260 Online Learning Statistics for 2026

How many US college students take online courses?

More than half take at least one. Online course-taking at US colleges fell back after the fall 2020 pandemic spike but has stayed well above where it was before, on federal enrolment counts.

View data table

Share of students at US degree-granting institutions taking any distance education course, fall 2017 to fall 2024. Fall 2020 is the pandemic year and fall 2024 is provisional. Each year comes from the release for that year; the link is to the latest one.

Period	Value
2017	33.5%
2018	35.3%
2019	37.0%
2020	74.7%
2021	60.0%
2022	54.2%
2023	54.2%
2024	54.7%

• In fall 2024, 10,761,754 of the 19,659,633 students at US degree-granting postsecondary institutions took at least one distance education course. The fall 2024 data are provisional. (NCES (Digest 2025, Table 311.15), Mar 2026)
• In the same term, 5,547,429 students (28.2%) took some but not all of their courses at a distance. (NCES (Digest 2025, Table 311.15), Mar 2026)
• Before the pandemic, in fall 2017, 33.5% of US students took any distance education course. (NCES (Digest 2019, Table 311.15), Dec 2019)
• The share was 35.3% in fall 2018 and 37.0% in fall 2019, then 74.7% in fall 2020, during the pandemic. Each year comes from that year's edition of the table; the link is to the latest one. (NCES (Digest 2025, Table 311.15), Mar 2026)
• It fell to 60.0% in fall 2021 and was 54.2% in both fall 2022 and fall 2023. Each year comes from that year's edition of the table; the link is to the latest one. (NCES (Digest 2025, Table 311.15), Mar 2026)
• By level, 54.6% of US undergraduates and 55.3% of postbaccalaureate (graduate) students took any distance education course in fall 2024. (NCES (Digest 2025, Table 311.15), Mar 2026)
• At 4-year institutions 54.1% of students took any distance education course in fall 2024; at 2-year institutions 56.9% did. (NCES (Digest 2025, Table 311.15), Mar 2026)
• At public 2-year colleges, 57.5% of students took any distance education course in fall 2024 and 29.2% studied exclusively online. (NCES (Digest 2025, Table 311.15), Mar 2026)
• At public institutions 55.2% of students took any distance course in fall 2024 and at private institutions 53.4% did: 46.2% at private nonprofits and 81.9% at private for-profits. (NCES (Digest 2025, Table 311.15), Mar 2026)
• At private for-profit 4-year institutions, 90.5% of students took any distance education course in fall 2024. (NCES (Digest 2025, Table 311.15), Mar 2026)

NCES counts students enrolled at Title IV degree-granting institutions at the fall census (IPEDS). A distance education course is one delivered online or remotely on the IPEDS definition. These are enrolment counts, not survey shares of adults, and are not comparable with the survey figures elsewhere on this page.

How many US college students study entirely online?

A large minority, and far more at for-profit colleges and among graduate students than elsewhere. The federal count covers all Title IV institutions; the NC-SARA count covers its member institutions only, so the two are not compared.

View data table

Share of students at US degree-granting institutions enrolled exclusively in distance education courses, fall 2017 to fall 2024. Fall 2020 is the pandemic year and fall 2024 is provisional. Each year comes from the release for that year; the link is to the latest one.

Period	Value
2017	15.7%
2018	16.6%
2019	17.5%
2020	46.7%
2021	30.3%
2022	26.4%
2023	26.2%
2024	26.5%

• In fall 2017, 15.7% of US students were enrolled exclusively in distance education courses. (NCES (Digest 2019, Table 311.15), Dec 2019)
• The exclusively distance share was 16.6% in fall 2018 and 17.5% in fall 2019, then 46.7% in fall 2020, during the pandemic. Each year comes from that year's edition of the table; the link is to the latest one. (NCES (Digest 2025, Table 311.15), Mar 2026)
• It was 30.3% in fall 2021, 26.4% in fall 2022 and 26.2% in fall 2023. Each year comes from that year's edition of the table; the link is to the latest one. (NCES (Digest 2025, Table 311.15), Mar 2026)
• At US public institutions, 3,300,299 students (23.0%) were exclusively online in fall 2024; at private institutions, 1,914,026 (35.9%) were. (NCES (Digest 2025, Table 311.15), Mar 2026)
• The exclusively online share in fall 2024 was 28.7% at private nonprofit institutions and 64.4% at private for-profit institutions. (NCES (Digest 2025, Table 311.15), Mar 2026)
• 23.7% of US undergraduates (3,888,681) were exclusively online in fall 2024, against 40.5% of postbaccalaureate students (1,325,644). (NCES (Digest 2025, Table 311.15), Mar 2026)
• By type of institution, 25.9% of students at 4-year institutions (3,876,823) and 28.5% at 2-year institutions (1,337,502) were exclusively online in fall 2024. (NCES (Digest 2025, Table 311.15), Mar 2026)
• At private for-profit 4-year institutions, 76.1% of students were exclusively online in fall 2024. (NCES (Digest 2025, Table 311.15), Mar 2026)
• 143 US degree-granting institutions, 3.7% of 3,894, were primarily online in fall 2022. NCES defines these as institutions with more than 90 percent of students enrolled exclusively in distance education courses in the fall term. (NCES (Digest 2023, Table 311.33), 2023)
• Primarily online institutions enrolled 1,124,509 students in fall 2022, 6.1% of all enrolment. (NCES (Digest 2023, Table 311.33), 2023)
• Primarily online institutions awarded 6.6% of US bachelor's degrees (132,274) and 10.7% of master's degrees (94,159) in 2021-22. (NCES (Digest 2023, Table 311.33), 2023)
• Institutions that take part in SARA, an agreement between states on distance education, reported 4,442,246 students enrolled exclusively in distance education in fall 2024, up 5.0% from 4,232,210 in fall 2023. The count covers member institutions only. (NC-SARA (2024 data report), Sep 2025)
• In-state students were 2,602,686 (58.6%) of the exclusively distance enrolment that SARA institutions reported for fall 2024. (NC-SARA (2024 data report), Sep 2025)
• Public institutions accounted for 58.9% (2,614,880) of the exclusively distance enrolment reported by SARA institutions in fall 2024, private nonprofits for 27.3% (1,214,184) and private for-profits for 13.7% (610,244). (NC-SARA (2024 data report), Sep 2025)
• Western Governors University, an online nonprofit university, had 192,613 students enrolled at 30 June 2025, on its own figures. (Western Governors University (FY25 annual report), 2025)
• WGU awarded 59,358 degrees to 58,625 graduates in its fiscal year 2025: 31,024 master's and 28,334 bachelor's. (Western Governors University (FY25 annual report), 2025)
• WGU reports on-time completion rates of 34% in FY21, 33% in FY22, 29% in FY23, 29% in FY24 and 32% in FY25, on its own definition. (Western Governors University (FY25 annual report), 2025)

NCES (IPEDS) covers all Title IV degree-granting institutions; NC-SARA covers only the institutions that take part in SARA, so its total is a different count and is not set against the NCES figure as a change. WGU's figures are the institution's own, and its on-time completion rate is not comparable with federal graduation rates.

