apps as learning tools: a systematic reviewapple and android app stores opened), and written in...

16
Apps As Learning Tools: A Systematic Review Shayl F. Grifth, PhD, Mary B. Hagan, BA, Perrine Heymann, BPhil, Brynna H. Hein, BA, Daniel M. Bagner, PhD abstract CONTEXT: Young children have increasing access to interactive applications (apps) at home and at school. Existing research is clear on the potential dangers of overuse of screens, but there is less clarity around the extent to which interactive apps may be helpful in supporting early learning. OBJECTIVE: In this systematic review, we present a narrative synthesis of studies examining whether children ,6 years can learn from interactive apps. DATA SOURCES: The PsycInfo, PubMed, ACM Digital Library, and ERIC databases were searched. STUDY SELECTION: Studies were included if the study design was randomized or nonrandomized controlled (quasi-experimental), the sample mean age was ,6 years, the intervention involved children playing with an interactive app, and academic, cognitive, or social-emotional skill outcomes were measured. DATA EXTRACTION: Of 1447 studies, 35 were included. RESULTS: Evidence of a learning benet of interactive app use for early academic skills was found across multiple studies, particularly for early mathematics learning in typically developing children. Researchers did not nd evidence of an intervention effect for apps aiming to improve social communication skills in children with autism spectrum disorder. LIMITATIONS: Risk of bias was unclear for many studies because of inadequate reporting. Studies were highly heterogenous in interventions, outcomes, and study design, making comparisons of results across studies difcult. CONCLUSIONS: There is emerging evidence to suggest that interactive apps may be useful and accessible tools for supporting early academic development. More research is needed to evaluate both the potential of educational apps to support early learning, and their limitations. Department of Psychology and Center for Children and Families, Florida International University, Miami, Florida Dr Grifth conceptualized and designed the study, developed the data extraction instrument, adapted the risk of bias instrument, collected and coded the data, and drafted the initial manuscript; Ms Hagan participated in developing the risk of bias instrument, collected and coded the data, drafted portions of the initial manuscript, and reviewed and revised the manuscript; Ms Hein and Ms Heymann participated in developing the risk of bias instrument, collected and coded data, and reviewed and revised the manuscript; Dr Bagner conceptualized and designed the study, participated in the development of the data extraction instrument, and reviewed and revised the manuscript; and all authors approved the nal manuscript as submitted and agree to be accountable for all aspects of the work. DOI: https://doi.org/10.1542/peds.2019-1579 Accepted for publication Aug 30, 2019 Address correspondence to Daniel M. Bagner, PhD, Department of Psychology and Center for Children and Families, Florida International University, 11200 SW 8th St, Miami, FL 33139. E-mail: dbagner@u.edu PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275). Copyright © 2020 by the American Academy of Pediatrics To cite: Grifth SF, Hagan MB, Heymann P, et al. Apps As Learning Tools: A Systematic Review. Pediatrics. 2020;145(1):e20191579 PEDIATRICS Volume 145, number 1, January 2020:e20191579 REVIEW ARTICLE by guest on March 16, 2021 www.aappublications.org/news Downloaded from

Upload: others

Post on 15-Oct-2020

1 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

Apps As Learning Tools: ASystematic ReviewShayl F. Griffith, PhD, Mary B. Hagan, BA, Perrine Heymann, BPhil, Brynna H. Heflin, BA, Daniel M. Bagner, PhD

abstractCONTEXT:Young children have increasing access to interactive applications (apps) at home and atschool. Existing research is clear on the potential dangers of overuse of screens, but there isless clarity around the extent to which interactive apps may be helpful in supporting earlylearning.

OBJECTIVE: In this systematic review, we present a narrative synthesis of studies examiningwhether children ,6 years can learn from interactive apps.

DATA SOURCES: The PsycInfo, PubMed, ACM Digital Library, and ERIC databases were searched.

STUDY SELECTION: Studies were included if the study design was randomized or nonrandomizedcontrolled (quasi-experimental), the sample mean age was ,6 years, the interventioninvolved children playing with an interactive app, and academic, cognitive, or social-emotionalskill outcomes were measured.

DATA EXTRACTION: Of 1447 studies, 35 were included.

RESULTS: Evidence of a learning benefit of interactive app use for early academic skills wasfound across multiple studies, particularly for early mathematics learning in typicallydeveloping children. Researchers did not find evidence of an intervention effect for appsaiming to improve social communication skills in children with autism spectrum disorder.

LIMITATIONS: Risk of bias was unclear for many studies because of inadequate reporting. Studieswere highly heterogenous in interventions, outcomes, and study design, making comparisonsof results across studies difficult.

CONCLUSIONS: There is emerging evidence to suggest that interactive apps may be useful andaccessible tools for supporting early academic development. More research is needed toevaluate both the potential of educational apps to support early learning, and their limitations.

Department of Psychology and Center for Children and Families, Florida International University, Miami, Florida

Dr Griffith conceptualized and designed the study, developed the data extraction instrument, adapted the risk of bias instrument, collected and coded the data, anddrafted the initial manuscript; Ms Hagan participated in developing the risk of bias instrument, collected and coded the data, drafted portions of the initial manuscript,and reviewed and revised the manuscript; Ms Heflin and Ms Heymann participated in developing the risk of bias instrument, collected and coded data, and reviewedand revised the manuscript; Dr Bagner conceptualized and designed the study, participated in the development of the data extraction instrument, and reviewed andrevised the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

DOI: https://doi.org/10.1542/peds.2019-1579

Accepted for publication Aug 30, 2019

Address correspondence to Daniel M. Bagner, PhD, Department of Psychology and Center for Children and Families, Florida International University, 11200 SW 8th St,Miami, FL 33139. E-mail: [email protected]

PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).

Copyright © 2020 by the American Academy of Pediatrics

To cite: Griffith SF, Hagan MB, Heymann P, et al. Apps As Learning Tools: A Systematic Review. Pediatrics. 2020;145(1):e20191579

PEDIATRICS Volume 145, number 1, January 2020:e20191579 REVIEW ARTICLE by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 2: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

Touchscreen mobile devices (eg,smartphones, tablets) have becomeubiquitous for young children.1

Interactive applications or “apps”considered “educational” for youngchildren have similarly gained inpopularity2 and are increasinglybeing integrated into early childhoodclassrooms as learning tools becauseof perceived advantages for childengagement and active learning.3 Theintegration of interactive apptechnology into children’s lives athome and school has outpacedresearch needed to informcomprehensive recommendations forits use. Recommendations have thusfar focused on preventing overuse ofscreens4 rather than opportunities formaximizing learning. Research onwhether young children can learnfrom interactive apps; the academic,cognitive, or social-emotional skilldomains that may be best supportedby interactive apps; and theconditions under which this learningmay be maximized; is still emerging.

In 2016, the American Academy ofPediatrics (AAP) recommendedscreen time (including touchscreens)for 2- to 5-year-olds be limited to ,1hour per day of “high-qualityeducational programming” and thatparents coview screens withchildren.5 Given the scarcity ofstudies on interactive screen media,the AAP statement largely referencedstudies from the more-extensiveliterature on television, in whichresearchers documented thedetrimental effects ofoverconsumption of television onsleep,6 obesity,7 and cognitive andlanguage development8 but alsoshowed that high-quality anddevelopmentally appropriateprogramming (eg, Sesame Street)9

can support vocabulary, math, andsocial-emotional skills in children.30 months.10,11

Theories on learning suggest childrenlearn best when activities promoteactive “minds-on” learning, areappropriately scaffolded, and are

engaging.2 Additionally, researchshows that young children benefitfrom repeated and varied practice offoundational skills.12,13 Interactiveapps have potential to support activeengagement with learning materialthrough embedding learning conceptsinto gamelike activities, scaffoldingchildren’s learning through adaptivelearning technology, providingfeedback and rewards throughgameplay, and promoting repeatedpractice of important foundationalskills.2,14 Additionally, the ubiquity oftouchscreen devices could facilitatemore equitable and convenient accessto educational materials in the home,in contrast to other types oftechnology (eg, computers), whichstill show marked income gaps inaccess.1 Research on whetherinteractive educational touchscreenmedia can support early learning istherefore essential for updating ourunderstanding of “high-qualityeducational programming” for thenew digital age.

