[MUSIC] Hello, and welcome to milestone three of your project. Now you've done a lot of thinking and planning, it's time to really dig into your analysis and start using GIS to answer your questions. By the time you're done with this milestone, which I expect will take most of you anywhere between one and three weeks to complete, you'll have your analysis mostly complete, and your data ready to share. You'll also look at your data and interpret your results, then write up a short summary of your conclusions. In this milestone, you'll focus entirely on the analysis of your data and interpreting the results, but not yet on how to present that data to others as a map. Once you complete your analysis, you'll submit your resulting data layers, along with your executive summary and a short document where you explain your interpretation of the results. You will have more detailed instructions on what to look for when writing up your interpretation of your results. But I want to discuss a few items with you now. First, it's often easy to interpret that the data as showing you what you expected, that it proves some preconception you have, consciously or subconsciously about what the result was likely to be. This is why you shouldn't just look at the data from the angle of whether or not it proves your hypothesis. Instead, you should also look for a trends or values in your results that disprove your hypothesis, or simply support alternative interpretations or hypotheses. Include these interpretations in your results. It's also worth mentioning that it's perfectly fine if your data doesn't support your hypothesis or supports an alternative hypothesis. Remember that a hypothesis is a starting point. You're proposing a potential answer to your analysis questions. But if you find out that your hypothesis turns out to be incorrect, that doesn't mean your analysis was wrong. In most cases, it was probably right, and your job becomes trying to figure out why the results show what they do. What factors did you not consider or did you misunderstand when you formed your hypothesis? This is the chance to really learn what your results are telling you. So do spend some time on it. As you explore your results and attempt to learn from them, try to separate out the objective information from the subjective. That is, separate out the data and what it directly tells you, such as that car crashes occur more frequently on certain stretches of road from why that might be the case. Include them as separate sections in your short writeup. One for the results and one for your discussion of the results and potential interpretation that leads to future analysis work. Also, don't force explanations on your data. It's easy for me to tell you this process, but each analysis and industry is different. It's possible that your analysis won't lend itself to serious discussions of why the results are the way they are. Think critically and make sure that you're always using your results as your guide. Now, as I said, you don't yet to make maps for display. Those are for the next milestone. But you might find yourself needing to make some simple maps to help you explore your data. By all means, do so. But don't worry about the coloring, labeling, presentations, map elements, colors, etc, accept the extent that it helps you find out what your data means. Mapping is always an important part of discovery, but the finesse and design comes in the next milestone. Once you completed your analysis, you'll upload your resulting data for review, along with your executive summary and your interpretation of the results. What data to include in your resulting package is also somewhat subjective, but here's a simple rule. If the data is reasonably needed in order to understand or verify the interpretation of your results, include it. That means you don't need to incorporate intermediate data products, and temporary test layers, but should probably include any layers that you would use in the future to show your results on a map. Before you upload your data, there's one last thing you'll need to do. Think about it for a second and see if you can guess what I'm about to say. Yes, you need to add metadata to all of your final data products. Any layer that you call a data result, and that you distribute should always have metadata. The official assignment for this milestone will have a full listing of everything you should include. But remember now that metadata should answer someone else's questions about how data were generated. This includes processing steps in general at least where the source data was obtained from and when it was obtained if the dataset changes. It should also include your name, your organizational affiliation, and contact information. You can include a fake email in the metadata for this course if you like. Because someone who downloads the data may have additional questions about how the data can be used. We can include much more, but those are core components of good metadata. Okay, now it's time for you to jump into your analysis. Remember to ask your peers for help in the discussion forum if you run into problems or have questions on how you're approaching a problem. As you work with your data and start seeing the results, don't forget to start thinking of interesting themes or results to display as maps. Good luck and have fun.