Who takes online classes and fully online degrees?

Older and part-time undergraduates. In the federal student survey, undergraduates aged 30 or older were far more likely than those aged 15 to 23 to be in a degree programme taught entirely online.

1. 1 Aged 15 to 23 15.6%
2. 2 Aged 24 to 29 30.5%
3. 3 Aged 30 or older 43.7%

• The share of US undergraduates taking any distance education or online classes was 15.6% in 2003-04, 20.6% in 2007-08, 32.0% in 2011-12, 43.1% in 2015-16 and 66.0% in 2019-20, a year that includes the spring 2020 move online. (NCES (Digest 2022, Table 311.22), Jun 2023)
• The share of US undergraduates whose entire degree programme was online was 4.9% in 2003-04, 3.8% in 2007-08, 6.5% in 2011-12, 10.8% in 2015-16 and 23.9% in 2019-20. (NCES (Digest 2022, Table 311.22), Jun 2023)
• In 2019-20, 43.7% of US undergraduates aged 30 or older were in an entirely online degree programme, against 30.5% of those aged 24 to 29 and 15.6% of those aged 15 to 23. (NCES (Digest 2022, Table 311.22), Jun 2023)
• 31.0% of undergraduates who studied part-time for part of the year were in an entirely online programme in 2019-20, against 17.0% of full-time, full-year students. (NCES (Digest 2022, Table 311.22), Jun 2023)
• 25.7% of female undergraduates and 21.5% of male undergraduates were in an entirely online programme in 2019-20. (NCES (Digest 2022, Table 311.22), Jun 2023)
• By race and ethnicity, 27.6% of Black undergraduates, 24.1% of White, 22.3% of Hispanic and 18.5% of Asian undergraduates were in an entirely online programme in 2019-20. (NCES (Digest 2022, Table 311.22), Jun 2023)

These figures come from the National Postsecondary Student Aid Study, a student survey that covers a full academic year, not the fall enrolment census, so its shares are not comparable with the IPEDS counts above. NCES flags a change of question wording in 2011-12, and 2019-20 includes the pandemic spring term.

What do colleges and students think of online study?

Colleges are planning for more of it and most students surveyed want the option, but online-only students report a weaker sense of belonging. These are surveys of officials and students, not enrolment data.

1. 1 Graduate students 74%
2. 2 Adult-age undergraduates 66%
3. 3 Traditional-age undergraduates 60%

• 88% of US colleges and universities plan to expand their online offerings in the next three years, according to the tenth CHLOE survey of chief online learning officers. (Encoura (CHLOE 10 release), Aug 2025)
• Adjuncts teach nearly half (47%) of online courses at US colleges, the same survey found. (Encoura (CHLOE 10 release), Aug 2025)
• Chief online learning officers at US colleges report increased online interest from graduate students (74%), adult-age undergraduates (66%) and traditional-age undergraduates (60%). (QM, Eduventures and EDUCAUSE (CHLOE 10 report), Aug 2025)
• Just 28% of faculty are considered fully prepared for online course design and 45% for online teaching, the officers report. (QM, Eduventures and EDUCAUSE (CHLOE 10 report), Aug 2025)
• 83% of institutions in the survey offer online-specific student orientations, down from 93% in 2021. (QM, Eduventures and EDUCAUSE (CHLOE 10 report), Aug 2025)
• 65% of chief online learning officers report some or major investment in nondegree offerings such as certificates, micro-credentials and bootcamps, up from 29% in 2018-19; major investment went from 3% to 15%. (QM, Eduventures and EDUCAUSE (CHLOE 10 report), Aug 2025)
• In spring 2021, 57% of US college students surveyed were more optimistic about online learning than before the pandemic, and 48% were more optimistic about hybrid courses. (Bay View Analytics (Digital Learning Pulse, spring 2021), 2021)
• In the same spring 2021 survey, 46% of students strongly agreed and 27% somewhat agreed that they wanted some courses delivered fully online after the pandemic. (Bay View Analytics (Digital Learning Pulse, spring 2021), 2021)
• In fall 2022, 56% of US community college students and 52% of their faculty were more optimistic about online learning than before the pandemic. (Bay View Analytics (Digital Learning Pulse, fall 2022), 2022)
• In a spring 2024 survey of 5,025 US two- and four-year students, 54% of online-only students chose online asynchronous courses as a preferred way to learn, while 74% of in-person students and 52% of those mixing both preferred in person. (Inside Higher Ed and Generation Lab (Student Voice survey), May 2025)
• Only 31% of online-only students in that survey rated their sense of social belonging good or excellent, against 48% of in-person students. (Inside Higher Ed and Generation Lab (Student Voice survey), May 2025)
• 75% of the 2,365 US college students in a WGU Labs survey reported positive attitudes toward more online learning in the future. 53% of respondents were enrolled at a primarily online institution. (WGU Labs (College Innovation Network brief), Dec 2024)
• 40% of in-person students in the WGU Labs survey reported technology fatigue, against 27% of online learners. (WGU Labs (College Innovation Network brief), Dec 2024)

The CHLOE figures are reports by chief online learning officers at US two- and four-year institutions, not student data, and the report has vendor sponsors. The Bay View Analytics, Inside Higher Ed and WGU Labs figures are those organisations' own student surveys. The WGU Labs sample over-represents online learners and is not nationally representative. The 2021 and 2022 attitudes date from the pandemic period and are not the current level.

How many Americans take classes or training online?

About a quarter of US internet users, on the federal internet use survey. The young, the employed and college graduates are well ahead of everyone else.

View data table

Share of US internet users aged 15 or over who take classes or take part in job training online, 2001 to 2023, on NTIA's survey. The item has no data for 2005 to 2013.

