Skip to content

ASER 2023 · Selected rural districts · Published-base reconstruction

A successful share is the last step: rebuilding a mobile resource route from ASER 2023

In ASER 2023's pooled sample from twenty-eight selected rural districts, 89.3% of youth who found the specified video could share it. That is a conditional success rate at the final step. Combining the report's explicitly nested phone, finding and sharing rates reconstructs approximately 48.9% of the full source group completing the observed route. The calculation describes a particular assessment protocol among fourteen-to-eighteen-year-olds. It is not an estimate of ebook readership, permanent internet access, true digital literacy or what would happen if everyone were supplied with an accessible device.

Published
Share
Share this page
EmailCite & reuse this study
Key findings as text
Reusable imagesA successful share is the last step: rebuilding a mobile resource route from ASER 2023 — 48.9%: Reconstructed complete path in the pooled source group. 67.1% bring × 81.6% find among bringers × 89.3% share among finders. Observed task path, not ebook access.Download sharing image (JPG) ↓

Key findings

48.9%

Reconstructed complete path in the pooled source group

67.1% bring × 81.6% find among bringers × 89.3% share among finders. Observed task path, not ebook access.

89.3%

Sharing success among successful video finders

A valid conditional final-stage rate; it does not describe all youth.

−16.0 pp

Female-minus-male complete-path product difference

Rounding-only enclosure −16.2012 to −15.8245 pp. No sampling confidence interval or significance claim.

3 factors

Symmetric arithmetic accounting

Bringing −7.9 pp; finding −4.4 pp; sharing −3.7 pp. Contributions are not causal effects or intervention priorities.

The result and the reader decision

In ASER 2023's pooled sample from twenty-eight selected rural districts, 89.3% of youth who found the specified video could share it. That is a conditional success rate at the final step. Combining the report's explicitly nested phone, finding and sharing rates reconstructs approximately 48.9% of the full source group completing the observed route. The calculation describes a particular assessment protocol among fourteen-to-eighteen-year-olds. It is not an estimate of ebook readership, permanent internet access, true digital literacy or what would happen if everyone were supplied with an accessible device.

The distinction matters to anyone who distributes reading resources through digital routes. A final task can work well among the people who reach it while many intended recipients remain outside its denominator. This reference rebuilds the observed route for all seven source groups displayed in two ASER tables, retains uncertainty from rounded inputs, and separates three descriptive contributions to the reported male–female product difference. It does not measure an author's marketing funnel or recommend replacing books with video. Its purpose is to make a public assessment's changing bases usable without turning a successful last step into a claim about everyone.

Three stages, three bases

The source first reports the proportion of youth who could bring a smartphone with good connectivity to perform the digital tasks. Among those able to bring a phone, it reports the proportion who could find the specified video. Among those who found it, it reports the proportion able to share it. These are deliberately nested bases. The final rate concerns people who have already crossed the first two recorded stages. It therefore cannot be read as the percentage of all youth who demonstrated the whole route.

For the source total, the three reported rates are 67.1%, 81.6% and 89.3%. The calculation is 0.671 × 0.816 × 0.893, or 0.488949648. Approximately 48.9% of the full source group therefore completed the observed phone–find–share path under the report's routing. Approximately 54.8% reached the finding stage on that full base. These products use the source's displayed percentages, so they reconstruct reported-rate quantities rather than exact respondent counts. Multiplying them by the full survey sample would create false precision.

The identity does not require independence between the stages. It requires the opposite kind of information: each later percentage must explicitly use the preceding successful group as its base. In probability notation, the product is P(B) × P(F given B) × P(S given F and B). If the finding and sharing figures had been unrelated marginal percentages, multiplying them would not establish joint reach. The source table headings provide the nested relation here. The downloadable input table preserves that relation instead of leaving a reader to infer it from three unlabeled numbers.

Bringing a phone is not owning one

The first stage is easy to misname. Respondents were asked to bring a smartphone with good connectivity for assessment; it could be their own, a family member's or a neighbor's. Failing to produce such a phone at that encounter is not the same as never having access to a smartphone. Conversely, having a smartphone in the household does not guarantee that the respondent could bring it, use it privately or rely on it whenever a resource is needed. Those are different observations and should remain different variables.

