Device access
88% of public schools
Reported a device for every student through a 1:1 computing program in 2024–25.
NCES school surveyWhat’s allowed? What’s restricted? What are schools required to teach?
Explore state and district policies by grade, compare requirements, and see what could change next.
October 2026 editionUpdated Oct 8, 2026Scope & limitations
Device access, limits on use and required learning
Limits on use
9 states with findings
Adopted limits on school-issued devices, instructional screen time or student AI. Personal-phone rules are separate.
Topics and states
School-issued devices: 5
Arizona, California, Iowa, Tennessee, Utah
Screen time: 6
Alabama, Connecticut, Iowa, Minnesota, Tennessee, Utah
Student AI: 1
Oklahoma
Topic counts overlap. Each state counts once in the total.
Required learning
21 states with findings
Require instruction or demonstrated competency in computer science, AI or digital literacy for at least one grade or cohort.
Subjects and states
Computer science: 16
Alabama, Arkansas, Iowa, Idaho, Indiana, Louisiana, North Carolina, North Dakota, Nebraska, Nevada, Rhode Island, South Carolina, Tennessee, Texas, Virginia, West Virginia
AI literacy: 0
No matching reviewed finding.
Digital literacy: 11
Alabama, Idaho, Illinois, Kentucky, Maine, New Hampshire, New Jersey, Nevada, Rhode Island, Texas, Virginia
Subjects overlap. Standards alone, course offerings and optional resources are excluded.
Where do limits and learning overlap? 3 of the 9 states with documented limits also have a documented learning requirement. For the other 6, this tracker has no matching statewide learning requirement recorded. Requirements may cover different grades and subjects.
Dates, coverage and what this comparison means
The device figure measures schools in 2024–25. The policy figures count states in the research release of Oct 8, 2026, including adopted requirements starting later. They exclude proposals and D.C.
Device access does not measure classroom use. Required learning does not establish what students have learned, and computer science or digital literacy requirements do not automatically include AI literacy. These figures do not form a time series or establish that device adoption caused restrictions.
Limits and required learning: Alabama, Iowa, Tennessee.
Limits found; matching required learning not documented: Arizona, California, Connecticut, Minnesota, Oklahoma, Utah. Missing findings are research gaps, not proof that schools do not teach these subjects.
Start with the national picture
What changed in 2026?The map compares limits on use with requirements to teach.
All K–12 · Published findings; coverage is partialLimits and learning · All K–12
Passed, including requirements starting laterSmall states & D.C.
Limits + required learning documented21Limits found; learning not documented24No matching policies documented6
1 Number of selected districts with matching evidence. State fill describes statewide evidence. District markers describe local findings separately.
Select a state for grades, dates and sources. Legend totals include D.C. A missing finding is a research gap, not proof that no policy exists. Learning means required CS, AI or digital-literacy instruction or demonstrated competency; standards and course offerings are separate. Coverage & method
State rules and learning policies are the focus. Device provision and approved AI tools are documented only where verified. Gray means an evidence gap, not no policy. Check coverage
How to read the map
A statewide action may cover a specific grade, school type or cohort. The overview includes all affected K–12 grades; details show the actual scope. A current action takes priority over an upcoming action. District evidence is a selected sample and is never treated as a statewide rule.
Counts include adopted future duties. “Upcoming only” means there is no current matching state requirement in the published findings; a state with both current and upcoming requirements counts once under current. The combined view adds personal-device rules to required student instruction. The school-technology view covers instructional screens, school-issued devices and student AI access conditions, including requirements for local policies. Required learning covers instruction or demonstrated competency in computer science, AI or digital literacy; it excludes course offerings, standards alone and optional resources.
Learning categories distinguish instruction, course availability, standards alignment and resources. “Varies by grade” means the personal-device rules differ across grade bands. Proposals appear separately from adopted requirements. Policy details distinguish documented funding from related support programs.
Map geometry: US Atlas / US Census. Attribution
Definitions and national totals: phones, instructional technology and required learning
Personal-device rules
43 states with findings
Statewide restrictions or duties to set local rules; includes adopted future starts.
Which states count?