Studies in diverse fields, includingpsychology, education, and computerscience, have begun to investigatechildren’s learning from apps, withmost studies published in the past3 years. The findings of these studieshave not yet been well synthesized. Inexisting review articles, authors haveprovided commentaries on children’suse of mobile technology, rather thansystematic reviews of theliterature.15,16 In one recent meta-analysis, authors reporteda significant positive effect acrossstudies of young children’s learningfrom touchscreen devices.17 However,there was significant clinical andmethodological heterogeneity in theincluded studies, which encompassedvaried types of interventions (eg,electronic books, interactive gameapps), outcomes targeted,comparison groups, and studydesigns (randomized andnonrandomized, controlled andnoncontrolled). Statisticalheterogeneity in this study was high

(I2 .90%), limiting interpretability ofa meta-analytic effect size. Instead,a narrative synthesis of results is therecommended approach whenstudies are highly variable ininterventions, outcomes, studydesign, and risk of bias.18,19

Thus, in this study, we aimed tosystematically review the rapidlyexpanding but heterogeneous body ofevidence for young children’slearning from touchscreen interactiveapps. Given the popularity ofinteractive educational apps foryoung children; the importance ofearly learning and school readinessfor later academic, behavioral, andemotional functioning20,21; and theexisting questions about whetheryoung children can apply informationlearned on a screen to real life15; wefocused on children younger than6 years in this review. The specificresearch questions were as follows:

1. Is there evidence that children,6 years can learn from usinginteractive touchscreen apps?

2. What content subjects or skills areparticularly well suited totouchscreen app learning?

3. How does learning fromtouchscreen interactive app gamescompare with children’s learningin other instructional contexts,including in-person instruction oreducational video?

METHODS

Data Sources

An electronic database search wasconducted in 4 databases popular tothe fields of psychology, pediatrics,education, and computer science(PsycINFO, PubMed, ERIC, and ACMDigital Library) by using the followingBoolean search phrase: (tablets ORiPads OR touchscreen OR apps ORinteractive media) AND (learning ORliteracy OR mathematics OR skills ORdevelopment OR functioning ORoutcomes OR achievement) AND(toddlers OR young children OR

2 GRIFFITH et al by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 3: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

preschoolers). Articles publishedbefore January 1, 2008 (the yearApple and Android app storesopened), and written in languagesother than English were excluded.The database searches wereconducted on January 22, 2019.Additionally, the reference sections ofincluded articles and relevant reviewarticles were examined to identifyany other studies meeting inclusioncriteria (n = 2).

Study Selection

Study inclusion and exclusion criteriaare presented in Table 1. The searchinitially identified 1447 uniquerecords. After initial screening, 115records were subject to full textreview by 2 independent coders froma group of 4 (authors S.F.G., B.H.H.,M.B.H., P.H.). Disagreements werediscussed, and consensus wasreached. Randolph’s k22,23 was 0.84,indicating excellent interraterreliability. Thirty-four articles,describing 35 studies, were included(see Fig 1 for Preferred ReportingItems for Systematic Reviews andMeta-analyses diagram).

Data Extraction and Synthesis

The following information wasextracted from each study: design,sample size, location of study,language of participants, length ofintervention, type of control group,name of app(s), study context,targeted subject(s) of intervention,outcome measures, and results. Given

the number of articles (n = 12) forwhich there was not sufficientinformation to calculate an effect sizeand the clinical, methodological, andstatistical heterogeneity of theincluded studies in terms of design,outcomes measured, and comparisongroup used, formal meta-analyticstrategies were not used to

summarize overall effects.18,19

Instead, we presented a narrativesynthesis of the included studiesaccording to the targeted subject ofthe intervention.19 Additionally, toprovide context as to clinicalsignificance of study findings, wherepossible, we presented ranges ofeffect sizes (using Cohen’s d)

TABLE 1 Population, Intervention, Comparison, Outcome, and Study Design Chart of Inclusion and Exclusion Criteria

Category Inclusion Criteria

Population Children 5 y and under (mean age of sample ,72 mo or identifiable subgroup with mean ,72 mo). Studies conducted in any country, withsamples of any language background, were included, as long as the article was published in English. Clinical and nonclinical samples wereincluded.

Intervention Intervention involves use of an interactive game app on a touchscreen. “Interactive” suggests that children’s interaction with the screen has aneffect on the material presented. Studies were excluded if they involved only interventions with electronic books, computers, noninteractivevideo, or multimedia or transmedia interventions. Studies in which a touchscreen device or app was used only as an assistive technology (eg,iPad version of a Picture Exchange Communication System) were excluded.

Comparison Any comparison group type (eg, treatment or instruction as usual, no treatment, attention control, waitlist control, noninteractive video control,alternative app condition control, etc).

Outcome Academic achievement, executive functioning skills, socio-emotional skills, and task-specific learning (eg, learning nonsense names, learning tofind an object, learning facts). Studies were excluded if only process outcomes (eg, engagement) were measured.

Study design Randomized controlled designs (including laboratory experiments and field trials) and quasi-experimental designs. Case studies, reviews,multiple baseline or multiple probe studies, correlational studies, and descriptive studies were excluded.

FIGURE 1Preferred Reporting Items for Systematic Reviews and Meta-analyses flowchart of the studyselection.

PEDIATRICS Volume 145, number 1, January 2020 3 by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 4: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

representing post-test differencesbetween the app intervention andcontrol groups for the randomizedstudies.

Assessment of Risk of Bias

Because most included studies hada randomized design, risk of bias wasassessed with the Cochrane’s risk ofbias tool for randomized trials24 thatwas adapted to include onlycategories relevant to psychologicaland educational research. Studieswere rated in the followingcategories: randomization (selectionbias), blinding of outcome assessment(detection bias), missing data(attrition bias), and selectivereporting (reporting bias). Quasi-experimental studies, for whichassignment to group was not random(eg, administrative convenience; n =5), were also evaluated by using thistool to capture risk of bias due tononrandomized assignment in therandomization category. Ratings of“high,” “low,” or “unclear” risk wereassigned for each domain. Each studywas coded for risk of biasindependently by 2 coders froma group of 4 (coders stated above).Kappa values were 0.91, 0.61, 0.51,and 0.96 for randomization, missingdata, outcome assessment, andselective reporting, respectively,indicating moderate to excellentagreement.25 All disagreements werediscussed, and a consensus wasreached between coders.

RESULTS

Description of Studies

The total number of participantsacross studies was 4639. Eighteenstudies were conducted in the UnitedStates, 5 in Australia, 4 in the UnitedKingdom, 2 in Canada, and 1 in eachof the United Arab Emirates, Italy,Greece, Croatia, the Netherlands, andGermany. The primary language ofmost of the samples (n = 30) wasEnglish (for studies conducted in theUnited States, Australia, the United

Kingdom, Canada, and the UnitedArab Emirates). The 5 remainingstudies included samples in which theprimary language spoken was notEnglish (Italian, Greek, Croatian,Dutch, and German, respectively); ineach case, the interventions andassessments were delivered in theprimary language of the sample.Samples included typicallydeveloping children, except for thestudies in which researchersexamined social communication (n =3), which all consisted of childrendiagnosed with autism spectrumdisorder (ASD). In 12 studies,researchers used apps developed bythe study authors (noted in Table 2);in 4 of these 12 studies, the app wascommercially available, suggestingpotential financial profit for theauthor.26–29 In the remaining studies,researchers used existing appsdeveloped by nonrelated companies.

