This workshop
This is an introduction to jamovi. Jamovi is free, and built on the powerful stats language that is R. There’s a lot to like.
It’s point-and-click, like SPSS, so there’s an easy transition. Lots of the decisions on how analyses run in jamovi are based on how they were set up in SPSS, so it will be familiar. It’s useful for students, especially those who don’t have a license for SPSS, but who want to keep up their stats skills.
It’s also useful for those who want to use R but find data prep and cleaning in R to be difficult, because you can’t really see your data. I have other content on R if you would like to explore this too (R workshop 1 on the basics like opening and manipulating data and running basic analyses, and R workshop 2 showing some cool features like writing a paper in the same program you use for your stats, text analysis, web scraping as an alternative way of getting data, and making amazing plots in ggplot2).
I’ve been teaching stats since 2007, and have had a close eye on different statistics programs over the years. I’ve used lots of them – Splus, R, various Excel add-ons, SPSS, PSPP, Gretl, JASP, Jamovi, MPlus. I still use a variety of them depending on who I’m working with. There’s a lot that I like about Jamovi, but I also acknowledge that different software is aimed at different people.
Most importantly, this page is not a stats class. This is me showcasing some relatively new software and how it can be used, so you can think about whether it’s useful for your needs. There is an assumption that you know some basic stats analyses.
Finally, and also very importantly, thank you to Professor Tania Signal (DDR-HMAS) for providing the resources to facilitate this workshop and the following R workshops, and to Associate Professor Grace Vincent in particular for helping make it happen.
WHAT’S ON THIS PAGE?
You can click these links to skip to a section.
How to download and install Jamovi
Setting up variables (like “variable view” in SPSS)
Exploring data (descriptive stats like means, visuals like figures)
Running analyses like ANOVA, Correlations, Linear Regression
Installing extensions (including running a mediation analysis)
What is Jamovi?
(from https://www.jamovi.org/about.html)
The jamovi project was founded to develop a free and open statistical platform which is intuitive to use, and can provide the latest developments in statistical methodology. At the core of the jamovi philosophy, is that scientific software should be “community driven”, where anyone can develop and publish analyses, and make them available to a wide audience.
jamovi aims to be a neutral platform, and takes no position with respect to competing statistical philosophies. The project was not founded to promote a particular statistical ideology, instead wanting to serve as a safe space where different statistical approaches might be published side-by-side, and consider themselves first-rate members of the jamovi community.
Used jamovi for your publication? Please consider adding the following reference:
The jamovi project (2024). jamovi (Version 2.5) [Computer Software]. Retrieved from https://www.jamovi.org
We appreciate that some publications require a city and country of publication, which doesn’t really make sense in the context of an internationally developed open-source project, but if they insist, Sydney, Australia, is probably the best option.
You’ll find additional, analysis specific citations inside jamovi.
Downloading AND INSTALLING Jamovi:
Go to the Jamovi Website: Visit https://www.jamovi.org to download the latest version of Jamovi.
Choose Your Operating System: Jamovi is available for Windows, Mac, and Linux. Select the version compatible with your computer and click “Download. Jamovi Desktop is free. Cloud is free for anonymous, limited use, otherwise $9.90 a month. I suggest you use Desktop. Cloud, though, may be useful if you use something like an iPad a lot.
Install Jamovi: Open the downloaded file and follow the installation prompts. Once installed, open Jamovi to get started.
Opening and Saving Data
Opening Your Own Data in Jamovi:
In Jamovi, you can open data files, including SPSS data files. I particularly like this functionality, especially for those who are working with other software that can output SPSS files (e.g., Qualtrics).
Start a New Session: Open Jamovi, and you’ll see a blank spreadsheet by default.
Import Your Data: Go to File > Open and select your file type. Jamovi supports CSV, TSV, ODS, and SAV (SPSS) files.
Select Your File: Locate your dataset and open it. Jamovi will load it directly into the spreadsheet view.
Quick Adjustments: Once loaded, you can click on each column to adjust variable properties, which we’ll cover in the next section.
Using Built-In Datasets:
Jamovi also has built-in datasets that we can use. We’ll use one of these today.
Access Built-In Datasets: Go to Data Library in the File menu to explore datasets provided by Jamovi.
Select a Dataset: Choose a dataset to load. This is great for practising analyses without needing an external file.
