抄録
The paper deals with some problems of lunnar tidal anlyses of ionospheric height, which are treated from the standpoint of mathematical statistics. The semi-amplitude of lunar semi-diurnal tide of ZmF2 (height of maximum electron density of F2 region) analyzed from Japanese data is tested by statistical hypothesis and its results are described. Next the error caused by a rough scaling of height is estimated in connection to scaler's skill (or carefulness), and the results are given.
Some statistical measures are introduced for the criterion of the validity of analyzed results in the above treatments, that is, the level of significrnce for the former case (test of statistical hypothesis by non-parametric method) and the coefficient of confidence for the later case (estimation of error).
