RegressIt is a powerful Excel add-in which performs multivariate descriptive data analysis and regression analysis with high-quality table and chart output in native Excel format. It was developed at Duke University and has been distributed on this web site since 2014, and you can download it for free. The linear regression version of the program runs on both Macs and PC's, and there is also a separate logistic regression version for the PC with highly interactive table and chart output. All of them feature a 32-button ribbon interface (pictured below) which provides innovative model space and worksheet navigation tools, 10,000 words of built-in documentation and teaching notes, use of cell colors and fonts to highlight significant coefficients and correlations, interactive sorting of coefficient and residual tables, removal of insignificant variables directly from the coefficient table, a multi-threaded audit trail, and the ability to exchange regression code with R. If it looks overly complicated to you, don't worry: the only buttons that you need to use to perform analysis are the "Descriptive Statistics" and "Linear Regression" buttons at the left. The others are there just to help you to navigate your output in files that contain lots of models. RegressIt makes it easy to run and compare alternative models and follow best practices of regression analysis.
RegressIt is an excellent tool for online teaching of regression in electronic classroom environments such as Zoom and in recorded video presentations, as well as in classroom demonstrations and interactive exercises. See the instructors page for details.
The ribbon interface greatly streamlines the analysis, minimizing the number of mouse clicks or finger taps that are needed for careful well-labeled work, and it makes it largely unnecessary to use any of Excel's own buttons or menus. The accompanying RegressIt slide show and user manual and the following pages on this site describe the use of these features. The videos page on this site includes several videos that provide demonstrations of them in action.
RegressIt is an excellent tool for online teaching of regression in electronic classroom environments such as Zoom and in recorded video presentations, as well as in classroom demonstrations and interactive exercises. See the instructors page for details.
The ribbon interface greatly streamlines the analysis, minimizing the number of mouse clicks or finger taps that are needed for careful well-labeled work, and it makes it largely unnecessary to use any of Excel's own buttons or menus. The accompanying RegressIt slide show and user manual and the following pages on this site describe the use of these features. The videos page on this site includes several videos that provide demonstrations of them in action.
The output format and navigation tools are designed to take advantage of all 4 dimensions of the Excel environment—stacks of worksheets, rows and columns of cells, and layers of information within cells--while avoiding the tedium of scrolling and paging and searching with Excel's own controls. These tools greatly simplify the process of exploring multivariate data, fitting and refining and comparing regression models, sharing and evaluating work performed by others, and leaving a comprehensive and easily navigated audit trail.
- For students, these features make it easy to learn principles of regression analysis (and thoughtful data analysis in general) while working efficiently in an Excel environment. The software is good for team exercises and is more fun to use than the standard alternatives, and its presentation quality tables and charts are ready-made for inclusion in reports. It is also self-instructing, with pop-up documentation and teaching notes displayed in message boxes and cell comments.
- For instructors, these features make it easy to teach both statistical concepts and hands-on data analysis, to evaluate student modeling efforts (the entire analytical process, not just the end results), and to verify the originality and authenticity of submitted work. The program is good for interactive demonstrations in the classroom as well as for assignments. A gradeable list of the contents of all worksheets in hundreds of student files can be exported to a single Excel worksheet by using the History tool. See the features for instructors document for more details. Even if use of RegressIt is not required in a course, it can be recommended to students as a self-teaching supplementary tool, and many will find their way to this web site anyway. Customization of the built-in teaching notes could be arranged to suit a program. RegressIt is also an excellent platform for online teaching: it makes great videos and is very efficient for use in workshops and webinars. If you would be interested in developing or co-developing an online course that uses it (e.g., at Coursera), please contact us.
- For practitioners, these features make it easy to explore a complex model space, to maintain a comprehensive and easily navigable audit trail, to collaborate and share results with colleagues and clients who may have different levels of technical expertise, and to brush up on one's knowledge while working.
- For R users (or would-be R users) it reads and writes R code for linear and logistic regression, so that models whose variables are selected in RegressIt can be run in RStudio, with nicely formatted output produced in both RStudio and Excel, allowing you to take advantage of the output features of both and to get a gentle introduction to R (or perhaps Excel) if you need it. You can use Excel as a front end for running analyses in R and/or use R as a computational back end for analysis in Excel. This tool allows very large data sets to be handled and includes more analysis options such as stepwise variable selection and out-of-sample validation. See the R interface pages on this site and the accompanying pdf file and video.
- For touchscreen and tablet computer users, the interface makes it easy to drive a fully detailed analysis by tapping one's finger at the top or bottom or side of the screen. The output files also look good on smartphones and are easily navigated with Excel viewing apps, a great way to share results with others.
- For anyone who otherwise uses Excel for data analysis (or who has expressly avoided doing so until now), it is a free, major upgrade from what is provided in its own data analysis add-in (Analysis Toolpak), whose dreadful regression tool has not changed since it was introduced in 1993. See what's-wrong-with-the-Analysis-Toolpak-for-regression.