Search

Objective: Both computerised and telephone surveys have potential advantages for research data collection. The current study aimed to determine the: (i) feasibility, (ii) acceptability, and (iii) cost per completed survey of computer tablet versus telephone data collection for clients attending an outpatient drug and alcohol treatment clinic.Design: Two-arm randomised controlled trial.Method: Clients attending a drug and alcohol outpatient clinic in New South Wales, Australia, were randomised to complete a baseline survey via computer tablet in the clinic or via telephone interview within two weeks of their appointment. All participants completed a three-month follow-up survey via telephone.Results: Consent and completion rates for the baseline survey were significantly higher in the computer tablet condition. The time taken to complete the computer tablet survey was lower (11min) than the telephone condition (17min). There were no differences in the proportion of consenters or completed follow-up surveys between the two conditions at the 3-month follow-up. Acceptability was high across both modes of data collection. The cost of the computer tablet condition was $67.52 greater per completed survey than the telephone condition.Conclusion: There is a trade-off between computer tablet and telephone data collection. While both data collection methods were acceptable to participants, the computer tablet condition resulted in higher consent and completion rates at baseline, therefore yielding greater external validity, and was quicker for participants to complete. Telephone data collection was however, more cost-effective. Researchers should carefully consider the mode of data collection that suits individual study needs. [ABSTRACT FROM AUTHOR]

Aim: To estimate the alcohol-attributable crime and traffic accidents for rural communities in Australia, controlling for potential bias., Method: For 20 rural communities in New South Wales, Australia, crime and traffic accident data was obtained from police records along with risky alcohol use estimated from a postal questionnaire. The relationship between community levels of risky drinking and crime and traffic accidents that occur in alcohol-related times is analysed controlling for the underlying level of crime by using the rate of incidents that occur in non-alcohol-related times., Findings: For the 20 rural communities, it was estimated that risky alcohol use is likely to have attributed to between 1.4 and 7.7 common assaults per 1000 population and between 0.6 and 1.8 serious traffic injuries or fatalities per 1000 population, every year., Conclusions: Rural communities in Australia are experiencing a sizeable amount of potentially avoidable harm due to risky alcohol use. Reducing the population levels of those drinking at risk of acute harm or improving the settings in which drinking takes place may have benefits for these communities, especially in terms of crime and traffic accidents., (2009 Elsevier Ltd. All rights reserved.)

Adolescent smoking prevalence is usually assessed via self-complete questionnaires. However, concern has been expressed about the validity of such self-report. One approach to increase validity involves the threat of biological validation, known as the bogus pipeline method (BPL).This study aimed to assess the effects of BPL, using an expired air carbon monoxide monitor, and of questionnaire anonymity on student smoking self-report data. High school students (n=801) were randomly allocated to one of four conditions: anonymous questionnaire+BPL, named questionnaire+BPL, anonymous questionnaire without BPL and named questionnaire without BPL. Overall, 37% of students agreed that questionnaires were a good way to obtain honest answers. In a logistic regression analysis, students in the BPL condition had significantly higher odds of reporting weekly smoking (OR=1.83 95% CI 1.27-2.65) and monthly smoking (OR=1.66 95% CI 1.21-2.28) but not of lifetime smoking compared with non-BPL students. Students in the named questionnaire condition had a significantly higher odds of reporting lifetime smoking (OR=1.49 95% CI 1.08-2.04) compared with anonymous students. Studies assessing current smoking patterns in adolescents should consider incorporating a BPL method.