Alabama, Alaska, Arizona, Arkansas, California, Colorado, Delaware, Florida, Georgia, Hawaii, Idaho, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Maine, Maryland, Michigan, Minnesota, Missouri, Nebraska, Nevada, New Hampshire, New Jersey, New Mexico, New York, North Carolina, North Dakota, Ohio, Oklahoma, Oregon, Rhode Island, South Carolina, Tennessee, Texas, Utah, Vermont, Virginia, West Virginia, Wisconsin, Wyoming.
50-state denominator; D.C. has its own profile. Reviewed findings in this collection, not a certification of exhaustive coverage. Categories overlap.
Instructional-tech limits
10 states with findings
School devices, instructional screens or student AI. Separate from ordinary phone rules.
Which states count?
Alabama, Arizona, California, Connecticut, Florida, Iowa, Minnesota, Oklahoma, Tennessee, Utah.
50-state denominator; D.C. has its own profile. Reviewed findings in this collection, not a certification of exhaustive coverage. Categories overlap.
Required learning
21 states with findings
CS, AI or digital-literacy instruction or demonstrated competency for a covered grade or cohort; includes future starts.
Which states count?
Alabama, Arkansas, Iowa, Idaho, Illinois, Indiana, Kentucky, Louisiana, Maine, North Carolina, North Dakota, Nebraska, New Hampshire, New Jersey, Nevada, Rhode Island, South Carolina, Tennessee, Texas, Virginia, West Virginia.
50-state denominator; D.C. has its own profile. Reviewed findings in this collection, not a certification of exhaustive coverage. Categories overlap.
National, all-grade totals. Different questions, different counts. Phone restrictions do not automatically limit school-issued technology. Learning requirements can cover specific grades or cohorts. Definitions & counting method
Selected district examples · local detail
What the detail reveals · selected district examples
Access and understanding are separate decisions.
These districts show why “allows AI” or “bans AI” cannot tell the whole story.
Fairfax County, VA
Teach about AI without student chatbot use
Elementary generative-AI use is prohibited on district devices; secondary exceptions need written authorization.
K–12: one digital-literacy lesson each quarter. Grade 2 examines how data affects AI responses; grade 8 evaluates bias in teacher-provided output. Students do not use AI tools in these lessons.
See grades, exceptions and sources →Sources checked 2026-10-05
New York City, NY
Restricted access, required AI learning
Student-facing generative AI is prohibited through grade 8. High-school access is limited to approved programs.
Grades 9–12: two required 45-minute modules cover how AI works, privacy, bias and school use.
See grades, exceptions and sources →Sources checked 2026-10-04
Jeffco, CO
Approved tools, with classroom conditions
Grades 6–12 have Gemini access, but teachers decide assignment use. Other named chatbots are blocked. Device access and digital-citizenship lessons have their own grade-specific rules.
See grades, exceptions and sources →Source checked 2026-10-05
Compare Fairfax and NYC side by side · Find another district
Examples, not national prevalence. Required teaching is not evidence of classroom delivery or student understanding.
More research: learning, policy history and public views
Teaching requirements
What must schools teach?
Check whether schools must teach a subject, offer a course, or follow learning standards.
What does the policy require?
Grade view: All K–12. A grade-band finding may apply to only some grades; each source gives the precise scope.
Timing: adopted, including future starts. Counts show reviewed findings; a missing state may still have a requirement. Coverage & gaps.
Required learning includes instruction and graduation competency routes. A competency requirement specifies what students must demonstrate; it does not necessarily prescribe a separate course.
Where is computer science documented?
Required learning · All K–12 · Adopted, including future starts. National view; state policies only.
Current findingFuture start onlyNot established in this collection
Gray means a research gap, not that a state has no requirement. A finding may cover only particular grades or cohorts. District examples remain separate in the table below.
Break down by subject · overlapping counts
Explore graduation requirements with CS content · 13 verified states
Includes adopted future cohorts where selected. Requirements may allow proficiency routes or approved embedded instruction. Idaho is included because its digital-literacy credit explicitly requires algorithms and programming. This is broader than a count of courses titled computer science.
