1. Jane B. Sprott, Department of Sociology and Anthropology, University of Guelph. "Work (soon to be) In Progress: Using the National Longitudinal Survey of Children and Youth to Understand the Development of School Bonds."
   
   
2. Tom Perks, Ph.D. Program, Department of Sociology, University of Waterloo. "Levels of Involvement in Extracurricular Activities as a Predictor of Positive Outcomes in Canadian Children."
   
   
3. Leanna Leader (Master's program in the Department of Child and Youth Studies), Tom O'Neill (Child and Youth Studies) and Mohammed Dore (Child and Youth Studies and Economics), all at Brock University. "Placing the Dimensions of Multiple Deprivation of Canadian Children."
   
   
4. John Goyder, M.E. Thompson and Shane Dixon (UW Departments of Sociology, Statistics and Actuarial Science and Sociology, respectively). "Scaling the National Occupational Classification (NOC) Major Groups: Using the SLID for Validation."

1. Vinay Kanetkar, Department of Consumer Studies, UG, "Factor Structure of Measurement Scales: Higher Order, Hierarchal or Explanatory Analysis."
   
   
2. Norberto Pantoja, Ph.D. Program, Department of Statistics and Actuarial Science, UW. "Pregnancy and Smoking Cessation."
   
   
   
   -------------------------------------Coffee Break------------------------------------
   
   
   
3. Timothy Gawley, Ph.D. Program, Department of Sociology, UW. "A Repeated Cross-Sectional Analysis of Training and Labour Market Outcomes in Canada During the 1990s: Results from the Adult Education and Training Surveys."
   
   
4. Christian Boudreau, Ph.D. Program, Department of Statistics and Actuarial Science, UW. "Analysis of Jobless Spells in the Survey of Labour and Income Dynamics."
   
   
5. Jiri Zuzanek and Roger Mannell, Department of Recreation and Leisure Studies, UW. "Emotional well-being and mental health as a function of life course transitions: a longitudinal perspective."
   
   

ABSTRACTS

Session 1, Presentation Number 1
Jane B. Sprott
jsprott@uoguelph.ca
Department of Sociology and Anthropology, University of Guelph

Title: Work (soon to be) In Progress: Using the National Longitudinal Survey of Children and Youth to Understand the Development of School Bonds

Abstract:

Research has consistently identified a weak school bond as an important risk factor for later delinquency. That is, children who are not committed to, or interested in school are generally at higher risk for involvement in delinquency than children with stronger bonds to school. Previous work using the National Longitudinal Survey of Children and Youth (Sprott, Jenkins and Doob 2002) found that a strong school bond was also an important "protective" factor. That is, a strong school bond at age 10 and 11 was found to decrease levels of violence two years later (age 12 and 13) among high risk children (those who displayed high levels of early aggression) but had little or no effect on lower risk children. Building on this previous work, the current study uses two cycles of data from the National Longitudinal Survey of Children and Youth and investigates factors that are important for developing a strong school bond between children displaying high levels of early aggression and those who are not.

Session 1, Presentation Number 2
Tom Perks
taperks@artsmail.uwaterloo.ca
Department of Sociology, University of Waterloo

Title: Levels of Involvement in Extracurricular Activities as a Predictor of Positive Development in Canadian Children

Abstract:

Using data from the National Longitudinal Survey of Children and Youth, this study will investigate, in the Canadian context, the relationship between children''s involvement in extracurricular activity and three indicators of positive development: namely, academic achievement, prosocial behaviour, and psychological well-being. It is hypothesized that a positive relationship will exist between level of extracurricular involvement and each of the outcome measures, while controlling for social background factors. Interpretations of the results will be offered, including the interpretation that children's involvement in extracurricular activity is one aspect of the "accumulation of social capital".

Session 1, Presentation Number 3
Leanna Leader, Tom O'Neill, Mohammed Dore
leanna_l@hotmail.com
Department of Child and Youth Studies, Brock University

Title: Placing the Dimensions of Multiple Deprivation of Canadian Children

Abstract

The initiative for the current proposal is based on the notion that child poverty is a relative phenomenon that is best assessed by understanding the dimensions of multiple deprivation that face the poor. One research objective that we plan to explore is the evaluation of distributions of child health deprivation in Canada, specifically the identification of significant variations between provincial, and between urban and rural areas. We have a particular interest in several suppressed variables included in the NLSCY (National Longitudinal Survey of Children and Youth) data set that will allow us to precisely place these group variations. The second research objective is the assessment of how the local and social policy environment may account for some of this variation. Using the SWORDC database, we plan to conduct a univariate analysis of multiple deprivation, concentrating on those variables which can be used to clearly define patterns of chronic respiratory infection, exposure to asthma trigger factors, the mental health of each child, and the family and community health context. We hope to show that the local social and policy environment does play a role in either aggravating or ameliorating health deprivation for impoverished children.

Session 1, Presentation Number 4
John Goyder, M.E. Thompson, Shane Dixon, Departments of Sociology, Statistics and Actuarial Science, Sociology, respectively, University of Waterloo
jgoyder@watarts.uwaterloo.ca

Title: Scaling the National Occupational Classification (NOC) Major Groups: Using the SLID for Validation.