Twenty-nine studies had randomizeddesigns, including 18 randomizedcontrolled trials (RCTs) with appinterventions delivered over anextended period (ranging from 5 daysto 1 academic year) and11 laboratory-based experimentalstudies, which included interventions(ie, learning phases) delivered overa short time span (ranging from 5 to20 minutes). For simplicity, studieswith randomized designs andinterventions delivered over anextended time are referred to asRCTs. Laboratory-based experiments(all of which also have randomizeddesigns but a short-term interventionphase) are referred to asexperimental studies. The remaining6 studies included in this review werequasi-experimental (ie,nonrandomized controlled) trials.Comparison groups varied acrossstudies, including in-personinstruction (eg, classroom-based, in-person, using manipulatives), videoinstruction, or other non–app-basedinstruction (eg, pen and paperpractice); attention controls receivingapps not targeting the identified

outcome; or groups receiving notreatment. Five studies includedinterventions delivered at home, 18included interventions delivered inthe school, and 12 were delivered ina laboratory. The studies had a varietyof outcomes in targeted educationaldomains, including mathematics (n =15), language arts (LA) (n = 11),executive functioning (n = 5), healthand science knowledge (n = 3), andsocial communication (n = 3). Onestudy included targeted outcomes inboth mathematics and science,62 andanother included math and LAoutcomes.60 Each of the studiesexamined potential learning or skillbenefits of interactive apps. Althoughnot specifically excluded from thesearch strategy, no identified studywas designed to examine harmfuleffects of interactive app use onfunctioning. See Table 2 fora summary of studies.

Risk of Bias Within Studies

Overall, 19 of the 35 studies (54%)were at high risk of bias in at least 1domain. High risk of bias ratings weregiven for nonrandom assignment togroups, substantial missing data anddata not missing at random (eg,imbalances in missing data betweengroups), and lack of blinding ofoutcome assessors. Additionally, riskof bias was judged to be unclear dueto inadequate reporting of methods inat least 2 domains for 19 of the 35studies. Only 3 studies had low risk ofbias across the randomization,missing data, and outcomeassessment domains, and no studieswere given low risk ratings across all4 domains. The results of the studiesshould be interpreted with cautiongiven the literature lacks rigorouslydesigned RCTs.

Narrative Synthesis of Studies byLearning Outcomes

Early Math Learning

In 12 randomized studies (9 RCTs)and 3 quasi-experimental studies,researchers examined children’s math

4 GRIFFITH et al by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 5: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

TABLE2StudiesIncluded

intheSystem

aticReview

Author,Year

StudyPopulation

StudyDesign

Length

ofIntervention

AppIntervention

Condition

andNameof

App

ControlCondition

Outcom

e(M

easure)

InterventionCondition

Outperform

edControl

Condition?

Learning

domain:

earlymath

learning

Aladéet

al,30

2016

aN=60,m

eanage=

58.06mo,42%

male

Experimentalb

3trials

Measure

That

Animal

Noninteractivevideo

ofgameplayin

Measure

that

Animal

Measurement(m

easuring

taskssimilarand

less

similarto

appgameplay)

Yes(for

measuring

task

similarto

app)

Han,312018

N=40,m

eanage=

42.08,50%

male

Experimentalb

12training

trials

Mathappdeveloped

forstudyc

Notreatm

ent

Quantityknow

ledge(quantity

recognition

andGive-N

tasks)

No

KoskoandFerdig,32

2016

aN=73,m

eanage=

51.34mo

RCTd

3wk

Zorbit’sMath

Adventure

Notreatm

ent

Broadmathskills(19-item

measure

created

forstudy)

No

Mattoon

etal,33

2015

aN=24,range

=4–5y,

62.5%

male

RCTe

6wk

DinoKids

–MathLite,

Hungry

Fish,

MotionMath,

Math

Express,Maths

Age

3–5,Park

Math

Instructionwith

traditional

math

manipulatives

Broadmathskills(TEM

A-33

4 )No

Miller,35

2018

aN=13,range

=4–5y,

53.8%

male

RCTe

10d

Packageof

15math

apps

Classroom

instruction

Broadmathskills(22-item

measure

created

forstudy)

No

Outhwaite

etal,36

2019

aN=389,meanage=

60.64mo,49.6%

male

RCTe

12wk

Maths

3–5,Maths

4–6

Classroom

instruction

Broadmathskills(ProgressTest

inMath

Level53

7 )Yes

Outhwaite

etal,38

2017

aRange=50–61

mo,

48.1%

male

Quasi-experimentale

16wk

Maths

3–5,Maths

4–6

Classroom

instruction

Broadmathskills(48-item

and50-item

measurescreatedforstudy)

Yes

Papadakiset

al,39

2018

aN=365,meanage=

62.0mo,48.5%

male

RCTe

14wk

16mathapps

designed

forthe

studyc

Classroom

instruction

Broadmathskills(TEM

A-3)

Yes

Park

etal,40

2016

aN=103,meanage=

4.87

y,51.5%

male

RCTe

10sessions

over

2–3

wk

Mathappdesigned

for

studyc

Mem

orygameapp

Broadmathskills(TEM

A-3)

Yes

Schacter

andJo,26

2016

aN=227,meanage=

56.00mo,46.3%

male

Quasi-experimentale

15wk

MathShelfc

Classroom

instruction

Broadmathskills(iPadassessmentcreated

forthestudy)

Yes

Schacter

andJo,27

2017

aN=433,meanage=

4.55

y,50.6%

male

RCTe

22wk

MathShelfc

Classroom

instruction

Broadmathskills(iPadassessmentcreated

forthestudy)

Yes

Schacter

etal,35

2016

aN=100,meanage=

55.85mo,48%

male

RCTe

6wk

MathShelfc

Top5mathapps

intheappstore

No.sense

abilities(iPadassessmentcreated

forthestudy)

Yes

Spencer,4

12013

Range=4–5y

RCTe

5d

Know

NumberFree

Classroom

instruction

Numeracy(correctlywritten

No.1–10)

Yes

Learning

domain:

EnglishandLA

Blackw

ell,4

22015

N=352,meanage=

5.56

y,51.1%

male

Quasi-experimentale

1y

Shared

iPad

inclassroom

with

set

of70

apps

Classroom

instruction

Earlyliteraryskills(StarEarlyLiteracy

Assessment43 )

Yes

Brow

nand

Harm

on,44

2013

aN=20,range

=48–59

mo

RCTe

1wk

3educationaliPad

apps

targeting

upper-and

lowercase

alphabet

know

ledge,

matching,andNo.

concepts

Educationalapps

unrelatedto

target

academ

iccontent

Alphabetknow

ledge(PALSPre-K4

5Upperand

Lower

Case

Alphabet

Know

ledgesubtests

andmeasurescreatedforstudy)

No

PEDIATRICS Volume 145, number 1, January 2020 5 by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 6: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

TABLE2

Continued

Author,Year

StudyPopulation

StudyDesign

Length

ofIntervention

AppIntervention

Condition

andNameof

App

ControlCondition

Outcom

e(M

easure)

InterventionCondition

Outperform

edControl

Condition?

Cubelic

and

Larw

in,46

2014

aN=281,kindergarten,

“relativebalanceof

boys

andgirls”

Quasi-experimentale

Oneacadem

icyear

Apps

targetingfirst

soundfluency(ABC

Phonics,Build

aWord,

Magic

Penny,Starfall),

phonem

esegm

entation

fluency(First

Word

Sampler,W

ordWall

HD,PocketPhonics,

SkillsBuilder

Spelling,Phonic

Monster

1),and

nonsense

word

fluency(ABC

Touch

andLearn,

Bee

Sees,Kindergarten

Lite,S

tarfall,Super

Why)

Classroom

instruction

Phonological

awareness(DIBELS4

7Next

AssessmentPhonem

eSegm

entation

FluencyandNonsense

Fluencysubtests)

Yes(for

Phonem

eSegm

entation

Fluencyand

Nonsense

Fluency)

Gilliveret

al,48

2016

aN=30,m

eanage=

56.5mo,50%

Male

RCTe

6wk

Phonological

awarenessapp

createdforthe

studyc

Vocabulary

app

createdforthe

study

Phonological

awarenessandvocabulary

(measure

createdforstudy)

