Toronto RDC Events

The Toronto RDC hosts a number of events each year. Please visit our Calendar Page for upcoming events.

Quick Links

Workshops

Currently, the Toronto RDC conducts two workshops on a regular basis. Please visit our calendar page for dates and times.

1. Introductory SAS Workshop

This workshop is open to RDC researchers as well as members of the larger research community and is presented at least twice a year.

This workshop is meant to help current researchers at the RDC acquaint or re-acquaint themselves with the SAS programming language. In this workshop, the basics of the SAS programming language will be covered, including creating variables, reading in raw text files, modifying SAS data sets, and merging SAS data sets. Finally, a highly interactive section on group processing rounds out the workshop.

2. Practical Bootstrap Estimation Workshop

This workshop focuses on the practical aspects of running bootstrap estimations using SAS Survey Procedures and the statistical software package Stata. We will review why we bootstrap, what procedures can be replicated using bootstrap estimation techniques, and what to expect from bootstrapped results. Using a subset of the NLSCY Synthetic files, researchers will then use the RDC's workstations to calculate estimates of their own with RDC staff guidance.

Back to top

Courses

The Toronto RDC has hosted the following courses in the past few years.

1. Panel Data Methods Course

The course provides a rigorous introduction to statistical methods for the analysis of panel data with specific application to the major Canadian longitudinal data sets. This course is offered in collaboration with the Toronto RDC. The RDC provides secure access to Canada's preeminent panel data sets for public policy analysis as well as variety of other Statistics Canada data. The course will take place within RDC providing students hands on experience with these important sources of information on public issues. The RDC offers both lecture space and a computer lab for tutorials.

While the specific goal of this course is to introduce students to empirical methods for the analysis of longitudinal data, an important by product is their exposure to the RDC data. These data are increasingly "the basis" for new survey based research in health, education, economics and other social sciences in Canada.

Instruction includes a combination of lectures and tutorials. In tutorials, students will complete series of problem sets that provide an introduction to the RDC panel data sets and practice in their analysis. The statistical methods reviewed will be drawn from a variety of disciplines to promote the inter-disciplinary study of public policy.

Certain topics of particular relevance to the RDC panel data (e.g., cluster sampling, bootstrapping) will also be covered.

The course is intended for a) students from the School of Public Policy and Governance, b) students from departments, schools and faculties where small numbers preclude a similar course being offered, or that desire instruction in the use of data housed in the Toronto RDC.

2. Ryerson University Statistical Analysis in Social Science Research Course

The Statistical Analysis in Social Science Research course is sponsored by Ryerson University and held at the Toronto RDC. This course offers graduate students from Ryerson a hands‐on research experience in advanced quantitative methods for generating and analyzing a variety of large social science datasets (Aboriginal People’s Survey, Canadian Community Health Survey, Census of Population, Ethnic Diversity Survey, General Social Survey, Labour Force Survey, Longitudinal Survey of Immigrants to Canada, National Population Health Survey, Participation and Activity Limitation Survey, Survey of Household Spending, Survey of Labour and Income Dynamics, and the Youth in Transition Survey). Students from Ryerson included doctoral students from the Policy Studies Program, the first and only interdisciplinary doctoral program in policy studies with three fields of study (Public Policy and Administration; Immigrant Settlement and Diaspora Policies; and Social Policy).

Back to top

Past Events

2015 CRDCN National Conference

Wednesday, November 4, 2015
Robarts Library, University of Toronto
Thursday, November 5 and Friday, November 6, 2015
Marriott Bloor Yorkville Hotel, Toronto ON

The CRDCN's 2015 National Conference was held for the first time in Toronto in the fall of 2015 at the Toronto Marriott Bloor Yorkville Hotel. The theme was "Research and Public Policy: Health, Economic and Social Perspectives" and the conference was attended by 150 participants and included 11 paper sessions with 43 papers and an additional 23 posters in a dedicated poster session. Session themes included health, immigrant well-being, work and well-being, wealth and inequality, education and human capital, economics, families and well-being, educational and health capital, health and health care access, as well as two sessions dedicated to the Ministry of Community and Social Services (MSCC) data. The first MCSS session was a panel chaired by Jean-Pierre Voyer (Social Demonstration and Research Corporation) and the panelists included Norm Helfand (Director of Income Security and Pension Policy, Ministry of Finance) and Didem Proulx (Director of the Centre of Excellence for Evidence Development, Treasury Board). They discussed the challenges that governments face in assessing the impacts of social policy interventions, and how government and academia can work together to address these challenges. The second MCSS session included two MCSS data presentations by RDC researchers involved in a pilot use of the data, and followed by a round table discussion. Both MCSS sessions were the result of meetings held with Ministry officials over the past two years and illustrated the mutual desire on the part of the Ministry and the CRDCN Executive to bring administrative data to the RDCs. Accessing both federal and provincial administrative data files and being able to cross-link these files with existing longitudinal data files currently in use at RDCs is a significant step in the future of the RDC network.

