Is it possible to run ANOVA on this data?

Question

I am requesting your guidance on analysing some data. This is data that was collected to determine the levels of concentration of Toluene in soil samples at a petroleum refinery. The hypothesis of the study was to determine whether there was a significant difference between the Toluene concentration level detected in soil and the Dutch Intervention Value (DIV). I used SPSS to do an independent sample t test and obtained the results. However, my supervisor is insisting that an ANOVA test should be run to determine the significance of the concentration between the sampling points. Looking at the data, I see that it is not suitable for ANOVA test. I have attached the data. Kindly, can you advise me on the way forward. Thanks

SP1- 1035.36

SP2- 1278.75

SP3- 1079.44

SP4- 1200.49

SP5- 694.9

SP6- 728.65

SP7- 673.74

SP8- 1005.14

SP9- 1117.02

SP10- 24.05

Answer 1

Hello,

To address your request regarding the data analysis in SPSS, let's clarify the steps you should take based on the scenario you've described:

1. Review of Statistical Tests:

1. Independent Samples t-Test: This test is appropriate when you are comparing the mean of a single sample to a known value (in this case, the Dutch Intervention Value or DIV) to see if there is a significant difference. This seems to be the correct approach if you are testing against a single reference value.
2. ANOVA (Analysis of Variance): This test is used to compare the means of three or more groups to see if there are significant differences between them. It’s not appropriate for comparing a single sample's mean to a reference value. In your case, if you are only comparing the concentration levels to the DIV, ANOVA is not suitable unless you have multiple groups or sampling points with different conditions.

Recommendations:

1. Stick with the Independent Samples t-Test if your primary aim is to compare the mean concentration of Toluene in soil to the DIV. Ensure you have the DIV value correctly entered as the test value.
2. Use ANOVA only if you have data from multiple groups or conditions that need comparison.

I hope this helps. If additional questions are needed, please feel free to reach out. Take care

-Dr. Christal-Joy Turner

Answer 2

I have questions: How many measurements were made at each sampling point? Is it reasonable to assume that the mean value of many sets of samples of a given size N, taken at a particular point, would have a fairly "normal" distribution?

If only 1 measure was taken at each sampling point, then 1-way ANOVA using "sampling point" as the classifying variable is not going to be possible.

Let me know,

Phil C.