# Graduate Center Digital Initiatives annual reports

Eight distinct reports provide evidence for **10 academic years from 2012–2013 through 2024–2025**. The supplied 2012–2014 Digital Fellows report covers two years, and the 2015 report explicitly describes 2014–2015 activity. The supplied 2015–2017 file is byte-identical to the supplemental report already found through WordPress. Five additional reports are listed in the May 1, 2026 archived About page. [Coverage by year](years.csv) · [Source register](sources.csv) · [Archived About page](https://web.archive.org/web/20260501182353/https://gcdi.commons.gc.cuny.edu/about/)

**No report located:** 2017–2018, 2018–2019 and 2019–2020. **2025–2026:** PDF supplied earlier and available, with extraction still reserved for manual addition. The early reports focus on the Digital Fellows program, while later reports cover more of GCDI. A year with report evidence can still lack a separate annual roster or an annual activity total.

## Activity by academic year

Counts are taken from the annual reports and linked to physical PDF pages, starting at the cover. Summer activities are marked separately from fall and spring. A blank or different measure means the sources provide different evidence.

| Academic year | Workshops reported | Consultations or closest reported measure | Digital Research Institute | Listed GC Digital Fellows |
| --- | ---: | --- | --- | ---: |
| 2012–2013 and 2013–2014 | 9 on a reported schedule, with annual totals unavailable | More than 30 faculty and 250 hours over three semesters | No figure in these reports | 8 in a single two-year roster |
| 2014–2015 | More than 200 students served in the Python series | More than 12 digital projects supported during spring office hours, as reported in the 2015 snapshot | No figure in this report | 10 in the 2015 report snapshot |
| 2015–2016 | 31 | 26 faculty members consulted | January and June institutes reported separately by date | 11 |
| 2016–2017 | 27 | 14 faculty consultations | January institute | 11 |
| 2020–2021 | 16 online | 54 student requesters and more than 100 hours of one-to-one and group support | 28 students | 8 |
| 2021–2022 | 15 online | 40 or more sessions | 40 students | 8 |
| 2022–2023 | 23 | 40 sessions | 20 participants | 8 |
| 2023–2024 | 14, plus 5 summer Python workshops listed separately | 37 sessions | 23 participants and 25 seats | 6 |
| 2024–2025 | 17 fall–spring, plus 8 summer Python workshops listed separately | 29 sessions | 37 participants and 37 seats | 8 |
| 2025–2026 | Manual addition pending | Manual addition pending | Manual addition pending | Manual addition pending |

[Activity rows and page numbers](activity.csv) · [Staffing rows and page numbers](staffing.csv)

## What changed

- **Workshops remained a core service, with changing formats and reporting.** The reports list 16 online workshops in 2020–2021 and 15 in 2021–2022. The 2022–2023 report lists 23 Digital Fellows workshops and 383 registrations, which can include repeat sign-ups. It separately lists 17 workshops within the Digital Research Institute. The next two reports separate summer Python workshops from their main annual workshop count. That distinction matters when comparing 14 and 17 across the last two years. [Activity evidence](activity.csv)
- **Reported one-to-one consultations fell from 40 to 37 to 29 between 2022–2023 and 2024–2025.** That is 11 fewer sessions, or 27.5%, using the three exact community session totals. The 2020–2021 figures describe requesters and hours. The 2015–2017 report alternates among people, faculty-only sessions and a two-year headline. The report figures alone cannot explain the change in demand or availability. [Activity evidence](activity.csv)
- **Institute participation varied.** The Graduate Center Digital Research Institute moved online in 2020–2021 and offered Python and R tracks in 2021–2022. Reported cohort sizes then went from 40 to 20 to 23 to 37. The 2023–2024 report gives 25 seats and 23 participants on the same page. The 2024–2025 report uses 37 for both. The undergraduate Digital Project Institute started with 18 CUNY undergraduates in 2022–2023 and had a distinct semester-long mentoring component. [Program evidence](programs.csv) · [Activity evidence](activity.csv)
- **Core staffing continued while people moved between roles.** The five main reports list 8, 8, 8, 6 and 8 GC Digital Fellows, with 3, 5, 4 and 4 shared names between successive rosters. The last pair shares 4 of 10 distinct listed names. The 2022–2023 report lists Chelsea Lane as a postdoctoral Fellow and Param Ajmera and Di Yoong as Digital Scholarship Center Fellows after listing them as GC Digital Fellows a year earlier. Sam O’Hana appears as a GC Digital Fellow in 2023–2024 and a Data Visualization Advising Fellow in 2024–2025. These observations describe reported roles, with no inference about appointment dates or workloads. [Roster evidence](staffing.csv) · [Overlap counts](staffing_overlap.csv)
- **The reports describe overlapping routes into digital scholarship.** More than half of 2022–2023 Institute participants joined additional workshops. The 2023–2024 and 2024–2025 Digital Showcases gave Praxis students and Provost’s Digital Innovation Grant recipients a shared venue. The 2024–2025 report records the launch of Conversations in Digital Scholarship and a Humanidades Digitales blog, plus the public debut of the DHRIFT teaching platform and an April halt to its federal funding. [Program evidence](programs.csv) · [Activity evidence](activity.csv)

