., Stickgold,2005). A second reason to consider daily sampling privileged is that the present study andothers (e.g., Ganger and Brent’s study of the vocabulary explosion and Lampl and colleagues’work on physical growth) serve as demonstration proofs of important day-to-day changes inmultiple domains of development. Sampling less frequently than every day risks losing theshape of those trajectories. Sampling multiple times each day may provide additional orconverging insights into development, as in the cases of infant walking and physical growth.However, multiple samples per day also introduce variability that may not be meaningfulbecause circadian rhythms affect patterns of performance by changing children’s behavioralstate, motivation, and opportunity for performance.

(4). Look before the onset—To satisfy the objective of describing the entire trajectory,especially the shape of the acquisition period, researchers will need to focus attention on theages when the skill is first expressed. A preliminary investigation using coarse sampling shouldbe useful for obtaining an initial estimate of an onset age, but as the findings of the presentstudy show, estimates of onset ages based on infrequent sampling are likely to produce largedelay errors, and such errors increase with larger sampling intervals. In trajectories based onmonthly samples, for example, estimates of onset age are three times more likely to occur afterthe actual onset age. Therefore, the earliest expression of the skill is likely to occur before theearliest onset age identified by relatively infrequent sampling. As a consequence, more densesampling efforts should include ages prior to the crude estimate of onset.

5. Look for changes in variability—For skills indexed by binary data, trajectories maybe step-like or variable. The latter will show fluctuations prior to attaining a stable level ofperformance. If the base rate of occurrence is high but < 1 during the period of stableperformance (> .8), then a variable acquisition period will likely consist of an increased numberof transitions. In contrast, if the base rate is low (< .5), then a variable acquisition period shouldshow a lower number of transitions relative to the later period of stability. As shown in Figure5B-C, application of a simple moving average or similar smoothing procedure can revealperiods of enhanced variability regardless of the base rate. After smoothing, variable periodsappear as a lower rate of occurrence relative to later ages. Thus, smoothing techniques can be

Adolph et al. Page 20

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

useful in demarcating changes in the level of variability of performance, which can helpresearchers to verify that they have distinguished the acquisition and stable periods.

Concluding RemarksHistorically, much of developmental research has resembled the old saw about the man wholost his keys in a dark alley and turned his attention to searching for them on the street, “wherethe light was much better.” Understanding child development has benefited from descriptionsof age-related differences and demonstrations of the surprising abilities of infants. Butunderstanding the process of developmental change requires more. It requires solid empiricalfoundations built upon accurate depictions of change over time. The implications of ouranalysis of sampling intervals would appear to offer a bleak view of methodological difficulties,even greater than those already recognized by researchers engaged in longitudinal andmicrogenetic research. The payoff for dealing with the thorny methodological difficulties ofsampling rate is that accurate descriptions of developmental trajectories will be instrumentalto advancing theories of development. It is simply necessary for understanding the shape ofdevelopmental change.

Appendix 1. Skills Analyzed from Daily Diaries

Skill Description

Sits (propped on hands) Sits on floor for ≥ 30 s, with legs outstretched, using hands for support.*Sits (hands free) Sits on floor for ≥ 30 s, with legs outstretched, without using hands for support.

Sitting to prone Shifts from sitting position with legs outstretched to prone position.

Prone to sitting Shifts from prone or crawling position into sitting position with legs outstretched.

Kneel to stand (holding) Shifts from kneeling, sitting, or crawling position to standing position by holding onto furniture to pullbody upright.

*Squat to stand (hands free) Shifts from kneeling, sitting, or crawling position into a squat, and then stands up without pulling uprighton furniture.

Stands (holding) Balances upright for ≥ 3 s by holding onto furniture for support.*Stands (hands free) Balances upright for ≥ 3 s without holding onto furniture for support.

Stand to sit (holding) Shifts from upright to sitting position while holding onto furniture for support.*Stand to sit (hands free) Shifts from upright to sitting position without holding onto furniture for support.

Rolls front to back Shifts from lying prone to lying supine.

Rolls back to front Shifts from lying supine to lying prone.

Torso raised (propped on arms) Pushes head and chest off floor by propping on forearms or hands while lying prone.*Torso raised (1 arm free) Pushes head and chest off floor by propping on 1 arm and using the other hand or arm to reach or manipulate

objects.

Rocks on hands and knees Rocks ≥ 2 oscillations while balanced on hands and knees.

