What one would expect: That in a search for "mirror neuron" many results would be items which had "mirror neurons." At one time, there were 572 hits for "mirror neuron," and 502 hits for "mirror neurons." So, were all of those 502 plurals, included among the 572 singulars? No; actually, nearly two thirds of the plurals (327/502) were NOT in the singular list.

Here's a PubMed summary of what was sought, with the singular and plural terms (plus some text and a few red lines inserted with M.S.Paint).



44.2% (397) were uniquely singular, 36.4% (327) were uniquely plural, 19.4% (175) were duplicates. 64% of the plural hits were unique. One would expect that zero percent of the plural hits would be unique; that is, some "mirror neuron" hits would not match "mirror neurons" hits, but all the hits for "mirror neurons" should be included among those for "mirror neuron," because "neuron" is part of "neurons."

(Here's a clue: For PMID=24563573, which is #1 below in the "unique plural" list, I don't know whether "mirror neuron(s)" is in the full text or not, but on the abstract page, at the bottom is a list of keywords:

"KEYWORDS:
"EEG mu rhythm, action perception, infants, mirror neurons"

It may be that the grammatical fact of "neuron" being part of "neurons" doesn't apply for specifically designated keywords. Thus this item would come up in a search for "mirror neurons" but not "mirror neuron.")

Anyway--

I used an Excel spreadsheet, combining the results from singular and plural PubMed searches, then sorting on the leading phrase- The PMID index number- and then applied this formula in a parallel column:

=IF(AND(LEFT(E2,8)<>LEFT(E1,8),LEFT(E2,8)<>LEFT(E3,8)),IF(RIGHT(E2,13)="singular list","UNIQUE SINGULAR","UNIQUE PLURAL"),"")


(Actually, the PMID number is a little more complicated than that. See footnote1).

Click here for the result, sorted into 3 categories- Unique singular, Unique plural, and Duplicates.