Inverse-variance weight: what it means and how to read it
In the report: weight — n / variance; empty for a column without spread or with fewer than 2 values.
How to read it
1 / SE² of the mean: the weight of the column in inverse-variance weighting (as in a meta-analysis) — larger for columns with more values and less spread.
In plain words
A researcher had the average marks of two classes and wanted to join them into one average. Class A has 40 students whose marks are close together, and class B has 10 students whose marks are more spread out. She asked for weight. It came out 10 for class A and 2 for class B. Now she knows that when the two averages are put together, class A should count five times as much as class B, since its average is known more surely. A class where every mark is the same, or with only one student, gets an empty cell.
Where you see it
STATS_DESCRIPTIVE— Column-wise descriptive statistics with flexible metric selection and optional column filtering.
Related values
- z-score of a value — The z-score of the value v: (v − mean) / std; 0 for a column without spread; empty without v.
- Weighted mean — The weighted mean with the weights in the column weightsCol, row by row; a zero, negative or missing weight leaves the row out, as SPSS WEIGHT. Empty when no row has a usable weight.
- Standard error — The standard error of the estimate: how much it would vary from sample to sample.
- Number of values — The number of values (rows) the result is computed on.
- Sum — The sum of the values; empty for a column without a single number (the sign . in a STATS_COMPUTE row where no item is answered), not 0.
Get it in your sheet
One formula in a cell gives the whole report, Inverse-variance weight included. ALS Stats is a free add-on for Google Sheets™.
Install ALS Stats from Google Workspace Marketplace (opens in a new tab)