Back to top / Back to Past Events

Communicating Quantitative Social Science Research: A workshop for new scholars in economics, sociology, applied health studies, psychology, and other social science disciplines

Heather Juby PhD
Knowledge Transfer Coordinator for the Canadian Research Data Centre Network
Experienced researcher in family demography

Wednesday, March 16th and Thursday, March 17th, 2011
Toronto RDC Conference Room

Workshop objective: Through formal presentations, practical exercises and individual feedback, this two-day interactive workshop aims to help quantitative researchers (graduate students, post-docs, new faculty) build the skills for successful research communication. Working on their own material, participants learn about communicating quantitative social science research effectively to themselves, to academic audiences and to those outside their field - and gain new insights into their research in the process.

Workshop content: The workshop explores three key steps in the communication process...

  1. Working out what to communicate: how to identify the key elements of a research story; how to extract these elements from your research (or create them if they are not there) and structure them in a clear, concise and interesting way.
  2. Adapting the story to the audience: how to prepare a conference proposal or a scientific article abstract; how to develop presentation or poster content for specialist and non-specialist audiences; how to transform the story into a press release.
  3. Using visual tools to clarify the message: how to create a slide show that complements (rather than competes with) and clarifies (rather than complicates) an oral presentation; how to design a poster that attracts attention and communicates a clear message.
Back to top / Back to Past Events

CCHS 2.2 SIDE/Bootstrapping Workshop

Didier Garriguet
Senior Research Analyst
Health Statistics Division,
Statistics Canada

Wednesday, September 30 and Thursday, October 1, 2009

Wednesday: 1:00 pm to 5:00 pm - An overview of CCHS 2.2 (open to all RDC researchers)
Thursday: 10:00 am to 5:00 pm - SIDE/Bootstrapping workshop (open to CCHS 2.2 researchers)

This workshop is a live tutorial on the use of Software for Intake Distribution Evaluation (SIDE) and the bootstrapping technique. The use of both of these macros will be demonstrated using data from the Canadian Community Health Survey 2.2 (CCHS 2.2). The use of SIDE will be demonstrated to estimate the prevalence of inadequacy of a few micronutrients in some age/sex groups in the population. Likewise, the bootstrapping technique will be demonstrated to estimate the standard errors associated with the prevalence estimates used in the SIDE workshop.
This talk is open to RDC Researchers only, and will be held in the Toronto RDC at 130 St. George Street, on the 7th Floor in Room 7032.

Back to top / Back to Past Events

National Longitudinal Survey of Children and Youth (NLSCY) Workshop

Stephanie Lalonde
Rochelle Garner
Claude Girard
Yves Lafortune
Statistics Canada

Tuesday, September 30th, 2008 and Wednesday, October 1st, 2008, 9:30 am to 4:30 pm

This 2-day workshop will begin with an overview of the NLSCY, including a discussion of the content and changes over the cycles, sampling across the cycles, response rates (both cross-sectional and longitudinal), issues that could affect comparisons across cycles, and the sampling distribution of ages across cycles. NLSCY researchers will benefit from a discussion of the NLSCY's survey weights, including how they are created, how to use them for both longitudinal and cross-sectional research, how they can be used to make inferences about the population of children, how to pool samples and calculate new weights for the pooled sample, information about normalized/standardized weights, and why bootstrap weights should be used to generate final variance estimates. A presentation on variance estimation will explain sampling error and how it indicates if the sample size is large enough to reliably report a survey estimate, what confidence intervals and coefficients of variation are, and how bootstrap weights are used to estimate sampling error. Nonresponse issues will be detailed, including how the NLSCY weights account for unit nonresponse, how to deal with partial nonresponse, and how to evaluate nonresponse bias. A presentation on general research issues, such as sample size, sampling error and other errors in the survey will round out the presentations. A question-and-answer session will be held on Wednesday afternoon, and NLSCY researchers are encouraged to submit questions before the workshop.

NLSCY Workshop Presentations Slides (.pdf files):

1. Overview of the NLSCY (Stephanie Lalonde)
2. Statistical Concepts: Weighting and Variance Estimation (Claude Girard)
3. Nonresponse, Normalized/Standardized Weights, Pooling Data from Several Cycles, An application of the statistical concepts introduced on Day 1 using NLSCY data, "Use of Computers by Teenagers" (Yves Lafortune)
4. Things to Know and Question and Answer Session (Rochelle Garner, with Stephanie Lalonde, Claude Girard, and Yves Lafortune)

Back to top / Back to Past Events

Some Survey Data Sets Available from Statistics Canada's Social and Aboriginal Statistics Division

Cathy Connors
Michael Wendt
Susan Stobert
Kathy Marshall
Christina Jaworski
Statistics Canada

Tuesday April 1st, 2008, 1:00 pm to 5:00 pm
Room 230, the Fields Institute, 222 College St., 2nd Floor.

