If copyright is respected, so there is no problem. Try to use plagiarism software. It will help you. About data of other studies, it should be allowed based on the related authors. In general, you should read the rules of a journal which have an high impact factor value.
This has been done by many psychometrics researchers (including myself). What you must do is to cite the origin sources of the survey items (regardless of whether they were published or not) and mention the specific changes you made the items at hand. Presuming that those items are indeed ungrammatical, changing their grammar will not be plagiarism at all.