Yes

Neum

ann,492018

aN=48,m

eanage=

45.19mo,52.1%

male

RCTe

9wk

EndlessAlphabet,

Letter

School,the

Draw

App

Classroom

instruction

Letter

nameandsoundknow

ledge,name

writing(m

easure

createdforstudy)

Yes

Patchanand

Puranik,502016

N=170,meanage=

51.9mo,43%

male

RCTe

8wk

Writingpracticewith

iPad

app(finger

draw

ing)

Writingpracticewith

penandpaper

Alphabet

know

ledge(nam

ing26

lettersand

writing26

uppercaseletters)

Yes

Russo-Johnson

etal,51

2017

aN=71,m

eanage=

41.05mo,48.2%

male

Experimentalb

61-min

trials

Wordlearning

app

designed

forthe

studyc

Noninteractivevideo

ofgameplayin

wordlearning

app

Wordlearning

(No.correctresponsesin

object

naming)

No

Teepeet

al,52

2017

aN=71,m

eanage=

40.05mo,40.0%

male

RCTb

210-min

intervention

sessions

over

2wk

Jeffy’sJourney

Notreatm

ent

Expressive

andreceptivevocabulary

(measure

createdforstudy)

No

Vatalaro

etal,53

2018

aN=63,range

=3–5y,

52.5%

male

Quasi-experimentale

8wk

EndlessAlphabet,

Noodle

Words,

GoodnightABC,ABC

GO,B

eckandBo,

Draw

andTell,

Don’tletthePigeon

RunthisApp,Alien

Assignment

Apps

chosen

andused

bythestudy

preschool(Letter

School,G

azzili

Science,Yumilo

RainbowPower,

FacesIMake,

CountingBear)

Receptivevocabulary

skills(PPVT54 )and

expressive

vocabulary

skills(EVT

55)

No

Walter-Laager

etal,56

2017

aN=98,m

eanage=

27.3mo,53.1%

male

Experimentalb

20min

Lingua

Kidz

used

with

anadult

Vocabulary

picture

cards,used

with

anadult

Language

developm

ent(abridgedversionof

thestandardized

“Languageand

Developm

entTest

for2-Year-Olds”

57)

No

Learning

domain:

scienceandhealth

Kwok

etal,58

2016

aExperimentalb

8training

trials

No.correct

answ

erson

4animal

facts

No

6 GRIFFITH et al by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 7: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

TABLE2

Continued

Author,Year

StudyPopulation

StudyDesign

Length

ofIntervention

AppIntervention

Condition

andNameof

App

ControlCondition

Outcom

e(M

easure)

InterventionCondition

Outperform

edControl

Condition?

N=86,m

eanage=

66.93mo,62.8%

Appdevelopedfor

studyto

teach

animal

factsc

Face-to-face

instructionon

animal

facts

Putnam

etal,59

2018

aN=114,meanage=

4.99

y,50%

male

RCTd

5d

Appteaching

about

healthyand

unhealthyfoods

(D.W.’s

Unicorn

Adventure)

Notreatm

ent

Recall(No.tim

escorrectly

names

healthy

andunhealthyfoods)

Yes

Learning

domain:

mathandELA

Bebelland

Pedulla,60

2015

aN=∼750,

kindergarten

Quasi-experimentale

1y

Setof

commercially

availableapps

targetingELAand

mathused

inthe

classroom

(during

school-wideiPad

implem

entation

year)

Classroom

instruction

asusual(inthe

year

before

the

iPad

implem

entation)

Earlyliteraryandbroadmathskills(CPAA6

1

ELAandMathSubtests)

Yes

Learning

domain:

mathandscience

Schroederand

Kirkorian,62

2016

a

N=100,meanage=

55.85mo,48%

male

Experimentalb

Don’sCollections:15

questions

total;Life

Cycles:5

trials

Twominigam

esinthe

appMesozoicMath

Adventures

(Don’s

Collections

[Math]

andLife

Cycles

[Science])

Noninteractivevideo

ofgameplayin

wordlearning

app

Quantity;know

ledgeof

grow

thcycle

(measurescreatedwith

itemssimilarto

appgameplay)

Yes(for

know

ledgeof

thegrow

thcycle

forolder

preschoolers);no

(for

quantitytask

foryounger

preschoolers)

Learning

domain:

executive

functioning

Choi

and

Kirkorian,63

2016

a

N=75,range

=23.5–25.6mo,44%

male

Experimentalb

4trials

Appthat

required

childrento

tapon

specificlocationto

reveal

thehiding

placeof

acharacterc

Appthat

show

edanoncontingent

videoof

thehiding

placeof

acharacter

Responseson

object

retrievaltrials

Yes(for

younger

toddlers)

Huberet

al,64

2016

aStudy1:N=18,m

ean

age=5.4y,50%

male;study2:N=

50,m

eanage=5.1

y,46%

male

Experimentalb

45-min

trials

Practicewith

theapp

ExtraTower

ofHanoi

Practicewith

3DTower

ofHanoi

Model

No.extra

moves

on3D

Tower

ofHanoitask

No

Huberet

al,65

2018

aN=96,m

eanage=

36.6mo,56.5%

male

Experimentalb

9min

ofscreen

time

Earlymathapp(Shiny

Party)

Cartoon(Penguinsof

Madagascar)

Delayedgratification(gift-delay

task)and

working

mem

ory(working

mem

orytask)

Yes

Tarasuik

etal,66

2017

aN=49,m

eanage=

4.8y,45%

male

Experimentalb

4trials

Practicewith

theapp

ExtraTower

ofHanoi

Practicewith

3DTower

ofHanoi

Model

No.extra

moves

on3D

Tower

ofHanoitask

No

PEDIATRICS Volume 145, number 1, January 2020 7 by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 8: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

TABLE2

Continued

Author,Year

StudyPopulation

StudyDesign

Length

ofIntervention

AppIntervention

Condition

andNameof

App

ControlCondition

Outcom

e(M

easure)

InterventionCondition

Outperform

edControl

Condition?

Learning

domain:

social

communication

Esposito

etal,67

2017

aN=30,m

eanage=47

mo,90%

male

RCTd

4wk

Standard

ABA1

3apps

createdfor

thestudyto

prom

oteattention,

imitation,

and

vocabulary

c

Standard

ABA

Attention,

vocabulary,and

imitation

(masteredtargetsin

standard

ABA)

No

Fletcher-Watson

etal,29

2016

aN=54,m

eanage=

49.63mo,79.6%

male

RCTd

2mo

Apptargetingsocial

communication

skills,attendingto

people

and

followingsocial

cues

(FindM

e)c

Treatm

entas

usual

Social

communicationskills(M

CDI68 ;CSBS-

DP69;BOSCC

70),(),and

presence

ofAutism

(ADO

S-27

1 )

No

Whitehouseet

al,72

2018

aN=80,m

eanage=

34.83mo,79.6%

male

RCTd

6mo

Community

therapy1

Appdesigned

toenhanceABA

therapysocial

communication

outcom

es(Therapy

Outcom

esby

You)

Community

therapy

Socialcommunicationskills(M

CDI,CSBS-DP),

restricted

andrepetitivebehaviors(RBS-

R73 ),b

ehavioralflexibility

(BFRS7

4 ),

developm

entalskills(ATEC7

5 ),and

adaptiveskills(VABS-27

6 )

No

ABA,AppliedBehavior

Analysis;ADO

S-2,Autism

DiagnosticObservationSchedule;ATEC,Autism

Treatm

entEvaluationChecklist;BFRS,B

ehaviorFlexibility

RatingScale;BOSCC,

BriefObservationof

Social

CommunicationChange;C

PAA,Children’s

Progress

Academ

icAssessment;CSBS-DP,CommunicationandSymbolic

Behavior

Scale,Developm

entalProfile;DIBELS,DynamicIndicators

ofBasicEarlyLiteracy

Skills;ELA,EnglishLanguage

Arts;EVT,ExpressiveVocabulary

Test;M

CDI,MacArthur

CommunicativeDevelopm

entInventory;PALS

Pre-K,Phonological

AwarenessLiteracy

Screener,Preschool;PPVT,PeabodyPictureVocabulary

Test;R

BS-R,Repetitive

Behavior

Scale,Revised;TEMA-3,Test

ofEarlyMathematicsAbility,Third

Edition;

VABS-2,VinelandAdaptiveBehavior

Scales,S

econdEdition.

aPeer-reviewed.

bStudycontextislaboratory.

cAppwas

developedby

studyauthors;italic

script

indicatestheappiscommerciallyavailable.

dStudycontextishome.

eStudycontextisschool.