Members of the Social and Aboriginal Statistics Division at Statistics Canada will provide a series of 20 minute presentations on some of our key data sets. These rich sources of social data are or will soon be available for researchers to use within Statistics Canada's Research Data Centres Program. Additionally, a representative from one of Statistics Canada's refereed journals, Perspectives on Labour and Income, will discuss publishing opportunities at Statistics Canada.

The presentations will be of interest to researchers and graduate students in the social sciences. Please feel free to attend one or more of the presentations. We will also be available to answer any questions you have on the surveys. A representative of the Toronto RDC will also be on hand to answer any questions you might have on how to apply to the Research Data Centres Program.

Schedule (session title links to .pdf of presentation materials):

1:00 - 1:10: Introductory Remarks
1:10 - 1:30: Aboriginal Children's Survey: is a national survey of First Nations, Inuit and Métis children under the age of 6 years which collects information on the development and well-being of Aboriginal children.
1:30 - 1:50: Aboriginal People's Survey: provides data on the social and economic conditions of Aboriginal people in Canada. Its purpose is to identify the needs of Aboriginal people focusing on issues such as health, language, employment, income, schooling, housing, and mobility.
1:50 - 2:10: General Social Survey: is an annual cross-sectional household survey that gathers information on the social conditions of Canadians.
2:10 - 2:30: General Social Survey 20th Anniversary Project: We are constructing a coherent, easily used historical database containing data from all the 20 annual cycles in harmonized form.
2:30 - 2:50: Break: Feel free to talk with any of the representatives.
2:50 - 3:10: Participation and Activity Limitations Survey: is Canada's principal national survey focusing on persons with disabilities.
3:10 - 3:30: Statistics Canada Publication Opportunities: Statistics Canada provides a range of publishing options for RDC research. We encourage submissions on topics of relevance to citizens and policy-makers.
3:30 - 3:50: Vitality of Official-Language Minorities Survey: is the first survey that pertains specifically to Canada's official-language minorities. It contains a vast and rich source of data on these groups.
3:50 - 4:15: Discussion: Feel free to talk with any of the representatives.

The Toronto RDC would like to thank the Fields Institute for providing the venue for this presentation.

Back to top / Back to Past Events

Working with Canadian Census Data at the Toronto RDC

Martine Grenier
Mokili Mbuluyo, PhD
Jean-René Boudreau, PhD
Statistics Canada

Thursday, December 13th, 2007, 12:15 pm - 1:00 pm

This presentation is open to current researchers at the Toronto RDC only. Lunch will be provided.

Please see this poster for more information. This PowerPoint presentation is also available for download.

Back to top / Back to Past Events

Toronto RDC CCHS Mini-conference

Friday, October 26, 2007, 9:30am to 12 noon

The Toronto RDC is inviting RDC researchers to attend a mini-conference to be held on Friday, October 26, 2007. The focus of this conference will be the Canadian Community Health Surveys (CCHS). The conference is intended to allow researchers an informal environment in which to present their research issues or findings.

Didier Garriguet, Senior Analyst in the Health Statistics Division at Statistics Canada will be presenting an overview of the CCHS 2.2 survey from 9:30 to 11:00 on the morning of our CCHS Conference.

Following Didier's presentation, Valerie Tarasuk, Sharon Kirkpatrick and Sandra Fitzpatrick of the Department of Nutritional Sciences at the University of Toronto will give a presentation on "Using CCHS 2.2 to Describe Nutrition Disparities in Canada" from 11:15 to 12 noon.

The conference will take place in the Toronto RDC Conference Room, and is open to current RDC researchers only.

Download the .pdf version of the CCHS 2.2 presentation

Back to top / Back to Past Events

Toronto RDC Brown Bag Lunch Series 2007-2008

Prof. Charles Jones
Department of Sociology,
University of Toronto

Wednesday, October 10, 2007, 12:00 noon - 1:00 pm

"Effects of maternal smoking while pregnant on child outcomes in adolescence: Results from a prospective national longitudinal study."

There is a large body of research on the effects of fetal and neonatal events on adolescent and adult outcomes. This paper uses data from a recent prospective and nationally representative longitudinal study in Canada (NLSCY). Focusing on outcomes reported by ten to fourteen-year old children our results clearly demonstrate significant long-term effects of maternal smoking during pregnancy. Adolescents whose mothers had smoked while pregnant with them had worse scores on numerous self-report scales measuring both externalizing and internalizing disorders, including anxiety, self-image, general mood and various forms of aggression. These effects remain significant after controlling for birth weight, family structure and region but are attenuated when the socio-economic status of the child's household is controlled through statistical adjustments. Further analysis demonstrates significant regional differences in maternal smoking during pregnancy as well as a secular trend towards a lesser incidence over the 1990s. Implications for federal and provincial health policies are discussed.

