MOLS 2006

Submissions are invited for contributed papers to be presented at the conference, Methodology of Longitudinal Surveys (MOLS 2006). The conference is sponsored by the International Association of Survey Statisticians (IASS) and the UK Economic and Social Research Council (ESRC). The contributed papers programme will complement the programme of approximately 20 monograph papers, which have already been chosen following an earlier call for submissions. Preliminary titles of monograph papers are listed on the conference website.

Contributed papers should address important methodological topics in the design, implementation or analysis of longitudinal surveys. The focus is primarily on surveys that involve collecting data from subjects on multiple occasions. The subjects could be individuals, households, businesses or other establishments. Note that the conference is concerned with the methods used by such surveys, not with the substantive findings. Topics of interest include, but are not limited to, the following:

• Sampling for longitudinal surveys, including sample rotation, refreshment sampling and methods for dealing with population births and deaths;
• Sample management procedures and methods for tracing sample members over time and maintaining contact with sample members;
• Design methods for meeting dual requirements for longitudinal and cross-sectional estimates;
• Instrument design for longitudinal surveys, including the use of dependent interviewing;
• The effects of using different modes of data collection at different waves, multiple modes or changing modes;
• Issues in the collection of data from multiple members of a sampling unit (e.g. household, business), including the use of different respondents at different waves;
• Methods for minimising sample attrition;
• Assessment of non-response bias (unit and item) on longitudinal
  surveys;
• Adjustment methods for non-response and attrition, including weighting;
• Imputation and other methods for dealing with item missing data, especially wave non-response;
• Respondent conditioning, seam effects and other measurement error issues specific to longitudinal surveys;
• Methods for estimating measurement error that are unique to panel data;
• Use of administrative data to enhance longitudinal surveys;
• Methods for disclosure avoidance and privacy protection in the release of longitudinal data files;
• Ethical issues on longitudinal surveys, including confidentiality and consent;
• Variance estimation for panels, rotating panels and other complex designs;
• Analysis methods specific to repeated measures or duration and event data;
• Analysis of longitudinal survey data under complex designs and informative sampling;
• Nonsampling and sampling errors in panel data analysis;
• Other issues in the analysis of longitudinal survey data

Contributed papers may provide a review of research and practice in a particular methodological area or present the findings of new methodological research. Submissions will be screened by the MOLS 2006 scientific committee, who will select papers that address relevant topics and
collectively provide a balanced programme. There is also an opportunity for suitable contributed papers to be considered for inclusion in a special issue of the Journal of Official Statistics (JOS). Details of the process for submission to JOS will be made available once the conference papers have been chosen.

Submissions should consist of an abstract of approximately 500 words, describing clearly the topic that will be addressed, the material that will be drawn upon and the key messages of the paper. Please include the names, affiliations and email addresses of all authors. Abstracts should be submitted to mols-abstracts@isermail.essex.ac.uk by 30 November 2005, preferably as a Word document. Informal enquiries regarding possible submissions can be made to plynn@essex.ac.uk.

Authors will be informed whether or not their submission has been accepted by 31 January 2006. Authors of accepted contributed papers will be required to register for the conference and to present their paper orally.