Statistical method to compare datasets, Help!

Question

Hello JMP!

I´m am currently writing my thesis where I am looking at environmental pollution during construction of marine structures. From four locations/construction projects I have datasets from continuous measurments of pollutant x concentration (ex. /hour for 3-10 weeks depending on location). The data consists of samples within the construction boundaries and representative reference sample data measured in the same way.

Example:

Location1: 1.3, 2,1, 1,3, 4,1, 2,8, 2,2, 3.4....
Location1reference: 2.1, 2.3, 4.6, 1.1, 2.2, 3.0.....

(over time)

I´m looking to test whether the samples at the construction site is statistically different from the reference samples.

I´m now seeking input to how to approach this problem and what statistical test that can suit this question.

Any help will be much appreciated!

/envMikael

envMikael · Answer

Many thanks! @Bill_Worley

Bill_Worley · Answer

@envMikael a good place to start is with the Distribution platform under Analyze.  More than likely though your best option will be Oneway Analysis in the Fit Y by X platform under Analyze.  Location will be X and you may need to stack the data to get the best overall comparison.  Within the oneway analysis you can use Compare Means (All Pairs, Tukey HSD) and Means ANOVA.  To help interpret these results use JMP documentation.  You can use Google to get to the documentation just make sure whatever you search under that you include JMP in the search text.

HTH,

Bill