## What the earlier reports change

### Work visible outside the blog

The 2012–2014 report documents a portfolio that includes faculty consultations, Digital Praxis teaching, collaborative website builds and ongoing maintenance. It names Andrew McKinney as consultation lead on PDF p. 4, and Keith Miyake and Laura Kane as leads for building and maintaining the Advanced Research Collaborative Commons on p. 17. The 2015 report names the work of curating, organizing and co-teaching the workshop series on p. 5 and the website redesign on p. 24. These are specific examples of labor that a blog post count cannot measure. [Program evidence](programs.csv)

The earlier report also describes Tagging the Tower as part of teaching and outreach. It points to the *Learn Bootstrap* tutorials and reproduces a screenshot of the post *tech versus culture*. The Islamic History Commons is a second bridge across sources. Both early reports name Erin Glass as project lead, and the export contains a launch announcement. The project lead designation and the announcement’s posting-account credit are separate evidence about separate work. [Cross-source matches](archive_matches.csv)

### Earlier demand, familiar reporting limits

The 2015 report says 96% of workshops sold out and 65% had a waitlist across its first five workshop series. Its Introduction to Python example shows more than 130 students expressing interest in 60 slots. This gives early evidence of demand exceeding capacity. It leaves open how many people attended or went on to other programs. The repeated faculty-consultation figures in the two early reports describe the same three-semester claim and appear once in the dataset. Likewise, the website visitor figure repeated in the 2015 and 2015–2017 reports is a site-wide snapshot with unspecified dates. [Activity evidence](activity.csv)

### Rerun of archive and roster queries

- **Discovery by title.** Among 83 early local published records from 2012–2015, including any summer records, 12 titles mention a workshop or recap and none match the consultation, office-hours or drop-in screen. The all-archive counts remain 34 and 4 among 325 local records. This is a title test. The absence of a title match leaves open what a post discusses in its body. [Title screen and search patterns](archive_title_visibility.csv)
- **Publication mix.** The three early fall–spring totals are 38, 27 and 18, all local or unflagged in the saved ledger. The later contrast remains: the drop from 42 records in 2021–2022 to 24 in 2024–2025 reflects curated records falling from 25 to zero, while local or unflagged records rise from 17 to 24. Report coverage and blog publication are distinct. [Publication mix](archive_publication_mix.csv)
- **Roster comparison.** All eight names on the 2012–2014 report roster match the union of the two directory years under normalized exact spelling or existing directory variants. The report supplies one combined list. The 2015 snapshot has seven exact matches to the 2014–2015 directory under the same rule. Kenneth U. Ezirim / Kenneth U. Ezrim and Laura Wildemann Kane / Laura Kane or Laura Wildermann Kane remain visible for review, along with Michelle Johnson’s report-only listing under this comparison rule. No new identity or appointment assignment enters the project roster. The five exact matches and two disagreements among the later annual rosters persist. [Roster comparison](roster_source_comparison.csv)
- **Credits and membership.** The 325 local records now include 93 source-supported, 26 declared and 206 unresolved-account credits. Among 147 local posts with an original signed roster match, 105 still have unresolved-account credit. Ian’s clarification about migration and archived bylines supported 43 historical corrections. Reported project roles still describe a separate kind of contribution. [Coverage audit](../coverage/coverage_audit.json) · [Post membership matches](../membership/post_membership_matches.csv)

The earlier evidence shifts the emphasis toward **making and maintaining digital scholarship**, alongside teaching it. A useful follow-up would compare reported project roles with surviving announcements, guides and project pages. Which people appear as leads or collaborators in the reports but have little searchable evidence of that work in the blog? Which of the sites and teaching resources remained usable after the initial launch? The materials collected here support questions about visibility and maintenance, with appointment timing, undocumented contributions and later project status still open.

