Skip to main content
SheetTutor.
← All practice

Plot paired numbers and inspect a fitted relationship

A marker pairs one numeric X with its numeric Y. Unequal X gaps remain unequal. Fit each named series separately, then inspect the data and assumptions behind a predicted value.

1. Choose or edit the observed pairs

Fictional numeric observations, with no causal claim. IDs are unique; series labels are literal and case-sensitive. Each row needs both finite numbers, including measured zero. Use at most 50 rows.

IDSeriesXYRemove
Import local paired data

Use exactly ID,Series,X,Y headers. Preview validates the entire input before replacing the current source. Text and local files stay in this tab.

2. Inspect the numeric relationship

  • Bench A
  • Bench B
-5.05-0.40.7251.86.5412.2756.218.058.4Numeric XNumeric YBench A-1 / Bench A: X 1, Y 3Bench A-2 / Bench A: X 2, Y 5Bench A-3 / Bench A: X 4, Y 9Bench A-4 / Bench A: X 8, Y 17Bench B-1 / Bench B: X 0, Y 10Bench B-2 / Bench B: X 3, Y 4Bench B-3 / Bench B: X 7, Y -4
SeriesPairsSlopeInterceptCorrelation rR²Predicted Y
Bench A421119
Bench B3-210-112

Bench A: Prediction X lies within this series’ observed range; this alone does not validate a forecast.

Bench B: Prediction X lies within this series’ observed range; this alone does not validate a forecast.

Markers are not connected in input order. Repeated X values are separate observations. Fits use a free intercept, minimize squared Y residuals and are calculated per series, never by pooling unlike series. A high R² is an in-sample fit measure, not evidence of causation or future accuracy; no uncertainty interval is estimated.

Microsoft scatter-chart guidance ↗ · SLOPE · INTERCEPT · CORREL · RSQ

Inspect X and Y uncertainty separately

Each series has independent horizontal X and vertical Y amounts. Sample-SD bars use that coordinate’s series mean; SE uses each observation. These intervals do not fit a model, estimate a prediction interval or alter bubble size. A zero-size selected bubble still has a numeric coordinate and can carry error marks.

X coordinate

Error bars for the selected series

Supply lengths on the value axis. For a point at 20, a negative length of 2 and a positive length of 5 gives endpoints 18 and 25. These bars describe your chosen amounts; they do not establish a confidence interval.

Apply the uncertainty draft, then verify current sources. Definition history and worksheet history are separate. Changing error amounts or numeric coordinates requires a new check.

Y coordinate

Error bars for the selected series

Supply lengths on the value axis. For a point at 20, a negative length of 2 and a positive length of 5 gives endpoints 18 and 25. These bars describe your chosen amounts; they do not establish a confidence interval.

Apply the uncertainty draft, then verify current sources. Definition history and worksheet history are separate. Changing error amounts or numeric coordinates requires a new check.

Custom ranges assign one literal length to each plotted point in the order shown below. Error amounts C/D/E/F are X positive/X negative/Y positive/Y negative. The worksheet keeps all original IDs and series; a range’s row labels do not silently regroup its values. Blank active lengths require correction; measured zero is a valid length.

Check the statistics in a real worksheet

The source occupies A:D. Guarded SLOPE, INTERCEPT, CORREL, RSQ and FORECAST.LINEAR formulas appear in H. This copy does not change the chart until you import its source values. No model or chart is fitted automatically to a new causal claim.

3. Use the current analysis in Excel

The file keeps all original Source records, a selected/grouped Plot data snapshot, statistics and an actual native scatter chart with two numeric axes. Edit Plot data to update existing chart pairs. Re-export after adding pairs or changing selection. Frozen statistics remain fixed even if you edit the chart’s numeric data. Native Excel recalculation/edit/save and other clients remain separately unverified.