Super Toy: a statistical issue

Some Indonesian people were surprised by the appearance of Super Toy case. It is the name of rice variety that is claimed could yield 15 ton per hectare. What a fantastic amount! The name was called after its inventor Toyong (his original name is Supriyadi). President of Republic of Indonesia, his expert staffs as well as ministry and governor were convinced that this new variety is the answer for our food crisis. Thanks God.

Farmers in several districts agreed to plant this variety replacing their custom one. Unfortunately, the harvest yields were not as high as stated by the inventor. Even, in some districts (eg. Yogyakarta, Klaten, Purworejo, Madiun) the yields is extremely bad. No farmer got the expected amount. What was going on? How come the results are so much different from the inventor’s claim?

I just want to say that probably the inventor (Toyong and team) did not apply an appropriate statistical experimental design. By doing good experimental design with scientific method, what was done by inventor should be repeatable. He should also do some multi-location trial to assess its stability. He should also try with enough number of replication. It’s a pity that the inventor has never described his design, so we could not evaluate and suggest better analysis. Knowledge of robust design might be necessary for current and future inventor to avoid some misleading conclusion due to the different environment setting between the experiment and the real farm condition.