Period	Value
2001	4.0%
2003	6.4%
2015	21.0%
2017	19.1%
2019	20.1%
2021	25.7%
2023	24.6%

• About 56.7 million US internet users aged 15 or over took classes or took part in job training online in November 2023. (NTIA (Internet Use Survey), 2024)
• The share of US internet users doing so was 4.0% in 2001, 6.4% in 2003, 21.0% in 2015, 19.1% in 2017, 20.1% in 2019 and 25.7% in 2021. (NTIA (Internet Use Survey), 2024)
• By age in November 2023, 37.7% of internet users aged 15-24 took classes or job training online, against 29.8% of those aged 25-44, 21.9% of those aged 45-64 and 8.5% of those aged 65 or over. (NTIA (Internet Use Survey), 2024)
• For internet users aged 15-24 the share was 29.8% in 2019 and 43.4% in 2021; for those aged 25-44 it was 25.1% and 29.7%; for those aged 45-64, 17.0% and 22.1%; and for those aged 65 or over, 6.2% and 9.4%. (NTIA (Internet Use Survey), 2024)
• By education in November 2023, the share was 31.9% for college graduates, 28.0% for people with some college, 13.3% for high school graduates and 16.8% for those with no diploma. (NTIA (Internet Use Survey), 2024)
• 30.3% of employed internet users took classes or job training online in November 2023. The share was 31.4% in households with income of $100,000 or more against 15.9% under $25,000, and 25.3% in metropolitan areas against 18.9% outside them. (NTIA (Internet Use Survey), 2024)
• By state, the November 2023 share was 36.4% in Utah, 35.8% in the District of Columbia, 34.2% in Colorado, 30.1% in Massachusetts, 27.2% in Texas, 27.0% in California, 22.3% in New York, 19.0% in Florida, 17.5% in Mississippi and 14.3% in West Virginia. (NTIA (Internet Use Survey), 2024)
• 73% of US adults considered themselves lifelong learners in a Pew survey published in 2016. (Pew Research Center (Lifelong Learning and Technology), Mar 2016)
• 63% of working US adults, or 36% of all adults, had taken a course or additional training in the past 12 months to improve job skills, Pew found in 2016. (Pew Research Center (Lifelong Learning and Technology), Mar 2016)
• In the same 2016 survey, 80% of US adults had little or no awareness of MOOCs, 79% of Khan Academy and 61% of distance learning. (Pew Research Center (Lifelong Learning and Technology), Mar 2016)

NTIA's Internet Use Survey is a supplement to the Current Population Survey, fielded by the Census Bureau, and covers internet users aged 15 or over. Shares are shown to one decimal place, and state estimates carry sizeable standard errors. It is a survey share, not an enrolment count, and uses a different age range and time window from Eurostat's EU measure. Pew's figures come from its own survey of US adults and are shown with their year, 2016.

How many people in the EU learn online?

About a third of internet users, up from about a fifth in 2019. Take-up differs widely by country, age and education.

View data table

Share of EU27 internet users aged 16-74 who did an online course or used online learning material in the three months before the survey, 2017 to 2025. The item has no 2018 data.

Period	Value
2017	19.1%
2019	21.4%
2020	26.6%
2021	30.9%
2022	28.4%
2023	30.1%
2024	33.4%
2025	34.8%

• The 2025 share of EU internet users who took an online course or used online learning materials was up from 33.4% in 2024 and 21.4% in 2019. (Eurostat (online learning news item), Mar 2026)
• 17.3% of EU internet users took an online course in 2025 and 30.5% used online learning materials. (Eurostat (online learning news item), Mar 2026)
• The Netherlands had the highest share of internet users taking an online course or using online learning materials in 2025 (60.2%), followed by Ireland (59.7%), Finland (50.7%) and Sweden (50.6%). Online learning was less prevalent in Romania (11.8%), Bulgaria (18.4%) and Cyprus (21.0%). (Eurostat (online learning news item), Mar 2026)
• For online courses alone, the highest shares of internet users in 2025 were in Ireland (29.6%), Finland (29.3%) and the Netherlands (28.5%). (Eurostat (online learning news item), Mar 2026)
• Among EU internet users aged 16-74, the share doing an online course or using online learning material was 19.1% in 2017, 26.6% in 2020, 30.9% in 2021, 28.4% in 2022 and 30.1% in 2023. (Eurostat (isoc_ci_ac_i), Apr 2026)
• The share of EU internet users doing an online course was 8.0% in 2017, 9.8% in 2019, 15.0% in 2020, 20.9% in 2021, 18.3% in 2022, 16.8% in 2023 and 17.9% in 2024. (Eurostat (isoc_ci_ac_i), Apr 2026)
• Elsewhere in Europe, the share of internet users who took an online course in 2025 was 30.9% in Norway, 28.3% in Spain, 25.8% in Hungary, 23.6% in Sweden, 21.7% in Switzerland, 21.4% in Italy, 18.5% in Austria, 18.1% in Belgium, 11.9% in France, 11.6% in Poland, 10.7% in Germany and 2.8% in Romania. Norway and Switzerland are outside the EU. (Eurostat (isoc_ci_ac_i), Apr 2026)
• In 2019 the share of internet users doing an online course or using online learning material was 24.4% in the Netherlands, 26.5% in Ireland, 37.1% in Finland and 38.9% in Sweden. (Eurostat (isoc_ci_ac_i), Apr 2026)
• Between 2019 and 2025 that share went from 30.8% to 46.0% in Spain, from 17.8% to 36.7% in France, from 20.4% to 35.0% in Italy, from 20.2% to 27.8% in Germany and from 14.5% to 21.3% in Poland. (Eurostat (isoc_ci_ac_i), Apr 2026)

Eurostat's survey covers people aged 16-74 and asks about the three months before the survey. In the figures above the base is internet users, meaning people who used the internet in the last three months. Dataset values are shown to one decimal place. This is a survey share and is not comparable with US enrolment counts or with the US survey, which uses a different age range and window.

How big are the online course platforms?

Large on sign-ups and much smaller on paying customers. Each company reports its own measure, and registered learners, learners reached and paid subscribers are not the same thing.

View data table

Coursera's cumulative Registered Learners at each year end, 2021 to 2025, in millions. These are sign-ups, not active users. Each year comes from the release for that year; the link is to the latest one.

Period	Value
2021	97m
2022	118m
2023	142m
2024	168.2m
2025	197.3m

• Coursera's Registered Learner total at the end of 2025 was up 17% from 168.2 million a year earlier. (Coursera (2025 Form 10-K), Feb 2026)
• 29.1 million new Registered Learners joined Coursera in 2025, against 26.3 million in 2024 and 23.7 million in 2023. (Coursera (2025 Form 10-K), Feb 2026)
• Coursera ended 2021 with a total of 97 million registered learners and revenue of $415.3 million, up 41% from $293.5 million in 2020. (Coursera (Q4 2021 results), Feb 2022)
• It ended 2022 with a total of 118 million registered learners and revenue of $523.8 million, up 26%. (Coursera (Q4 2022 results), Feb 2023)
• It ended 2023 with a total of 142 million registered learners and revenue of $635.8 million, up 21%. (Coursera (Q4 2023 results), Feb 2024)
• Coursera revenue was $757.5 million in 2025, up 9% from $694.7 million in 2024, with Consumer revenue of $502.2 million and Enterprise revenue of $255.3 million. (Coursera (2025 Form 10-K), Feb 2026)
• By region, Coursera's 2025 revenue was $384.7 million in the United States, $185.9 million in Europe, the Middle East and Africa, $108.9 million in Asia Pacific and $78.0 million elsewhere. No single country other than the US accounted for 10% or more. (Coursera (2025 Form 10-K), Feb 2026)
• Coursera reported 205 million Total Registered Learners at 31 March 2026, up 17% from 175 million a year earlier, with 7.6 million new Registered Learners in the quarter. (Coursera (Q1 2026 results), Apr 2026)
• On metrics that combine Coursera and Udemy, the company had 1,687,000 Paid Subscribers at 30 September 2026, up 30% from 1,299,000 a year earlier. (Coursera (Q3 2026 results), Oct 2026)
• The combined company had 11,897 Enterprise Customers at 30 September 2026, down 4% from 12,338, and third-quarter revenue of $376.7 million: $200.9 million Enterprise and $175.8 million Consumer. (Coursera (Q3 2026 results), Oct 2026)
• Generative AI course enrolments on Coursera rose 195% year over year to more than 8 million in total, the company said in June 2025. (Coursera (2025 Global Skills Report), Jun 2025)
• Coursera's 700 GenAI courses averaged 12 enrolments per minute in 2025, up from 1 per minute in 2023 and 8 in 2024. (Coursera (2025 Global Skills Report), Jun 2025)
• Coursera's Entry-Level Professional Certificates passed 15.4 million enrolments, with 37% growth in North America. (Coursera (2025 Global Skills Report), Jun 2025)

