Skip to content
Omnikalc

Dixon Q test calculator

Enter the suspect value, its nearest neighbor, the value at the other end of the sorted data and the number of measurements to get the Q statistic, the critical value at the chosen confidence and the verdict.

Updated on
Your inputs

Value out of range

Value out of range

Value out of range

Value out of range

Results

Q statistic

0.7750

Critical value Q
0.710
Verdict
1
Outlier: the suspect value can be rejected at this confidence level
Gap to the nearest neighbor
0.3100
Range of the data
0.4000
Margin over the critical value
0.0650

The Dixon Q test calculator tells you whether a suspect value is an outlier in a small sample of 3 to 10 measurements, from the Q statistic and its critical value at 90, 95 or 99 % confidence.

How the calculation works

  • Enter the suspect value, its nearest neighbor and the value at the other end of the sorted data.
  • Enter the number of measurements and the confidence level.
  • The calculator gives Q, the critical value, the verdict, the gap, the range and the margin.

Example

Five titrations give 10.12, 10.15, 10.18, 10.21 and 10.52 mL. The suspect value is 10.52, its nearest neighbor is 10.21 and the other end is 10.12. The gap is 0.3100 and the range 0.4000, so Q = 0.3100 / 0.4000 = 0.7750. At 95 % with 5 measurements, the critical value is 0.710. Q is above it by 0.0650, so the value is an outlier and can be rejected.

Why it matters

The calculator helps chemistry and biology students, laboratory technicians who run replicate measurements, and quality analysts who must decide whether to keep a doubtful result.

Good to know

Sort the data first, then test the largest value with the gap to the second largest, or the smallest with the gap to the second smallest. The test is meant for data that are otherwise normally distributed. It has low power with three measurements, where only an extreme value is detected, so it is better to collect a few more. Never reject a value only because it looks odd: look for a cause such as a spill, a misread burette or a calculation slip, and document the decision. The critical values come from the standard published table for the single-gap form of the test. For larger samples, the Grubbs test is preferred.

The Grubbs test calculator handles larger samples, the outlier calculator uses the interquartile range, the average calculator gives the mean and median and the limit of detection calculator covers the lowest measurable level.

Frequently asked questions

What is the Dixon Q test?

It is a simple test for a single outlier in a small sample of 3 to 10 measurements. The statistic is Q = gap / range, where the gap is the distance from the suspect value to its nearest neighbor and the range is the distance between the largest and smallest values. If Q is above the critical value for the sample size, the value is rejected.

Which confidence level should I use?

The 95 % level is the usual default in analytical chemistry. Use 99 % when you want to be very sure before throwing a measurement away, and 90 % only for a first screening. A higher confidence level makes rejection harder.

Can I reject several values?

Avoid it. The test is designed for one suspect value per data set. Applying it again on the remaining data inflates the risk of wrongly discarding good measurements. If you find more than one suspect value, repeat the experiment or use a robust statistic such as the median.

Was this calculator helpful?

Thanks for your feedback!

Results are estimates provided for information purposes. They do not replace professional advice.