Cart

How to enter t test data

Enter each group into its own column. The calculator will compare the means to ask whether the observed differences are likely to be due to coincidence. Enter either raw data (enter each value) or averaged data (enter mean, N and SD or SEM). If you enter averaged data, you cannot choose a paired t test, which requires raw data. When entering raw data, simply leave a blank spot in the table to denote missing values. If you enter averaged data, you must enter the mean, N and SD (or SEM) for each column. It is okay if N differs among columns, but you must enter mean, N and SD (or SEM) for each column; you can't leave any of those values blank.

This calculator is set up for you to enter data in a format that is natural to many scientists. For example, to compare the blood pressure of a group of men and a group of women, enter the men's blood pressure in one column and the women's blood pressure in another. Some other statistics programs expect you to arrange data differently, putting all of the data into one column and using another column to define group. For the blood pressure example, you would enter all the blood pressure values (for both groups) in one column. In another column you would enter a code or index (perhaps 1 for men and 2 for women). Don't arrange data in this stacked or indexed format when using this calculator.

Before comparing columns, consider whether you should first transform the values. The t test assumes that your data are sampled from a population that follows a Gaussian distribution. If your data do not follow a Gaussian (normal) distribution, you may be able to transform the values to create a Gaussian distribution. If you know the distribution of your population, transforming the values to create a Gaussian distribution is a good thing to do, as it lets you use a t test, which has more power than a nonparametric test. If you plan to use a nonparametric test, transforming the data will make no difference (but this calculator does not perform nonparametric tests).