Choose an analysis that answers your question
Choose a method that matches your data and the question you want to answer. Select the range, run the analysis and review its report; you do not need to build a separate formula model for each test.
| Question | Useful workflow | What it summarizes |
|---|---|---|
| What does this numeric data look like? | Descriptive statistics or histogram | Counts, central tendency, spread, or frequency distribution. |
| Do two numeric variables move together? | Correlation or covariance | Direction and strength of an association; this alone does not establish causation. |
| How is an outcome related to several predictors? | Regression | Model fit, coefficient estimates, uncertainty statistics and ANOVA output. |
| Do group means differ? | ANOVA or a suitable t-test | A statistical comparison of group means under the method’s assumptions. |
| How does a series change over time? | Moving average or exponential smoothing | Smoothed values that help inspect a time series. |
Regression and ANOVA reports
For regression, select the outcome (Y) range and one or more predictor (X) columns with matching observations. The report includes regression statistics, an ANOVA table, coefficient estimates and related diagnostic values. Use it to investigate relationships and fit; do not interpret an association as proof that one variable caused another.
- Use single-factor ANOVA to compare group means.
- Use two-factor ANOVA to examine row and column factors, with or without replication as appropriate.
- Include or exclude headers consistently, and confirm that selected observations align across input ranges.
Check the data before you run an analysis
Keep the source data intact while you validate an analysis. Select a clearly bounded range, tell the tool whether the first row contains labels, check any second input range, and write results to a new worksheet. Then inspect the report and confirm that missing values, sample size and test assumptions make sense for your data.
When built-in Google Sheets™ features are enough
For everyday summaries, sorting, filters, pivot tables, duplicate removal or trimming spaces, the built-in features in Google Sheets™ may be all you need. Sheet Analyst is intended to complement that workflow with guided statistical tests, regression reports and other analysis methods.
Common questions
Can Sheet Analyst run regression in Google Sheets™?
Yes. The add-on provides regression reports with model statistics, ANOVA output and coefficient estimates. Choose aligned Y and X ranges and review the assumptions and output before using the model.
Does ANOVA prove that one factor caused a result?
No. ANOVA tests whether group means differ under its assumptions. Study design and domain knowledge are needed to make causal claims.
Where are the analysis results written?
You can review the output destination in the tool. The standard workflow creates a result worksheet and leaves your selected source range unchanged.