Page 66 - PhD GT

the way forward is to enhance the accuracy of self-report by examining the interactions of the four main factors which affect it. These are: the interview situation, how the information is elicited, what is going on with the respondent at the time and the context of the interview.
3.3.3 Enhancing the validity of self-report
Citing Lessler and Sirken (1985), in a (US) National Center for Health Statistics study, Midanik (1988) proposed the following key questions needed to be addressed in consideration of the validation of information gathering methods. Some of these questions have been addressed in the present study: i) the effects of bounded and unbounded reference periods on the temporal judgements of survey respondents. This question is addressed with reference to each of the measurement instruments used in the present study; ii) the effects of complex concepts on the comprehension of survey questions. This issue is discussed with reference to the Leeds Dependence Questionnaire, the testing of its readability and participants’ responses to it in Chapter 4. Questions relating to the complexity of the instruction given for completion of the Impaired Control Scale are discussed in Chapter 5. iii) The effects of conditioning on responses to survey questions. Differences between questionnaires and time lag between responses would not allow for conditioning of the response to occur, though problems of response set relating to one of the instruments used (namely the Impaired Control Scale) are discussed in Chapter 5.
3.3.4. Maximising the reliability and validity of self-report
In conclusion, verbal report procedures have become the dominant form of obtaining clinical data on alcohol and drug problems. The question of whether or not they are reliable is less easy to address than the question to what extent do the instruments used and the circumstances of their administration enhance their reliability. There is evidence that reliability and validity vary depending upon the context and timing of information gathering, the characteristics of the person gathering the information, the state, both physical and motivational, of the respondent and the sensitivity of the information sought. Rather than pursue categorical answers about reliability and validity, the question of when bias does occur and what might be the best ways for minimising it are the key issues to be addressed.
Table 3.1 summarises the sources of bias and the proposed remedial action to minimise their effect which have been discussed above.
54