See the 13 states and their start dates
- Alabama
Law effective October 1, 2026; graduation provision begins with class of 2032. Detailed approved coursework and embedded-standards guidance is to follow. Requirement and source
- Arkansas
Applies to students entering grade 9 in 2022–23 and later; the first standard graduating cohort is 2026. Requirement and source
- Idaho
Rule effective July 1, 2025; one-credit requirement applies to students graduating after January 1, 2028. Requirement and source
- Indiana
First graduating cohort: class of 2029. This does not apply to all current graduating students. Requirement and source
- Louisiana
Ninth-grade entrants in 2027–28 and later. Requirement and source
- Nebraska
Graduation requirement begins in academic year 2027–28. Requirement and source
- Nevada
Current half-credit graduation requirement. The 2023–24 guidance supersedes earlier versions; a qualifying high-school-level course may be completed in middle school. Requirement and source
- North Carolina
Graduation requirement begins with grade 9 entrants in 2024–25. SL2026-41 §7.53 extends implementation-delay notifications for Wake/Johnston only through 2026–27; affected units must submit a course-offering plan by January 1, 2027. Older DPI pages stating no 2026–27 waiver do not reflect this later local exception. Requirement and source
- North Dakota
Graduation rule: 2025–26, after July 31, 2025. Board-approved integration-plan alternative remains distinct. Requirement and source
- Rhode Island
Applies beginning with the graduating class of 2028. Requirement and source
- South Carolina
Current policy; the cited source does not provide a confirmed start date. Requirement and source
- Tennessee
Incoming freshmen from fall 2024; first graduating cohort is the class of 2028. Requirement and source
- West Virginia
Beginning with grade 9 entrants in 2027–28, one full CS credit is required. May be completed in grades 8–12; approved courses and Policy 2510 implementation are governed by WVDE/WVBE guidance. Requirement and source
Show graduation requirements on the map
Cross-check: Code.org’s 2025 report reports 12 states using data finalized December 1, 2025. Our broader definition includes Idaho’s digital-literacy requirement. This difference is a classification correction, not an additional law passed in 2026. Nationwide amendment review remains open.
How we count learning requirements
District counts reflect reviewed local requirements, not every district covered by state law. Course access duties include offering alternatives: Kansas permits an implementation plan, which does not prove that a course is available now.
Each state counts once per subject and requirement type. D.C. is displayed in its own profile, not counted as a state. Subjects overlap. Required instruction or demonstrated competency can apply to specific grades or graduation cohorts; it does not mean every child takes a separate course. A required offering does not mean every pupil takes it. Standards alone do not establish a must-take course or actual classroom delivery.
North Carolina’s AI-literacy standards and CS course alignment appear under standards, not as a separate AI course for all pupils.
Explore selected district learning examples
National examples: what learning looks like in district documents
Kansas · Learning opportunities required
Wichita Public Schools
Requires support for students and staff in developing AI literacy, including foundational concepts and when and how to use AI. Supports teachers in adapting instruction. It does not prescribe a separate course or graduation credit.
Read the commitment & sourceAlabama · Learning opportunities required
Talladega County Schools
Directs teaching about AI limitations, bias, critical evaluation and citation, with subject-area integration and staff professional learning. This is a district commitment; classroom delivery has not been independently measured.
Read the commitment & sourceIllinois · Student instruction required
Chicago Public Schools
CPS requires one computer-science credit for high-school graduation. Grade 9 is the recommended placement, not the only permitted grade.
Read the commitment & sourceNew York · Optional support / resources
New York City Public Schools
CS4All supports schools in expanding computer-science experiences and computational thinking. The program page does not establish that every student currently receives a specified course or number of hours.
Read the commitment & sourceRequiring every student to learn a subject is different from making a course available. Neither alone shows what happens in classrooms. State and district briefs identify which is documented.
Read about learning & student agencyUpcoming changes
What could change next?
See requirements that have passed but start later, followed by proposals that are still being considered.
Proposed: not yet required
United StatesIntroduced / referred to committee
S. 5620 · UNPLUGGED Schools Grant Act of 2026
Would fund state policies restricting student possession or use of personal electronic devices during school hours. Eligible policies would cover public elementary and secondary schools, include health and disability exceptions, and be developed with district and community input. School-authorized instructional laptops and tablets meeting specified safeguards are excluded.
What happens next: No next official date announced.
Status checked Oct 8, 2026
Status, scope & official sourceNorth CarolinaFiled / committee
SB 948 · Screen Free Schools Act
Would prohibit most K–5 classroom device use, limit daily device time in grades 6–12 and restrict individually assigned or take-home school devices by grade band.
What happens next: No next official date announced.
