If you’re new to Julius, welcome! We are excited to have you here! Julius has grown rapidly, and is trusted by individuals and organizations across many fields – whether you are a researcher, business professional, healthcare provider, or new to data analysis, Julius is a great tool to help hone your data analysis skills.
This guide will show you how use Julius as a data expert and copilot to perform advanced statistical tests like ANOVA, t-tests, regression analysis, and chi-square tests — all without the hassle of coding!
1. Selecting the Right Statistical Test
Before starting any data analysis, it is always important to know what test aligns with your goals. Below are some of the common tests and a quick description of their use:
- ANOVA (Analysis of Variance): Use when comparing means across three or more groups. Learn more about the ANOVAs here
- T-Test: Best for comparing the means of two groups. Learn more about T-tests here
- Chi-Square Test: Ideal for examining relationships between categorical variables. Learn more about Chi-Square Tests here
- Linear Regression: For modeling relationships between continuous variables. Learn more about Linear Regression here.
2. Uploading and Preparing Your Dataset
After choosing a test that fits your goals, you can start by uploading your dataset into Julius. Julius can handle multiple file types including .csv, .xlsx, google docs., and docx to name a few. You can follow this walkthrough to get a better understanding of the process of uploading datasets and a general walkthrough on Julius here.
Before you upload your dataset, you should also check to and make sure your data is in an easy to read format. Don’t know what that entails? Check out this guide on Formatting Column Headers.
Once you have properly formatted your columns, and uploaded your dataset into Julius. You should then ask Julius to ‘Preview the dataset’. Below is a screenshot of me prompting Julius to do so:
The image also depicts the ‘sidekick’ option, which gives the user a quick breakdown of what the file entails.
After bringing the file in, and proceeding to ask Julius to preview it, you should then ask Julius to examine the dataset and perform any data cleaning deemed necessary. Sometimes datasets may have missing information that could hinder the process of data analysis. Thus, handling missing values, checking for outliers and determining if the format is suitable for data analysis is of upmost importance when processing any dataset in Julius.
The image above shows Julius examining my dataset to determine if it is ready for analysis.
3. Running a Statistical Test with AI
So far we have done the following:
- Reviewed the different types of statistical tests available and align them to the goals of our data analysis.
- Made sure our dataset was in a file that we could upload to Julius.
- Checked our headers to ensure proper format.
- Uploaded our dataset to Julius, previewed the dataset and then performed any necessary data cleaning.
Our next steps:
- Performing a one-way ANOVA with Julius.
- Performing a post-hoc test.
- Reporting the Results.
- Creating a visual.
This example is from the Guide: One-way ANOVA. You can read more on the scenario, and the assumptions there. For this, I’m showing the results based off of the test. You just simply prompt Julius to run a one-way ANOVA and it’ll do the rest for you!
Aside: If you are unsure of the independent or dependent values, you can ask Julius to help you determine them based on your research question. I highly recommend figuring that out as well as determining the characteristics of your dataset before performing any test.
Quick help: Unsure on what test you should be using for your data? Visit parametric or non-parametric testing to see what type of test you should be using! (hint, hint, you can also ask Jullius to help)!
One-way ANOVA
Post Hoc
In the image above, we have both the p-value and the post hoc test to confirm the statistical differences between the groups. Now we can report the results.
4. Understanding the Results
Julius has provided us with the following values:
- P-Value: Tells you the probability of getting a particular result. If under 0.05, it is considered significant. For us, it was 0.004.
- F-Statistic (for ANOVA): Helps determine whether there are significant differences between group means. A larger F value indicates more variation between group means relative to the variation within groups. If it’s close to 1, it means that the variation between groups is similar to the variation within and no significant difference can be detected. For us, it is larger than 1.
- Degrees of Freedom: In a one-way ANOVA, there are two types of degrees of freedom: between group, and within group, which we report as F(42,2). See below for more explanation.
5. Visualizing the Results
We can visualize our results via a table and a graph. When showing any results, remember to denote significance with an asterisk or using lettering. Below is an example of how to format a table and figure.
Table 1. A one-way ANOVA analysis for the effect of fertilizer on height of plants. MS (Mean Square), F-value and significant value are displayed. Asterisks denote significant effects (*p<0.05, **p<0.01).
| Measurement | MS | F | Sig |
|---|---|---|---|
| Between Groups | 60.000 | 6.462 | 0.004** |
| Within Groups | 9.286 |
Figure 1. Boxplot depicting the effect of three different fertilizer treatments on plant height. Asterisks denote statistically significant differences between groups (here there would be a double asterisk above boxplot 1 and 3).
6. Reporting Findings
Remember to always report your findings! This is the most important aspect of research.
"A one-way ANOVA showed statistically significant differences between fertilizer treatment and plant height (F(42,2) = 60.000, p<0.05) (Table 1). Further post hoc tests revealed statistically significant differences between in plant height between treatment 1 (25.00±3.05) and treatment 3 (29.00±3.05)(p=0.0024), but no statistically significant difference was found between any other comparisons (Figure 1).
Conclusion:
Performing statistical analyses doesn’t need to be scary. With Julius, you can perform statistical analyses, and generate beautiful graphs without the hassle of coding! Try it out for yourself!