This is why the study does not put household smartphone availability before the route and multiply again. The report's household rate and bringing rate are not supplied here as nested stages for the same respondents. A youth might bring a neighbor's phone despite having none at home; another might have one at home but be unable to bring it at that time. Subtracting the two aggregate rates can describe a difference between reported indicators, but it cannot identify the exact people lost between those states. The reconstructed path starts only where the source establishes a proper conditional chain.

The same restraint applies to people who did not complete a later task. An observed path can stop because of a device, connectivity, task execution, a recorded response or the assessment circumstances. This secondary analysis does not separate those causes. It also does not recode every stop as lack of ability. The product measures completion along the recorded protocol, not an underlying trait that the protocol perfectly reveals. A more ambitious interpretation would require respondent-level information and an explicit treatment of missingness and task conditions.

Seven groups, not seven independent populations

Tables 42 and 43 provide three sex/total rows and four enrollment rows. The present calculation retains all seven. For males, the source rates of 72.9%, 85.2% and 92.5% produce an observed complete-path percentage of approximately 57.5%. For females, 62.0%, 77.9% and 85.8% produce approximately 41.4%. The source total reconstructs to approximately 48.9%. These are distinct descriptions of overlapping groups, not components to be summed into a larger count.

Enrollment rows show why a generally successful final task can coexist with different total reach. Youth at Standard X or below have source rates of 62.5% bringing a phone, 79.1% finding among bringers and 86.2% sharing among finders. Their reconstructed complete-path rate is approximately 42.6%. For Standards XI or XII, the corresponding product is about 63.3%. For the undergraduate-or-other group it is about 73.4%. For respondents not enrolled, it is about 30.4%. The underlying conditional sharing rates alone span a much narrower range than those full-path products.

These comparisons do not isolate an effect of school enrollment. Ages, circumstances, prior education and other characteristics differ between the groups. The analysis neither standardizes their composition nor treats them as interchangeable intervention groups. It also avoids a ranking of districts: the selected tables supply pooled groups, not a comparable state-level or all-India league table. A distributor can recognize that later-stage task success and broad route reach are different planning quantities without claiming to know what caused a particular group's result.

Accounting for the male–female product difference

The reconstructed female-minus-male complete-path difference is approximately minus 16.0 percentage points. By comparison, the final conditional sharing-rate difference is minus 6.7 points. These numbers describe different quantities. The latter compares sharing among successful finders; the former incorporates differences in all three reported stages. This does not mean that expanding the denominator caused an additional disadvantage. It means that the wider-base result contains information that a last-step comparison leaves outside its definition.

To show how the three factors enter the arithmetic, the reference includes a symmetric replacement decomposition. Start with the three male source factors and replace them with the female factors in every possible order. For each order, record the change in the product when one factor is replaced. Average each factor's contribution across the six possible orders. This is a Shapley-style accounting identity. It prevents the answer from depending on whether the analyst happens to replace phone availability, finding or sharing first.

On the displayed inputs, the bringing factor contributes approximately minus 7.9 percentage points, finding among bringers approximately minus 4.4 points, and sharing among finders approximately minus 3.7 points. They sum to the complete-path difference. These are allocations of a mathematical difference, not causal estimates of access barriers, training effects or discrimination. Replacing a number in an equation does not establish that changing the corresponding real-world condition would leave the other factors unchanged. The decomposition is useful precisely when its limited accounting purpose is stated openly.

One should therefore resist turning the largest contribution into a claim about the most effective intervention. Providing a phone, changing connectivity, modifying a task and changing a distribution channel could affect several stages together. Selection into the later bases could also change. The present tables do not identify those responses. The chart explains which reported factors contribute to the existing product difference under a symmetric rule; it does not forecast how the difference would respond to a program or a publishing decision.