Back to top / Back to Past Events

Toronto RDC Brown Bag Lunch Series 2006-2007

Sharon Kirkpatrick, Ph.D. Candidate
Department of Nutritional Sciences,
University of Toronto

Tuesday, December 12th, 2006, 12:00 noon - 1:00 pm

"Food access among Canadian households: Insights from national health and expenditure surveys"

Ever since food banks began to proliferate in Canada in the early 1980s, public attention has been drawn to the food shortages experienced by low-income families. The experience has come to be known as food insecurity. Responses to this problem to date have been dominated by community-based food initiatives, with little attention paid to potential policy directions. Data from national health and expenditure surveys are examined to provide evidence to inform the development of policy responses to foster food security among Canadian households. Analyses of food security indicator questions included in recent cycles of the Canadian Community Health Survey are drawn upon to elucidate the sociodemographic correlates of food insecurity, as well as associations between food insecurity and physical health, mental health and dietary intakes among adults and children. In addition, data from the 2001 Survey of Household Spending is used to examine factors associated with adequacy of household food expenditures. The findings of these analyses highlight the inextricable link between food insecurity and markers of poverty, including inadequate household incomes, reliance on welfare, lack of home ownership, and female lone-parenthood. Further, the findings indicate that the experience of food insecurity is marked by vulnerability to poor health and compromised dietary intakes. This research highlights the need for a broadening of current responses to food insecurity to address factors that constrain food purchasing among low-income households.

Amanda Sacker, Ph.D.
Department of Epidemiology and Public Health
University College London

Peggy McDonough, Ph.D.
Department of Public Health Sciences
University of Toronto

Wednesday, October 25th, 2006, 12:00 noon - 1:00 pm

"Modelling Health and Poverty Dynamics"

The seminar discusses some of the methods and results from several studies with the common theme of taking a life-course epidemiological approach to self-rated health and poverty. Health and poverty changes during the 1990s in two countries are examined using the British Household Panel Survey (BHPS) and the US Panel Study of Income Dynamics (PSID). Four case studies are presented which share these common data sets and have a common philosophy to make the best use of self-reports over time. The studies had different research quests: determining patterns of change in individuals' health over time and the ways in which such patterns are structured by continuity and change in poverty experiences; an investigation of whether different experiences of poverty have distinct consequences for long-term health dynamics; a comparison of dynamic health and labour force participation in the US and the UK; and a study of the joint interrelationships of poverty and health over time. The case studies also use different research strategies. The first uses growth curve methodology in a multilevel framework. This will not be described in detail but serves to illustrate the background to the more complex theoretical and statistical models in the other case studies. The second study estimates a general growth mixture model to assess the relationship between longitudinal courses of poverty and health. The model estimates latent poverty classes and determines their effects on initial health status and changes in health. The third case study uses a latent transition model to represent dynamic health as movement across discrete states involving stable periods, unidirectional change and intermittent deterioration and/or improvement. The final study presents some preliminary findings which extend the work from the previous case to a latent mover-stayer model of poverty and a joint latent transition model of health and poverty. The different methodologies give distinct insights into individual and population poverty and health dynamics. We summarise with a discussion of the conceptual, methodological and contextual challenges that have arisen while carrying out this work.

Tony Fang
Assistant Professor
School of Administrative Studies,
York University
and
Research Associate
Centre for Industrial Relations and Human Resources,
University of Toronto

Tuesday, October 3rd, 2006, 12:00 noon - 1:00 pm

"Employment Dynamics of Non-Standard Workers: Evidence from SLID"

Non-standard employment has attracted increased research interest not only because of its growth, but also because of its association with a variety of current policy issues. The objective of the paper is to help policy-makers to understand the characteristics associated with such work and various factors determining the crucial transitions between non-standard and permanent employment. Making use of the wealth of information on non-standard employment that is available in the six waves (1999-2004) of the Survey of Labour and Income Dynamics (SLID), a profile of the individual, job, and firm characteristics of workers in various types of non-standard employment is first provided. This is followed by an analysis of employment transitions between non-standard and permanent workers by pooling five two-year panels of the SLID (1999-2004). Prior to the empirical analysis, the data and empirical methodology are discussed. The paper ends with summary and concluding observations, linking the evidence back to the policy issues.

Back to top / Back to Past Events

The Whys and Hows of Analyzing Complex Survey Data

David Binder, Ph.D.
Georgia Roberts, Ph.D.
Statistics Canada

Friday, September 8th, 2006, 1:00 pm - 4:00 pm

The workshop will begin with a general talk about why typical survey data are different from data collected by simple random sampling and why these differences need to be accounted for when doing analyses. The advantages of a 'design-based approach' to analysis to account for survey complexities will be given. One particular artificial example will be used to illustrate the points made, along with some references to real examples. Situations will be discussed where the design-based approach for analyzing data from a complex survey may be either inappropriate or may need further investigation.

