I'm reviewing a scientific article to be publish. I tried to evidence that pathogens collected from different geographic areas clustered. I used Nearest Neighbour test to prove it. However, I've noticed that no p-value is assigned in my output. So, how could I get the p-value from my analysis? It will be for a publication. Otherwise, how could I explain the significance of the result?

This is the output obtained from my analysis using QGIS3:

Observed mean distance: 495.1212145321877

Expected mean distance: 654.1975182291224

Nearest neighbour index: 0.7568375005036007

Number of points: 28

Z-Score: -2.46153773279898

Could anyone help me? Thank you in advance

  • It's unfortunate. Any chance you'll try Whitebox-GAT's Nearest-Neighbour Analysis tool? – Kazuhito Apr 25 '20 at 9:58
  • Unfortunately the native tool in QGIS does not output a p-value. Your Z-Score would indicate clustering in your dataset. You can infer a confidence level by looking at a critical value and your sample size but I would like to point out that a sample size of 28 is below what is recommended. – AWGIS Apr 25 '20 at 11:27
  • My sample size is 45. I think a solution might be the following: since Z-score is based on the standard normal variate, I can calculate the p under the curve. For α=0.05, Z would take values +/-1.96, so if |Z| > Zα/2 indicates a significant clustering pattern, otherwise my sample is overdispersed. – Agar Apr 25 '20 at 22:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.