Rounding sensitivity is separate from sampling uncertainty

Each displayed input has one decimal place. The reproducible calculation assumes an enclosure of plus or minus 0.05 percentage points around each published value, clipped to the valid percentage range. The product is monotone over these nonnegative intervals, so multiplying all lower endpoints gives a conservative lower enclosure and multiplying all upper endpoints gives an upper enclosure. For the source total, the reconstructed path lies between approximately 48.8013% and 48.9888% under that tolerance. The central product is displayed as 48.9% for readability.

The paired gender difference uses the female product's lower bound minus the male product's upper bound for its lower endpoint, with the opposite endpoints for the upper bound. That produces an enclosure from about minus 16.2012 to minus 15.8245 percentage points. Ordinary input rounding within the assumed tolerance does not remove the difference in these published-rate products. This is not evidence that the underlying population difference is statistically significant. The enclosure includes no sampling variation, respondent-level covariance or uncertainty from the assessment itself.

The selected tables do not provide enough information for the new product's design-based standard error or the decomposition's sampling distribution. This study therefore supplies no confidence intervals, p-values or significance tests. It does not infer that a small difference is equivalent to zero or that a narrow rounding enclosure means high survey precision. Statistical uncertainty and deterministic input precision answer separate questions. Both should be visible when an attractive chart turns several rounded percentages into a newly calculated result.

What this says about reading-resource distribution

A publisher may place an explanation behind several actions: obtain a suitable device, locate a resource and then share or open it. This study is a concrete example of why the final action's success rate is insufficient to describe the entire route. It does not show that an ebook platform has the same rates, that young people prefer video to text or that the specified assessment is a valid test of reading a digital book. Those applications would need their own observed paths and clearly defined recipients.

The practical lesson is to define the quantity before choosing the headline. If the decision concerns usability after a resource has been located, a conditional success rate can be appropriate. If it concerns observed reach among the intended source group, the earlier stages belong in the denominator accounting. If it concerns underlying ability when device and connectivity barriers are removed, neither of these products answers it. Keeping those questions separate helps avoid overstating what a public statistic can support.

This reference also suggests a discipline for future measurement. Record who was eligible for each step, which steps were actually attempted, how nonresponse was handled and whether the later base is nested within the earlier one. Preserve a route that works without assuming private device ownership, and test it with the actual audience rather than transplanting percentages from this assessment. These are design questions prompted by the analysis, not evaluated interventions or promises of improved participation. They make the evidence useful while leaving the unmeasured outcomes unclaimed.

A bounded and reusable reference

A reader checking the recipe should inspect both the displayed rates and the connecting words in the table. “Of these” identifies the device-ready base for finding, while the sharing column narrows it again to successful finders. The same row must supply every factor. Taking a phone rate for all youth, a finding rate for enrolled students and a sharing rate for males would not estimate any defined source group. Likewise, averaging the seven reconstructed rows would give overlapping totals arbitrary weight. The provided total is reconstructed from its own three source inputs, not from an unweighted average of the subgroup results.

ASER 2023 surveyed 34,745 fourteen-to-eighteen-year-olds across twenty-eight selected rural districts in twenty-six states. Its pooled source group should not be relabeled all Indian teenagers, all rural India or the current audience for a digital product. The observations belong to 2023, even though the source was checked in October 2026. There is no before-and-after comparison, international ranking or claim that these rates apply to younger children, adults, cities or later versions of any platform. The selected task involved one specified video and a particular sharing procedure.

The public recipe contains the twenty-one numeric inputs, seven route reconstructions, rounding enclosures and a fully specified three-factor accounting. Original charts and accessible tables share the same bases. The source report, assessment instruments, photographs and original graphics are not redistributed. ASER Centre receives credit for the survey; Alisson Bento receives credit for this secondary calculation and explanation. No endorsement, new survey or universal digital-literacy scale is implied. The contribution is a transparent way to carry changing denominators through a real public example, so a successful last step does not silently stand in for the whole path.

The complete mobile route has a wider denominator

Seven original source groups, with each path stage reconstructed on that group’s full base.