All figures in this section are the companies' own, from their filings, results releases and reports. Coursera's Registered Learners are cumulative accounts and the company says the count is not a measure of active engagement; its enrolment figures count enrolments, not people or completions. Its 2025 annual figures are for Coursera alone, while the September 2026 figures combine Coursera and Udemy. Udemy's learners reached, its paid consumer subscribers and Chegg's subscribers are different measures again and are not compared.

How many people use language learning apps?

Duolingo, which publishes user numbers every quarter, counts tens of millions of daily users, of whom a minority pay. These are one company's figures for its own app, not a count of all language learners.

View data table

Duolingo's daily active users in the fourth quarter of each year, 2021 to 2025, in millions, on the company's own figures. Each year comes from the release for that year; the link is to the latest one.

Period	Value
Q4 2021	10.1m
Q4 2022	16.3m
Q4 2023	26.9m
Q4 2024	40.5m
Q4 2025	52.7m

• Duolingo's daily active users in the third quarter of 2026 compared with 50.5 million in the third quarter of 2025. (Duolingo (Q3 2026 shareholder letter), Nov 2026)
• Duolingo had 153.1 million monthly active users in the third quarter of 2026, up 13% from 135.3 million a year earlier. (Duolingo (Q3 2026 shareholder letter), Nov 2026)
• Duolingo had 13.1 million paid subscribers at the end of September 2026, up 14% from 11.5 million. (Duolingo (Q3 2026 shareholder letter), Nov 2026)
• Duolingo's revenue in the third quarter of 2026 was $309.0 million, up 14% from $271.7 million. (Duolingo (Q3 2026 shareholder letter), Nov 2026)
• In the fourth quarter of 2025 Duolingo had 52.7 million daily active users, up 30% from 40.5 million a year earlier. (Duolingo (Q4 2025 shareholder letter), Feb 2026)
• Its monthly active users in the fourth quarter of 2025 were 133.1 million, up 14% from 116.7 million. (Duolingo (Q4 2025 shareholder letter), Feb 2026)
• Duolingo had 12.2 million paid subscribers at the end of 2025, up 28% from 9.5 million. (Duolingo (Q4 2025 shareholder letter), Feb 2026)
• Duolingo's full-year 2025 revenue was $1,037.6 million, up 39%. (Duolingo (Q4 2025 shareholder letter), Feb 2026)
• In the fourth quarter of 2022 Duolingo had 16.3 million daily active users (10.1 million a year earlier), 60.7 million monthly active users (42.4 million) and 4.2 million paid subscribers at the quarter end (2.5 million). (Duolingo (Q4 2022 results), Feb 2023)
• In the fourth quarter of 2023 Duolingo had 26.9 million daily active users, 88.4 million monthly active users and 6.6 million paid subscribers at the quarter end. (Duolingo (Q4 2023 results), Feb 2024)
• Duolingo's chess course reached more than seven million daily active users in less than a year after launch, the company said in February 2026. (Duolingo (Q4 2025 shareholder letter), Feb 2026)
• Duolingo names its nine most learned languages as English, Spanish, French, German, Italian, Japanese, Korean, Chinese and Portuguese. (Duolingo (Q4 2025 shareholder letter), Feb 2026)
• English was the most studied language on Duolingo in 154 countries, 79% of all countries, in 2025, an increase of 14% over 2024. (Duolingo (2025 Language Report), Dec 2025)
• In 2025 Japanese became the fourth most popular language to study on Duolingo worldwide, passing German, and Korean the sixth, passing Italian. (Duolingo (2025 Language Report), Dec 2025)

All figures are Duolingo's own and cover its users worldwide. Daily and monthly active users are measures of activity over a quarter; paid subscribers are counted at the period end. Chess users are part of the daily active total, not extra users. The language rankings describe learners on Duolingo, not language learning in general.

Do online learners finish their courses, and do they do as well?

Mostly not on free open courses, and the independent studies here find weaker results for fully online study than for in-person study. Platform surveys of people who completed a course give a far rosier picture, and the two kinds of evidence should not be mixed.

View data table

Completion rate of all participants in HarvardX and MITx MOOCs on edX, 2013-14 to 2017-18, from the supplementary tables of a 2019 paper in Science.

Period	Value
2013-14	4.96%
2014-15	5.91%
2015-16	4.13%
2016-17	3.98%
2017-18	3.13%

• The completion rate for all participants in HarvardX and MITx MOOCs on edX was 4.96% in 2013-14, 5.91% in 2014-15, 4.13% in 2015-16 and 3.98% in 2016-17. (Reich and Ruipérez-Valiente (Science, 2019), Jan 2019)
• Among learners who paid for the verified track, completion was 46.02% in 2017-18, against 44.23% in 2013-14. Among those who said they intended to complete, it was 15.42% in 2017-18. (Reich and Ruipérez-Valiente (Science, 2019), Jan 2019)
• Of those who register for a HarvardX or MITx MOOC, 52% never enter the courseware. (Reich and Ruipérez-Valiente (Science, 2019), Jan 2019)
• Countries with very high human development supplied 59.3% of HarvardX and MITx MOOC enrolments in 2012-13 and 55.6% in 2017-18, and 68.7% of certifications in 2017-18. Countries with low human development had 3.2% of enrolments and 1.43% of certifications in 2017-18. (Reich and Ruipérez-Valiente (Science, 2019), Jan 2019)

The MOOC figures cover HarvardX and MITx courses on edX only, with learners worldwide, up to 2018. The paper's second-year retention figures are left out because its text and its table give different values.

How many children are homeschooled?

A small minority in both the US and England, though the counts have risen. The US figures come from three different measures, of children, of households and of parents, and cannot be read as one trend.

View data table

Share of US students aged 5-17 who were homeschooled, 1999 to 2019, from NCES household surveys. NCES cautions that the enrolment cutoff and the survey mode changed over the period. The 2022-23 estimate comes from a later NCES report and is not plotted.