The second portion of the workshop will focus on some practicalities and will use examples from Statistics Canada surveys. Some of the questions to be covered are:

  • How are weights generated, so that weighted estimates are approximately unbiased for the quantities under study in the analysis?
  • What is design based variance estimation and what are approaches for carrying it out?
  • How are survey bootstrap weights used for variance estimation?
  • What are the software tools available for carrying out design-based analysis, with emphasis on tools that can make use of survey bootstrap weights?
  • Why might there be different survey weight variables in a data file?
  • Why and how might you consider combining data from more than one survey in a single analysis?
  • How might you deal with repeated measures from a longitudinal survey?
Discussion will be encouraged throughout the workshop.

This workshop is open to all and will be held in Room 20 in the Woodsworth College Residence, 321 Bloor St. W. (at St. George).

Back to top / Back to Past Events

Toronto RDC Brown Bag Lunch Series 2005-2006

Staff at the Toronto RDC would like to thank the presenters and attendees for making our 2005-2006 Brown Bag Lunch Series a success! The 2006-2007 Series will begin in October, 2006.

Rahim Moineddin, PhD
Department of Family and Community Medicine
University of Toronto

Flora Matheson, PhD
Centre for Research on Inner City Health
Department of Public Health Sciences
University of Toronto

Dr. Richard Glazier, MD, MPH
Centre for Research on Inner City Health
Department of Family and Community Medicine
University of Toronto

Tuesday, April 4th, 2006, 12:00 pm - 1:00 pm
Toronto RDC Conference Room

"A Simulation Study of Sample Size for Multilevel Logistic Regression Models"

This talk will focus on methodological issues concerning varying sample size at both the group and individual level on the accuracy of the parameter estimates and variance components of multilevel logistic regression models using simulation studies. We consider situations where data is sparse at level 1, the prevalence of the outcome is low, and where there are a relatively large number of level 2 groups. This situation is typical for many large national surveys. We also examine the effect of increasing the number of individuals at level 1 by aggregating similar level 2 units. The Canadian Community Health Survey Cycle 1.1 is used to further explore the findings and conclusions from the simulation studies. For the random intercept model, the estimates of the fixed effect parameters and variance components remain unbiased when the number of groups at level 2 is 100 or more, and remain so even when the model is mis-specified. For the model, including random intercept and random slope, the estimates of the fixed effect parameters are unbiased; however the estimates of the variance components have relatively large standard errors when the number of level 2 groups is 100. Estimates of the variance components are unbiased when the number of level 2 groups is at least 1000 and remain so even with model misspecification. Aggregating level 2 groups does not change the parameter estimates or the variance components when the number of groups approaches 1000. Moreover, when the number of groups is 100 and the minimum group size is either 1 or 10 the estimates of the true non-zero cross-level interaction is not statistically significant. Previous research indicates that samples of 30 to 50 groups are sufficient to produce reliable parameters for linear regression. Our findings suggest that the number of level two groups should be in excess of 100 to produce reliable parameter estimates for multilevel logistic regression.

Curtis Breslin, Ph.D.
Jason Pole, Ph.D. Candidate
Ryan Zhao, MSc
Institute for Work and Health

Tuesday, March 7, 2006, 12:00 pm - 1:00 pm
Toronto RDC Conference Room

"Using the SLID to examine the antecedents and consequences of work disability absences among young people" With work injury rates of adolescents (16 to 19 years old) and young adults (20 to 24 years old) being 1.2 to 2.0 times higher than older adults (1-4), occupational health among young people is a public health concern. Using the longitudinal components of the Survey of Labour and Income Dynamics, we sought to identify the individual, job, and temporal factors associated with work disability absences. Our longitudinal analyses show that job characteristics are the strongest predictors of work disability absence. In addition, we evaluated the subsequent earnings of young workers who had a work disability absence. We found that young workers experiencing a work disability absence had significantly less earning than their controls in the year after the absence. We will discuss the methods used for these analyses and potential implications.

Rupa Banerjee, Ph.D. Candidate
Centre for Industrial Relations

Tuesday, February 7, 2006, 12:00 pm - 1:00 pm
Toronto RDC Conference Room

"Career Progression of Immigrants in Canada: Evidence from SLID"

Previous studies have found that new immigrants to Canada start their careers with significant wage disadvantage, but that those immigrants who came to Canada in the 1970s or earlier gradually increased their earnings over time to reach or even surpass the wages of the native born. However, more recent studies have discovered that immigrants arriving in Canada since the 1980s may not fully catch up to the wages of their native-born counterparts even after ten or more years. A wage gap persists several years after immigration, although it is generally smaller than the gap at the time of entry. This smaller wage gap a few years after entry suggests that there is a period during which new immigrants’ earnings actually grow faster than the earnings of the native-born. However, after the brief catch-up period, new immigrants’ earnings slow down for reasons that are not entirely clear. These findings suggest that there are three characteristics of the immigrant wage disadvantage that need to be understood: the initial wage disadvantage (the entry effect), the rate at which immigrant wages rise during the first few years (the catch-up effect), and the longer-term wage disadvantage (the persistence effect). In this study, I examine the entry effect, the catch-up effect and the persistence effect experienced by new immigrants, using the longitudinal Survey of Labour and Income Dynamics (SLID), which allows the same cohort of immigrants (and their native-born counterparts) to be followed over time. This analysis also focuses on the effect of visible minority status on these effects.