8 GRIFFITH et al by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 9: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

outcomes after using math interactiveapps. In 11 of the 15 studies,researchers examined school-basedinterventions; a home-basedintervention was examined in 1; and3 were laboratory-based experiments.Outcomes included measures of earlymath skills (including numberrecognition, number naming, andbasic addition and subtraction).School and home interventionsranged from 5 days to 1 academicyear. Overall, 7 of the 12 randomizedstudies (including 1 of the3 laboratory-based experiments) andall 3 quasi-experimental studiesshowed better outcomes in the mathapp intervention group comparedwith the control group used.

Specifically, of the 9 studies (6 RCTs)that compared a math appintervention to in-person classroominstruction, 7 (4 RCTs) reportedsignificantly greater math learninggains on early math measures (n = 8)and a number-writing task (n = 1) forgroups using interactive mathapps,26,27,36,38,39,41,60 whereas theremaining 2 RCTs reportedstatistically equivalent gains betweenthe groups.33,35 Of the 4 studies (3RCTs) in which researchers compareda math app intervention to controlgroups using apps not targeting math(ie, attention controls) or notreceiving treatment, 3 reported betteroutcomes on early math measures forapp intervention groups comparedwith controls,28,31,40 and 1 reportedno significant differences betweengroups.32 Effect sizes, when reported(n = 9, from 9 RCTs with interventionperiods between 5 days and 22weeks), ranged from 0.04 to 0.94.

In 2 experimental studies,researchers examined math learningfrom interactive apps compared withnoninteractive video, with mixedresults. Younger preschoolers’ (3–4.5years) performance on a quantitymath task improved after watching aninstructional video, but not afterplaying a math app, whereas olderpreschoolers (4.5–5.5 years)

improved in neither condition.62 Inthe other study, preschoolers’(3.5–5.5 years) performance ona measuring task improved in bothinteractive app and noninteractivevideo conditions.30

LA Learning

In 7 randomized studies (5 RCTs) and4 quasi-experimental studies,researchers examined children’s LAoutcomes (letter knowledge,phonological awareness, letterwriting, or vocabulary). Eight of the11 studies were based in schools,whereas 3 were laboratory-basedexperiments. Interventions rangedfrom 1 week to 1 academic year, andmost used commercially availableapps. Overall, 3 of the 7 randomizedstudies (including 0 of the 3experimental studies) and 3 ofthe 4 quasi-experimental studiesshowed better outcomes in theapp intervention group comparedwith the control group, whereasthe remainder showed statisticallyequivalent outcomes betweengroups.

Specifically, in 3 studies (2 RCTs),researchers examined alphabetknowledge and phonologicalawareness after children used appstargeting these skills, compared withapps not targeting these skills. Nosignificant differences in learningwere found between groups for 2studies,44,53 whereas the other foundchildren’s phonological awarenessand vocabulary skills improved moreafter using a phonological awarenessapp compared with a vocabularyapp.48 In 5 studies (2 RCTs),researchers found that children’sletter naming and writingskills50,60,77,78 and phonologicalawareness46,60,78 improved afterusing a set of LA apps compared witha classroom or in-person instructioncontrol group. In 2 experimentalstudies, researchers examinedchildren’s vocabulary learning usingan app compared with nonappinstruction (noninteractive video and

picture card instruction) and foundno differences in learning.51,56 In 1final study, researchers found thatchildren’s receptive and expressivevocabulary did not improve after2 10-minute sessions usinga storytelling app with their parentscompared with no treatment.52 Effectsizes for the RCTs, when reported(n = 12, from 3 school-basedinterventions over 1–9 weeks),ranged from 0.07 to 0.94. Effectsizes for the experimentalstudies, when reported (n = 3,from 2 studies), ranged from 0.41to 0.60.

Health and Science Facts

In 3 randomized studies (1 home-based RCT and 2 experimentalstudies), researchers examinedchildren’s learning of health orscience concepts. In the studies,researchers examined children’slearning of healthy versus unhealthyfoods, the growth cycle, and factsabout animals. Overall, the 3 studiesindicated children learned from theapps, with some exceptions. In 1study, researchers found nodifference in learning facts aboutanimals between an interactive appgroup and a group receiving in-person instruction,58 whereasresearchers in another found the appgroup learned more about healthyfoods compared with a no-treatmentcontrol.59 In a third study, researchersfound that older preschoolers(4.5–5.5 years) learned about thegrowth cycle better from interactiveapp play compared withnoninteractive video instruction, butyounger preschoolers (3–4.5 years)were not able to learn in eithercondition.62

Executive Functioning

In 4 experimental studies,researchers examined children’sperformance on executive functioningtasks. In 2 studies, researchers foundthat children improved theirperformance on the traditional three-dimensional (3D) version of the

PEDIATRICS Volume 145, number 1, January 2020 9 by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 10: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

Tower of Hanoi task after practicetrials with a touchscreen app versionof the task, and this improvementwas not significantly differentfrom children who both practicedand were tested with the 3Dversion.64,66 In the third study,researchers found that youngtoddlers (mean age = 25 months)were better able to learn how toretrieve objects in a real-life situationwhen they received training with aninteractive app compared witha noninteractive video; however,this effect was reversed for oldertoddlers (mean age = 33 months).63

In the fourth study, researchersfound that children were morelikely to pass a gift-delay task andperformed better on a workingmemory task after playing aneducational app game comparedwith watching noneducationaltelevision.65

Social Communication

In 3 RCTs, researchers examinedwhether social communication skillsimproved after using apps designedto improve social skills, such asimitation, attention to social stimuli,and following social cues. Each studyincluded a clinical sample of childrenwith ASD; no studies with socialdevelopment outcomes in typicallydeveloping children met inclusioncriteria for this review. In these 3studies, researchers comparedoutcomes in a treatment groupreceiving the app at home in additionto standard applied behavioralanalysis or other communitytherapies to a control group receivingonly treatment as usual. For 2 ofthe 3 studies, children and parentsused the app together,67,72 whereasin the third study, children usedthe app alone.29 None of the 3studies reported significantimprovement in the primary socialcommunication skills outcomemeasures for the app treatmentgroup compared with the controlgroup. Effect sizes for gains inthe app groups on social

communication outcomes rangedfrom 0 to 0.40.

DISCUSSION

In this article, we systematicallyreviewed 35 studies on youngchildren’s learning from interactiveeducational apps. The studieswere heterogeneous in targetedoutcomes, length of intervention,context of intervention, andexperimental design. Evidencefor learning from touchscreeninteractive app use was strongestfor math-related app interventionsand weakest for interventionstargeting social communicationskills. Study quality was mixedoverall, with the risk of biasfor many studies unclear inmore than 1 domain, oftenfrom deficits in methodsreporting.

Can Young Children Learn FromInteractive Touchscreen Apps?

A majority of studies comparing mathor LA (letter naming and writing,phonological awareness) outcomes inchildren using interactive apps tocontrols receiving equivalent regularclassroom or small-group instructionfound more favorable outcomes forthe interactive app condition. Inaddition, although study effect sizesvaried widely, large effects werereported for school-based studies ofmath and LA learning. For example,the largest effect sizes in math werefrom a school-based 22-weekintervention using the app MathShelf,27 whereas the largest effectsizes in LA were from a school-based6-week intervention usinga phonological awareness app.48

These findings are notable becausethey suggest the addition ofinteractive touchscreen technologymay afford a learning advantage forearly academic learning exceedingtraditional modes of instruction andare consistent with the theory thatinteractive apps are well suited to

support active, repeated, and variedpractice of skills.2,16

Results of other studies reviewedsuggest touchscreen app gameshave the potential to support learningin other skill areas, includingscience knowledge and executivefunctioning. However, there werea small number of studies targetingthese skill areas, and further researchis needed to establish whetherchildren’s understanding of scienceconcepts could be promoted usinginteractive apps, and similarly,whether improvement in executivefunctioning task performance istask specific or would generalize toother measures of executivefunctioning.

