Now that we have talked about linear regression, we’ve reached a level of mathematical sophistication that we can do some basis modelling and analysis. Find some data on the internet that is either relevant to your major or just interesting for its own sake. Make sure that the data contains a pair of numerical variables which can be reasonably ordered. If you don’t know where to start to look for such data, consider skimming these data sources to get ideas.
Once you find some data you’d like to model, copy or transcribe it into a Microsoft Excel spreadsheet† and plot the data: click and drag to select the data and then click “Insert” → “Charts” → “Scatter” to insert a scatterplot of the data. Next fit a linear model to the data: select the scatter plot points, then click “Chart Design” → “Add Chart Element” → “Trendline” → “Linear”. Excel calls this regression line a trendline. Double-clicking on the trendline, an options panel will open; select the option to “Display Equation on Chart”.
After creating and customizing your chart in Excel, copy it over to a Microsoft Word document to more easily draft a report to analyze the data and interpret your model. Save your files for this assignment since we’ll be revisiting it later in the term, and be sure to include a link to the source of your data as a Reference in your report.
This assignment is designed to be simple, having only a few requirements, to afford you plenty of time to research data sources online that interest you, and plenty of time to play with Microsoft’s Excel and Word software to discover their capabilities.
Deliverable
Hand in a printed copy of your report at the start of the next midterm exam. Typeset your report to fit on a single sheet of paper. Design your chart to be clean and self-contained, and write your analysis to be pleasant to read. Your report must contain a chart with the scatter plot of the data and the trendline and the trendline’s equation, an analysis of the data in reference to the trendline, and a reference linking to your data source. Address at least the following questions in your analysis: Does the trendline tell you if the data is increasing or decreasing overall? What does the slope of the formula for you model specifically mean in the context of your data? Do you think a linear function is an appropriate choice of model for your data? What does your linear model predict about the future of the phenomenon your data measures? In the report, refer to Excel’s computation as an "embryonic linear regression analysis".