Graham Bean, Ph.D.
Department of Psychiatry
Lake Ridge Health Hospital, Oshawa

Tuesday, January 10, 2006, 12:00 pm - 1:00 pm
Toronto RDC Conference Room

"Risk Factors for Binge Eating and Purging in 12- and 13-year old Girls"

Utilizing the National Longitudinal Survey of Children and Youth (NLSCY), this study examined risk factors for binge eating and purging in 12-and 13-year-old girls. Measures of pubertal development, body mass, dissatisfaction with appearance, general self-esteem and perceptions of parental discipline and nurturance were examined at 10 and 11 years of age as potential risk factors for binge eating and purging. A group of 10-and 11-year-old girls served as a symptom control group.

The results revealed girls who binged and purged were more likely have early puberty and high body mass at 10 and 11 years of age than girls with high anxiety and low mood. After taking into account pubertal development, logistic regression analyses revealed a high body mass at 10 and 11 years of age increased the risk for binge eating and purging at 12 and 13 years of age and this effect was potentiated by low appearance-related self-esteem. In addition, parental depression and inconsistent discipline at 10 and 11 years of age made significant and unique contributions to the prediction of binge eating and purging at 12 and 13 years of age.

The results will be discussed in term of the importance of physiological and psychological factors in the development of binge eating and purging. A putative developmental pathway to binge eating and purging will be presented and recommendations made for early prevention programs that focus specifically on the identified risk factors.

Amber Bielecky, MSc Candidate
Department of Public Health Sciences
University of Toronto

Tuesday, December 6, 2005, 12:00 pm - 1:00 pm
Toronto RDC Conference Room

"The association between socioeconomic status and the prevalence of major depressive disorder: A result of incidence, duration, and/or recurrence?" An inverse association has repeatedly been observed between socioeconomic status (SES) and the prevalence of major depressive disorder (MDD). It is unclear if the higher prevalence observed among low SES individuals is a function of a higher incidence rate, longer episode duration, and/or more frequent rate of recurrence. Both age and gender have been shown to modify the relationship between SES and MDD prevalence. The objective of this study was to determine the age-/gender specific associations between SES and each of MDD prevalence, incidence, episode duration, and rate of recurrence to better understand the nature of the relationship between SES and MDD. The data from this study came from the 2002 Canadian Community Health Survey, Mental Health and Well-being, Cycle 1.2. The findings show that SES differences in each of MDD incidence, episode duration, and recurrence may all influence the association between SES and MDD prevalence: no single element is driving the observed association.

Back to top / Back to Past Events

The Ethnic Diversity Survey (EDS): Content and Data Availability

Kelly Tran
Analyst of Social and Aboriginal Statistics Division (SASD)
Statistics Canada

Tuesday, June 21, 2005, 2:00pm - 3:00pm

This workshop will cover an overview of the EDS design, and provide information about data availability and use. The presentation will be about 45 minutes with time for questions afterwards, All those researching ethnicity in Canada are welcome.

Download the .pdf version of the EDS presentation.

Back to top / Back to Past Events

Analysis Issues for Data from Complex Surveys

David Binder, Senior Advisor to the Assistant Chief Statistician
Analysis and Development Field
Statistics Canada

Monday, September 27, 2004, 12:30 pm - 2:50 pm

This talk focuses on the general approaches in modelling and making inferences. Mr. Binder will use examples from SLID and NPHS to demonstrate the statistical justification for the design-based approach to fitting a model to the data. Design-based methods such as linearization and bootstrap will also be discussed:

Suppose that we are interested in fitting a model to data obtained from a complex survey. These data are assumed to be a 'representative' sample from a population of interest. The proposed model is also assumed to be at least approximately true for this population. There are generally three approaches to using such models with survey data and then making inferences - such as tests of hypotheses and estimation of confidence intervals. These are:

  1. Assume that the model that is valid for the population is equally valid for the sample, and proceed in the traditional manner
  2. Augment the model by incorporating relevant design information, assuming that this would compensate for any deficiencies in Approach 1
  3. Use a design-based approach, known to provide correct inferences for finite population quantities related to the parameters of the models of either of the previous two approaches
One of the key assumptions in Approaches 1) and 2) is that, after having included all the relevant explanatory variables in the model, the sample design used to collect the data is 'ignorable'. Through a simple artificial example, we show how it is possible for the model to be valid for the population, and for the sample design to be 'non-ignorable'. We show when Approach 3) can lead to valid inferences for the model being fitted, even though the approach is based on methods for finite population inference. It turns out that Approach 3) is a generalization of the Liang-Zeger robust method for analyzing longitudinal data. It is also related to the to the Huber-White variance estimator often used by economists.

