504

To Mine Data

  1. Question the overall situation for data mining.
    • Subject: What topics am I dealing with? What traits or quantities do I want to analyze? Where can I find data to gather about this topic?
    • Purpose: Why am I analyzing the topic? Which data is most important?
    • Audience: Who besides me could benefit from this data analysis? What do they need to know?
  2. Plan your approach, using the table-making features of your spreadsheet or word-processing program. (Go to thoughtfullearning.com/h504 for assistance.)
  3. Research your topic.
    • Gather information about the topic from reliable sources.
    • Use analytics built into your own Web applications to find and graph data.
    • Organize your collected data in a table or spreadsheet.
  4. Create your graphic or graphics. (See also “Tips for Data Mining” on page 506.)
    • Choose the data you will analyze.
    • Insert that data into a spreadsheet or table.
    • Select the best graphic (pie chart, bar graph, line graph, and so on) from your spreadsheet or the best arrangement for your table.
    • Include any necessary labels and keys.
  5. Improve your understanding.
    • Evaluate the data you’ve presented.

      What patterns can you see? Do they accurately predict results from other data?

    • Revise your graphic or table.

      Remove nonessential information.

      Rearrange elements that are out of order.

      Redo any part that is unclear or confusing.

      Add missing information.

    • Perfect your graphic or table, making it clean, clear, and accurate.
  6. Present your data analysis to others online, in a report, or during a presentation.

Note: Graphs and tables are often used to present information, but they can also be used to gather and analyze information—a simple type of “data mining.” True data mining employs computer programs to analyze patterns in large databases. For example, a spam blocker searches for similarities between spam emails in order to filter new spam messages; an online shop analyzes customer purchases in order to suggest other products based on what you’ve put in your shopping cart. You can use data-mining techniques to deepen your understanding of a topic by gathering information into a spreadsheet or table, creating graphics that present the data, and analyzing the results.

 
505

Data-Mining Example

In the example below, a student reports information gleaned from mining data from a number of reports released by the United States Bureau of Labor Statistics. The information appears both in text and graphic form.

In D.C., Life Ain’t Cheap

The beginning captures the reader’s attention with an interesting title and a catchy opening.

It should come as no surprise that it takes a lot more money to get by in Washington, D.C., than in small-town America. Economists measure this difference by tracking a number of statistics, for example, the consumer price index, the cost of housing (renting or buying), the cost of groceries, and access to other necessities of life.

The middle provides data both in words and in images.

The following graph presents a cost-of-living index for each of four basic expenses. Note that, in some places such as California, New York, and the District of Columbia, housing costs far exceed those in other places, while other costs may be the same or slightly lower. For example, though a typical apartment in New York City is astronomically expensive, a typical restaurant meal isn’t. It’s one reason wait staff have a tough time living off low wages and tips. In Michigan, though, the same server would be able to afford better lodgings. . . .This stacked bar graph emphasizes the overall cost of living in each location.

U.S. Cost of Living Index, 2011. Source: United States Bureau of Labor Statistics
 
506

Features Table

Another data-mining technique uses a features table, which allows a side-by-side analysis of details for two or more subjects. In the following table, a high school student compares three possible career-training paths.

Emergency Medical Training Possibilities Table
 

Tips for Data Mining

Use data-mining techniques to analyze data and search for patterns.

  • Gather your data.
  • Consider your purpose. What questions are you hoping to answer? What other questions are suggested by the data you have gathered?
  • Create a spreadsheet or table that accommodates your data.
  • Arrange your data clearly. Try different organizations and graphics.
  • Look for patterns. What does the data suggest?
  • Test your patterns by applying them to new data sets.
  • Write an analysis of your graphic, explaining what it reveals.
  • Give your graphic a title that reveals its scope and purpose.
  • Use subtitles and keys to clarify parts.