Period	Value
1999	1.7%
2003	2.2%
2007	3.0%
2012	3.4%
2016	3.3%
2019	2.8%

• About 1,765,000 US children aged 5-17 were homeschooled in 2022-23, on NCES's weighted estimate from its parent survey. (NCES (Parent and Family Involvement 2023), Sep 2024)
• 5.2% of US students aged 5-17 received instruction at home in 2022-23, either homeschooled or in full-time virtual education, up from 3.7% in 2018-19. (NCES (Parent and Family Involvement 2023), Sep 2024)
• The same survey put 2.5% of US students in full-time virtual education in 2022-23. Some children are in both groups, so the two shares do not add up to the at-home total. (NCES (Parent and Family Involvement 2023), Sep 2024)
• Earlier rounds of the NCES household survey estimated 850,000 homeschooled students aged 5-17 in 1999, 1,096,000 in 2003, 1,520,000 in 2007, 1,773,000 in 2012, 1,690,000 in 2016 and 1,457,000 in 2019. (NCES (Digest 2022, Table 206.10), Dec 2021)
• By locale, the US homeschooling rate in 2022-23 was 5.2% in rural areas, 5.1% in towns, 3.1% in suburbs and 2.1% in cities. (NCES (Parent and Family Involvement 2023), Sep 2024)
• By race and ethnicity, the 2022-23 homeschooling rate was 5.1% for White non-Hispanic students, 4.0% for students of two or more races, 1.8% for Hispanic students, 1.7% for Black students and 0.7% for Asian students, an estimate NCES says to interpret with caution. (NCES (Parent and Family Involvement 2023), Sep 2024)
• By grade, the 2022-23 homeschooling rate was 4.0% in kindergarten to grade 2, 3.6% in grades 3-5, 3.0% in grades 6-8 and 3.1% in grades 9-12. (NCES (Parent and Family Involvement 2023), Sep 2024)
• 83% of US parents of homeschooled children in 2022-23 named concern about the environment of other schools as an important reason for homeschooling, and 28% named it the most important. (NCES (Parent and Family Involvement 2023), Sep 2024)
• Other reasons, as important and as most important: moral instruction 75% and 6%, family life together 72% and 8%, dissatisfaction with academic instruction at other schools 72% and 17%, religious instruction 53% and 11%, a nontraditional approach 50% and 8%, and a child's other special needs 21% and 7%. COVID-19 related reasons were named by 3%. (NCES (Parent and Family Involvement 2023), Sep 2024)
• In a US cohort of 12,288 children of nurses followed from adolescence in 1999 into young adulthood, those who were homeschooled were less likely than public school attendees to attain a college degree (risk ratio 0.77) and more likely to attend religious services frequently (risk ratio 1.51). The study is observational. (Chen, Hinton and VanderWeele (PLOS ONE, 2021), Nov 2021)

NCES figures are parent-reported estimates from a household survey and count children aged 5-17. Homeschooled and full-time virtual students overlap. NCES cautions against close comparison across years because the survey mode and the enrolment cutoff changed. The cohort study covers children of nurses and is not nationally representative.

How many K-12 students attend virtual schools?

It depends on the definition. Federal school data count students in schools flagged as virtual, while the federal parent survey counts children whose parents say their schooling is full-time virtual and gives a much larger number. The two are not added together or read as one trend.

View data table

Enrolment in US public virtual schools, 2013-14 to 2022-23, on NCES school data. California, Illinois and Utah are missing in 2020-21 and California in 2021-22, so those two years are understated.

Period	Value
2013-14	200,343
2014-15	229,608
2015-16	234,148
2016-17	212,311
2017-18	278,783
2018-19	289,614
2019-20	293,689
2020-21	590,267
2021-22	566,188
2022-23	562,659

• About 1,340,000 US K-12-age students were in full-time virtual education in 2022-23, as reported by parents in the NCES household survey. (NCES (Parent and Family Involvement 2023), Sep 2024)
• 7% of US K-12 students, enrolled or homeschooled, took at least one virtual course in 2022-23: 3% took all their courses virtually and 5% took some but not all. 93% took none. (NCES (Parent and Family Involvement 2023), Sep 2024)
• 48% of US homeschooled students took at least one course virtually in 2022-23. (NCES (Parent and Family Involvement 2023), Sep 2024)
• 11% of students at chosen public schools took a virtual course in 2022-23, against 6% at assigned public schools and 3% at private schools; 9% of rural students did, against 6% of suburban students. (NCES (Parent and Family Involvement 2023), Sep 2024)
• Enrolment in US public virtual schools, on federal school data, was 200,343 in 2013-14, 229,608 in 2014-15, 234,148 in 2015-16, 212,311 in 2016-17, 278,783 in 2017-18, 289,614 in 2018-19 and 293,689 in 2019-20. (NCES (Digest 2023, Table 216.20), Dec 2023)
• Public virtual school enrolment was 590,267 in 2020-21, with data missing for California, Illinois and Utah, and 566,188 in 2021-22, with data missing for California. (NCES (Digest 2023, Table 216.20), Dec 2023)
• US public schools of all kinds enrolled 48,860,517 students in 2022-23. (NCES (Digest 2023, Table 216.20), Dec 2023)
• Virtual charter schools account for about 70% of students enrolled in US virtual schools, the Government Accountability Office reported in 2022. (GAO (GAO-22-104444), Jan 2022)
• 726 full-time virtual schools operated in the US in 2021-22 on the National Education Policy Center's inventory, up from 477 in 2019-20 and 653 in 2020-21. (NEPC (Virtual Schools in the U.S. 2023), May 2023)
• Districts created 245 new full-time virtual schools during the pandemic, while charter schools added a net four. (NEPC (Virtual Schools in the U.S. 2023), May 2023)
• Full-time virtual schools run by for-profit education management organisations enrolled an average of 1,483 students each in 2021-22, against 656 at schools run by nonprofit operators and 562 at independent virtual schools. (NEPC (Virtual Schools in the U.S. 2023), May 2023)
• Virtual school students in 2021-22 were 51.9% female, against 48.7% in US public schools nationally. (NEPC (Virtual Schools in the U.S. 2023), May 2023)
• One-quarter of US school districts planned to run a virtual school in 2021-22, a ninefold increase from the pre-pandemic level, and almost half of urban districts planned to, in a RAND survey of district leaders in June 2021. (RAND (The Rise of Virtual Schools), Aug 2021)
• 26% of US districts offered fully online courses before the pandemic and another 10% newly offered them in 2021-22. (RAND (The Rise of Virtual Schools), Aug 2021)
• In June 2021, one in five US school district leaders said parental demand for a fully remote schooling option in 2021-22 was strong. (RAND (The Rise of Virtual Schools), Aug 2021)
• 97% of districts planned a fully in-person option for fall 2021 and only 14% planned any temporary partial or fully remote option. Where a remote option was offered, leaders expected only 6% of students to enrol in it. (RAND (The Rise of Virtual Schools), Aug 2021)

Two enrolment measures appear here and they are kept apart. The NCES parent survey (2022-23) counts children whose parents report full-time virtual education, whether or not the parent considers the child homeschooled. NCES school data (Common Core of Data) count students enrolled in public schools flagged as virtual, where students and teachers are separated by time or location and interact through internet-connected computers. NEPC keeps its own inventory of full-time virtual schools on a third definition; only its school counts and averages are used. The RAND figures are district leaders' plans in 2021, not realised enrolment.

