It’s a list of the average interest rates on 30-year mortgages and the median home selling price from 1988 to 2003. In the Add-Ins dialog box, ensure that “Analysis ToolPak” is checked.įor Macintosh computers using Excel 20, you need to use external utility called StatPlus that provides similar capabilities to the Analysis ToolPak.Click on Excel logo at the top left-hand corner of your screen or go to the File menu.If your computer is a PC, it must have the Analysis ToolPak enabled: The data is courtesy of Sudhakar Raju of Rockhurst University.įirst, verify that Excel is set up to calculate regressions. We’ll use Microsoft Excel in our example, but free tools such as Open Office Calc have similar capabilities. To answer this question, you need some data and a spreadsheet program. (For the sake of simplicity, we’ll leave aside the “unconventional” mortgages that contributed to the housing boom and bust.) But what exactly is that relationship between mortgage rates and median house prices? If mortgage interest rates increase by 1 percentage point, how much would the median selling price of properties fall? In response to this, median housing prices frequently decline. Anyone who’s ever had a meeting with a bank loan officer knows that as interest rates rise, mortgages become more expensive and thus the maximum mortgage that one can afford declines.
Fortunately, this and other data-analysis programs come with the necessary tools built in, and it’s just a matter of your getting access to the numbers, and then properly using the program.Ī basic example to use for creating a regression is mortgage rates and housing prices. You can do this visually, just by eyeballing the data, but it can be done with much greater accuracy and certainty by using a spreadsheet application such as Excel.
At its most basic, this involves plotting data points on a X (horizontal) and Y (vertical) axes - for example, car weight and crash fatality rates - and looking for a trend line and, potentially, a relationship. Simply put, regression analysis is a way to determine if there is or isn’t a correlation between two (or more) variables and how strong any correlation may be. Still, there’s a lot to be said for getting more familiar with this somewhat obscure yet ubiquitous form of data analysis. The technique is well known to data journalists, but even savvy reporters may feel a measure of discomfort when they come across it - they seldom have the expertise or time needed to understand advanced mathematics or dig into a study’s original methods and data. When journalists read academic studies, whatever the subject - the relationship between wages and well-being, the effectiveness of flu vaccines or the benefits of political incumbency - they frequently encounter the term “regression,” a mathematical tool for establishing the relationship between two or more variables.
0 kommentar(er)
