[Tristan Lall]

A few comments on survey methodology:

• It's conventional to put the demographic questions at the end, because a lot of respondents are somewhat wary of providing it. If they feel that way, and quit before completing it, at least you have their other substantive responses.¹ (If the demographics are a core part of the analysis, this isn't always valid.)
• Questions like
  

  How did you get involved with mentoring robotics?
  Parent of a student? Sponsor involvement? Experience as a student yourself in robotics competition? Teacher at a school?

  lend themselves to multiple choice responses, because that yields categorical data straight from the survey. (Compared to free-form data, categorical data is more powerful for making generalized conclusions, because there are relatively unambiguous ways to compare it with other data.) By instead providing a text box, you end up having to parse the data and code it for key concepts, which is time-consuming. If you want to hear the stories people tell in answering that, consider multiple choice first, plus an optional text box.
• Alternatively, sometimes you don't know what categories you need...so in that case, a trial survey can be conducted where you gather a small (statistically insignificant, but likely sufficient to guide your intuition) sample, and generate your full list of categories based on the ideas people provide.
• There are some subtle assumptions that you (probably) shouldn't make in crafting the questions. For instance, how should a mentor for an all-year school respond to the questions about summer and the school year? And should a New Zealand team member respond in terms of their summer, or the North American summer that more closely defines a robotics season?
• Since you're looking for trends over time, you may want to ask about how a person's responses would have changed over the years. Personally, my involvement is quite different than it used to be, but that's not captured with the present-focused questions.
• A simple privacy policy would be nice. (Say how the data will be used and shared. Will the data be available for peer review?)
• You may prefer to use a survey service like SurveyMonkey, because it's designed to implement the sorts of refinements that social scientists use when they do research.

Despite those comments, this is a very good survey, and I look forward to hearing what you conclude.

¹ Though I'm not sure that's actually the case with Google Forms; it looks like if they quit, you probably get no data. Traditional telephone surveys and most online surveys will still count partial responses.