We give a brief overview of the statistical justification for the design-based approach to fitting a model to the data. Design-based methods for variance estimation such as linearization and the bootstrap are described. Examples are presented using logistic regression modelling with data from the Survey of Labour and Income Dynamics (SLID) and the National Population Health Survey (NPHS). Finally, we address some common questions raised by researchers who have analyzed longitudinal or cross-sectional survey data.

This workshop is open to anyone who is interested, and will be held in the Ramsay Wright building at 25 Harbord Street in Room 117.

Back to top / Back to Past Events

Toronto RDC NLSCY Conference

NLSCY Researchers
Toronto RDC

Tuesday, June 29, 2004, 1:00pm to 4:30 pm
Toronto RDC Conference Room

The Toronto RDC is inviting RDC researchers to a one-day conference to be held on Tuesday, June 29, 2004. The focus of this first conference will be the National Longitudinal Survey of Children and Youth (NLSCY). Future conference days will focus on other core RDC datasets.

The conference will take place in the Conference Room of the Toronto RDC, and run from 1:00pm to 4:30pm. Presentations will follow a general format of approximately 15-25 minutes for the presentation, plus approximately 15 minutes for discussion.

Researchers may present the results from their NLSCY research, and/or describe and discuss issues relating to their research on the NLSCY in order to invite advice from the conference attendees.

The conference is open to current RDC researchers only. All presentations should follow normal RDC disclosure rules, or utilize pre-vetted output. Please contact Glenn Stalker if you have any questions about the results you wish to present.

Conference Presentations
If you would like to present, please contact Dave Haans for submission instructions.

In your submission, please also include the contact information (institution, email address, phone number) of those presenting, as well as any technical equipment (LCD projector, laptop computer, overhead projector, etc.) you may require.

Poster presentations may also be accepted, depending on the size of posters and available space in the Toronto RDC.

Conference Attendees
Please contact Dave Haans if you would like to attend the conference, and the hours during which you would like to attend. Due to space constraints, we may have to limit the number of participants, however we will do all we can to ensure that as many researchers as possible may attend.

Deadline
The deadline for submitting abstracts is Tuesday, June 22. Researchers will be notified regarding their submission on Thursday, June 24.

Back to top / Back to Past Events

An Overview of the Longitudinal Survey of Immigrants to Canada (LSIC)

Jessie-Lynn MacDonald, Project Manager
Owen Phillips, Senior Methodologist
Sylvain Tremblay, Content Manager
Statistics Canada

Friday, April 16, 2004, 8:30 am to 12:00 pm
Ryerson University, Room EPH 433 Eric Palin Hall, 87 Gerrard Street East

On April 16, 2004, Ryerson University hosted a presentation by Statistics Canada on the new Longitudinal Survey of Immigrants to Canada (LSIC). Presentations were made by Sylvain Tremblay, Content Manager with LSIC at the Labour and Household Surveys Branch of the Special Surveys Division and Owen Phillips, Methodologist with LSIC at the Methodology Branch of the Social Survey Methods Division. Also in attendance was Jessie-Lynn MacDonald, Project Manager with LSIC.

The presentation was specifically meant to benefit those researchers at Ryerson University who will take part in Ryerson's unique Master of Arts in Immigration and Settlement Studies graduate program commencing this fall. Among the 28 in attendance were researchers from Ryerson, York and the University of Toronto, as well as 4 Toronto RDC staff members.

Sylvain Tremblay presented a detailed account of the survey content and change in the data collection instrument between cycles. Owen Phillips addressed methodological issues with respect to sampling design and provided a detailed account on the appropriate use of sample weights and variance estimation and additionally addressed necessary considerations when assessing the quality of estimates. Glenn Stalker, Toronto RDC Analyst, presented information on accessing the RDC, to assist in the recruitment of new researchers.

Download the .pdf version of the LSIC presentation
Download the .pdf version of the LSIC coefficient of variation handout

Back to top / Back to Past Events

Overview of the NLSCY

André Cyr, Senior Methodologist
Statistics Canada

Friday, October 31, 2003, 9:30 am to 12:00 pm
Toronto RDC Conference Room

These presentations features an overview of the NLSCY, methodological issues related to survey design, the use of final survey and replicate weights, and understanding non-response in the sample.

Download the .pdf version of the NLSCY Overview presentation
Download the .pdf version of the NLSCY Weights presentation
Download the .pdf version of the NLSCY Non-response presentation

Back to top / Back to Past Events

A 2-Day Workshop on Multilevel Modeling

Blair Wheaton, Academic Director, Toronto RDC and Professor of Sociology, University of Toronto
Lisa Strohschein, Post-Doctoral Fellow, University of Toronto

Monday, April 28 and Tuesday, April 29, 2003, 1:00 pm - 4:00 pm

These sessions will provide participants with a overview of hierarchical linear models (i.e., multilevel models), growth curve models, and the use of SAS to estimate most types of models in multi-level situations.

