Getting started with GraphHelix

This guide walks you through your first analysis: creating a project, importing data, running a statistical test, and getting publication-ready results.

Step 1: Create a project

After signing in, you'll land on your Dashboard. Click New Project, type a name (e.g., "Drug Response Study"), and press Enter. You'll be taken to your new project immediately.

Each project holds one or more datasets and all the analyses you run on them. Think of a project as one study or experiment.

Step 2: Import your data

Click Import in the project header (or the "Import Data" button if the right panel is empty). A modal will appear where you can drag and drop your file or browse to select it.

Supported formats:

• CSV (.csv) — comma, tab, or semicolon delimited
• Excel (.xlsx, .xls) — first sheet is imported by default
• SPSS (.sav) — variable labels and value labels are preserved
• Stata (.dta) — variable labels are used as column headers by default

GraphHelix automatically detects column types (number, text, date) and shows a preview. You can adjust column types before importing if needed.

Once imported, GraphHelix's AI automatically scans your data and suggests which statistical tests might be appropriate based on your data structure.

Step 3: Run a statistical test

You have two ways to run a test:

Option A: Ask the AI (recommended)

Type your research question in the chat panel. For example:

• "Compare the treatment and control groups on blood pressure"
• "Is there a correlation between age and recovery time?"
• "Run a one-way ANOVA on pain scores across the three drug doses"

The AI will evaluate your data, check which assumptions are met, and recommend the most appropriate test. It explains why it chose that test, so you can make the final call.

Option B: Select a test manually

Use the test selector in the analysis panel to choose a specific test. Select your variables (outcome, grouping, etc.) and run it. GraphHelix still checks assumptions automatically even when you select the test yourself.

Step 4: Review your results

Every test in GraphHelix produces the complete statistical output:

• Test statistic — the value that determines significance (e.g., t, F, U)
• Degrees of freedom
• p-value — formatted per APA convention (no leading zero)
• Effect size — with magnitude label (small, medium, large)
• 95% confidence interval
• APA-formatted string — ready to copy into your manuscript

Below the statistics, the AI provides a plain-language interpretation of what the results mean. If something isn't clear, click Ask follow-up to continue the conversation.

Step 5: Export your results

GraphHelix offers several export options:

• Copy APA string — one-click copy of the formatted result for pasting into your paper
• Export chart — download figures as PNG (150/300/600 DPI) or SVG, with journal presets for Nature, Science, PLOS ONE, JAMA, and Cell
• Generate report — create a full Methods & Results section in APA format from the "Export & History" dropdown
• Export analysis — download individual analysis results as Markdown or JSON from the History tab

What's next

• Supported tests reference — full list of all 30+ tests with details on what each reports
• Data format requirements — detailed specifications for each file format
• Statistical guides — educational articles on choosing tests, checking assumptions, and reporting results