Correct data provided for ungraded surveys. The CSV file formats the data exactly like quizzes with correct answers. Ungraded Surveys require data for each answer in a multiple answer question, not just a score for the question. Survey data requires the total count for each possible answer. Currently the ungraded survey data included in the CSV is assume that the first item in a multiple choice survey is the correct answer and reports the score for the assumed correct answer, but not the total for each option within a question. An average total score (grade) for the entire survey is meaningless in a survey.
Instead we need to be able to calculate basic statistics for each item/question in the survey. In order for that to happen, the CSV data file needs to be reformatted. The CSV needs to provide a response count for each potential answer, for each question/item.
Example ungraded survey with data from respondents correctly/usefully presented (all answer options are correct):
Same ungraded survey example with data as currently provided in CSV file, incorrectly presented (assumes first answer option is correct):