Starting a Project

• For the next few sets of notes we will be working with the Super Heroes Dataset from kaggle.
• A local copy is available.
  • Download this.
  • Save it as an excel file somewhere you can get access to it.
• Worksheets
  • Take a look at the specification for the worksheets.
  • Build an about, Raw Data and Cleaned Data worksheet.
  • We will not construct the ABOUT sheet, but it should be there for the final project.
• Preparing to start.
  • Take a look at the Methods Document rubric right now.
  • My document needs some reorganization
    • I will use the rubric,
    • but the following will be a good guide to creating your final document.
  • Right now open a word document.
  • Add the following sections
    • TOC
    • Introduction
      • Describe the basics of your project, what question will you try to answer and what data will you look at.
      • This section should
        • Introduce anything the reader needs to understand to read the rest of the document.
        • Draw the reader in, invite them to continue reading.
      • Don't discuss meta-information about the project or paper. (ie this stuff)
    • Data
      • Overview
        • Introduce your dataset.
        • Where did you get this data?
        • How much data is there (rows and fields)
        • If you will filter your data (remove some of the original dataset)
          • Briefly describe why you will filter the data.
          • Provide an overview (size in rows and fields) of the remaining data.
        • Describe the fields of interest.
        • Define any specialized terms associated with this dataset.
      • Data Dictionary
        • Provide a description of every field in the dataset.
        • This should be a well formatted table or a bulleted list.
      • Example Data
        • Give a few rows from the dataset
        • These should contain examples of valid data in all fields.
        • This should be one or more well formatted tables.
    • Data Cleaning
      • If you decide to filter the data (not due to problems)
        • Discuss the reasons for filtering data
        • Discuss the rational for selecting the data you did.
        • Describe how you filtered the data (technical excel discussion)
      • For each problem you encounter with the data
        • Describe the problem.
        • Provide an example of bad/missing data
        • Describe how the problem will be mitigated.
        • Give a rational for fixing this problem the way you did.
        • Describe the actions taken in excel to mitigate the problem.
    • Analysis of Fields Used.
      • This will depend on the field.
        • Numeric data should have at least a five number summary.
        • Some form of chart/graph to depict the data.
        • If you performed some computation on this data:
          • Describe that computation in sufficient details that a knowledgeable excel user could reproduce the computation.
          • This will include a written discussion.
          • It should include screen shots for any out of the ordinary computations.
    • Methods
      • What did you do beyond basic analysis of the data?
        • Include a description of what you did.
        • Discuss why you did it.
        • Provide a description of how you did it in excel
          • Description and screen shots as required.
        • What did it yield.
      • Problems Encountered
        • What went wrong?
        • Why did it go wrong?
        • This is not a whining section.
          • You want to provide your reader with a guide to problems.
          • You want to document that you tried something that didn't work so they know not to try it
            • Or at least not try it the way you did.
        • Don't include
          • A discussion of how you didn't know how to use excel.
            • You will never completely know how to use a tool
            • They are too large
            • There are constantly new pieces added.
            • People discover new ways to do things.
          • A discussion of running out of time.
          • A discussion of what you did not know how to do.
      • Discoveries and Conclusions
        • Restate what you have learned/discovered.
        • State the question and discuss the answer based on what you have discovered/learned.
      • Future Work
        • Talk about things you would like to do but did not have time to accomplish
          • But don't say "I ran out of time, but I would like to ..."
        • New data you would like to collect to strengthen your study.
        • Other areas you would like to investigate.
      • Bib
    • This is a general set of guidelines.
      • Start with your proposal document.
      • Go through the rubric and see what it asks for.
      • If you don't have information for a section
        • State that you didn't encounter the issue/problem the section discusses
        • State why this is the case.
        • For example
          • If your dataset was in great shape when you got it your Data Cleaning section could consist of the following:
          • "After basic analysis it was found that the data contained no errors. It was not necessary to filter or clean to raw dataset to obtain a usable version of the data."
      • Create lots of chart (graphs) and include them in your presentation.
      • Keep your worksheets labeled and clean as you go
        • You won't have to go back and redo the work when you produce your final report.
      • As we learn a new technique, apply it to your dataset.