Flag of Canada
Government of Canada Government of Canada
 
Français Contact Us Help Search Canada Site
About Us Services Where You Live Policies & Programs A-Z Index Home
    Home >  Programs and Services > Policies, Planning and Reporting
Services for you

Bullying and Victimization Among Canadian School Children - October 1998

  What's New Our Ministers
Media Room Forms
E-Services
Publications Frequently Asked Questions Accessibility Features

  Services for: Individuals Business Organizations Services Where You Live
 

2. Method

PreviousContentsNext

2.1 Participants

Respondents were parents of children aged four to 11 who were attending school and participating in the Cycle 1 of the National Longitudinal Study of Children and Youth (NLSCY). Table 1 presents the number of boys and girls in each of the three age groups, their mean age, and the standard deviation of their age.

Table 1: Ages of Each Cohort
  Age 4 to 6 Age 7 to 9 Age 10 to 11
N M SD N M SD N M SD
Boys 2418 4.99 0.82 2221 7.99 0.82 1023 10.49 0.5
Girls 2309 4.95 0.82 2236 7.98 0.81 1101 10.51 0.5

2.2 Measurement Model

Latent variables are those representing theoretical constructs (i.e., abstract). Parent rating scales of family demographics, family functioning, and child behaviours were used to form the six latent composite variable: (1) family demographics (parents' education, household income, and age of parents); (2) family functioning (Positive interaction, hostile interaction, consistency, and punitive practises); (3) externalizing behaviour problems (physical aggression, indirect aggression, property offenses, hyperactivity, and prosocial behaviour); (4) internalizing behaviour problems (emotional problems), (5) victimization (parent ratings of victimization), and (6) bullying (parent ratings of bullying). The measures are briefly described in Table 2.

2.3 Models and Statistical Analysis

The bullying and victimization model was estimated with LISREL 8 (Jöreskog & Sörbom, 1996), using weighted least squares (WLS) estimation. The bullying and victimization model is an application of the full LISREL model that comprises two major parts: (a) the measurement submodel which specifies the relationship of directly observed to latent variables (the hypothetical underlying constructs, such as family demographics, family functioning, internalizing and externalizing problems, bullying and victimization); and (b) the structural submodel which specifies the relationship among the latent variables (e.g., the relationship of bullying to victimization). See the figure on page 9 for the model of bullying and victimization that was tested.

Table 2: Summary of Measures
Measure NLSCY Measure Description
Family demographics 5 items were included in the construct: parent education, household income, and the age of each of the parents. Parent's education Household income (In dollars), Age of parents (in years)
Family functioning Parent Questionnaires: Positive interaction (5 items), hostile interaction (7 items), consistency (5 items), and punitive practises (5 items). Interval measure of certain aspects of parenting behaviours. Items are responded to on 5 point-Likert types scale. For the positive interaction and consistency scales higher scores indicate more positive behaviours. For the remaining scales higher scores indicate more negative interactions.
Externalizing behaviour problems Parent Questionnaires: physical aggression (6 items), indirect aggression (5 items), property offenses (6 items), hyperactivity (8 items), and prosocial behaviour (10 items). Interval measure of children's behavioural functioning. Items responded to on a 3-point Likert type scale. With the exception of the prosocial scale, higher scores indicating more behavioural problems. For the prosocial scale a higher score indicates more social skills.
Internalizing behaviour problems Parent Questionnaire: emotional problems (7 items) Interval measure of children's anxiety, depression, and happiness. Items are responded to on a 3 point Likert type scale, with higher scores indicating more internalizing problems.
Victimization Parent ratings of victimization Single items asking if a child is bullied. Items are responded to on a 3 point Likert type scale, with higher scores indicating more victimization.

Child ratings of victimization Single item asking if a child is bullied by others.
Bullying Parent ratings of bullying Single items asking if a child bullies others. Item is responded to on a 3 point Likert type scale, with higher scores indicating more bullying.

Child ratings of bullying One item asking if a child is cruel, bullies, or is mean to others.
Model of Bullying and Victimization
PreviousContentsNext
     
   
Last modified : 2005-01-11 top Important Notices