Status checked Oct 8, 2026
Status, scope & official sourceKansasDiscussion / review
Instructional technology policy · State Board draft
The draft would require local technology plans to specify devices and shared/individual assignment, permitted activities, take-home conditions and AI use by grade or grade band. It would require publication and submission of plans by the end of 2026–27, with parental acknowledgment before individual device access from 2027–28.
What happens next: Regular board meeting; no item-specific vote verified: 2026-10-13
Status checked Oct 8, 2026
Status, scope & official sourceAlready adopted measures with future start dates are listed separately in the policy watch.
See what is already scheduledSupporting context / Over time
School technology policy is moving in more than one direction.
Follow dated decisions: tighter limits, wider access and new conditions on use. These examples do not measure national growth.
How we arrived here: access expanded before today’s policy debate.
Before the pandemic → March 2021
A device for every student became widespread during the pandemic
| Grade level | Before* | March 2021 |
|---|---|---|
| Elementary | About 40% | 84% |
| Middle & high | About two-thirds | 90% |
*Leaders recalled pre-pandemic provision. These are survey respondent percentages, not percentages of all U.S. schools.
EdWeek Research Center · published May 17, 20222024–25 · Later national context
Public schools reporting a 1:1 computing program
88%
U.S. public schools reporting a program that provides every pupil a school-issued device. This does not measure time on screens or require take-home use.
NCES · published February 19, 2025Earlier NCES comparison: devices assigned to take home
Public-school principals reported devices assigned to every student to take home at 23% before the pandemic (2019–20), rising to 45% in spring 2020. This narrower measure excludes other in-school access.
NCES technology-support comparisonDifferent surveys, populations and measures: these figures are not one continuous trend. Measures, access notes & data. They cannot establish when states adopted 1:1 policies or whether device access caused later restrictions.
Sep 21, 2026
Restriction / duty adoptedSeattle Public Schools
Moves K–2 from assigned one-to-one iPads to shared classroom devices, with a general daily screen-time range of zero to 30 minutes.
The district letter and FAQ give January 4, 2027 for shared classroom devices. A Key Takeaways bullet instead says January 1; confirm the precise start with the district. The change remains upcoming.
Read the original sourceSep 10, 2026
Restriction / duty adoptedNew York City Public Schools
Moratorium on student-facing GenAI in 2K–8; no routine 1:1 screen time in 2K–2; older-grade daily limits are recommendations.
2026–27 school year; September 10 school start is the implementation reference, not a verified board adoption date.
Read the original sourceJul 16, 2026
Restriction / duty adoptedFairfax County Public Schools
Ends FCPS-issued tablets and laptops in pre-K and kindergarten, makes student generative AI unavailable in elementary school and requires principal authorization for secondary use pending policy.
Board action applies beginning in fall 2026.
Read the original sourceThese are selected decisions, not a measured national growth rate. Historical gaps prevent a reliable 2024–2026 total.
Explore the full timeline & earlier evidenceViews & experiences
What do different groups want from school technology?
Explore reported views on limits, use and learning. People can support boundaries and technology education at the same time.
These are selected studies, not a single measure of public opinion. Audience filters identify who was studied; each finding names whose response it reports. These studies do not follow your state or grade filters.
8 studies shown. Publication year and fieldwork year can differ. Mixed-audience studies may report different questions for different groups.
Survey · Parents · Published 2026-09-28
Parents on school AI, screens and instruction
National Parents Union / Echelon Insights · Source checked 2026-10-05
71%Support banning student chatbot use during the school day
85%Support schools teaching AI safety
77%Support teaching effective AI use
78%Prefer limiting K–2 learning screen time; 11% prefer a ban
58%Prefer laptops on carts for designated use; 38% prefer desks all day
Useful for: Pair age-specific limits with the separate question of teaching safe and informed use.
Read with care: Responses to separate questions cannot identify which individual parents support both. Different May wording prevents a simple trend. These are national views, not local district opinion or evidence of learning effects.
Sample, questions and methods
- Fieldwork
- September 7–11, 2026
- Who was surveyed
- 1,528 U.S. parents of K–12 public-school pupils
- Question / measure
- Separate policy-support items and screen/device questions. The publisher provides wording and response tables. Support combines strong and somewhat support where applicable.