What happened when schools moved online during the pandemic?

Almost every US household with schoolchildren switched to some form of distance learning in spring 2020, and the studies that followed link long spells of remote teaching to bigger learning losses and more absence. Everything here describes emergency teaching in 2020 to 2022, not today's level and not purpose-built online courses.

1. 1 Separated from classmates and teachers 45%
2. 2 Child's attention span or motivation 44%
3. 3 Balancing a job and helping with school 43%
4. 4 Knowing how to teach the child 28%
5. 5 Technical issues with computer or internet 10%
6. 6 Unable to access educational websites 4%

• In spring 2020, nearly 93% of people in US households with school-age children reported that the children were engaged in some form of distance learning from home. (US Census Bureau (Household Pulse, distance learning), Aug 2020)
• From 28 May to 2 June 2020, 80% of people living with children in distance learning said the children were using online resources, and about 20% said they were using paper materials sent home by the school. (US Census Bureau (Household Pulse, distance learning), Aug 2020)
• Among US students who were ninth-graders in fall 2022, 86.8% said they received all or some of their instruction online in 2020-21. (NCES (Digest 2023, Table 218.95), Feb 2024)
• Of those taught online in 2020-21, 49.0% said they struggled to keep up with schoolwork, against 40.6% of those taught fully in person; 47.0% said they were motivated to take part in learning assignments, against 58.7%; and 62.5% were confident they had learned the math skills, against 76.7%. (NCES (Digest 2023, Table 218.95), Feb 2024)
• 70% of US 9-year-olds assessed in 2022 recalled learning remotely during the 2020-21 school year. (NAEP (Long-Term Trend 2022, age 9), Sep 2022)
• Average long-term trend scores for US 9-year-olds fell 5 points in reading and 7 points in mathematics between 2020 and 2022, and higher-performing students were more likely than lower performers to have had a computer or tablet available all the time. (NAEP (Long-Term Trend 2022, age 9), Sep 2022)
• Within US school districts that were remote for most of 2020-21, high-poverty schools had 50% more achievement loss than low-poverty schools, 0.46 against 0.30 standard deviations in math, in a study of test data from 2.1 million students in 10,000 schools. Math gaps did not widen in areas that stayed in person. (Goldhaber, Kane and others (NBER Working Paper 30010), May 2022)
• Across 11,017 US school districts, chronic absenteeism rose from 15.9% in 2018-19 to 29.4% in 2021-22. (Evans, Muchnick and Rosenlund (JAMA Network Open), Aug 2024)
• Districts with 100% virtual instruction in 2020-21 had chronic absenteeism rates 6.9 percentage points higher in 2021-22 than fully in-person districts, and 10.6 points higher in the highest-poverty districts. Hybrid instruction was not linked to higher absenteeism. (Evans, Muchnick and Rosenlund (JAMA Network Open), Aug 2024)
• Learning poverty in low- and middle-income countries rose from 57% before the pandemic to an estimated 70% of 10-year-olds unable to understand a simple written text, after prolonged school closures, the World Bank and its partners estimated in 2022. (World Bank (learning poverty 2022 release), Jun 2022)
• The same release said this generation of students risks losing $21 trillion in potential lifetime earnings in present value. (World Bank (learning poverty 2022 release), Jun 2022)

All figures are for the pandemic period and are labelled with their dates. The Census Bureau figures come from its experimental Household Pulse Survey. The ninth-grader figures are students' own reports, and the test-score and absenteeism studies show links, not proven cause: the fall in scores covers the whole pandemic period, not remote learning alone. The NBER figures are from a working paper. The World Bank figure is a simulation-based estimate for low- and middle-income countries, not a measured test result.

Who is shut out of online learning?

Children in poorer households and poorer countries. When schools closed in 2020, at least a third of the world's schoolchildren could not be reached by remote learning at all, on UNICEF's minimum estimate.

1. 1 East and Southern Africa 49%
2. 2 West and Central Africa 48%
3. 3 Middle East and North Africa 40%
4. 4 South Asia 38%
5. 5 Eastern Europe and Central Asia 34%
6. 6 East Asia and the Pacific 20%
7. 7 Latin America and the Caribbean 9%
8. 8 Global 31%

• Globally, 72% of the schoolchildren unable to access remote learning in 2020 lived in their countries' poorest households. (UNICEF (remote learning reachability release), Aug 2020)
• At least 70% of schoolchildren of pre-primary age, 120 million children, could not be reached by remote learning in 2020. (UNICEF (remote learning reachability release), Aug 2020)
• The minimum share of schoolchildren who could not be reached during the 2020 closures was 49% in East and Southern Africa (67 million), 48% in West and Central Africa (54 million), 40% in the Middle East and North Africa (37 million) and 38% in South Asia (147 million). (UNICEF (remote learning reachability release), Aug 2020)
• It was 34% in Eastern Europe and Central Asia (25 million), 20% in East Asia and the Pacific (80 million) and 9% in Latin America and the Caribbean (13 million). (UNICEF (remote learning reachability release), Aug 2020)
• Half of all learners kept out of the classroom by the pandemic in April 2020, some 826 million students, had no household computer, and 43% (706 million) had no internet at home. (UNESCO (digital divides in distance learning), Apr 2020)
• In sub-Saharan Africa, 89% of learners had no household computer and 82% lacked internet access in April 2020. (UNESCO (digital divides in distance learning), Apr 2020)
• Globally, only 40% of primary schools, 50% of lower secondary schools and 65% of upper secondary schools are connected to the internet, UNESCO reported in 2023. (UNESCO (GEM Report 2023), Jul 2023)
• An estimated 6.3 billion people, 76% of the world's population, use the internet in 2026, up from a revised 6.1 billion (74%) in 2025. 2 billion people remain offline. (ITU (Facts and Figures 2026 release), Nov 2026)
• Globally, 82% of urban residents have internet access in 2026, against 63% in rural areas. (ITU (Facts and Figures 2026 release), Nov 2026)

UNICEF's 2020 figures are minimum estimates based on household access to television, radio and the internet and on whether a remote curriculum was offered, not on actual use. UNESCO's April 2020 figures were compiled at the height of the closures. The ITU and UNESCO connectivity figures describe internet access in general and are context for online learning, not measures of it.

How much workplace training happens online?

More US training hours are delivered through online or computer-based tools than by a classroom instructor, in Training magazine's 2025 survey. Trade and vendor surveys differ in sample and definition, so their spending figures are not comparable with each other.