Results from the 3 RCTs in whichresearchers examined appinterventions targeting socialcommunication skills for childrenwith ASD suggest these skills may notbe best supported by touchscreenapps. This finding is particularlyimportant to note becausetouchscreen technology is popular inASD treatment, including electronicaugmentative exchangecommunication systems.79 Socialcommunication skills may be lesseasily generalized to the real world,compared with foundational earlymath and preliteracy skills, which isconsistent with research showingimprovements in app gameplay butnot real-world social communicationskills.29 However, given the smallnumber of studies examining thisoutcome, further research in this areais needed.

Finally, findings from laboratoryexperimental studies on children’slearning from interactive appscompared with noninteractive videowere mixed. In general, researchers inthese studies examined learning inyounger children who have difficultytransferring learning from media.80

Although in some studies researchersfound that task performanceimproved with an interactive learning

10 GRIFFITH et al by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 11: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

phase,63 others found the interactivecomponent may have interfered withlearning because of overtaxedattentional resources.62 Althoughother studies show older childrenmay benefit from learning throughinteractive app play, further researchis needed to establish the lower agelimit of this potential benefit.

Future Research and ClinicalImplications

Important gaps were identified andshould be addressed in futureresearch. First, more large-scalerandomized trials of apps are needed,with focus on clarifying what appfeatures and content may bestsupport learning. Thousands of appsare added to app stores everymonth,81 so large-scale trials of eachnew educational app may not befeasible. Understanding featuresof interactive apps supportinglearning in different contentareas will allow clearer standardsregarding apps to be deemed“educational” by appmakers.

Second, most studies reviewed wereconducted in schools and examinedmath and LA outcomes. Giventouchscreen devices are widelypresent in homes and the AAP’srecommendation that screen time athome be educational, it will beimportant to examine educationalbenefits of touchscreen media use athome. More studies examining otherlearning content areas are alsoneeded. Third, almost all appinterventions reviewed involvedchild solo use of apps. Childrenlearn best when coviewingtelevision with parents,82 andparent behaviors while co-usingapps have been linked to childpositive engagement and affect.83

Thus, examining children’slearning from app games withparental co-use is an importantarea for future research.

Fourth, all but 5 of the reviewedstudies included app interventions

delivered in English to English-speaking populations, and mostincluded studies based in the UnitedStates, making it difficult to evaluatethe impact of country or language onlearning from apps. There was noapparent pattern to the results ofstudies according to country orlanguage; for example, of the 4studies in which researchersexamined children’s math outcomesconducted outside of the UnitedStates, 3 showed better performancein the app condition, includingthe 1 study conducted in Greek.Further research is needed onthe extent to which educationalapps can support learning ina child’s nonpreferred languageas well as on the generalizabilityto other languages of the findingspresented on LA learning fromapps.

Fifth, further research is needed toevaluate important moderators oftreatment effects. There is emergingevidence from the studies reviewedthat age may be an importantmoderating factor, and childrenbelow preschool age may havemore difficulty learning fromscreen media. Future researchersshould more closely examinemoderating factors that mayimpact learning from touchscreenapps, including children’s pretestability and demographiccharacteristics, such as sex,socioeconomic status, andprevious experience andfamiliarity with technology.

Finally, in a separate body of researchon television, researchers haveidentified negative effects of excessivescreen use on children’sdevelopment.6–8 Although theauthors of the articles in this reviewdid not explicitly examine negativeeffects of use or overuse, as thepopularity and diversity of digitalmedia for children increases, it will benecessary for researchers andclinicians to balance the potentialpositives of interactive screen media

for learning with the imperative toprevent overuse. Further researchexamining the relative advantagesand disadvantages of various types ofscreen media (eg, television versusinteractive touchscreen) will beimportant to updaterecommendations to better reflect thenew digital context.

Limitations

There were some limitations to thissystematic review. First, risk of biaswas unclear for many studies, andmany studies had incompletereporting of study methods. Second,although the study includedunpublished research, it is possibleadditional unpublished research wasnot identified. Unpublished studiesmay have fewer positive outcomesas a result of publication bias.84

Third, because of the nascentstate of the literature, thesummarized studies were highlyheterogeneous. Thus, it was notpossible to compare results acrossstudies or investigate moderatorsof these effects.

CONCLUSIONS

In this systematic review, weexamined the literature oneducational app learning in youngchildren. Evidence of learningbenefits from educationaltouchscreen apps was found acrossmultiple studies, particularly for earlymathematics. Conversely, multiplestudies did not find evidence of anintervention effect for apps aiming toimprove social communication skillsin children with ASD, providing somepreliminary evidence that concreteacademic skills may be bettersupported by interactive apps thanmore-complex social-emotional skills.Findings were mixed in studies inwhich researchers examinedchildren’s short-term learning frominteractive touchscreen media,particularly for toddlers. With theresults of this review, we highlightboth the potential for interactive apps

PEDIATRICS Volume 145, number 1, January 2020 11 by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 12: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

to play a useful role in promotingsome types of early skills as well asthe substantial gaps in knowledgethat still exist regarding their use.Continued research in this area willbe critical to inform the debatearound young children’s screentime, as clinicians and researcherstry to strike a balance betweentaking advantage of the potential

benefits of new technology,while encouraging limits inscreen time.

ACKNOWLEDGMENTS

We thank David H. Arnold, PhD, forhis helpful feedback on a draft of thisarticle.

ABBREVIATIONS

AAP: American Academy ofPediatrics

App: applicationASD: autism spectrum disorderLA: language artsRCT: randomized controlled trial3D: three-dimensional

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.

FUNDING: No external funding.

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

COMPANION PAPER: A companion to this article can be found online at www.pediatrics.org/cgi/doi/10.1542/peds.2019-3503.

REFERENCES

1. Rideout V. The Common Sense Census:Media Use by Kids Age Zero to Eight.San Francisco, CA: Common SenseMedia; 2017. Available at: https://www.commonsensemedia.org/sites/default/files/uploads/research/csm_zerotoeight_fullreport_release_2.pdf. AccessedMarch 13, 2019

2. Hirsh-Pasek K, Zosh JM, Golinkoff RM,Gray JH, Robb MB, Kaufman J. Puttingeducation in “educational” apps:lessons from the science of learning.Psychol Sci Public Interest. 2015;16(1):3–34

3. Haßler B, Major L, Hennessy S. Tabletuse in schools: a critical review of theevidence for learning outcomes.J Comput Assist Learn. 2016;32(2):139–156

4. McCarthy C. Pediatricians andtelevision: it’s time to rethink ourmessaging and our efforts. Pediatrics.2013;131(3):589–590

5. Council on Communications and Media.Media and young minds. Pediatrics.2016;138(5):e20162591

6. McDonald L, Wardle J, Llewellyn CH, vanJaarsveld CHM, Fisher A. Predictors ofshorter sleep in early childhood. SleepMed. 2014;15(5):536–540

7. Cox R, Skouteris H, Rutherford L, Fuller-Tyszkiewicz M, Dell’ Aquila D, Hardy LL.Television viewing, television content,food intake, physical activity and bodymass index: a cross-sectional study of

preschool children aged 2-6 years.Health Promot J Austr. 2012;23(1):58–62

8. Zimmerman FJ, Christakis DA, MeltzoffAN. Associations between mediaviewing and language development inchildren under age 2 years. J Pediatr.2007;151(4):364–368

9. Anderson DR, Bryant J, Wilder A,Santomero A, Williams M, Crawley AM.Researching Blue’s Clues: viewingbehavior and impact. Media Psychol.2000;2(2):179–194