Abstract:

From the Kitchener-Waterloo Metropolitan Area Survey (KWMAS) collected in 2000, we have prestige ratings for the major groups of the National Occupation Classification (NOC). Heretofore, these groups have been un-scaled, simply existing as discrete or un-ranked categories at the nominal level of measurement. A scaling is useful for the many occasions when the analyst needs to conceive of occupation in terms of socio-economic status (SES). At present, the only option is to treat the 26 major groups as a set of 25 dummy variables plus a reference category. The need for a more concise and conceptually informed scaling is all the more acute given that public release versions of many Statistics Canada data sets include only the first two digits of the NOC (i.e., the groups). The RDC work proposed concerns the validation of the KWMAS scale. Validation is understood in two senses here: (i) Using Panel II of the Survey of Labour and Income Dynamics (SLID) to examine correlations among SES indicators such as education, income and occupation for the Kitchener-Waterloo metropolitan area on the one hand and the rest of Canada on the other. The "rest of Canada" shall be broken down into components classified by city size and region; (ii) Comparing the same correlations across successive waves of the SLID (in which cumulative response rate declines) and the KWMAS2000, in order to gauge the consequences of the rather low KWMAS response rate of 45%.

Session 2, Presentation Number 1
Vinay Kanetkar
vkanetka@uoguelph.ca
Department of Consumer Studies, University of Guelph

Title: Factor Structure of Measurement Scales: Higher Order, Hierarchical or Exploratory?

Abstract:

Over the last 100 years, use of exploratory factor analysis in scale development is documented. Researchers, however, find that the exploratory factor model does not lead to well fitting confirmatory factor model. In this paper, several plausible confirmatory models are tested to improve goodness of fit as well as estimated loading interpretation. Two alternative scales (price sensitivity and sense of coherence) using two different domains (Marketing and Psychology) are used to demonstrate alternative model comparisons. The paper presentation will focus on three different datasets using same measurement items. Present investigation is proposed using the National Population Health Surveys of 1994 and 1997 using 13 items involving three sub-domains of sense of coherence scale.

Session 2, Presentation Number 2
Norberto Pantoja
npantoja@math.uwaterloo.ca
Department of Statistics and Actuarial Science, University of Waterloo

Title:Pregnancy and Smoking Cessation

Abstract:

Pregnancy may be viewed as an ideal time for women to make several health behavior changes. One of the most notable examples is smoking cessation. However, previous studies indicate that often this health attitude adopted by some pregnant women is not continued in the postpartum period. Therefore, once pregnancy has been successfully completed, it is not surprising that resumption of smoking occurs. In this context, the longitudinal data from the National Population Health Survey could be used in order to have an analysis of how the duration of smoking cessation depends on whether the motivation was pregnancy.

Session 2, Presentation Number 3
Timothy Gawley
tgawley@watarts.uwaterloo.ca
Department of Sociology, University of Waterloo

Title: A Repeated Cross-Sectional Analysis of Training and Labour Market Outcomes in Canada During the 1990s: Results from the Adult Education and Training Surveys

Abstract:

Transformations in the Canadian labour market during the 1990s such as the on-going emergence of post-industrial service or knowledge-based occupations, the growth of nonstandard employment and increasing levels of educational attainment have prompted critical analyses about the relationship between training and labour market outcomes. Post-industrial and human capital perspectives on training and labour market security propose that the accruement of skills by Canadians over time will lead them to the "good jobs" that entail such characteristics as secure income attainment, employment security and adequate education-job match. But despite the participation of Canadians in training activities, the structural segmentation of the Canadian labour market suggests that training has inequitable returns; these disparities can be seen in the form of income polarization, increasing unemployment and underemployment. Using secondary data from a series of surveys called the Adult Education and Training Survey, this presentation discusses preliminary findings regarding the influence of training participation on the labour market outcomes of Canadians during the 1990s.

Session 2, Presentation Number 4:
Christian Boudreau
cboudreau@math.uwaterloo.ca
Department of Statistics and Actuarial Science, University of Waterloo

Title: Analysis of Jobless Spells in the Survey of Labour and Income Dynamics.

Abstract:

Currently, there is much interest in the use of longitudinal, panel, or cohort surveys to understand the different events that individuals experience over time. Examples of such events include marriage, divorce, fertility and spells of unemployment. Data collected through longitudinal surveys involve the use of complex survey designs, with clustering and stratification. Statistical analysis of such data must account for intra-cluster dependence and possibly for response-selective sampling. In this talk, we discuss and illustrate some of the unique features of longitudinal surveys. These include analytical objectives, attrition, top-up samples, seam effect, left-truncation, right-censoring, missing data and measurement errors. Some of these features will be exemplified using the Survey of Labour and Income Dynamics (SLID) from Statistics Canada. We also present one of the many possible analyses of the SLID dataset. In that analysis, we look at the duration of jobless spells and some of the factors influencing it.