Data

I guess there are no strict rules about this.
Google search shows 12.000.000 matches for "data was" and 16.000.000 matches for "data were".

In Latin, "data" (plural) is a collection of "datum" (singular). But nobody uses the word "datum" in modern English.

data is plural. the data are supporting my hypothesis.

Data is a non-countable noun, like food.

You don't say "All the food were great at the buffet."

It's slightly sad that what was once wrong is now acceptable. Data is plural, but since no one knows that anymore, it's become okay to treat it as if it were something else. No one says "I have two data" anymore, even though that's perfectly correct. They would say "I have two data points."

It's like using "they" as the genderless version of "he" or "she".

Jon: food (and linen, cutlery etc.) is a collective noun and has no singular form - unlike data.

Data is the plural of datum (L, 2nd declension) and should be used as such.

But we should distinguish use in computing (and related disciplines) where it is strictly a singular noun (by overwhelming preponderance of evidence).

Since 'data' is less and less frequently used outside the field of computing, though, 'datum' will likely disappear from use altogether and 'data' will finally become a true collective noun.
But I don't think we're quite there yet.

[Interestingly, 'agenda' (sing. agendum) took the other path and became a true singular noun.]

I admit to a fondness for "the data are", but then I have a fondness for "the media are". Both were indisputably the correct form in the past, but I think this battle's completely lost (at least in American English; not sure about the British form). I'd go so far as to say that "the data are" (just like "the media are") has a vaguely quaint and obsolete sound to it.

It matters not one iota what the plural or singular forms are in another language. When a word is borrowed it doesn't come with rules attached. The rules that guide are those of the "new" language.

Studies show that both 'is' and 'are' are used with about the same frequency.

Data is generally an uncountable noun.

"water was drunk"
"data was handled"