Question about statistical analysis, and discarding data

[Jeff7]

I am doing a speech soon, and I took a survey of the class:
1. What was NASA?s entire budget for 2007? (just a simple dollar figure is needed)
_________________________

2. Which is greater?
O Total spending on NASA since its formation
O Total spending on the Iraq War

3. Which is largest?
O 1 million
O 1 trillion
O 1 billion

4. What is the yearly spending of the United States government? (for one of the past 3 years)
_________________________

5. US consumer spending on tobacco during 2005 was:
O One-half of NASA?s budget
O 5.66 times NASA?s budget
O 1.13 times NASA?s budget

6. Do you view spending on the space program and NASA as a waste of taxpayer money?
O Yes
O Neutral
O No



I was hoping to do a quick analysis to see if there was a correlation between people thinking NASA's spending was a waste, and their idea of its yearly cost, as in, if you think it's really really expensive, you'd be more likely to think it's a waste.

Problem is, the numbers I got vary incredibly. Based on a glance at the papers as I was getting them back just a few minutes ago, for NASA's yearly spending, I got numbers anywhere from $100,000 to $500 billion.
I was figuring on just using Excel's statistical analysis functions to do this, and some short notes on statistics I have (somewhere) from another class. Do I discard data that's off by, I don't know what it'll be, 10 standard deviations? Do I just keep everything? Or wait and see what the results really are? I wasn't expecting such a crazy spread of data.


And tell me how this makes sense. One form:
NASA's total yearly budget: $15 billion
Yearly government spending: $150 million

Okay then.

Another one:
NASA's total yearly budget: $500 billion
Yearly government spending: $250 million

Maybe I should have asked, "Which is larger, 1 billion or 1 million?"

Quite a few have NASA's budget being greater than the government's total spending. Maybe I should have mentioned, "NASA spending is included in the budget."
Or said, "Please give reasonable answers."

One person did say 1 billion was the largest.



Early update:
NASA spending guesses (3 surveys had no reply)
500,000,000,000
15,000,000,000
2,000,000,000
100,000,000
150,000,000,000
500,000
1,000,000,000
5,000,000,000
500,000,000,000
2,500,000
25,300,000
200,000,000
10,000,000,000
20,000,000,000
20,000,000,000
32,000,000
5,000,000,000
1,000,000,000


Standard deviation is crazy: 160,828,453,475

So what the heck do I do with that kind of data?

 

[Jeff7]

Bump for the nightly mathematicians.

Should I just skip the standard deviation stuff, and just look for a correlation between "Yes/Neutral" responses for the "Is NASA spending a waste?" replies, without bothering with Excel? Seems this might not be a large enough sample size to get decent data from. I did a histogram, and, well, look at the range: 500,000 - 500,000,000,000. Doesn't make for a good chart.

Maybe I just didn't do a good job on the survey. I might simply be better off covering my bases, and assuming that everyone pretty much knows nothing about NASA or government spending. If there were a correlation between saying NASA's a waste, and thinking that it's really really expensive, well, I could just assume that it's there in at least some of my audience, and address it.

Perhaps sometime I'll figure out how to create a form, and open up this survey to ATOT, just for the heck of it. I'll do it in several months, so that everyone's forgotten the answers.

 

[fishmonger12]

Your survey is just poorly designed for it's intention. Instead of seeing if people think the government should spend money on certain programs and the correlation between those views, you're testing whether individuals read the news, know how to count, know the heirarchy of government, etc.

If all you're trying to do is test whether there's correlation between whether people think Nasa/the war/government spending in general is necessary, have a question that asks:

What would be your best guess as to the total cost of funding NASA per year? (multiple choice range answers)

What would be your best guess as to the total cost of funding the Iraq war per year?

Do you believe the United States government should fund or continue funding a space program such as NASA?

Do you believe the United States government should continue funding the war in Iraq?

Do you believe the United States government should fund nationalized health care?

Have the potential responses for the above be -2 through 2, -2 being strongly disagree, -1 being disagree, 0 being neutral, 1 being agree, and 2 being strongly agree. It keeps the useless responses to a minimum.

 

[Jeff7]

Yeah. Well hey, I've never designed a survey before. Oh well.

At least I do know for sure that some people in the class think that NASA is a waste of money.

Maybe I'll look at the "spending on tobacco" question, see if there's any consensus there among the class.

And at this point, it's too late for a do-over. My speech is Tuesday, and we're already behind in terms of the speech schedule, so the professor probably won't want to be losing any more time to surveys.

I'll just do what I did for my previous speech on the Mars Rovers - aim for the lowest common denominator, so to speak. For that, my baseline knowledge was "Mars is red." That's what I had to build on.
I'll have to build on a similar level for this speech, make it another all-encompassing sort of thing.