Abstract
By using the STAI developed by Spielberger et al. we have investigated the validity and reliability of two scales, that is, State Anxiety (A-State) and Trait Anxiety (A-Trait), and at the same time have examined them under various conditions.
The results obtained are as follows:
1) As a result of factor analysis concerning 40 items of the STAI used in this research, we have confirmed that both A-State and A-Trait have independent factor structures of their own, and that the items of the scales also carry their own validity.
2) After due consideration of the test-retest reliability of the two scales, we have found that A-Trait has rather high stability. Moreover, we have noticed that Cronbach's alpha coefficients, which show the reliability of the two scales, are high. In consequence, we have confirmed the high reliability of the two scales.
3) In comparing the scores of A-State and A-Trait obtained from young healthy people with those from healthy aged people, we have noticed that the aged get low scores on each of the two scales, and that each score distribution shows an excellent fit to the normal one.
4) We have found that A-State scores go up significantly when people are in a condition of emotional stress, but that there is not any change of A-Trait scores.
5) We have observed a significant increase of A-State scores at each noise level over 75dB(A).
From the above-mentioned results, we have reached the conclusion that the STAI is very valid and reliable, and that the A-State scale, especially, is a very good indicator of men's psychological states under conditions of temporary stress.