Editing Datasets: You can make edits to built-in datasets just like your own, which is handy for practising data wrangling.
Please open up the “Anderson’s Iris Data” dataset from the data library.
Saving Data:
Note that saving data in jamovi also saves things like your analyses, output, etc. Contrast this with SPSS where you need to save your data in a .sav file, and your output in a .spv file, and syntax in a .sps file. With jamovi, everything is in one place, which makes sharing your work with others much easier. But, if you need to, you can also export your data as a csv or sav file. This can be useful for clients.
Saving in Jamovi Format: Save your work by going to File > Save As and choosing the .omv format, which preserves your variables, analyses, and data structure.
Exporting to CSV or SPSS: If you want to use your dataset in another program, you can export it as a CSV or SAV file. Go to File > Export and select your preferred format.
Save your data somewhere using the jamovi .omv format. Up to you where you save it.
Setting Up the Output Display:
Before we play with data, let’s look at what’s available in terms of setting up output.
Locate the Settings Menu: In the top right corner of the Jamovi window, click on the settings icon (a small gear symbol, or three vertical dots – see image below)
Adjusting Decimal Places and Significant Figures:
Under “Output Settings,” you can adjust the decimal places (dp) and significant figures (sf) for the results displayed.
For example, you might set decimals to 2 places to make results easier to read, or increase them for more precision.
Figure formats
There are built-in formats, like I love SPSS, Hadley (after Hadley Wickham who is a big name in the R community), and others. Feel free to explore! We’ll have more customisation in R later, and also look at how to adjust these a bit too.
Updates
If updated versions are available, you cah select them here. Updating jamovi is very easy.
Comparing to SPSS:
In SPSS, changing decimal places or significant figures requires modifying individual output settings or reformatting tables manually, whereas Jamovi allows quick adjustments in a single menu.
Jamovi interface
Up the top, you’ll see five main menus. Think of this like the ribbon in Word, where when you select things like Review and the options change.
Jamovi file management - New/file/open/save/import/export (the three horizontal lines)
Variables – this is where you might set up some things about variables, e.g., variable labels. Note that you can also click on a variable and select “Setup” to access more variable setup choices (e.g., value labels, scale of measurement, missing values)
Data – this is where you can see and edit your data. Similar to data view in SPSS
Analyses – where you can conduct analysis.
Edit – where you can edit how your results section/output looks (e.g., headings, formatting, etc).
Setting Up Variables (The Equivalent of “Variable View”)
Assigning Variable Types in Jamovi:
Increasingly in SPSS, you need to define whether your variable is nominal, ordinal or scale (i.e., continuous). In R, you need to do this a lot too. Here, in Jamovi, we’ll find that it’s useful to have this set up too.
Access the Variable Setup: In the main data view, click on the column name at the top of each column to open the variable properties menu.
Choose Variable Type:
Continuous: Use this for numeric variables that represent measurements or amounts (e.g., age, income).
Ordinal: Select for variables with a meaningful order but without consistent spacing (e.g., satisfaction ratings).
Nominal: Use for categorical variables without order (e.g., gender, colour).
Setting Variable Levels:
For categorical variables, you can specify levels by clicking on the variable and selecting Levels. You can name each level, which is useful for clear outputs in analyses.
Editing Variable Properties:
If needed, you can adjust variable names and missing values.
Rename Variables: Click on the column header, and you can rename the variable to something more intuitive.
Set Missing Values: If there are missing values, you can define them here by choosing a specific code (like 999) and setting it as missing.
Comparing to SPSS:
In SPSS, variable setup is done in the “Variable View” tab, where you define types, labels, and missing values in a table format. Jamovi’s approach allows you to set these directly in the data view, which can feel quicker and more intuitive.
Variable levels are easier to set in Jamovi, where they’re handled visually within the variable setup menu. In SPSS, defining categories often requires extra steps in the Value Labels section.
Exploring Data – Means, Standard Deviations, Counts, and Visual Elements
Getting Descriptive Statistics (Means, Standard Deviations, Counts):
Let’s play around with the iris data.
Load the Dataset: For this example, open the iris dataset (or load your own data) to explore some classic car performance stats.
Navigate to Descriptives:
Go to the Analyses tab at the top.