Turns 180° prone Pivots in place ≥ 180° while on belly or hands and knees.

Crawls belly (< 3 m) Crawls forward < 3 m, before stopping, with belly resting on floor for duration of each crawling cycle.*Crawls belly (≥ 3 m) Crawls forward ≥ 3 m without stopping, with belly resting on floor for duration of each crawling cycle.

Crawls intermittent belly (< 3 m) Crawls forward < 3 m, before stopping, with belly alternately raised in air and resting on floor during eachcrawling cycle.

*Crawls intermittent belly (≥ 3 m) Crawls forward ≥ 3 m without stopping, with belly alternately raised in air and resting on floor during eachcrawling cycle.

Crawls hands and knees (< 3 m) Crawls forward < 3 m, before stopping, balancing on hands and knees for duration of each crawling cycle.*Crawls hands and knees (≥ 3 m) Crawls forward ≥ 3 m without stopping, balancing on hands and knees for duration of each crawling cycle.

Adolph et al. Page 21

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Skill Description

Crawls hands and feet (< 3 m) Crawls forward < 3 m, before stopping, balancing on hands and feet for duration of each crawling cycle.*Crawls hands and feet (≥ 3 m) Crawls forward ≥ 3 m without stopping, balancing on hands and feet for duration of each crawling cycle.

Cruises 2 hands (< 3 steps) Takes < 3 upright steps, torso sideways, holding onto furniture for support with both hands.*Cruises 2 hands (≥ 3 steps) Takes ≥ 3 upright steps, torso sideways, holding onto furniture for support with both hands.

Cruises 1 hand (< 3 steps) Takes < 3 upright steps, torso frontward, holding onto furniture for support with 1 hand.*Cruises 1 hand (≥ 3 steps) Takes ≥ 3 upright steps, torso frontward, holding onto furniture for support with 1 hand.

Walks supported (2 hands held) Walks with both hands held by caregiver, supporting own weight.*Walks supported (1 hand held) Walks with 1 hand held by caregiver, supporting own weight.

Walks (< 3 m) Walks independently < 3 m.*Walks (≥ 3 m) Walks independently ≥ 3 m.

*Denotes skill with stricter definition than preceding skill.

ReferencesAdolph KE. Learning in the development of infant locomotion. Monographs of the Society for Research

in Child Development 1997;62(3)Serial No. 251Adolph, KE. In: Lockman, J.; Reiser, J., editors. Learning to learn in the development of action; Action

as an organizer of learning and development: The 32nd Minnesota Symposium on Child Development;Hillsdale, NJ: Lawrence Erlbaum Associates. 2005; p. 91-122.

Adolph, KE.; Berger, SE. Motor development. In: Kuhn, D.; Siegler, RS., editors. Handbook of childpsychology (6th ed., Vol. 2: Cognition, Perception, and Language. John Wiley & Sons; New York:2006. p. 161-213.

Adolph, KE.; Badaly, D.; Garciaguirre, JS.; Sotsky, R. 15,000 steps: Infants’ locomotor experience. 2008.Manuscript in preparation

Adolph KE, Vereijken B, Shrout PE. What changes in infant walking and why. Child Development2003;74:474–497.

Badaly, D.; Adolph, KE. Walkers on the go, crawlers in the shadow: 12-month-old infants’ locomotorexperience; Poster presented to the International Society on Infant Studies; Vancouver, Canada. 2008,March;

Baltes, PB.; Reese, HW.; Nesselroade, JR. Life-span developmental psychology: Introduction to researchmethods. Wadsworth Publishing Company; Belmont, CA: 1977.

Bates E. Comprehension and production in early language development: Comments on Savage-Rumbaugh et al. Monographs of the Society for Research in Child Development 1993;58(34)SerialNo. 233

Boker SM, Nesselroade JR. A method for modeling the intrinsic dynamics of intraindividual variability:Recovering the parameters of simulated oscillators in multi-wave panel data. Multivariate BehavioralResearch 2002;37:137–160.

Bloom, P. Myths of word learning. In: Hall, DG.; Waxman, SR., editors. Weaving a lexicon. MIT Press;Cambridge, MA: 2004. p. 205-224.

Blumberg MS, Seelke AMH, Lowen SB, Karlsson KÆ. Dynamics of sleep-wake cyclicity in developingrats. Proceedings of the National Academy of Sciences 2005;102:14860–14864.

