Is It Really Rising? Test the Trend

How it works

The series is ordered by the time column and every pair of time points votes: is the later value higher or lower? The vote total is the Mann-Kendall S statistic, standardized against its tie-corrected variance with a continuity correction to give a two-sided p-value. Sen's slope — the median of all pairwise slopes — states the trend's pace in the value's own units per period, with the classical rank-based confidence interval, and is carried across the window to give the implied total change. Finally the residuals from a simple linear fit are checked for lag-1 autocorrelation; above 0.3 the analysis warns that serial correlation inflates the test's confidence.

Use it whenever a chart looks like it's drifting up or down — a KPI, a cost, a sensor reading — and you need to know whether the trend is statistically real and how fast it is moving.

Not for seasonal or rise-then-fall patterns (it tests monotonic trend only), and not a substitute for a forecasting model when you need future values rather than a verdict on the past.

What data do you need?

Any table with a time column and a measured value. For example, monthly revenue:

Minimum 8 rows · Best with 20-500 time points

What's in the report?

Standard-library analysis: the Mann-Kendall trend test with Sen's slope — the classical nonparametric pair for deciding whether an ordered series has a real monotonic trend or is just noise. Computes the tie-corrected Mann-Kendall statistic with a two-sided p-value, Sen's slope (the median pairwise slope) with its classical confidence interval, the implied total change over the window, and an honest serial-correlation caution when autocorrelation would inflate the test's confidence. Works on any dataset: map a date (or sortable index) column and the numeric value to test.