Full chart and source
Bringing, finding and complete finding-and-sharing paths are shown on the full base for all seven groups.

Scroll horizontally to inspect the chart. ↔

Chart and values

Scroll horizontally to see all columns.

Every reconstructed mobile route and its bases
Source groupCould bring (%)Find among bringers (%)Share among finders (%)Find on full base (%)Complete path (%)Rounding lower (%)Rounding upper (%)
Male72.985.292.562.1157.4557.348457.5567
Female62.077.985.848.3041.4441.355641.5239
All youth67.181.689.354.7548.8948.801348.9888
Standard X or below62.579.186.249.4442.6242.529442.7009
Standards XI–XII76.789.192.668.3463.2863.171763.3936
Higher education / other83.292.895.177.2173.4373.304173.5487
Not enrolled56.761.787.034.9830.4430.367030.5050

Observed task path, not ebook access or true digital ability. Arithmetic accounting is not causal. ASER 2023 · India · 28 selected rural districts · ages 14–18. Alisson Bento original analysis.

Observed task path, not ebook access or true digital ability. Arithmetic accounting is not causal. ASER 2023 · India · 28 selected rural districts · ages 14–18. Alisson Bento original analysis.

Download chart (SVG) ↓ · Download chart (PNG) ↓

Three reported factors account for the product gap

Symmetric three-factor arithmetic accounting for the female-minus-male product difference.

Full chart and source
The three arithmetic contributions sum to about minus 16 percentage points; no cause or intervention effect is established.

Scroll horizontally to inspect the chart. ↔

Chart and values

Scroll horizontally to see all columns.

Three symmetric arithmetic contributions
Reported factorFemale-minus-male arithmetic contribution (pp)
Bringing a phone−7.9289
Finding, among phone bringers−4.3941
Sharing, given finding−3.6898

Average over all six factor-replacement orders. The sum is a descriptive identity, not a causal account or intervention ranking.

Observed task path, not ebook access or true digital ability. Arithmetic accounting is not causal. ASER 2023 · India · 28 selected rural districts · ages 14–18. Alisson Bento original analysis.

Download chart (SVG) ↓ · Download chart (PNG) ↓

Data & downloads

Scroll horizontally to see all columns.

Every reconstructed mobile route and its bases
Source groupCould bring (%)Find among bringers (%)Share among finders (%)Find on full base (%)Complete path (%)Rounding lower (%)Rounding upper (%)
Male72.985.292.562.1157.4557.348457.5567
Female62.077.985.848.3041.4441.355641.5239
All youth67.181.689.354.7548.8948.801348.9888
Standard X or below62.579.186.249.4442.6242.529442.7009
Standards XI–XII76.789.192.668.3463.2863.171763.3936
Higher education / other83.292.895.177.2173.4373.304173.5487
Not enrolled56.761.787.034.9830.4430.367030.5050

Observed task path, not ebook access or true digital ability. Arithmetic accounting is not causal. ASER 2023 · India · 28 selected rural districts · ages 14–18. Alisson Bento original analysis.

Scroll horizontally to see all columns.

Three symmetric arithmetic contributions
Reported factorFemale-minus-male arithmetic contribution (pp)
Bringing a phone−7.9289
Finding, among phone bringers−4.3941
Sharing, given finding−3.6898

Average over all six factor-replacement orders. The sum is a descriptive identity, not a causal account or intervention ranking.