Bodnarchuk JL, Eaton WO. Can parent reports be trusted? Validity of daily checklists of gross motormilestone attainment. Applied Developmental Psychology 2004;25:481–490.

Brainerd CJ. The stage question in cognitive-developmental theory. Behavioral & Brain Sciences1978;1:173–181.

Burchinal M, Appelbaum MI. Estimating individual developmental functions: Methods and theirassumptions. Child Development 1991;62:23–43.

Campos JJ, Anderson DI, Barbu-Roth MA, Hubbard EM, Hertenstein MJ, Witherington DC. Travelbroadens the mind. Infancy 2000;1:149–219.

Adolph et al. Page 22

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Case R, Okamoto Y. The role of central conceptual structures in the development of children’s thought.Monographs of the Society for Research in Child Development 1996;61(12)Serial No. 246

Chan, MY.; Lu, Y.; Marin, L.; Adolph, KE. A baby’s day: Capturing crawling experience. In: Grealy,MA.; Thompson, JA., editors. Studies in perception and action V. Lawrence Erlbaum Associates;Mahwah, NJ: 1999. p. 245-249.

Church RB, Goldin-Meadow S. The mismatch between gesture and speech as an index of transitionalknowledge. Cognition 1986;23:43–71. [PubMed: 3742990]

Clark EV, Hecht BF. Comprehension, production, and language acquisition. Annual Review ofPsychology 1983;34:325–349.

Clifton RK, Muir DW, Ashmead DH, Clarkson MG. Is visually guided reading in early infancy a myth?Child Development 1993;64:1099–1110. [PubMed: 8404258]

Cohen, P. A source of bias in longitudinal investigations of change. In: Collins, LM.; Horn, JL., editors.Best methods for the analysis of change. American Psychological Association; Washington, D. C.:1991. p. 18-25.

Collins LM. Analysis of longitudinal data: The integration of theoretical model, temporal design, andstatistical model. Annual Review of Psychology 2006;57:505–528.

Collins LM, Graham JW. The effects of the timing and spacing of observations in longitudinal studiesof tobacco and other drug use: Temporal design considerations. Drug and Alcohol Dependence2002;68:S85–S93. [PubMed: 12324177]

Corbetta D, Bojczyk KE. Infants return to two-handed reaching when they are learning to walk. Journalof Motor Behavior 2002;34:83–95. [PubMed: 11880252]

Craik FIM, Bialystok E. Cognition through the lifespan: Mechanisms of change. Trends in CognitiveSciences 2006;10:131–138. [PubMed: 16460992]

Crowley PH. Resampling methods for computation-intensive data analysis in ecology and evolution.Annual Review of Ecology and Systematics 1992;23:405–447.

Darwin, CR. A biographical sketch of an infant. In: Gruber, HE.; Barrett, PH., editors. Darwin on man.Dutton; New York: 1974. p. 464-474.Reprinted from Mind. A Quarterly Review of Psychology andPhilosophy, 2, 285-294, 1877

Dasen PR. The cross-cultural study of intelligence: Piaget and the Baoule. International Journal ofPsychology 1984;19:407–434.

Dromi, E. Early lexical development. Cambridge University Press; New York: 1987.Elman J. Development: It’s about time. Developmental Science 2003;6:430–433.Fischer KW, Silvern L. Stages and individual differences in cognitive development. Annual Review of

Psychology 1985;36:613–648.Flavell JH. Stage-related properties of cognitive development. Cognitive Psychology 1971;2:421–453.Flynn E. A microgenetic investigation of stability and continuity in theory of mind development. British

Journal of Developmental Psychology 2006;24:631–654.Frankenburg WK, Dodds JB. The Denver Developmental Screening Test. Journal of Pediatrics

1967;71:181–191. [PubMed: 6029467]Ganger J, Brent MR. Reexamining the vocabulary spurt. Developmental Psychology 2004;40:621–632.

[PubMed: 15238048]Garciaguirre JS, Adolph KE, Shrout PE. Baby carriage: Infants walking with loads. Child Development

2007;78:664–680. [PubMed: 17381796]Gingerich PD. Rates of evolution on the time scale of the evolutionary process. Genetica 2001;112:127–

144. [PubMed: 11838762]Gopnick A, Meltzoff A. The development of categorization in the second year and its relation to other

cognitive and linguistic developments. Child Development 1987;58:1523–1531.Gottlieb, G. The roles of experience in the development of behavior and the nervous system. In: Gottlieb,

G., editor. Studies in the development of behavior and the nervous system. Academic Press; NewYork: 1976. p. 1-35.