Select Exploration > Descriptives.
Select Variables: In the Descriptives panel, choose sepal lengths and put it into the variables box.
Viewing Results:
Jamovi will instantly display the descriptives for these variables. You’ll see the mean, median, standard deviation (SD), minimum and maximum values, as well as count (N).
If you want additional statistics like quartiles or range, check the relevant boxes in the Descriptives options.
Visualising Data:
Now let’s add a histogram for Sepal Length. Click the Plots dropdown and select Histogram.
Adding a Box Plot or Histogram:
In the same Descriptives panel, scroll down to Plots.
Select Violin plot to get a visual summary of your data, or select Box plot to show medians, quartiles, and outliers.
Choose Histogram to see the distribution of values for each selected variable. This is helpful for checking data normality.
Maybe you don’t like how the image looks. Go to settings and choose another style.
Now try descriptives for a categorical variable. Exploration > Descriptives. This time, put Species across. What do you note about the output? Now, try a Histogram plot. What happens? What plot can you run?
Scatterplots and Bar Graphs:
To compare relationships, choose Exploration > Plots > Scatterplot and select a pairing, like Petal Length and Petal Width. This will give insights into correlations.
Comparing across groups
Examine Sepal Length by Species, through Exploration > Descriptives
Note how there is a little image in the bottom right of the “Split by” box. It’ll only accept categorical variables. This is sensible, but means you need to set up your variables (See above).
Run the analysis, and try adjusting the dropdown menu for Descriptives – see what happens.
Comparing to SPSS:
SPSS requires running separate commands for descriptives and visuals, often navigating through multiple dialog boxes. In Jamovi, everything’s accessible in one panel, making exploration quicker and more intuitive.
Jamovi’s output is live, updating instantly as you select variables or change options, whereas SPSS needs to run each analysis individually.
ANALYSES
ANOVA
Setting Up an ANOVA
We’re going to run through ANOVA. T-tests are pretty self-explanatory, so I won’t worry about it.
Load Your Dataset: Use a dataset like iris, which includes continuous variables and categorical groupings.
Go to the ANOVA Menu:
In the Analyses tab, select ANOVA > One-way ANOVA.
Define Variables:
Dependent Variable: Select Sepal Length as your dependent variable, as it’s continuous.
Grouping Variable: Choose Species as your Grouping Variable, as it’s categorical.
Setting Up the Test:
Jamovi will display options like Welch’s F (default, does not assume equal variance) and Fisher’s F (the usual F-value that we know and love).
You can add descriptives, descriptive plots, assumption tests, and post-hoc tests. Ask for descriptives, plots, Q-Q plots, Tukey post-hoc and ask it to flag significant comparisons.
Interpreting Results:
Jamovi outputs the F-value, degrees of freedom (df), and p-value for your test. As usual, it’s up to you to determine what to compare your p-value to.
Checking Assumptions:
Lots of tests, like homogeneity tests, Q-Q plot, etc. Up to you to determine what you need, but the main point is that it’s super easy to run them.
Comparing to SPSS:
SPSS requires a separate menu to set up an ANOVA. Jamovi streamlines this by making it easy to adjust output, add to it, etc. Set up your analysis once and tweak it as you need.
If you need to adjust your analysis, you can just click in to it and adjust options again.
Correlations
Running a Correlation Analysis:
Let’s correlate all of our continuous variables. Of course, because the species variable has more than two levels, including it in a correlation doesn’t make sense.
Load Your Dataset: Use a dataset like iris to examine relationships between continuous variables.
Navigate to Correlations:
Go to Analyses > Regression > Correlation Matrix.
Select Variables:
Choose the variables you want to correlate (e.g., mpg, hp, wt). Jamovi will automatically generate a correlation matrix showing Pearson’s r values.
Adjusting Options:
Under Statistics, select options to display the p-values and confidence intervals for each correlation.
For non-parametric data, you can switch to Spearman’s correlation in the same options panel.
Check out a visualisation
As for a plot of your correlation matrix. Add in the densities and statistics. One plot to rule them all.
Now say you want to switch to Spearman instead of Pearson. Change the option in the setup. What happens to your output?
Visualising Correlations:
Scatterplot: To visualise relationships, go to Plots > Scatterplot and select pairs like Sepal Length and Width. This plot provides a visual representation of the correlation, with a line of best fit showing the trend. Or just do the plots in the correlation matrix.