10. Anderson DR, Pempek TA. Television andvery young children. Am Behav Sci.2005;48(5):505–522

11. Anderson DR, Hanson KG. Whatresearchers have learned abouttoddlers and television. Zero Three.2013;33(4):4–10

12. Whitehurst GJ, Lonigan CJ. Childdevelopment and emergent literacy.Child Dev. 1998;69(3):848–872

13. Purpura DJ, Baroody AJ, Lonigan CJ. Thetransition from informal to formalmathematical knowledge: mediation bynumeral knowledge. J Educ Psychol.2013;105(2):453–464

14. Neumann MM, Finger G, Neumann DL. Aconceptual framework for emergentdigital literacy. Early ChildhoodEducation Journal. 2017;45(4):471–479

15. Radesky JS, Schumacher J, ZuckermanB. Mobile and interactive media use byyoung children: the good, the bad, and

the unknown. Pediatrics. 2015;135(1):1–3

16. Radesky JS, Christakis DA. Increasedscreen time: implications for earlychildhood development and behavior.Pediatr Clin North Am. 2016;63(5):827–839

17. Xie H, Peng J, Qin M, Huang X, Tian F,Zhou Z. Can touchscreen devices beused to facilitate young children’slearning? A meta-analysis oftouchscreen learning effect. FrontPsychol. 2018;9:2580

18. Fletcher J. What is heterogeneity and isit important? BMJ. 2007;334(7584):94–96

19. Ryan R, Cochrane Consumers andCommunication Review Group.Cochrane consumers andcommunication review group: datasynthesis and analysis. Available at:https://cccrg.cochrane.org/sites/cccrg.cochrane.org/files/public/uploads/Analysis.pdf. Accessed April 13, 2019

20. Duncan GJ, Dowsett CJ, Claessens A,et al. School readiness and laterachievement. Dev Psychol. 2007;43(6):1428–1446

21. High PC, Klass P; Council on EarlyChildhood. Literacy promotion: anessential component of primary carepediatric practice. Pediatrics. 2014;134(2):404–409

22. Randolph J. Free-marginal multiraterkappa: an alternative to Fleiss’ fixed-

12 GRIFFITH et al by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 13: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

marginal multirater kappa. In:Proceedings from the JoensuuUniversity Learning and InstructionSymposium 2005; October 14–15, 2005;Joensuu, Finland.

23. Warrens MJ. Inequalities betweenmulti-rater kappas. Adv Data AnalClassif. 2010;4(4):271–286

24. Higgins JPT, Altman DG, Gøtzsche PC,et al; Cochrane Bias Methods Group;Cochrane Statistical Methods Group.The Cochrane Collaboration’s tool forassessing risk of bias in randomisedtrials. BMJ. 2011;343:d5928

25. Fleiss JL. Statistical Methods for Ratesand Proportions. Hoboken, JN: Wiley;1981

26. Schacter J, Jo B. Improving low-incomepreschoolers mathematicsachievement with Math Shelf,a preschool tablet computercurriculum. Comput Human Behav.2016;55:223–229

27. Schacter J, Jo B. Improvingpreschoolers’ mathematicsachievement with tablets: a randomizedcontrolled trial. Math Educ Res J. 2017;29(3):313–327

28. Schacter J, Shih J, Allen CM, et al. MathShelf: a randomized trial ofa prekindergarten tablet number sensecurriculum. Early Educ Dev. 2016;27(1):74–88

29. Fletcher-Watson S, Petrou A, Scott-Barrett J, et al. A trial of an iPadTM

intervention targeting socialcommunication skills in children withautism. Autism. 2016;20(7):771–782

30. Aladé F, Lauricella AR, Beaudoin-Ryan L,Wartella E. Measuring with Murray:touchscreen technology andpreschoolers’ STEM learning. ComputHuman Behav. 2016;62:433–441

31. Han A. How Training Set and PriorKnowledge Affect Preschoolers’Perception of Quantity and EarlyNumber Learning [dissertation]. WestLafayette, IN: Purdue University; 2018

32. Kosko K, Ferdig R. Effects of a tablet-based mathematics application for pre-school children. J Comput Math SciTeach. 2016;35(1):61–79

33. Mattoon C, Bates A, Shifflet R, Latham N,Ennis S. Examining computational skillsin prekindergarteners: the effects of

traditional and digital manipulatives ina prekindergarten classroom. EarlyChild Res Pract. 2015;17(1)

34. Ginsburg H, Baroody A. Test of EarlyMathematics Ability. Austin, TX: PRO-ED;2003

35. Miller T. Developing numeracy skillsusing interactive technology in a play-based learning environment. Int J STEMEduc. 2018;5(1):39

36. Outhwaite LA, Faulder M, Gulliford A,Pitchford NJ. Raising early achievementin math with interactive apps:a randomized control trial. J EducPsychol. 2019;111(2):284–298

37. Math Assessment Resource Service.Progress Test in Maths Level 5. London,UK: GL Assessment; 2015

38. Outhwaite LA, Gulliford A, Pitchford NJ.Closing the gap: efficacy of a tabletintervention to support thedevelopment of early mathematicalskills in UK primary school children.Comput Educ. 2017;108:43–58

39. Papadakis S, Kalogiannakis M, ZaranisN. The effectiveness of computer andtablet assisted intervention in earlychildhood students’ understanding ofnumbers. An empirical study conductedin Greece. Educ Inf Technol. 2018;23(5):1849–1871

40. Park J, Bermudez V, Roberts RC,Brannon EM. Non-symbolic approximatearithmetic training improves mathperformance in preschoolers. J ExpChild Psychol. 2016;152:278–293

41. Spencer P. iPads: improving numeracylearning in the early years. In: Steinle V,Ball L, Bardini C, eds. Proceedings fromthe 36th Annual Conference of theMathematics Education ResearchGroup of Australasia; July 7–11, 2013;Melbourne, Australia

42. Blackwell CK. Technology use in earlychildhood education: investigatingteacher access and attitudes towardtechnology and the effect of iPads onstudent achievement [dissertation].Evanston, IL: Northwestern University;2016

43. Renaissance Learning. STAR EarlyLiteracy Accessment. Available at:https://www.renaissance.com/products/assessment/star-360/star-early-literacy-skills/. AccessedNovember 1, 2019

44. Brown M, Harmon MT. iPad interventionwith at-risk preschoolers: mobiletechnology in the classroom. J LitTechnol. 2013;14(2):56–78

45. Invernezzi M, Sullivan A, Meier J.Phonological Awareness LiteracyScreening - Prekindergarten.Charlottesville, VA: University Press;2001

46. Cubelic CC, Larwin KH. The use of iPadtechnology in the kindergartenclassroom: a quasi-experimentalinvestigation of the impact on earlyliteracy skills. Compr J Educ Res. 2014;2(4):47–59

47. Good RH, Kaminski RA. DynamicIndicators of Basic Early Literacy SkillsAdministration and Scoring Guide, 6thed. Eugene, OR: Institute for theDevelopment of EducationalAchievement; 2003

48. Gilliver M, Cupples L, Ching TYC, Leigh G,Gunnourie M. Developing sound skillsfor reading: teaching phonologicalawareness to preschoolers withhearing loss. J Deaf Stud Deaf Educ.2016;21(3):268–279

49. Neumann MM. Using tablets and appsto enhance emergent literacy skills inyoung children. Early Child Res Q. 2018;42:239–246

50. Patchan MM, Puranik CS. Using tabletcomputers to teach preschool childrento write letters: exploring the impact ofextrinsic and intrinsic feedback.Comput Educ. 2016;102:128–137

51. Russo-Johnson C, Troseth G, Duncan C,Mesghina A. All tapped out: touchscreeninteractivity and young children’s wordlearning. Front Psychol. 2017;8:578

52. Teepe RC, Molenaar I, Verhoeven L.Technology-enhanced storytellingstimulating parent–child interactionand preschool children’s vocabularyknowledge. J Comput Assist Learn.2017;33(2):123–136