## WordPress cross-check

The November 1, 2026 WordPress export of Digital Initiatives at the Grad Center supplied publication dates and an older report link. Post **6779** points to the 2015–2017 report. That report, PDF p. 50, counts **28 Tagging the Tower posts** on the Digital Fellows site in 2016–2017. The project coverage ledger and export dates yield **21 published archive records** for that fall and spring (9 fall and 12 spring). The counts come from different sites and inclusion rules. A seven-post difference cannot establish which items were omitted or whether all 21 are in the report’s 28. The 21 comparison records and the link-bearing post have stable IDs in the dataset. [Archive matches](archive_matches.csv) · [Coverage methodology](https://github.com/GCDigitalFellows/TTT-Archive-Explorer/blob/main/analysis/coverage/README.md)

The join used published post IDs marked included in the project coverage ledger and original local publication dates from the export. Fall is August–December and spring is January–May. June–July are summer and excluded from the fall–spring post counts. The report’s stated year governs its annual totals. Explicit summer and two-year items are labeled separately. Counts from the WordPress export measure publication evidence, independent of report coverage, so the pending year may still have archive counts. [Project calendar rule](https://github.com/GCDigitalFellows/TTT-Archive-Explorer/blob/main/analysis/analyze.py)

## Using the files

- [Years](years.csv) is the coverage index and includes gaps and the manual-addition row.
- [Sources](sources.csv) has the report IDs, original media URLs, file hashes, physical page counts and extraction methods.
- [Activity](activity.csv) keeps units and qualifiers with each value. “As stated” records the figure as printed or explicitly derived. “More than” and “At least” preserve the reported thresholds. The 2020–2021 partial workshop-registration sum uses the 14 counts displayed on pp. 8–11.
- [Staffing](staffing.csv) records a named person and role as transcribed from report roster pages. [Staffing overlap](staffing_overlap.csv) compares adjacent documented rosters for the GC Digital Fellow role only. Name matching is exact after case and typographic normalization as needed. Role counts do not represent distinct people across different roles.
- [Programs](programs.csv) records selected program evidence, milestones and links across activities. A program’s omission from these rows leaves its status open.
- [Archive matches](archive_matches.csv) is a selected export cross-check with IDs and URLs. Run `python3 analysis/annual_reports/compare_archive.py` from the project root to regenerate the publication-mix, title-screen and roster-comparison CSVs.

To add 2025–2026, update the source register with the supplied PDF’s page count and checksum, then add page-sourced activity, staffing and program rows. Update the year index and staffing overlap from the new roster. Keep any summer activity separate when the report distinguishes it.

## Archived website evidence

The [web archive source register](web_archive_sources.csv) records timestamped About, People, and program-page captures with their evidence uses. The [reviewed credit ledger](../attribution_revision/reviewed_credits.json) links the 43 original article captures used to restore writer credits. Roster and program pages document roles and activities. Article captures supply byline evidence.

## Evidence limits and contributions

The live About page returned a CAPTCHA. Two April 2026 captures and the May 2026 capture listed the same five PDFs. The downloads use the exact URLs in those captures. The separate 2015–2017 PDF was retrieved through the Commons media mirror from the path referenced in the export. Extracted text preserves PDF page breaks. OCR can scramble columns, names, figures and graphic labels. The 2015–2017 extraction has six pages without embedded text and the 2024–2025 OCR has one page without text. Consult the originals for graphics or precise transcription.

Ian G. Williams defined the scope, pointed to the About page and supplied the WordPress export, three early report files and the PDF reserved for manual addition. OpenAI Codex recovered sources, curated the dataset, drafted the comparison and checked selected values. Apple Vision and pypdf performed text extraction. The report writers retain credit for the original evidence. No independent human verification is recorded. Source reports retain their own rights.