Monday, 1:00pm - 4:00 pm: Introduction to the HLM model, Growth Curve models, and the use of SAS (as well as the HLM program).

Tuesday, 1:00pm - 4:00 pm: Generalized multi-level models (logistic and Poisson), Crossed-Random Effects Models, and Merging Census with Individual Level Data. All issues discussed in SAS.

This workshop will be held in Room 127, Rotman School of Management, 105 St. George St., Toronto.

This workshop is open to anyone who is interested, but space is limited.

Back to top / Back to Past Events

Presentation on the Survey of Labour and Income Dynamics (SLID)

Adam Wronski, Chief of Content and Analysis for SLID
Heather Lathe, Senior Analyst
Statistics Canada

Wednesday, December 11, 2002, 1:00 pm to 5:00 pm

Adam Wronski, Chief of Content and Analysis for the Survey of Labour and Income Dynamics, and Heather Lathe, Senior Analyst with Income Statistics Division, will be at the Toronto RDC the afternoon of Wednesday, December 11 to meet with current SLID users, and researchers interested in learning more about the survey.This presentation will be divided into two parts:

Part 1: (1:00 pm to 3:00 pm) Adam and Heather will provide researchers with a general overview of the survey methodology and content.

Part 2: (3:00 pm to 5:00 pm) SLID researchers working at the RDC will have an opportunity to make brief presentations about their current research followed by a discussion period. This will enable researchers to get feedback from others working with SLID on campus as well as subject matter specialists at Statistics Canada.

Back to top / Back to Past Events a name="WES">

Presentation on the Workplace and Employee Survey (WES)<

Marie Drolet, Senior Research Economist and WES Research Manager
Statistics Canada

Thursday, October 24, 2002, 1:30 pm to 3:30 pm

Marie Drolet, Senior Research Economist and WES Research Manager at Statistics Canada will be coming to the RDC October 24th from 1:30 to 3:30 pm to discuss research opportunities using the Workplace and Employee Survey (WES). In this presentation, Marie will discuss the survey methodology and content of WES, as well as data access possibilities at both the Toronto RDC and through remote data access. The RDC currently has WES employer and employee data files for 1999 and 2000.

Back to top / Back to Past Events

A Presentation on the Canadian Community Health Survey (CCHS)

Yves Beland, Household Survey Methods Division
Statistics Canada

Monday, May 27, 2002, 1:00 pm to 3:00 pm

This session will provide participants with an overview of the Canadian Community Health Survey (CCHS); describe the methodological challenges surrounding the two-year cycle plan for the survey focusing on the sample design of the health region-level survey. Sample allocation, frame, selection, oversampling of sub-populations, data collection, imputation, estimation and dissemination will be discussed. Details on how to use the bootstrap weights to compute sampling variances will also be given.

Back to top / Back to Past Events

NPHS Bootstrap Hands-on Workshop

Francois Brisebois, Senior Methodologist
Household Survey Method Division
Statistics Canada

Friday, March 22, 2002, 9:30 am to 4:30 pm

Francois Brisebois, Senior Methodologist for Statistics Canada's Household Survey Method Division, will be here on Friday March 22 to offer its users the "NPHS Bootstrap Hands-on Workshop". The primary goal of this workshop is to allow NPHS data users to understand the variance estimation process using the bootstrap method, the recommended method of the survey.

This workshop is only open to researchers who have been approved to access NPHS master data files at the Toronto RDC.

Back to top / Back to Past Events

Official Toronto RDC Opening

Thursday, November 29, 2001, 4:00 pm to 6:00 pm

The official opening of the Canada Foundation for Innovation-funded Toronto RDC was held in the newly designed, state-of-the-art Centre on the 7th Floor of Robarts Library. The opening reception was attended by over one hundred guests representing the University of Toronto, York University, Ryerson University and Statistics Canada.

U of T's Vice-President of Research and International Relations, Dr. Heather Munroe-Blum, stated "The careful balance of extending access to these invaluable research resources, while at the same time, maintaining the security of privacy and integrity required to allow Statistics Canada to operate with the full support of Canadians has been a very large task. This Centre will be a great addition to the research and learning experiences of faculty and students at our three great universities in Toronto."

Blair Wheaton, Academic Director of the Toronto RDC, put forward his vision of the Centre: "1) new and better evidence to shape public policy, using certainly the best data ever we have had for this purpose in Canada; 2) the redirection of training of Ph.D. students and of social science in Canada via the diffusion of analytical skills though the current and future pool of researchers in Canada; 3) higher rates and densities of collaboration across disciplines; and 4) the enhancement of the visibility and influence of Canadian research in the international disciplines and profession in which we take part. Now that we have been given this unprecedented resource, it is our responsibility to use it well."

Back to top / Back to Past Events