Comparing to SPSS:
SPSS requires selecting “Bivariate Correlation” from the Analyze menu, where Jamovi makes correlation setup simpler by presenting all options in one panel.
In Jamovi, scatterplots can be generated directly within the Correlation analysis, while SPSS often requires additional steps in separate menus.
Confidence intervals can be added easily. In SPSS, an add-on is required.
As always, the output is live, meaning it can easily be updated.
Linear Regression
Running a Linear Regression:
I am not a plant doctor, so I have no idea if any of these analyses make theoretical sense.
Load Your Dataset: Use iris to explore relationships.
Navigate to Linear Regression:
Go to Analyses > Regression > Linear Regression.
Select Variables:
Dependent Variable: Choose Petal Length
Add predictors. Covariates are continuous, Factors categorical.
Step 1 – let’s add Petal Width in Covariates. Check out your output.
Step 2 – add Species in factors. What do you notice about your output for species?
Model Options:
In the Model Builder, you can add or remove predictors and set up multiple regression models.
Let’s make this hierarchical. Remove Species from block 1 and add it to block 2. What happens to your output?
Let’s say we wanted a different reference level for Species. How can we adjust this?
Check some assumptions, like collinearity. Consider a Q-Q plot.
Standardised Estimates: Check this box (under Model Coefficients) to get beta values, making it easier to compare predictor impacts on the dependent variable.
Interpreting Results:
Coefficients: The output includes estimates, standard errors, t-values, and p-values for each predictor. Significant predictors (p < .05) indicate a meaningful contribution to explaining the dependent variable.
Model Fit:
R-squared: Shows the proportion of variance explained by the model. Higher values suggest better fit.
Adjusted R-squared: Adjusts for the number of predictors to provide a more balanced view of model fit.
Comparing to SPSS:
SPSS requires navigating through multiple dialog boxes to set up a regression model, while Jamovi’s interface keeps everything in a single, adjustable panel.
It’s super easy to check assumptions or update your analysis.
Installing Extensions and Running a Mediation Analysis
Installing Extensions in Jamovi:
Access the Jamovi Library:
Go to Modules > jamovi Library in the top menu.
Browse Available Extensions:
You’ll see a list of available extensions for extra analyses, including mediation, factor analysis, and advanced regressions.
Install the Mediation Module (medmod):
Find the medmod module and click Install. It will appear in your Analysis menu once installed.
Setting Up a Mediation Analysis:
Again, not a plant doctor, no idea if this makes theoretical sense. That’s not the point here.
Open the Mediation Module:
Go to Analyses > Mediation.
Define Variables:
Outcome Variable (Y): Sepal length
Mediator (M): Petal length
Predictor Variable (X): Sepal width
Run the Analysis:
Jamovi will display the indirect effect (through the mediator), direct effect (predictor to outcome), and total effect, along with their confidence intervals and p-values.
Adjust some settings
Bootstrap your SEs (it will take a moment)
Ask for path estimates
Ask for labels and confidence intervals
Interpreting Results:
Direct Effect: Shows the effect of the predictor on the outcome without mediation.
Indirect Effect: Indicates the effect that passes through the mediator, giving insight into the mediation’s strength.
Total Effect: The overall effect, combining direct and indirect paths.
Comparing to SPSS:
In SPSS, mediation requires either a custom plugin or syntax coding. Jamovi simplifies the process by offering a ready-to-use module, which is accessible without additional setup.
Jamovi’s output is immediate and includes bootstrapped confidence intervals for indirect effects, making results clearer and more accessible.
Note that many additional modules are available. Explore! There are even games.
Filtering, Recoding, and Computing Variables
Filtering Data:
Access Filter Options:
Go to Data > Filter. This will open a filtering panel.
Create a Filter:
Type your condition directly into the filter (e.g., Sepal.Length > 5)
You’ll see only rows that meet this condition, while others are temporarily hidden.
Note that the filter applies to the whole set of analyses. If you only want to filter certain analyses, you may need to do them in a separate .omv file.
Recoding Variables:
Note that this is a little tricky at first. I don’t find this all that intuitive.
Recoding in the Data Tab:
Select the variable you want to transform
Select Data > Transform
Note that it will automatically create a new variable that is identical to the previous one.