Gould SJ, Eldredge N. Punctuated equilibrium comes of age. Nature 1993;366:223–227. [PubMed:8232582]

Adolph et al. Page 23

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Granott, N.; Parziale, J. Microdevelopment: Transition processes in development and learning.Cambridge University Press; Cambridge: 2002.

Held R, Birch E, Gwiazda J. Stereoacuity of human infants. Proceedings of the National Academy ofSciences 1980;77:5572–5574.

Hertzog C, Nesselroade JR. Assessing psychological change in adulthood: A review of methodologicalissues. Psychology and Aging 2003;18:639–657. [PubMed: 14692854]

Johnson ML, Veldhuis JD, Lampl M. Is growth saltatory? The usefulness and limitations of frequencydistributions in analyzing pulsatile data. Endocrinology 1996;137:5197–5204. [PubMed: 8940335]

Kelso, JAS. Dynamic patterns: The self-organization of brain and behavior. MIT Press; Cambridge, MA:1995.

Kleven GA, Lane MS, Robinson SR. Development of interlimb movement synchrony in the rat fetus.Behavioral Neuroscience 2004;118:835–844. [PubMed: 15301609]

Kuhn D. Microgenetic study of change: What has it told us? Psychological Science 1995;6:133–139.Lampl M, Johnson ML, Frongillo EA. Mixed distribution analysis identifies saltation and stasis growth.

Annals of Human Biology 2001;28:403–411. [PubMed: 11459238]Lampl M, Veldhuis JD, Johnson ML. Saltation and stasis: A model of human growth. Science

1992;258:801–803. [PubMed: 1439787]Loreau M. On testing temporal niche differentiation in carabid beetles. Oecologia 1989;81:89–96.Majnemer A, Barr RG. Influence of supine sleep positioning on early motor milestone acquisition.

Developmental Medicine & Child Neurology 2005;47:370–376. [PubMed: 15934485]Marcus GF, Pinker S, Ullman M, Hollander M, Rosen TJ, Xu F, Clahsen H. Overregularization in

language acquisition. Monographs of the Society for Research in Child Development 1992;57(4)Serial No. 228

McArdle JJ, Epstein D. Latent growth curves within developmental structural equation models. ChildDevelopment 1987;58:110–133. [PubMed: 3816341]

McCartney K, Burchinal MR, Bub KL. Best practices in quantitative methods for developmentalists.Monographs of the Society for Research in Child Development 2006;71(3)Serial No. 285

McClelland, JL. Parallel distributed processing: Implications for cognition and development. In: Morris,RGM., editor. Parallel distributed processing: Implications for psychology and neurobiology.Clarendon Press; Oxford: 1989. p. 8-45.

McMurray B. Defusing the childhood vocabulary explosion. Science 2007;317:631. [PubMed:17673655]

McNeil N. U-shaped development in math: 7-year-olds outperform 9-year-olds on equivalence problems.Developmental Psychology 2007;43:687–695. [PubMed: 17484580]

Muller GB. Embryonic motility: Environmental influences and evolutionary innovation. Evolution &Development 2003;5:56–60. [PubMed: 12492410]

Munakata Y, McClelland JL, Johnson MH, Siegler RS. Rethinking infant knowledge: Toward an adaptiveprocess account of successes and failures in object permanence tasks. Psychological Review1997;104:686–713. [PubMed: 9337629]

Noonan KJ, Farnum CE, Leiferman EM, Lampl M, Markel MD, Wilsman NJ. Growing pains: Are theydue to increased growth during recumbency as documented in a lamb model? Journal of PediatricOrthopedics 2004;24:726–731. [PubMed: 15502578]

Nyquist H. Certain topics in telegraphic transmission theory. Proceedings of the IEEE 2002;86:280–305.Reprinted from Transactions of the AIEE, pp. 617-644, 1928

Perner J, Ruffman T, Leekam SR. Theory of mind is contagious: You catch it from your sibs. ChildDevelopment 1994;65:1228–1238.

Peterson CC, Siegal M. Representing inner worlds: Theory of mind in autistic, deaf, and normal hearingchildren. Psychological Science 1999;10:126–129.

