Source documents behind the assumptions
The NAAC, NIRF and regulator documents this analysis rests on are archived at
gs://avantifellows-private-data/analysis/nirf-college-employability/sources/Private bucket, deliberately. Two reasons. The per-student placement lists carry named students alongside their employer and salary, which is personal data and must not be republished. And the institutional filings are other organisations’ documents: we keep a copy so a broken link does not cost us the evidence, but we do not need to re-host them publicly. Verified private: the public URL returns 403.
Original URLs are below so anything here can be re-fetched from source.
per_student_lists/ — CONTAINS PERSONAL DATA, do not republish
Section titled “per_student_lists/ — CONTAINS PERSONAL DATA, do not republish”| File | Institution | What it evidences | Original URL |
|---|---|---|---|
dtu_5.2.1_list_of_placed_students_2019-2024.pdf |
Delhi Technological University | The only measured within-college salary distribution we have. 209pp, per-student name / employer / package, 2019-20 to 2023-24. Gives p25 ₹8.00L, median ₹10.50L, p75 ₹15.10L for 2023-24, implying σ = 0.471 against the 0.45 we assume. Also the only NIRF-vs-NAAC salary comparison available: ₹10.50L here against ₹13.25L filed to NIRF. | http://iqac.dtu.ac.in/naac/naac2/criteria5/5_2_1.rar |
placed_2014_15.pdf … placed_2018_19.pdf |
Lovely Professional University | Retrieved from the Wayback Machine after the live URLs 404’d. LPU’s SSR web pages publish no figures and its entire /files/ tree is gone. |
https://www.lpu.in/files/5.2.1/Supporting%20Documents/ (dead; via web.archive.org) |
institutional/ — SSRs, AQARs and regulator documents
Section titled “institutional/ — SSRs, AQARs and regulator documents”| File | What it evidences |
|---|---|
amrita_ssr_2020.pdf, amrita_aqar_2021-2022.pdf, amrita_aqar_2022_23.pdf |
Amrita placement 42–61% university-wide against 71% filed to NIRF for UG engineering. No salary anywhere: the university SSR template does not ask for it. |
srm_ssr_2024.pdf (+ text) |
SRM contradicts itself on the same academic year. SSR reports 7,906 placed for 2021-22; its own AQAR reports 5,571. Same regulator, near-identical denominators, 18 percentage points apart. |
kare_ssr3.pdf |
Kalasalingam 91.4% average placement 2017-22, consistent with the 92.1% it files to NIRF. Also the institution whose NIRF figures exposed the duplicate-row bug. |
dtu_aqar_19-20.pdf, dtu_aqar_21-22.pdf, dtu_aqar_23-24.pdf |
DTU’s AQARs disagree with its own SSR on placement counts for 2021-22 and 2022-23 (transposed), and 2019-20 records placement as “No Data Entered”. |
au_AQAR_2020_21.pdf, au_AQAR_2021-22_Submitted_29.07.2023.pdf |
Anna University placement 39–64% university-wide. No SSR published; the NAAC portal copy is unreachable. |
asci_oct.txt, asci_nov.txt, asci_aug2017.txt, asci_pdfs.txt, asci_pages.html |
ASCI Consumer Complaints Council rulings upholding placement claims as misleading against Chandigarh University, LPU, SRM, Adi Shankara and ~55 others. All private. The strongest evidence that private institutions’ placement advertising is misleading, as distinct from their NIRF filings. |
cag_od.pdf |
CAG Odisha: the state government conceding that placement figures “are not actual ones”, and 32 test-checked colleges maintaining no placement data at all. |
marisha.pdf |
Current Science 120(7), Monte Carlo analysis of India Rankings stability. Graduation Outcomes attains 0.26–0.28 effective weight against a nominal 0.20. |
Not archived, but load-bearing
Section titled “Not archived, but load-bearing”These are stable government URLs, re-fetchable, and were read directly rather than saved:
- NIRF framework PDFs,
nirfindia.org/nirfpdfcdn/2025/framework/*.pdf— placement and salary are marked “Primary Data: To be provided in the prescribed format” while research metrics say “Third-party sources”. Clause 5.3, the only misreporting penalty, appears only in the Overall framework, which scores no placement. - NIRF India Rankings 2025 report §6.5.2 — “the final responsibility for the accuracy of the submitted data lies with the concerned institutions”.
- NIRF scorecard PDFs for
IR-E-U-0458andIR-E-U-0436— the authoritative check that exposed the duplicate-row bug innirf_fact_aggregate. - Lok Sabha Q.1483 (09.02.2026) — the Ministry describing the figure as median salary “as declared by” institutions.
- EPFO contribution rates — the basis for the CTC decomposition behind the 20% salary haircut.
Re-fetching
Section titled “Re-fetching”Everything in institutional/ came from an institution’s own IQAC or NAAC page. Those move. The
NAAC portal itself, assessmentonline.naac.gov.in, where every per-student supporting file
officially lives, was unreachable throughout this work: DTU’s list was only obtainable because
DTU mirrors its own files.