- Method
- Online non-probability sample from Cint/Lucid; quotas and demographic weighting. Publisher reports ±2.6 points using a random-sample calculation; this does not capture all non-probability sampling error.
- Sponsor / research organization
- Commissioned by National Parents Union, an advocacy organization.
Survey · Teachers · Published 2026-09-16
Teachers’ views of classroom learning devices
Gallup / Walton Family Foundation / Joan Ganz Cooney Center · Source checked 2026-10-05
61%Want less student time on learning devices
59%Agree the digital tools they use help students learn effectively
34%Strongly support a daily maximum for device time
Useful for: Ask about purpose and classroom oversight alongside screen duration.
Read with care: Teacher perceptions and self-reported use do not establish causal effects. Learning devices are distinct from personal phones.
Sample, questions and methods
- Fieldwork
- May 6–29, 2026
- Who was surveyed
- 2,101 U.S. K–12 public-school teachers
- Question / measure
- Different items measure preferred device time, perceived effectiveness and strong support for a daily maximum. “Strongly support” is not total support.
- Method
- National probability-based teacher panel; see the linked study for weighting and questionnaire.
- Sponsor / research organization
- Walton Family Foundation–Gallup study with the Joan Ganz Cooney Center.
Survey · Students / Parents · Published 2026-02-24
Teen chatbot use and parents’ comfort with it
Pew Research Center · Source checked 2026-10-05
54%Of teens report ever using chatbots for schoolwork help
18%Of parents would accept chatbot emotional support for their teen
36%Of teens expect AI to affect their own lives positively over the next 20 years
15%Of teens expect AI to affect their own lives negatively over the next 20 years
Useful for: Distinguish homework assistance, information seeking and companionship.
Read with care: Fieldwork was in 2025 despite the 2026 publication. Comfort differs by use case. Neither statistic measures support for a school ban.
Sample, questions and methods
- Fieldwork
- September 25–October 9, 2025
- Who was surveyed
- 1,458 U.S. teens ages 13–17 and their parents
- Question / measure
- Separate teen-use, parent-acceptability and future-impact questions. Positive and negative personal-impact answers do not include the equally-positive-and-negative or unsure groups. Original wording and tables are linked.
- Method
- Online Ipsos KnowledgePanel recruitment through parents; weighted national estimates. Overall margin of error ±3.3 percentage points for each population.
- Sponsor / research organization
- Published by Pew Research Center; survey collection by Ipsos.
Original research · Questionnaire / methods source · Parent findings
Survey synthesis · Students / Parents / Teachers / Administrators · Published 2025-09-30
How students, parents and district leaders view AI
RAND · Source checked 2026-10-05
61% / 22%Parents / district leaders agreeing that more AI use could harm student critical thinking
35%District leaders reporting student AI training in spring 2025
Useful for: Surface differences between leadership assumptions and family/student concerns.
Read with care: These are beliefs and reports, not measured cognitive harm or verified policy coverage. Different populations and dates are not a single matched trend.
Sample, questions and methods
- Fieldwork
- October 2024–June 2025 across eight survey samples
- Who was surveyed
- Eight samples, including 984 parents of 12–17-year-olds and 232 district leaders; denominators differ by item
- Question / measure
- Item wording appears in figure notes; the Data Sources and Methods section lists eight samples. The parent–leader comparison uses different samples.
- Method
- National survey panels plus a small, nonrepresentative leader-interview sample. District sample weighting cannot eliminate unobserved selection differences.
- Sponsor / research organization
- See report acknowledgments for funding and partner details.
Survey · Teachers · Published 2026-05-26
Teachers’ reported access to AI guidance
Gallup / Walton Family Foundation · Source checked 2026-10-05
18%Report formal guidance on AI work tasks
48%Report informal guidance only
34%Report no guidance across the measured tasks
Useful for: Check whether adopted rules reach the educators expected to use them.
Read with care: A teacher reporting no guidance does not prove the district has no policy. These figures concern staff work, not student AI permission.
Sample, questions and methods
- Fieldwork
- February 9–March 2, 2026
- Who was surveyed
- 2,069 U.S. K–12 public-school teachers
- Question / measure
- Formal means written policy or official guidance; informal includes discussions and shared norms. Ten work tasks were measured.
- Method
- National probability-based teacher panel.
