The theory of testing hypotheses is quite hard to master.
To get acquainted the Test for two (binomial) alternatives is very suitable. The main concepts are shown in a clear and interactive way. It is illustrative to compare the test with different values of n, say n = 100 and n = 1000.
Settings and buttons
Hypotheses
The null hypothesis p0 and alternate hypothesis p1 can be filled in. The results on the screen are related to the null hypothesis, as is usual.
Parameters
The distributions are binomial. The value of n is the number of experiments. The decision line is at the number of successes where a decision voor p0 of p1 is taken..
In the graph this line is shown as a vertical line which can be dragged and switched on/off.
Choosing the decision line determines the value of the errors Type I and Type II.
Image
Decision line shows the line which can be dragged.
Errors shows the errors type I and II in the graph, the colors do match the colors in the table.
Decision region tshows under the graphs the region where the null hypothesis is accepted or rejected. Decision table shows the values of the errors type I and II, the colors do match the colors in the graph.
offers different output options