There were criminal charges against children (and their parents), and locking them in prisons - but that was after the protests.
Anyway beating was only one of consequences of other kinds of persecution against Polish culture, language, identity and people.
I don't think you can make the case that the situation of polish students was nearly as bad as that of First Nations children in the Residential Schools.
Probably not. But still there was enough persecution that the Germans themselves created entire generation(s) of German-hating people.
And another effect of those policies was of course that they partially succeeded in causing people to adopt German language and identity.
So in general they caused increasing polarization of citizens of eastern provinces from "neutrals" into Polish-haters and German-haters.
The division usually went along linguistic (Polish-German) and / or religious (Catholic-Lutheran) lines.
In the Upper Silesian Plebiscite of 1921 according to
my calculation ca. 90% of monoglot Polish-speaking residents voted for Poland:
% of monoglot Polish among eligible resident voters (> 20 years old) / % of resident votes for Poland / votes for Poland as % of monoglot Poles:
1) Counties with % of resident votes for Poland lower than 50% of monoglot Polish-speakers eligible to vote:
[
note: most of these counties were counties with significant
Lutheran majority, also among Polish-speakers]
Leobschuetz - 5,12% / 0,58% ========== 11,33%
Kreuzburg - 41,93% / 6,64% =========== 15,84%
Namslau - 23,23% / 4,83% ============ 20,79%
Neustadt - 39,52% / 14,37% =========== 36,36%
Kosel - 70,90% / 30,38% ============= 42,85%
Ratibor Stadt - 25,80% / 11,39% ======== 44,15%
2) Counties with % of resident votes for Poland amounting to 50% - 100% of eligible monoglot Polish-speakers:
Oppeln Stadt - 13,13% / 6,59% ========= 50,19%
Rosenberg - 77,23% / 41,71% ========== 54,01%
Oppeln Land - 71,77% / 38,17% ========= 53,18%
Lublinitz - 75,99% / 53,03% ============ 69,79%
Gross-Strehlitz - 75,58% / 58,70% ======= 77,67%
Toest-Gleiwitz - 72,40% / 62,91% ======== 86,89%
Koenigshutte Stadt - 29,54% / 27,15% ==== 91,91%
Pless - 83,17% / 80,24% ============== 96,48%
Beuthen Stadt - 28,51% / 28,10% ======== 98,56%
Kattowitz Land - 59,93% / 59,12% ======= 98,65%
Rybnik - 74,00% / 73,14% ============= 98,84%
3) Counties with % of resident votes for Poland higher than 100% of number of eligible monoglot Polish-speakers:
Beuthen Land - 57,93% / 61,72% ======== 106,54%
Tarnowitz - 62,16% / 67,11% =========== 107,96%
Ratibor Land - 44,36% / 49,05% ======== 110,57%
Hindenburg - 45,93% / 52,02% ========== 113,26%
Kattowitz Stadt - 10,98% / 13,85% ======= 126,14%
Gleiwitz Stadt - 12,25% / 23,01% ======== 187,84%
4) Average figures for the entire Plebiscite Area:
% of monoglot Polish among eligible resident voters (> 20 years old) / % of resident votes for Poland / votes for Poland as % of monoglot Poles:
TOTAL - 51,74% / 46,71% ========== 90,28%
So - if we trust the data from German 1910 census on the percent of Polish-speakers among the population -
around 90,28% of monoglot Polish-speaking inhabitants voted for Poland in the plebiscite. In those six counties in which % of votes for Poland was higher than % of monoglot Polish-speaking inhabitants, that was of course due to bilingual Polish-German speakers and monoglot German-speakers voting for Poland.
Notes:
According to the census of 1910, monoglot Polish-speakers had a different age structure than monoglot German-speakers. Among monoglot Polish-speakers only 448 out of 1000 were aged over 20 (= eligible to vote in the plebiscite), while among monoglot German-speakers as many as 561 out of 1000. German-speakers had a higher median age than Polish-speakers. That was of course disadvantageous for the Poles in the plebiscite, as they were a much lower % of the electorate than of the population. In my calculations I assumed that bilingual Polish-German speakers had 500 eligible voters per 1000 population. I also took the data on % of Polish-speakers among the population from the census of 1910 and extrapolated them to 1921, but in fact in 1921 the proportion of Polish-speakers could be lower than in 1910 since it was constantly decreasing (which is evident when we compare the censuses of 1890, 1900, 1905 and 1910). In my calculation I extrapolated the percent of Polish-speakers from 1910 to 1921 and also the age structure of inhabitants from 1910 to 1921.