- Sponsor / research organization
- Walton Family Foundation–Gallup research partnership.
Survey · Students · Published 2026-03-17
Student AI use and concerns about thinking
RAND · Source checked 2026-10-05
67%Agree greater AI schoolwork use could harm critical thinking
Useful for: Discuss when AI supports thinking and when it replaces students’ work.
Read with care: Not a K–12-only estimate and not a measured learning effect. Do not attach this percentage to all schoolchildren.
Sample, questions and methods
- Fieldwork
- December 2025; selected comparisons with February and May 2025
- Who was surveyed
- 1,214 youth ages 12–29; includes middle school, high school and college
- Question / measure
- Agreement with a belief statement about AI and thinking. The overall result includes college students; use grade-specific tables for K–12 analysis.
- Method
- RAND American Youth Panel; the report provides grade-specific results and methods.
- Sponsor / research organization
- Gates Foundation-funded research.
Survey · Parents / Students / Teachers / Administrators · Published 2020-09-29
Historical baseline: views of computer-science learning
Gallup / Google · Source checked 2026-10-05
69% / 40%Parents / students rating CS learning important or very important
Useful for: Ask whether learners see the same purpose in CS that adults do.
Read with care: Historical, pre-pandemic evidence—not current 2026 sentiment. Different respondent groups and recruitment routes need care.
Sample, questions and methods
- Fieldwork
- Late January–early March 2020
- Who was surveyed
- 2,092 parents, 1,402 grade 7–12 students, 979 teachers, 1,521 principals and 1,479 superintendents
- Question / measure
- Respondents were given a definition of CS that distinguishes computing concepts and programming from routine device use.
- Method
- Online Gallup Panel samples for families and teachers; a sample provider for school leaders. Methods and weighting are in the report.
- Sponsor / research organization
- Google-supported Gallup research.
Provider implementation report · Teachers / Administrators · Published 2024-06
How educators implement digital-citizenship lessons
Common Sense Media · Source checked 2026-10-05
49%Of sampled implementers report covering all six curriculum topic areas
Useful for: Check whether a curriculum listing translates into a complete learning sequence.
Read with care: Provider research and self-selected users. This statistic measures reported coverage, not causal effectiveness or national teaching prevalence.
Sample, questions and methods
- Fieldwork
- May 2023 implementation survey; report also combines other evidence
- Who was surveyed
- Implementation analysis of 1,749 experienced curriculum users within 2,097 educator respondents
- Question / measure
- Implementation section, pages 10–12: educator-reported lesson coverage. Do not combine with the separate quiz or annual-survey samples.
- Method
- Survey of registered educators, conducted with University of New Hampshire researchers. Selected curriculum users, not a representative national teacher sample.
- Sponsor / research organization
- Report published by the curriculum provider.
Using this research
Use this in your next planning meeting.
Start with your state, check the affected grades and dates, then follow the evidence to the original policy.
Take these questions to your next planning meeting
- Which grades and students does the rule cover, and when does it take effect?
- Does the learning policy require instruction, an available course, or standards only?
- Who will teach it, with what time, preparation and evidence of delivery?
- Which proposals should we monitor before committing budget or changing practice?
What this tracker can tell you
Documented rules, learning requirements, selected local approaches and proposed changes. Every policy brief links to its source and explains the type of requirement.
Compare states · Search & export policiesWhat the tracker cannot answer yet
Some topics, district histories and superseding policies remain incomplete. A missing finding is a research gap. Published requirements do not establish delivery in every classroom.
See scope and limitations · Suggest a correctionDates and accountability
Policy baseline: Oct 4, 2026. Latest release: Oct 8, 2026. Each published finding shows its official source and review date.
How changes are labelled · Download published dataTeaching resources & complementary research
Moving from policy to teaching
A small starting list, not a curriculum ranking or a guarantee of compliance. Check grade fit, accessibility, privacy and the actual learning goals.
- Day of AI : AI literacy curricula and professional learning.
- Code.org / CodeAI : AI curriculum and teacher learning.
Use this alongside specialist sources
This tracker brings limits and learning duties into one comparison. It does not replace these organisations’ research or claim to contain every finding they publish.
- State of AI & Computer Science Education : policy, access and participation data.
- FutureEd : AI-in-education legislation.
- TeachAI : AI guidance and policy resources.