53. Vatalaro A, Mcdonald A, Debbie C,Vaughn LH, Barnes AC. A quasi-experiment examining expressive andreceptive vocabulary knowledge ofpreschool head start children usingmobile media apps. Early Child Educ J.2018;46(4):451–466

54. Dunn L, Dunn D. Peabody PictureVocabulary Test. 4th Ed. Minneapolis,MN: NCS Pearson; 2007

PEDIATRICS Volume 145, number 1, January 2020 13 by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 14: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

55. Williams K. EVT-2. Minneapolis, MN:Pearson Assessments; 2007

56. Walter‐Laager C, Brandenberg K,Tinguely L, Schwarz J, Pfiffner MR,Moschner B. Media‐assisted languagelearning for young children: effects ofa word‐learning app on the vocabularyacquisition of two‐year‐olds. Br J EducTechnol. 2017;48(4):1062–1072

57. Grimm H, Aktas M, Frevert S. SETK-2.Göttingen, Germany: Hogrefe; 2000

58. Kwok K, Ghrear S, Li V, Haddock T,Coleman P, Birch SAJ. Children canlearn new facts equally well frominteractive media versus face to faceinstruction. Front Psychol. 2016;7:1603

59. Putnam MM, Richmond EM, Brunick KL,Wright CA, Calvert SL. Influence ofa character-based app on children’slearning of nutritional information:should apps be served with a side ofmedia characters? Games Health J.2018;7(2):121–126

60. Bebell D, Pedulla J. A quantitativeinvestigation into the impacts of 1:1iPads on early learners’ ELA and mathachievement. J Inf Technol Educ InnovPract. 2015;14:191–215

61. NWEA. Children’s Progress AcademicAssessment. 2009. Available at: https://www.nwea.org/cpaa/. AccessedNovember 12, 2019

62. Schroeder EL, Kirkorian HL. Whenseeing is better than doing:preschoolers’ transfer of STEM skillsusing touchscreen games. FrontPsychol. 2016;7:1377

63. Choi K, Kirkorian HL. Touch or watch tolearn? Toddlers’ object retrieval usingcontingent and noncontingent video.Psychol Sci. 2016;27(5):726–736

64. Huber B, Tarasuik J, Antoniou MN,Garrett C, Bowe SJ, Kaufman J. Youngchildren’s transfer of learning froma touchscreen device. Comput HumanBehav. 2016;56:56–64

65. Huber B, Yeates M, Meyer D,Fleckhammer L, Kaufman J. The effectsof screen media content on young

children’s executive functioning. J ExpChild Psychol. 2018;170:72–85

66. Tarasuik J, Demaria A, Kaufman J.Transfer of problem solving skills fromtouchscreen to 3D model by 3- to 6-year-olds. Front Psychol. 2017;8:1586

67. Esposito M, Sloan J, Tancredi A, et al.Using tablet applications for childrenwith autism to increase their cognitiveand social skills. J Spec Educ Technol.2017;32(4):199–209

68. Fenson L, Dale P, Reznick J, et al.Macarthur CommunicativeDevelopment Inventory: User’s Guideand Technical Manual. San Diego, CA:Singular Publishing Company; 1993

69. Wetherby A, Prizant B. Communicationand Symbolic Behavior Scales:Developmental Profile. Baltimore, MD:Paul H. Brooks Pub.; 2002

70. Grzadzinski R, Carr T, Colombi C, et al.Measuring changes in socialcommunication behaviors: preliminarydevelopment of the Brief Observation ofSocial Communication Change (BOSCC).J Autism Dev Disord. 2016;46(7):2464–2479

71. Lord C, Rutter M, DiLavore P, Risi S.Autism Diagnostic ObservationSchedule, Second Edition (ADOS-2). LosAngeles, CA: WPS; 2012

72. Whitehouse AJO, Granich J, Alvares G,et al. A randomised controlled trial ofan iPad-based application tocomplement early behaviouralintervention in Autism SpectrumDisorder. J Child Psychol Psychiatry.2017;58(9):1042–1052

73. Bodfish J, Symons F, Lewis M. TheRepetitive Behaviour Scale: TestManual. Morganton, NC: WesternCalifornia Center Research Reports;1998

74. Peters-Scheffer N, Didden R, Green V,et al. The behavior flexibility ratingscale-revised (BFRS-R): Factor analysis,internal consistency, inter-rater andintra-rater reliability, and convergent

validity. Res Dev Disabil. 2008;29(5):398–407

75. Rimland B, Edelson S. Autism TreatmentEvaluation Checklist: StatisticalAnalyses. San Diego, CA: AutismResearch Institute; 2000

76. Sparrow S, Cicchetti D, Balla D.Vineland Adaptive Behavior Scales, 2ndEdition Manual. Minneapolis, MN: NCSPearson; 2005

77. Neumann MM. Using tablets and appsto enhance emergent literacy skills inyoung children. Early Child Res Q. 2018;42:239–246

78. Blackwell CK. Technology Use in EarlyChildhood Education: InvestigatingTeacher Access and Attitudes TowardTechnology and the Effect of IPads onStudent Achievement [dissertation]. Vol76. Evanston, IL: NorthwesternUniversity; 2016

79. Alzrayer N, Banda DR, Koul RK. Use ofiPad/iPods with individuals with autismand other developmental disabilities:a meta-analysis of communicationinterventions. Rev J Autism Dev Disord.2014;1(3):179–191

80. Zack E, Gerhardstein P, Meltzoff AN, BarrR. 15-month-olds’ transfer of learningbetween touch screen and real-worlddisplays: language cues and cognitiveloads. Scand J Psychol. 2013;54(1):20–25

81. 42matters. Google Play Store stats.Available at: https://42matters.com/stats. Accessed July 12, 2019

82. Strouse GA, O’Doherty K, Troseth GL.Effective coviewing: preschoolers’learning from video after a dialogicquestioning intervention. Dev Psychol.2013;49(12):2368–2382

83. Griffith SF, Arnold DH. Home learning inthe new mobile age: parent–childinteractions during joint play witheducational apps in the US. J ChildMedia. 2019;13(1):1–19

84. Rosenthal R. The “file drawer problem”and tolerance for null results. PsycholBull. 1979;86(3):638–641

14 GRIFFITH et al by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 15: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

DOI: 10.1542/peds.2019-1579 originally published online December 23, 2019; 2020;145;Pediatrics 

BagnerShayl F. Griffith, Mary B. Hagan, Perrine Heymann, Brynna H. Heflin and Daniel M.

Apps As Learning Tools: A Systematic Review

ServicesUpdated Information &

http://pediatrics.aappublications.org/content/145/1/e20191579including high resolution figures, can be found at:

Referenceshttp://pediatrics.aappublications.org/content/145/1/e20191579#BIBLThis article cites 59 articles, 6 of which you can access for free at:

Subspecialty Collections

http://www.aappublications.org/cgi/collection/screen_time_subScreen Timehttp://www.aappublications.org/cgi/collection/media_subMediafollowing collection(s): This article, along with others on similar topics, appears in the

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtmlin its entirety can be found online at: Information about reproducing this article in parts (figures, tables) or

Reprintshttp://www.aappublications.org/site/misc/reprints.xhtmlInformation about ordering reprints can be found online:

by guest on March 16, 2021www.aappublications.org/newsDownloaded from

Page 16: Apps As Learning Tools: A Systematic ReviewApple and Android app stores opened), and written in languages other than English were excluded. The database searches were conducted on

DOI: 10.1542/peds.2019-1579 originally published online December 23, 2019; 2020;145;Pediatrics 

BagnerShayl F. Griffith, Mary B. Hagan, Perrine Heymann, Brynna H. Heflin and Daniel M.

Apps As Learning Tools: A Systematic Review

http://pediatrics.aappublications.org/content/145/1/e20191579located on the World Wide Web at:

The online version of this article, along with updated information and services, is

by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397. the American Academy of Pediatrics, 345 Park Avenue, Itasca, Illinois, 60143. Copyright © 2020has been published continuously since 1948. Pediatrics is owned, published, and trademarked by Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it

by guest on March 16, 2021www.aappublications.org/newsDownloaded from