Piaget, J. The construction of reality in the child. Basic Books; New York: 1954.Popper, K. The logic of scientific discovery. Harper and Row; New York: 1959.Price-Williams D, Gordon W, Ramirez M. Skill and conversation: A study of pottery-making children.

Developmental Psychology 1969;1:769.

Adolph et al. Page 24

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Raijmakers MEJ, Molenaar PCM. Modeling developmental transitions in adaptive resonance theory.Developmental Science 2004;7:149–157. [PubMed: 15320373]

Reznick JS, Goldfield BA. Rapid change in lexical development in comprehension and production.Developmental Psychology 1992;28:406–413.

Robinson SR, Smotherman WP. Fundamental motor patterns of the mammalian fetus. Journal ofNeurobiology 1992;23:1574–1600. [PubMed: 1487750]

Shultz TR. A computational analysis of conservation. Developmental Science 1998;1:103–126.Shannon CE. Communication in the presence of noise. Proceedings of the IEEE 1998;86:447–

457.Reprinted from Proceedings of the IRE, 37, 10-21, 1949Siegler RS. Three aspects of cognitive development. Cognitive Psychology 1976;8:481–520.Siegler, RS. Emerging minds: The process of change in children’s thinking. Oxford University Press;

New York: 1996.Siegler, RS. Microgenetic analysis of learning. In: Kuhn, D.; Siegler, RS., editors. Handbook of child

psychology (6th ed., Vol. 2: Cognition, Perception, and Language. John Wiley & Sons; New York:2006. p. 464-510.

Smotherman, WP.; Robinson, SR. Tracing developmental trajectories into the prenatal period. In:Lecanuet, J-P.; Krasnegor, NA.; Fifer, WP.; Smotherman, WP., editors. Fetal development: Apsychobiological perspective. Lawrence Erlbaum & Associates; Hillsdale, NJ: 1995. p. 15-32.

Stickgold R. Sleep-dependent memory consolidation. Nature 2005;437:1272–1278. [PubMed:16251952]

Thelen E. Learning to walk: Ecological demands and phylogenetic constraints. Advances in InfancyResearch 1984;3:213–260.

Thelen E, Corbetta D, Kamm K, Spencer JP, Schneider K, Zernicke RF. The transition to reaching:Mapping intention and intrinsic dynamics. Child Development 1993;64:1058–1098. [PubMed:8404257]

Thelen E, Fisher DM, Ridley-Johnson R. The relationship between physical growth and a newborn reflex.Infant Behavior and Development 1984;7:479–493.

Thelen, E.; Smith, LB. A dynamic systems approach to the development of cognition and action. MITPress; Cambridge, MA: 1994.

Thelen E, Ulrich BD. Hidden skills: A dynamic systems analysis of treadmill stepping during the firstyear. Monographs of the Society for Research in Child Development 1991;56(1)Serial No. 223

van der Maas HLJ, Molenaar PCM. Stagewise cognitive development: An application of catastrophetheory. Psychological Review 1992;99:395–417. [PubMed: 1502272]

van Rijn H, van Someren M, van der Maas H. Modeling developmental transitions on the balance scaletask. Cognitive Science 2003;27:227–257.

Vereijken B, Thelen E. Training infant treadmill stepping: The role of individual pattern stability.Developmental Psychobiology 1997;30:89–102. [PubMed: 9068964]

Vygotsky, LS. Mind in society: The development of higher mental processes. Cole, M.; John-Steiner,V.; Scribner, S.; Souberman, E., editors. Harvard University Press; Cambridge, MA: 1978.

Webb SJ, Long JD, Nelson CA. A longitudinal investigation of visual event-related potentials in the firstyear of life. Development Science 2005;8:605–616.

Wimmers RH, Savelsbergh GJP, Beek BJ, Hopkins B. Evidence for a phase transition in the earlydevelopment of prehension. Developmental Psychobiology 1998;32:235–248. [PubMed: 9553733]

Wohlwill JF. The age variable in psychological research. Psychological Review 1970;77:49–64.Wohlwill, JF. The study of behavioral development. Academic Press; New York: 1973.

Adolph et al. Page 25

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Figure 1.Idealized shapes of developmental change, with age shown on the X-axis and an index ofbehavioral expression or level of performance on the Y-axis. (a) Linear, (b) Accelerating, (c)Asymptotic, (d) Step-like, (e) S-shaped, (f), Variable, (g) Unsystematic, (h) Stair-climbing, (i)U-shaped, (j) Inverted-U-shaped.

