The DDS was a researcher-designed questionnaire which, of course, means that it was not an instrument for which there was prior psychometric testing. However, as noted in the dissertation itself, there were a number of procedures performed in an effort to boost the reliability and validity of this questionnaire.
The performed procedures involved giving the first draft of the instrument to a panel of experts and incorporating their comments into a second draft which was then given to a representative sample of older men. These older men were drawn from a local senior citizens club and varied in age from 63 to 78 years with a mean age of 67.54. Thus, the pilot subjects were drawn from the same subject pool as the subjects used in the final study and were roughly the same ages.
Reliability data (Hoyt Reliability Correlation Coefficients) were computed for all eight ego stages assessed by the Self-Description Questionnaire. These coefficients ranged from .61 to .80 with a median coefficient of 68.5. The coefficient observed for the final ego stage (the ego stage of interest in the conducted research) was .69, a value indicating at least a moderately good level of reliability.
Internal consistency coefficients were also computed for each of the eight ego stages. These coefficients ranged from .82 to .87 with a median value of .85. The value observed for the ego stage associated with old age was .84.
Regarding the sample used in this dissertation relative to samples used in other studies which have used the Self-Description Questionnaire, it can be noted that the instrument has been used with age groups across the entire lifespan. This includes school-age children (Boyd, 1964), young and middle-aged adults (Boyd, 1961; Boyd & Koskela, 1970), and samples of older and elderly adults (Boyd, Davie & Truax, 1970; Colley, 1966).
The LSI was developed by writing items which reflected each of the five components of the...