Data Collection Methodology - Overseas Trip Summary
In May 2010 I embarked for Germany to attend the Third International Business Data Collection Methodology Workshop hosted by the Federal Statistical Office (Destatis) in Wiesbaden. I presented two papers at this workshop, which was a unique opportunity to share DCM's work on questionnaire design and pre-testing with like-minded data collection methodologists. Discussions focused on learnings and challenges around questionnaire development and testing, instrument design guidelines, web surveys, communication and contact strategies, burden, use of paradata, use of administrative data and the future of statistical data collection. Workshop participants were particularly impressed with the number of centralised, coherent guidelines and standards the ABS has for questionnaire design and testing. I also learnt that other agencies face similar challenges to us in terms of pre-testing of business data collection instruments.
Following the workshop in Germany, I attended the Q2010 Quality Conference in Helsinki, Finland. At this conference, I presented two papers: one of my own regarding development of web form design standards, and one on behalf of my colleague, Stephen Cohen, regarding improving screening procedures for rare populations in the ABS. This conference covered broader quality issues than the workshop and was a great opportunity to learn from the wider methodological community about the quality issues affecting other statistical organisations. A memorable message from the closing keynote speaker, Professor Denise Lievesley, of King's College London, was that rather than focusing so much on reducing respondent burden, statistical agencies need to get better at explaining the reciprocal benefits and the necessity of statistics to the respondent.
Next I attended the 65th American Association for Public Opinion Research (AAPOR) Conference held in Chicago. With over 1,000 conference participants, and so many interesting and relevant parallel sessions I was spoilt for choice. A couple of sessions stood out, including one about mixed mode design and another on alternative data collection procedures (such as GPS and bio-specimens). The message from the mixed-mode design session was anything but mixed. In fact, researchers were presenting consistent findings that providing a choice of mode actually reduced response rates, which is contrary to the previous held thought that offering a choice of mode increased response rates. In terms of alternative data collection procedures, I heard a fascinating presentation about using Second Life (a virtual world) to research the polling of young adults.
During the trip, I also took the opportunity to meet individual methodologists or equivalent data collection methodology sections from various statistical organisations. The broad objectives of these visits was to learn more about the use of different modes of data collection, pre-testing methods, methodological training and instrument design guidelines and best practice. Meeting other experts in person and seeing demonstrations in their environment adds a world of value over e.g. reading their journal articles. I visited Statistics Netherlands, the Office for National Statistics, National Agricultural Statistical Service, Bureau of Labour Statistics and the US Census Bureau.
If you would like to know more about my travels or have a question about data collection activity at another agency please contact me, Kettie Hewett on (02) 6252 7295 or firstname.lastname@example.org.