Evidence & method

How we calculated it

  1. Inspect the actual source report, its task routing, table headers and footnotes. Verify all selected cells visually and preserve exact one-decimal source values as strings.
  2. Retain every row from the selected tables, including separate total rows. Record the full-group base and each later explicitly nested conditional base before multiplication.
  3. Calculate products using Decimal arithmetic. Convert percentages consistently; do not multiply unrelated marginals or assume independence to repair missing joint information.
  4. Enclose each one-decimal input within an explicit 0.05 percentage-point tolerance, clipped to [0,100]. Use monotone endpoint products and conservative paired-difference endpoints.
  5. Record the analysis specification after source inspection and informal arithmetic, before the complete scripted derivation. This is not blinded preregistration.
  6. Reproduce every CSV and JSON result from the supplied checksum-verified selected aggregate matrix using standard-library Python, without network access, accounts or respondent data.
  7. Use PDF page 63 (printed 57), Tables 42 and 43: male, female, total and four enrollment groups. P(path)=P(bring)×P(find given bring)×P(share given find and bring).
  8. Retain the source’s assessment device condition. A borrowed family or neighbor phone can qualify. Do not multiply household smartphone availability into this route.
  9. For the female-minus-male difference, average factor-specific product changes over all six orders of replacing male factors with female factors. The three contributions sum to the product difference; the decomposition is descriptive, not causal.

What the data cannot tell us

  • The source describes youth ages 14–18 in 28 selected rural districts of 26 states. The pooled source total is not a representative all-India, urban or all-age estimate.
  • Observations are from 2023. Inspecting the source in October 2026 does not make the measurements current, longitudinal or comparable to a different survey.
  • Products reconstruct explicitly nested published-base rates from rounded estimates. They are not exact respondent counts or independently validated microdata estimates. Underlying missing-response details are unavailable for the new products.
  • Input-rounding enclosures use an explicit ±0.05 percentage-point tolerance. They are not sampling confidence intervals, error guarantees, significance tests or evidence of equivalence.
  • Selected aggregate tables do not provide the covariance and cell-specific sampling uncertainty needed for the derived products or paired differences. No p-values or causal inference is supplied.
  • Enrollment and sex rows overlap with the separate total and are not summed or averaged to create another total. No age, education or other composition adjustment is performed.
  • This is a secondary calculation, not a new survey, first-ever finding, individual assessment, ASER endorsement or licence to reproduce assessment instruments.
  • The observed specified-video task route does not measure ebook access, reading comprehension, permanent connectivity, private device ownership or true digital ability. A stop in the path is not recoded as inability.
  • The source already documents digital and gender differences. The new contribution is limited to full-base products, rounding sensitivity and explicit symmetric accounting, not discovery of a digital divide.
  • Changing an earlier real-world condition can alter selection and later conditional rates. The arithmetic decomposition is not an intervention forecast or a ranking of causes.

Original sources

  1. ASER Centre, Annual Status of Education Report (Rural) 2023: Beyond Basics

    Tables 42 and 43 provide the 21 selected numeric inputs and explicit nested bases. The source permits a family or neighbor phone; household ownership is not an additional nested first stage.

    Accessed October 31, 2026

  2. ASER 2023: About the survey

    The source background describes the selected rural district sample, ages and fieldwork context. No national or all-age prevalence claim is made.

    Accessed October 31, 2026

  3. ASER Centre: ASER 2023 Beyond Basics source page

    The official ASER 2023 landing page identifies source reports and digital evidence assets; inspecting it later does not update the 2023 observations.

    Accessed October 31, 2026

  4. ASER 2023 Evidence Brief: Digital Readiness of India’s Youth

    The provider’s Digital Readiness evidence brief already explains access and gender differences. It does not supply the complete-path reconstruction or noncausal symmetric accounting used here; the original main report remains the sole numeric source for the selected three columns.

    Accessed October 31, 2026

Cite & reuse this study

Alisson Bento. 2026. A successful share is the last step: rebuilding a mobile resource route from ASER 2023. Secondary analysis of ASER Centre’s 2023 selected-district published aggregates. https://alissonbento.com/research/aser-mobile-resource-route-reach-2023/

Selected numeric source facts are credited to ASER Centre. Original calculations, explanation and figures are by Alisson Bento. The original charts and calculations may be reused with credit and a link to this study, retaining their scope and limitations. No licence to ASER report text, assessment instruments, photographs, source graphics or agency marks is inferred. No ASER endorsement is implied.

Corrections or editorial questions

Pass it on.

A successful share is the last step: rebuilding a mobile resource route from ASER 2023 | Alisson Bentoalissonbento.com

WhatsAppFacebookXTelegramRedditThreadsEmailSMS