Quickstart

Before you begin

You need:

• A Google or institutional account to sign in
• A CSV file of experimental data

No data yet? The sample dataset on the start screen works well for a first session — it covers a typical group comparison and is a good way to learn the workflow before bringing your own data.

If you want a quick overview of what Licklider is designed to do, and where its boundaries are, read What This Product Does first.

Step 1 — Sign in

Go to app.licklider.ai and sign in.

Step 2 — Start a session

The start screen offers two entry points:

Upload a CSV

Select your file. Licklider reads its structure and opens the workspace with your data loaded. Depending on how the file is organized, it may ask a clarifying question or two before proceeding — for example, which column identifies your groups, or what the observation unit is.

Start with a message

Describe your experiment or research question in plain language. Licklider will ask for your data when it needs it, and may ask follow-up questions to understand your analysis intent.

Step 3 — Confirm your data and intent

Once in the workspace, Licklider will surface what it understood about your data: the variables it found, the groups it identified, and the analysis it thinks you're after.

Review this before moving on. If something looks wrong — a column misidentified, a grouping missed — correct it here. The analysis that follows depends on this foundation.

Some important facts cannot be inferred from a table alone. Observation units, hidden nesting, undeclared pairing, and the exact research question may still require your input. If those are wrong or left implicit, Licklider may not detect the problem automatically, and later results can reflect the wrong effective sample size, analysis path, or interpretation.

Step 4 — Get your first figure and result

Ask for a figure in the Chat panel, or let Licklider suggest one based on your data structure. The Canvas shows the result. The first output typically includes a figure, a statistical result, and an initial set of quality check disclosures.

Step 5 — Review quality checks

Open the Inspector panel. The quality checks tab shows what Licklider verified before and during the analysis: assumption checks, outlier disclosures, sample size and replication structure. Read through these before drawing conclusions or sharing results.

These checks are important, but they do not replace study-design knowledge. Licklider can flag many structural problems, but it cannot infer every design fact or prevent every misuse on its own.

If you want to understand those limits in more detail, see Observation Unit Declaration and Statistical Independence Check.

Step 6 — Export

When you're ready, export from the Inspector or Chat:

• Figure — publication-ready image
• Results table — statistical summary
• Methods text — a draft for the methods section of your paper

The workspace at a glance

Open Licklider →