There are several confusing aspects to this survey.
First, the item numbering system uses a combination of Roman and Arabic numbers. This should be replaced with a simple numbering system (1,2,3...). You also may want to break two-part questions apart to give them separate numbers (see I.1 and II.10). For the items in a series (II.1-8), it is neater to eliminate numbers for each stem and simply use one number for the entire set. The subquestions then can be lettered.
A second problem is an inconsistent answer format. The survey has a combination of check-the-blank, check-the-box, and circle-the-numbers formats. Using only a check-the-box format would be better. Even for items II.1-8, boxes can be used as long as you create headings to clarify the meaning of each column. Furthermore, for items I.4 and II.9, the answer choices run horizontally. Running the answers vertically would improve the survey's visual flow and consistency.
A third problem is poor differentiation of text. The same unmodified font is used throughout and the spacing is cramped. At a minimum, there should be a difference between the size of the font used for questions versus answers, and more white space between items.