1. 1 Online or computer-based 34%
2. 2 Instructor-led classroom 28%
3. 3 Virtual classroom or webcast 24%
4. 4 Blended learning 22%
5. 5 Mobile devices 6%

• US companies spent $874 per learner on training in 2025, against $774 in 2024. (Training magazine (2025 Training Industry Report), Nov 2025)
• US employees received an average of 40 hours of training in 2025, against 47 hours the year before. (Training magazine (2025 Training Industry Report), Nov 2025)
• 34% of US training hours in 2025 were delivered through online or computer-based technologies, 28% by an instructor in a classroom and 24% through a virtual classroom or webcast. (Training magazine (2025 Training Industry Report), Nov 2025)
• Blended learning accounted for 22% of US training hours in 2025, down from 24%, and mobile devices for 6%, against 3% in 2024. (Training magazine (2025 Training Industry Report), Nov 2025)
• Organisations in ATD's benchmarking survey spent an average of $1,254 per employee on direct learning in 2024, a $29 decrease from 2023. (ATD (2025 State of the Industry), 2025)
• Employees in ATD's survey used an average of 13.7 formal learning hours in 2024, down from 17.4 in 2023. (ATD (2025 State of the Industry), 2025)
• 40% of organisations in ATD's survey used asynchronous or self-paced e-learning in 2024, down from 57% in 2023. Nearly 60% used live, instructor-led virtual classrooms and 54% used hybrid learning. (ATD (2025 State of the Industry), 2025)
• 91% of learning and development professionals agree that continuous learning is more important than ever for career success. (LinkedIn Learning (Workplace Learning Report 2025), 2025)
• 88% of organisations are concerned about employee retention, and providing learning opportunities is the No. 1 retention strategy among survey respondents. (LinkedIn Learning (Workplace Learning Report 2025), 2025)
• 49% of learning and development professionals agree that their executives are concerned employees do not have the right skills to carry out the business strategy, and 71% are exploring, experimenting with or integrating AI into their work. (LinkedIn Learning (Workplace Learning Report 2025), 2025)
• On barriers to career development, 50% of respondents say managers lack proper support, 45% say employees lack support and 33% say talent teams themselves lack support. (LinkedIn Learning (Workplace Learning Report 2025), 2025)
• The most common measures of success used by learning and development teams are employee engagement (72%), retention (64%), employees developing new skills (55%), promotions (48%) and internal mobility rate (32%). (LinkedIn Learning (Workplace Learning Report 2025), 2025)
• If the world's workforce were 100 people, 59 would need training by 2030, employers told the World Economic Forum: 29 could be upskilled in their current roles, 19 upskilled and redeployed, and 11 would be unlikely to receive the training they need. (World Economic Forum (Future of Jobs Report 2025), Jan 2025)
• 85% of employers surveyed plan to make upskilling their workforce a priority, and 50% of workers had completed training, reskilling or upskilling measures, against 41% in the report's 2023 edition. (World Economic Forum (Future of Jobs Report 2025), Jan 2025)
• Employers expect 39% of workers' existing skill sets to be transformed or become outdated over 2025-2030, down from 44% in the 2023 report and 57% in 2020. (World Economic Forum (Future of Jobs Report 2025), Jan 2025)
• 63% of employers see skill gaps as a major barrier to business transformation over 2025-2030. (World Economic Forum (Future of Jobs Report 2025), Jan 2025)

Training magazine surveys US-based organisations and projects its totals to the US. ATD's figures come from its own benchmarking survey of participating organisations, so its spending per employee is not comparable with Training magazine's spending per learner. LinkedIn's figures are a vendor's own survey of learning and HR professionals in several countries. The World Economic Forum figures are employers' expectations to 2030 from its global survey, published in January 2025, and are forecasts, not measurements; they cover training of all kinds, not online delivery alone.

How many students and teachers use AI for school?

A growing share on every survey here, but each asked a different question of a different group at a different date. The figures are stated one by one and are not ranked or averaged.

View data table

Share of US middle school, high school and college students using AI for homework in May and December 2025, on RAND's American Youth Panel. This is one survey's question and is not comparable with the other AI figures in this section.

Period	Value
May 2025	48%
December 2025	62%

• 26% of US teens aged 13-17 said in autumn 2024 that they had used ChatGPT for their schoolwork, up from 13% in 2023. 73% had not used the chatbot in this way. (Pew Research Center (teens and ChatGPT), Jan 2025)
• In RAND's survey of students on using AI for homework, 75% of female students said AI harmed critical thinking skills, against 59% of male students. (RAND (AI for homework release), Mar 2026)
• In 2025, 54% of US students and 53% of English language arts, math and science teachers indicated that they used AI for school, increases of more than 15 percentage points on survey results one to two years earlier. (RAND (AI Use in Schools report), Sep 2025)
• Over 80% of US students said teachers did not explicitly teach them how to use AI for schoolwork, and 45% of principals reported school or district policies or guidance on AI. (RAND (AI Use in Schools report), Sep 2025)
• 70% of US teens report using AI for schoolwork, in a Common Sense Media survey released in August 2026. (Common Sense Media (Teens in the AI Era release), Aug 2026)
• Among US teens who use AI, 63% say they get an answer to their schoolwork from AI: 25% use it as it is, 31% rewrite it to sound like them and 35% improve it with what they know. (Common Sense Media (Teens in the AI Era release), Aug 2026)
• Only 27% of US teens say a teacher has ever talked with them about what AI is and how it works. (Common Sense Media (Teens in the AI Era release), Aug 2026)
• 60% of US public K-12 teachers used an AI tool for their work during the 2024-25 school year, including 66% of high school teachers and 69% of early-career teachers. (Gallup and Walton Family Foundation (Teaching for Tomorrow), Jun 2025)
• Teachers who use AI at least weekly estimate that it saves them 5.9 hours a week, against 2.9 hours for monthly users. (Gallup and Walton Family Foundation (Teaching for Tomorrow), Jun 2025)
• Only 19% of US public K-12 teachers work at schools with an AI policy. (Gallup and Walton Family Foundation (Teaching for Tomorrow), Jun 2025)
• A quarter of US public K-12 teachers said in late 2023 that using AI tools in K-12 education does more harm than good. 32% saw an equal mix of benefit and harm, 6% more good than harm, and 35% were not sure. (Pew Research Center (teachers and AI), May 2024)
• 57% of chief online learning officers at US colleges say uneven access to AI tools is affecting at least some learners, and 72% expect AI to become very or extremely important within two years. (QM, Eduventures and EDUCAUSE (CHLOE 10 report), Aug 2025)

The student figures answer different questions. Pew asked US teens aged 13-17 in autumn 2024 whether they had used ChatGPT for schoolwork. RAND's youth panel covers middle school, high school and college students using AI for homework in May and December 2025; its separate spring 2025 panels asked students and teachers whether they used AI for school. Common Sense Media, an advocacy organisation, asked US teens about using AI for schoolwork and published in August 2026. The teacher figures come from a Gallup survey of public K-12 teachers in spring 2025, where time saved is the teachers' own estimate, and a Pew survey of public K-12 teachers in late 2023. None of these is ranked against or averaged with another.

