Wondering if your next flight will leave on time? You could use information such as the airline you fly with, your scheduled departure time, and the average precipitation level for your departure date to predict how long your flight may be delayed. In statistics, we call this kind of analysis “regression.”
Minitab's General Regression tool makes it easy to investigate relationships between a measurable response variable (like the length of a flight delay) and predictor variables that are both continuous (measurements such as departure time and average precipitation level) and categorical (such as the airline you use).
The General Regression tool in Minitab Statistical Software makes it easier than ever to perform regression analysis and understand your results, and lets you:
Minitab’s General Regression tool can help you answer a range of questions that commonly confront professionals in almost every walk of life. It can determine which variables are related to a response, and by how much. Strong regression models can even be used to calculate expected values and forecast the impact of future changes.
(But remember, a relationship between a predictor variable and a response does not mean the variable causes the response! For example, regression may find a positive relationship between individuals’ weight and the time they spend listening to music–but that relationship does not prove that listening to more music makes you gain weight! This is what we mean when we say “correlation does not imply causation.”)
Like any regression tool, Minitab’s General Regression tool can help you see how continuous factors affect a variable you’re interested in—but you can also use it to easily investigate your categorical factors.
For example, engineers at a car company use the General Regression tool to determine how various factors affect the distance it takes to stop a car. Some of these factors can be measured on a scale, such as tire width and tire pressure. But what if the engineers need to include a factor such as tire brand, which is categorical?
No problem! In the General Regression dialog box, the engineers simply include their ‘Tire Brand’ data in the model, add it into “Categorical predictors”, and Minitab does the rest!
Relationships between a response and its predictors can often be represented by a straight line. But sometimes the true relationship is a curve, and not a straight line. Minitab’s General Regression tool can model these relationships, too.
For example, a doctor studies how antibiotic dosage influences the number of bacteria in a throat culture. The data show that a strong relationship exists, but she cannot obtain a good fit with simple linear regression. However, Minitab’s General Regression tool lets her easily include quadratic, cubic, or other polynomial terms to find a model that fits her data and better explains the relationships between antibiotic dosage and the number of bacteria.
General Regression can also be used to explore interactions among factors. Interactions occur when the effect of one factor depends on the level of another factor. For example, a baker who is fine-tuning a cake recipe needs to include data about the oven temperature, baking time and sugar level. But he also needs to account for the interaction between time and temperature—the effect of cooking time on the cake will depend on how hot the oven is, and vice versa.
Minitab will automatically calculate the effects of the interaction between variables for you. To look at the interaction between two factors—Time and Temperature, for instance—just add them to the model with * between them, as shown below:
When we perform a regression analysis, we assume that the residuals follow a normal distribution, and the variance is constant. If these assumptions aren’t satisfied, we can’t trust the results of the analysis—but transforming our response data can give us better results.
When you need to transform a response variable because either the residuals do not follow a normal distribution or they exhibit nonconstant variance, Minitab’s General Regression tool can help you by transforming your data using the Box-Cox transformation.
For example, to improve customer satisfaction, the manager of a call center wants to estimate the time that customers wait on hold. He uses time of day and the number of operators as predictors. However, the residuals are extremely skewed. Using the General Regression tool, Minitab can find the most suitable Box-Cox transformation for the response, and the manager can find a sound model to predict hold time.
Regression isn’t new—but by making it easy to include continuous and categorical variables, specify interaction and polynomial terms, and transform response data with the Box-Cox transformation, Minitab’s General Regression tool makes the benefits of this powerful statistical technique easier for everyone.
If you’re not already using the power of Minitab to get the maximum value from your data, download a free, fully-functional 30-day trial of Minitab Statistical Software today.
Get our free monthly e-newsletter for the latest Minitab news, tutorials, case studies, statistics tips and other helpful information.
Data is the new gold: 5 ways to make sure your data is reliable
Advancing the Power of Analytics
A Statistical Analysis of Boston’s 2015 Record Snowfall
By using this site you agree to the use of cookies for analytics and personalized content in accordance with our Policy.