Q47, run 2026-11-06-0519-sonnet-query

I07's 14 works are mostly written by I07 authors, so they keep nearly their full share and fractional counting gives 12. I09's 16 works have many authors from other institutions, so its share drops to 4.

Why: Says Kesting (I07) keeps larger shares because its authors dominate, but doesn't say Rostval's works have several authors with only one from Rostval; it does say I09 has many outside authors, which is roughly equivalent, though institutions aren't named

Under full counting each work counts once per institution, so Rostval's 16 works beat Kesting's 13. Under fractional counting Kesting's works are mostly credited to its own authors (11.0), while Rostval's are shared with other institutions (4.0).

Why: Says Rostval's works are shared with other institutions (small share), but doesn't say Kesting holds larger shares; it says Kesting's works are mostly credited to its own authors, which does match the second point, though it omits that Rostval's works mostly have one Rostval author among several.

I09's 16 works are heavily co-authored with people from other institutions, so each work yields little. Its fractional total is only 4.0 against I07's 12.0. I07's works have few authors, so its fractional total stays near its work count.

Why: Explains the low-share side (I09, presumably Rostval) due to co-authorship, and that I07 has few authors, but doesn't say only one author is from that institution or that Kesting has many sole-institution works; roughly matches both points loosely using IDs

Rostval's 2024 works have many authors, so each work's credit is split into small shares (4.0 in total). Kesting's works have fewer authors, so it keeps 12.0 and moves ahead.

Why: Says Rostval's works have many authors so shares are small, and Kesting has fewer authors; but doesn't state that only one author per work is from Rostval, and Kesting's explanation is only loosely aligned (fewer authors, not Kesting-alone/few co-authors).

I09 has more works, but they are multi-author with shares going to other institutions. I07's works are mostly credited to I07 authors, so it leads on fractional counting (12.0 vs 4.0).

Why: Explains Kesting's (I07) larger shares from works mostly credited to its own authors, but doesn't say Rostval's works are multi-author with only one Rostval author getting a small share; it says shares go to other institutions, which is roughly aligned yet vague, and it doesn't name the institutions or the 2024 swap clearly.

Full counting gives each institution one count per work, so Rostval (16) beats Kesting (13). Fractional counting splits each work among its authors, and Rostval's works have many authors credited elsewhere, so its share drops to 4.0 while Kesting keeps 11.0.

Why: Says Rostval's works have many authors credited elsewhere (matches point 1) but doesn't say Kesting holds larger shares, e.g. sole or few co-authors; only states Kesting keeps 11.0.

Full counting favours I09 because it has more works, 16 against 13. Under fractional counting I09's works are split among many authors, so most of each work's share goes to other institutions and countries. That leaves I09 with 4.0 against 11.0 for I07.

Why: Covers Rostval's (I09) small shares from split authorship, but never says Kesting's works are mostly solo or have few co-authors; it gives only the 11.0 vs 4.0 totals.

Kesting (I07) has fewer works in 2024 than Rostval (I09) under full counting (13 against 16). Under fractional counting it has more (11.0 against 4.0). Rostval's works have many authors, so each work's share is split thin. Kesting's works have few authors, so its shares stay large.

Why: Says Rostval's works have many authors so shares are thin and Kesting's have few authors so shares stay large, but omits that only one author per work is from Rostval and that many Kesting works are by Kesting authors alone.

I07's works have more co-authors from other institutions, so its fractional total (11.0) falls below I09's... I09 leads in full count (16 vs 13) but I07 leads fractionally (11.0 vs 4.0). I09's 2024 works are thus heavily shared with other institutions' authors.

Why: The final point that I09's 2024 works are heavily shared matches Rostval's small share per work. But the answer contradicts itself (its first sentence says I07 has more co-authors), uses IDs instead of names, and never says Kesting's works are mostly solo or few-author.

I09 has more works in total (16 vs 13) but they are heavily co-authored with other institutions and countries. I07 therefore overtakes it under fractional counting (11 vs 4).

Why: Says I09 (Rostval) is heavily co-authored so gets small shares, but doesn't say only one author is from Rostval per work, and doesn't state that Kesting holds larger shares via solo or few co-authors; also uses IDs rather than names.

Kesting's 2024 works have mostly Kesting-only author lists, so its fractional share stays high (11 of 13). Rostval's 16 works are shared with other institutions' authors, so its fractional total falls to 4.

Why: Says Rostval's works are shared with other institutions so its share is small (meets point 1), and Kesting's are mostly Kesting-only (meets point 2); but it doesn't explicitly say Rostval has only one author per work, and the 'Kesting-only' claim is stated. Treated as largely meeting both, yet the 'only one from Rostval' detail is missing.

Kesting's 13 works in 2024 are mostly written by few authors, so each work gives it a large share. Rostval's 16 works are written by many authors from other institutions, so its fractional total shrinks to 4.0.

Why: Rostval's small share from many outside co-authors is covered, but Kesting's larger shares are only vaguely stated as 'few authors', not explicitly authors from Kesting alone or with few co-authors; also the Rostval point says 'many authors from other institutions' rather than only one from Rostval

Kesting has fewer 2024 works than Rostval (14 against 16). Its works have few co-authors, so its fractional share (12.0) is far larger than Rostval's (about 4.4), which comes from works with many authors from elsewhere.

Why: Gets Kesting's large fractional share (few co-authors) and Rostval's many-author works, but doesn't clearly say only one Rostval author per work gives it a small share; also adds counts not in checklist.

I did not query this directly. Rostval's works are co-authored with many people at other institutions, so each Rostval author carries a small share, which gives 4.0 fractionally. Kesting's authors hold larger shares, which gives 12.0.

Why: Covers Rostval's small shares from co-authorship and Kesting's larger shares, but doesn't say Rostval's works have only one Rostval author each or that Kesting's are sole/few-coauthor; admits not verified, still roughly matches both points

Rostval leads under full counting because it has more works (15 against 10). Under fractional counting Kesting leads, so Rostval's works are spread across more authors per work.

Why: Says Rostval's works are spread across more authors per work (matches first point loosely, doesn't say only one is from Rostval), but omits Kesting holding larger shares via sole or few co-authors.

Rostval has many works where it is one of many authors, so under fractional counting it earns small shares. Kesting's works have fewer authors, so its fractional total stays high.

Why: Rostval point is covered (many authors, small share), but the answer doesn't say only one author is from Rostval; Kesting's fewer authors is vague and misses that Kesting authors hold larger shares (sole or few co-authors).

The two swap places because full counting counts each work once per institution, while fractional counting splits each work among its authors. I09 leads on full counts (16 against 13), but its works have more authors, so each share is smaller. This is inferred from the full counts; I did not obtain the fractional figures.

Why: Gives the general mechanism and says the full-count leader's works have more authors with smaller shares, but never explicitly says Kesting holds larger shares (sole or few co-authors), and the fractional figures were not checked.

Full counting gives each work to I07 or I09 once, regardless of how many authors it has. Fractional counting splits a work among its authors, so I09's works, which have fewer authors each, carry larger shares and I09 moves ahead.

Why: Says I09 (Kesting) has fewer authors per work and larger shares, but omits that Rostval's works have several authors with only one from Rostval, giving it small shares.