Statistical Calculation of the minimum extra yield when using TerraCottem

Seen the various results in various places with various plants, we used a statistical method to calculate in what way we could be sure of results, without having done several tests in Burkina Faso ourselves.

To achieve this, we used 32 research documents of tests done with TerraCottem in various places around the world, both in reality and in controlled area's like test centres of universities.

We reduced the number of parameters to what we thought were relevant:

Could we consider the test as a stress situation in regards to:

• the soil
• watering / irrigation

Is the test done in a real situation or in lab conditions ?

If in a real situation, was the test done in Burkina Faso or not.

This last parameter dropped very quickly, as we only had 2 results of tests in Burkina Faso , which is a bit poor for decision making.

The results of this statistical study

We devised a log performance to bring our data down to normal dispersion.

A first survey showed no significant difference for soil stressed or not.

See first table, the values with 0 for stressed soil 1,9856, against 2,0698 for 1.

In the second table we can find that this does not make a significant difference. (0,256)

We did find a significant difference for real situations against lab conditions.

For 0 (for lab situation) we find 2,4665

For 1 (for real situation) we find 1,8594

As, in the end, we are only interested in the results in the field, we keep the 18 results for the real situation, with the factor water "stressed." (See Table 3 on page 2)

In this situation, we tried to find what range of results we can expect, on average and in the worst case.

The 95% confident interval gives us a worst case situation of 1,6783 in logarithm, or 47,68% more yield when using TerraCottem. (See Last table)

Conclusion:

From the results of these 32 tests, we can expect to have at least 47,68% more yield, when using TerraCottem. This in 95% of the cases.

This is the figure we will use to calculate the financial simulation for growing vegetables.