Example 1, applying the same transformation to all levels (e.g., adding 10 to each for some reason).
Make sure the Source variable is correct (i.e., the one you want to base your transformation on), and then note that so far it’s using transform none. From the Dropdown menu, select “Create new transform” and it asks for a recode condition. $source+10
Example 2, recoding categories, with labels.
If you want to make Sepal.Length into categories, use this condition: IF(Sepal.Length <= 4, "Small", IF(Sepal.Length <= 5, "Medium", "Large"))
Of course, you could combine conditions. I don’t find this syntax all that intuitive, but actually it’s just Excel syntax. This is one thing that can take some getting used to.
Note that transforms are saved, so if you need to do the same thing to multiple variables, you don’t have to type it each time.
Save the Recoded Variable:
Jamovi will add the new variable to the spreadsheet with values based on your recode criteria. The name of the new variable may not be accurate, so be careful!
Computing New Variables:
Creating a Computed Variable:
Go to Data > Compute.
Enter an expression, such as calculating overall length (e.g., Petal.Length + Sepal.Length), which I’m sure makes no sense, but it’s just a demonstration.
Using Functions:
Jamovi has built-in functions like MEAN and SUM. To compute the average of Sepal.Length and Petal.Length, use MEAN(Sepal.Length, Petal.Length).
Save the Computation:
Once entered, Jamovi will add the computed column with calculated values, which updates automatically if the original data changes.
Example of Recoding a Categorical Variable in Jamovi
Here’s an example from another dataset, where we’re recoding something into labels. This’ll make the output more clear.
Let’s say we want to recode the variable cyl (cylinder count) in mtcars to make it more descriptive:
Access the Transform Option:
In the data view, go to Data > Transform.
Select the Variable to Recode:
Choose cyl as the variable to recode.
Set Recode Criteria:
Here’s out plan
If cyl is 4 → recode as “Four Cylinders”
If cyl is 6 → recode as “Six Cylinders”
If cyl is 8 → recode as “Eight Cylinders”
Use the formula IF(cyl == 4, "Four Cylinders", IF(cyl == 6, "Six Cylinders", "Eight Cylinders")).
Save the Recoded Variable:
Jamovi will create a new column with the recoded categories.
Results:
Your cyl variable now has clear, descriptive categories, making output and visuals more intuitive.
Comparing to SPSS:
SPSS requires multiple dialog boxes to filter, recode, and compute, while Jamovi combines these functions within the Data menu, making it simpler and faster to use.
Jamovi’s formula syntax can be a bit different from SPSS syntax, but it provides a clear structure and updates instantly, reducing steps for each new computation.
Some recoding examples.
# Recode values above 50 as 1, others as 0
IF(score > 50, 1, 0)
# Recode into three categories: "Low", "Medium", "High"
IF(score <= 30, "Low", IF(score <= 70, "Medium", "High"))
# Flag cases where age > 18 and score > 50
IF(age > 18 & score > 50, 1, 0)
# Calculate a squared value of score
score^2
# Apply a log transformation to a variable
log(variable)
Exporting your work
Right click on a figure and Copy, then paste into a Word doc. Hopefully you’ll see that it looks good! Nice and easy to insert into your document.
Q-Q Plot
Right click on a table and paste into Word. How does that look? It can be a bit messy, but it’s editable. See below, where I’ll need to fix up some columns. Still, it’s MUCH better than typing it all out.
Hang on, my colleague has just come back to me and said they want all the output to 2dp, not 3sf. Uh oh. Easy! Go to settings, adjust number format to 2dp.
The output still needs a little bit of work, as values are wrapping over lines, but at least now you can put it into Word and edit it more easily there, rather than trying it all out by hand.
You can edit your output and add notes, code blocks and lots more. Try it out.
Next steps for you
There’s loads more to learn about Jamovi.
I’d suggest playing around with it, sharing datasets amongst yourselves (including your analyses) and seeing what you learn.
If you would like, I have set up a MS Teams team to explore Jamovi. I expect that it will be a bit quiet at first, but will hopefully build over time. It’s a great place to play around with it and explore how it works with each other. Everyone is welcome, including those who couldn’t be here today.
Jamovi - learning how it works and sharing tips | General | Microsoft Teams