(M)ANOVA or not to (M)ANOVA: Data Analysis Selection Confirmation/Help

#1
Let me first start off by saying, gosh I hate to admit this to you all, I'm a single-case researcher. Group design and statistical methodology are not my strong suit. I know enough to be dangerous--so please excuse my ignorance. I am here to learn though, so any help would be appreciated! I'm hoping someone can give me some guidance on a group design project we're proposing to run in the spring. Any thoughts would be greatly appreciated.

Questions for the forum (see background below):
Are the 5 activities actually considered separate IVs or they 5 individual DV's (I'd assume IV's)?
The DV appears to me to be continuous, but do the procedural constraints/the choice-paradigm make them more functionally a dichotomous/binary type measure?
Since we're attempting to order the activities based on the percentage of selection, would then cause our data to be ordinal? If so, how would this affect data anaylsis strategies?
Would a MANOVA be an appropriate data analytic strategy to parse out the proposed question (below)?
What alternative strategy(-ies) could or should be used?

Study Background:
Two groups: a "written description" control group; a "video description" treatment group (random equal assignment)
Measure(s): the percentage of item selection (when available) for 5 different activities (i.e., baseball, basketball, soccer, hockey, or lacrosse); For example, on average participants selected baseball 60% of the time when available. We're primarily interested in what was selected by respondents, we are not interested in their perceptions or a Likert scale rating for each activity, rather the particular percentage at which an item was selected over the available alternative. We would use the average percentage of selection for each activity to then establish a hierarchy within the group (i.e., which activity was selected first, second, third, so on for each group). We'd also compare the percentage of selection across the groups. Demographic information will also be collected and assessed in relation to the results (assuming a MANCOVA?)
Procedure: Participants will be presented with two simultaneously available activity option choices from which they will choose one. This will repeat until each activity is presented at least once with every activity for all participants (minimum of 10 choice trials per respondent; I'm assuming this will likely need to have more measurement iterations depending on the sample recruited). The activities will remain static in their content, but one group will experience all written choice options, the other will have to select from video options.
RQ: In the most basic of sense we're trying simply evaluate, does the format in which activity options are presented (i.e, written or video) affect the selection of an activity? We'd look at comparing the percentage of selection for each activity within a group to see which activity was selected most-to-least, then we would see how selection (possibly the hierarchy of) differed between the two groups. In further analyses, we might like to see if demographic info provided by respondents influence the choice

I originally thought a MANOVA would be appropriate because we have two groups and there is an individual measure for each of the 5 activities. However, given my questions below, I've been told that the measurement, while appearing continuous could be more accurately conceptualized as discrete or binary (e.g., it is either selected or not selected and the percentage is only a 'pseudo' continuous variable) which complicates our goal and the method to be used. Is it also possible to conduct statistical analyses on the rank order within and across groups using this method?

Any help would be GREATLY appreciated! Thank you!