Methodology and sources

Figures come from the public bodies, researchers and companies that produce each measure: the National Center for Education Statistics (IPEDS enrolment counts, the NPSAS student survey, the household parent survey, school data and NAEP), NC-SARA, Eurostat, NTIA, the US Census Bureau, the Department for Education in England, the ONS, UNICEF, UNESCO, the ITU and the World Bank; peer-reviewed and working-paper research; surveys by Pew Research Center, Gallup, RAND, Common Sense Media, Bay View Analytics, Inside Higher Ed, Training magazine, ATD, LinkedIn and the World Economic Forum; and the filings and reports of Coursera, Udemy, Chegg, Duolingo, Stride, Pearson, Western Governors University and Florida Virtual School. Each figure links to the page or document that published it, with the date, and source labels name the report or table. The sources measure different things and are kept apart: enrolment counts are not mixed with survey shares, the federal count of exclusively online students is not set against NC-SARA's member-only count, the three US homeschooling measures and the two virtual school enrolment measures are stated separately, the AI surveys are given with their own questions and dates, and company user numbers are named as each company names them. Eurostat and NTIA dataset values are shown to one decimal place. Pandemic-period figures are labelled with their dates. Figures that could not be checked on the producer's own page are left out, including Khan Academy's user counts, a global count of MOOC learners, state homeschool counts from the Johns Hopkins Homeschool Hub and distance-learner numbers for UK higher education. NC-SARA's state table and the MOOC paper's second-year retention figures are also left out. Forecasts and plans are labelled with who made them and when. The page is reviewed monthly.

• ACSES, Curtin University (mode of attendance, 2024 data) · Mar 2026
• ATD (2025 State of the Industry) · 2025
• Bay View Analytics (Digital Learning Pulse, fall 2022) · 2022
• Bettinger and Loeb (Brookings, 2017) · Jun 2017
• Chegg (Q3 2026 results) · Nov 2026
• Chegg (Q4 2024 results) · Feb 2025
• Chegg (Q4 2025 results) · Feb 2026
• Chen, Hinton and VanderWeele (PLOS ONE, 2021) · Nov 2021
• Common Sense Media (Teens in the AI Era release) · Aug 2026
• Coursera (2025 Form 10-K) · Feb 2026
• Coursera (2025 Global Skills Report) · Jun 2025
• Coursera (2025 Learner Outcomes Report) · Oct 2025
• Coursera (Q1 2026 results) · Apr 2026
• Coursera (Q3 2026 results) · Oct 2026
• Coursera (Q4 2021 results) · Feb 2022
• Coursera (Q4 2022 results) · Feb 2023
• Coursera (Q4 2023 results) · Feb 2024
• CREDO, Stanford University (Online Charter School Study 2015) · Oct 2015
• Department for Education (elective home education) · Jan 2026
• Duolingo (2025 Language Report) · Dec 2025
• Duolingo (Q3 2026 shareholder letter) · Nov 2026
• Duolingo (Q4 2022 results) · Feb 2023
• Duolingo (Q4 2023 results) · Feb 2024
• Duolingo (Q4 2025 shareholder letter) · Feb 2026
• Encoura (CHLOE 10 release) · Aug 2025
• Eurostat (isoc_ci_ac_i) · Apr 2026
• Eurostat (online learning news item) · Mar 2026
• Evans, Muchnick and Rosenlund (JAMA Network Open) · Aug 2024
• Florida Virtual School (2025-26 release) · Sep 2026
• Gallup (distance learning poll, 2020) · Jun 2020
• Gallup (K-12 parents poll, 2020) · Aug 2020
• Gallup and Walton Family Foundation (Teaching for Tomorrow) · Jun 2025
• GAO (GAO-22-104444) · Jan 2022
• Goldhaber, Kane and others (NBER Working Paper 30010) · May 2022
• Inside Higher Ed and Generation Lab (Student Voice survey) · May 2025
• ITU (Facts and Figures 2026 release) · Nov 2026
• LinkedIn Learning (Workplace Learning Report 2025) · 2025
• NAEP (Long-Term Trend 2022, age 9) · Sep 2022
• NC-SARA (2024 data report) · Sep 2025
• NCES (Digest 2019, Table 311.15) · Dec 2019
• NCES (Digest 2022, Table 206.10) · Dec 2021
• NCES (Digest 2022, Table 311.22) · Jun 2023
• NCES (Digest 2023, Table 216.20) · Dec 2023
• NCES (Digest 2023, Table 218.95) · Feb 2024
• NCES (Digest 2023, Table 311.33) · 2023
• NCES (Digest 2025, Table 311.15) · Mar 2026
• NCES (Parent and Family Involvement 2023) · Sep 2024
• NEPC (Virtual Schools in the U.S. 2023) · May 2023
• NTIA (Internet Use Survey) · 2024
• ONS (Coronavirus and homeschooling in Great Britain) · Jul 2020
• Ortagus, Hughes and Allchin (Third Way brief) · Oct 2024
• Pearson (2025 full-year results) · Feb 2026
• Pew Research Center (Lifelong Learning and Technology) · Mar 2016
• Pew Research Center (parents of K-12 students, 2020) · Oct 2020
• Pew Research Center (teachers and AI) · May 2024
• Pew Research Center (teens and ChatGPT) · Jan 2025
• Pew Research Center (teens and school, 2022) · Jun 2022
• QM, Eduventures and EDUCAUSE (CHLOE 10 report) · Aug 2025
• RAND (AI for homework release) · Mar 2026
• RAND (AI Use in Schools report) · Sep 2025
• RAND (The Rise of Virtual Schools) · Aug 2021
• Reich and Ruipérez-Valiente (Science, 2019) · Jan 2019
• Stride (Q4 FY2026 results) · Aug 2026
• Training magazine (2025 Training Industry Report) · Nov 2025
• Udemy (2025 Form 10-K) · Feb 2026
• Udemy (Q4 2025 results) · Feb 2026
• UNESCO (digital divides in distance learning) · Apr 2020
• UNESCO (GEM Report 2023) · Jul 2023
• UNICEF (remote learning reachability release) · Aug 2020
• US Census Bureau (Computer and Internet Use 2021) · Jun 2024
• US Census Bureau (Household Pulse, distance learning) · Aug 2020
• US Census Bureau (Household Pulse, homeschooling) · Mar 2021
• Western Governors University (FY25 annual report) · 2025
• WGU Labs (College Innovation Network brief) · Dec 2024
• World Bank (learning poverty 2022 release) · Jun 2022
• World Economic Forum (Future of Jobs Report 2025) · Jan 2025