Before you do anything else, we recommend that you visit the JASP video library, the JASP YouTube channel, and the ‘How to Use JASP’ page. Below is some general information about JASP that may be helpful as you analyze your first data set. We invite you to post any remaining questions on the JASP forum or on the JASP GitHub page.
Open Data & Example Analyses
JASP comes with a series of example data sets that can be accessed from the main menu with the 3 blue stripes, then ‘File → Open → Data Library’. Selecting an example data set will open either the data-only CSV (click on the upper part) or the JASP-file (click on the bigger lower part). The JASP-file besideds the data also holds analyses and interpretations. You can then edit, reorder or extend the analysis. In addition to .csv and .jasp, JASP can open many other formats, see the features page for the full list. Note that all files need to have a header row that contains names for each of the columns or variables. Missing values can either just be missing (i.e., an empty cell) or be denoted by “Nan”, “.” (period), or ” ” (space) or any other custom code, that you add to the missing value list. When opening a file, JASP makes a best guess to assign variable types, as explained below.
Variable Types
JASP distinguishes:
- Nominal
- Ordinal
- Scale
Nominal variables are categorical variables often without numeric value. An example might be a variable called “Favorite Dutch Soccer Club”, with three possible entries: “Ajax”, “PSV”, and “Feyenoord”. Even if the categorical variable is represented by numeric values (for example, a variable “Group” may have levels “1” and “2”) they do not imply an order, and the distance between them is not meaningful.
Ordinal variables are categorical variables with an inherent order. An example might be a Likert preference rating scale with levels 1 (“hate it”), 2 (“yuk”), 3 (“meh”), 4 (“cool“), and 5 (“in love”). Note that the distance between the numbers is not meaningful. If the values are not numeric, but like the labels also strings, JASP cannot determine the correct order and might throw an error message, for example in the correlation module.
Scale variables are continuous variables with values that allow a meaningful comparison of distance. Examples include income or weight.
Automatic Variable Type Assignment
If there are only numbers (apart from the missing values like NaN) JASP automatically assigns:
- Nominal for 2 unique integer values
- Ordinal for 3-10 unique integer values
- Scale if there are 1 or >10 unique integer values or if 1 value has a decimal.
If at least one of the values is not a number nor a missing value
- Nominal for 1 or 2 unique values or >10 unique values
- Ordinal for 3 to 10 unique values
You can customize some of these settings from the main menu, then ‘File → Preferences → Data → Import Settings’. There you can also change ‘missing value settings’ and the ‘threshold for scale’ (default 10).
Manually Changing Variable Types
If need be, the automatic variable type assignment can often be corrected manually. When you click the icon representing the variable type at the top of the column or within an analysis box, a menu is produced that allows you to choose a different variable type. JASP will try to respect this change and transforms the variable into the selected type. If a scale type is chosen in the data view, and there are non numerical values, these values will be set as missing values. However, variable types in analyses or computed columns are often enforced; for instance, if you use a t-test the type of the dependent variable will be coerced to Scale. This change is made transparent by a asterisk sign (*) next to the changed type symbol.
Analyses
Having loaded a data set, it is possible to run analyses. Only a few analysis are possible without loading data. After selecting an analysis from the ribbon along the top you will see input options for that analysis in the left panel, and the associated output in the right panel. As the input options are specified, the analysis results automatically update, providing immediate feedback.
You can return to an earlier analysis simply by clicking on the output of interest. This brings up the options that were used to generate that analysis, and allows you to make adjustments or additions. You can go back to your data by clicking on the bar on the very left that contains an arrow in the middle or by just clicking on the large ‘Edit Data’ button on the ribbon bar.
There is much more to say about JASP, its ability:
- to annotate, reorder, and quickly duplicate the output,
- to interact with the Open Science Framework (OSF.io) and many other databases,
- to use modules explicetly designed for teaching (e.g. ESCI) or to conduct sophisticated analyses, both accessible through the big blue plus sign on the top right,
- to recode variables and filter values with only one click,
- to add complex filters or compute variables with R code or through a drag and drop interface,
- to copy APA formated tables directly to a word file,
- to produce publication-ready figures, and much more.
These options are discussed in our videos and blog posts linked above or within our extensive list of JASP textbooks and manuals.