In 2013 the first Swedish national tests in science for school year 9 developed under the new curricula and syllabi were introduced. For these tests the standard setting is done using a modified Angoff method. Normally, the standard setting session begins with a part where the panellists are trained to use the Angoff method. However, there have been doubts if this training is enough and for the test in 2015, the panellists underwent an extended education about standard setting in general and how to use the modified Angoff method. The aim with the education was to find out if an extended education render more homogeneous and consistent estimations. Since the tests are developed from the same framework every year, the grading criteria in the syllabi are not changed and the panellists were almost the same for the test in 2014 and for the test in 2015 it might be possible to compare the standard setting between the two years. The preliminary results indicate slightly lower standard deviations which might be due to the education. In the presentation results from the study will be presented and discussed.