Adolph et al. Page 26

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Figure 2.Examples of developmental trajectories derived from daily data (black curves) for standing(balancing upright for ≥ 3s without holding a support) in two infants. (a) Trajectory that exhibitsabrupt step-function from absent to present from one day to the next. Simulated monthlysampling (gray curve) results in an error in identifying the skill onset age, but does not distortthe shape of the trajectory. (b) Variable trajectory, where skill vacillated 21 times betweenabsent and present over the course of several weeks. Simulated monthly sampling (gray curve)misrepresents both the shape of the variable trajectory and the estimated onset age.

Adolph et al. Page 27

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Figure 3.Effects of sampling interval on sensitivity to variability in developmental trajectories. (a) Thenumber of observed transitions between absence and presence for one skill (standing). Eachcurve represents data for one of the 8 infants for whom we had a complete time series. Opensymbols depict data when the skill was sampled daily; lines show data averaged across allpossible phases at each of the 1- to 31-day sampling intervals. Note that the data point nearestthe origin represents the stage-like data from infant #11, shown in Figure 2A. The other 7 datapoints show data for variable trajectories from other infants, including the top data pointdepicting infant #7, shown in Figure 2B. (b) Number of observed transitions, presented as inFigure 3A, for all 32 skills. The thick gray line represents the mean trajectory across all 261time series. (c) The same data presented in Figure 3B expressed as a percentage of observedtransitions recorded at daily intervals. The horizontal line at 100% represents the 41 time serieswith only 1 abrupt transition from absent to present (15.7% of all time series). Most time seriesconsisted of variable trajectories when measured daily, but more than 75% of transitions werenot detected when sampled at weekly intervals. (d) Distribution of R2 values for inverse powerfunctions fit to each of the 240 time series with multiple transitions. Most time series were best

Adolph et al. Page 28

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

described by an inverse power function, indicating that modest increase in small samplingintervals (< 1 week) resulted in a sharp decline in the ability to detect transitions.

Adolph et al. Page 29

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Figure 4.Effects of sampling interval on estimates of onset ages. (a) Neurally-inspired activationfunction and resulting estimate of the onset age applied to the daily data shown in Figure 1Bfor standing in infant #7. The onset age is determined by identifying the first instance of activitythat exceeds a criterion threshold, then tracing the function back to the preceding period ofinactivity. In this case, the function identifies an onset age at 501 days (shown as vertical dashedline). (b) Histograms showing errors in estimates of the onset age for one skill, standing, in all8 of the infants for whom time series were available. Y-axis is expressed as a percentage oftotal estimates. Note that larger sampling intervals result in a greater range of errors, a generalincrease in the magnitude of errors, and a tendency for errors to be shifted toward later ages.

Adolph et al. Page 30

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

(c) Number of days that estimates of onset ages deviated, either earlier or later, from estimatesderived from daily sampling. Data are presented for all available skills for each child (261 timeseries) as a function of the sampling interval; the superimposed gray line shows the meanabsolute error resulting from sampling at different intervals.

Adolph et al. Page 31

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Figure 5.Simulated developmental trajectories (dark lines) generated by a simple Markov switchingmodel. In each graph, the first 60 days represents a period where the behavior of interest is notyet expressed (p = 0), and the final 100 days represents a period of consistent expression inwhich the behavior occurs at a stable rate < 1. (a) A stage-like trajectory involving an abrupttransition from absence (extended through the first 120 days) to a high base rate of occurrence(p = .95) during the period of stable expression. (b) Trajectory involving an interveningacquisition period (from day 61 to day 120) before achieving a stable period with a high baserate (p = .95). During the acquisition period, behavior is generated by randomly switchingbetween the early regime (absence) and the later period of stability (high base rate). (c)

Adolph et al. Page 32

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Trajectory involving an intervening acquisition period before a stable period with a lower baserate (p = .5). Regime switching occurs in the same way as in (b). In all three graphs, the thickergray line shows a 15-day moving average that depicts the same data; in graphs (b) and (c), thissmoothing function visually demarcates the variable acquisition period from the later periodof stable expression.

Adolph et al. Page 33

Psychol Rev. Author manuscript; available in PMC 2009 March 11.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript