
Learn how regression links a dependent variable to multiple independent variables using scatterplots and trend lines to predict outcomes, with dummy variables for categorical information and time-series seasonality in Excel.
Learn how correlation measures association between variables and underpins regression, using the Pearson product moment r from -1 to 1. See how Excel's correlation function reveals relationship types.
Define the regression equation with an intercept and slope to predict the dependent variable from the independent variable, using Y hat to compare predicted and actual values and minimize error.
Discover how the method of least squares in Excel builds a trend line for scatter plots, predicts y from x, and evaluates model fit using adjusted R-squared and p-values.
Enable the Excel data analysis add-in (analysis toolpak) to access regression tools from the data tab for regression analysis.
Run regressions in Excel using trucking and installation datasets; interpret regression output, examine p-values, interpret slope and intercept, and predict y from x with a trend line.
Evaluate regression results by examining adjusted r-squared and p-values to gauge model accuracy, statistical significance of the independent variable, and the explained variance.
Improve regression accuracy by adding explanatory variables, as adding stops alongside miles raises adjusted r-squared from 0.61 to 0.876, with p-values < 0.05 confirming significance and a clearer prediction equation.
Learn to estimate price elasticity with log transformations in regression using Excel; interpret elastic and inelastic coefficients and assess how price and advertising affect log sales.
Use Excel to create binary dummy variables for categorical data, with a mechanical reference, to reveal how months since service and job type affect repair time and car price.
Model change over time in regression by using year as the independent variable to predict bike sales, and assess p-values and adjusted R-squared. Seasonality via dummy variables for umbrella sales.
Regression is an important statistical tool. Using regression, we can detect and quantify relationships within a data set. For example, you have a data set of truck distances driven and stops made. Using this information, we can construct an equation which allows prediction of duration given distance number of stops. This would be of great use to anyone having to give out quotations.
I show you how to check whether your regression actually works and how accurate it is.
Indicator or 'dummy' variables are an important source of information, and I show you how to convert textual data into dummy variables for inclusion in the regression analysis. We know how long repair jobs take and months since last service. Does including information about whether the job was electrical or mechanical make predicted repair time any more accurate?
Sales go up at certain seasons: being able to measure those increases and predict them is highly useful.
We also cover elasticity, a topic often missed out in regression courses. Using elasticity, we can predict the effect on sales volume in precent of a percent change in selling price.
I provide detailed explanations and provide the datasets so that you can follow along.
The pace of the course is measured and step by step, each section building on the last. The datasets I use in the examples are included so that you can run your own regressions and compare results.