Overview of the national longitudinal survey of children and youth l.jpg
This presentation is the property of its rightful owner.
Sponsored Links
1 / 47

Overview of the National Longitudinal Survey of Children and Youth PowerPoint PPT Presentation


  • 99 Views
  • Uploaded on
  • Presentation posted in: General

Overview of the National Longitudinal Survey of Children and Youth. N L S C Y . Methodologist. Why a methodologist and not an analyst today?. Not subject-matter expert. We do the sampling thing.

Download Presentation

Overview of the National Longitudinal Survey of Children and Youth

An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

Presentation Transcript


Overview of the national longitudinal survey of children and youth l.jpg

Overview of the National Longitudinal Survey of Childrenand Youth

N L S C Y


Methodologist l.jpg

Methodologist

Why a methodologist and not an analyst today?

  • Not subject-matter expert

  • We do the sampling thing

  • In view of all the analytic requirements, such as those discussed by SSD and HRDC, what type of sample will best meet all those goals?

  • Determine the sample size

  • How the sample is selected

  • Analysis done by the methodologist

    • is to assess data quality

    • adjust estimation methods accordingly

Themethodologist helps determinesthe survey vehicle.


Goal of the presentation l.jpg

Goal of the Presentation

  • The survey vehicle has an impact on the analysis.

  • Today we’re going to describe the vehicle in order to facilitate the analytical process.


The analytical issues l.jpg

The Analytical Issues

Being able to . . .

. . .make inferences

  • The data

    • strategy / concepts

    • issues:

      • partial non-response

      • inconsistencies

  • Sampling

    • Complex design

    • Impact on

      • estimation

      • precision

      • analysis

  • Type of analysis

    • longitudinal, cross-sectional

    • other and mixed


Slide5 l.jpg

NLSCY - National Longitudinal Survey of Children and Youth

N . L . S . C . Y

N L S C Y

NLSCY


Nlscy overview l.jpg

NLSCY - Overview

The survey is almost entirely funded by HRDC

NLSCY - National Longitudinal Survey of Children and Youth


Nlscy overview7 l.jpg

NLSCY - Overview

  • Complex data structure

    • the lives of children are complex

    • dual child/household structure

    • new content in each cycle

    • some changes in old content

  • Other constraints

    • limit on quantity of information

    • limited resources


Analytical framework l.jpg

Analytical Framework

Social

Temporal

Transitions:

Health/injuries

Accidents

Mortality in the family

Periods of poverty

Starting school

Adolescence

Graduation

First job

Marriage

First child

Outcomes

  • Life-long learning

  • Effective worker

  • Effective parent

  • Involved citizen

Physical

health

Context

Resources Family School Community Work Public programs

Emotional

Social

Language/

communication

Cognitive/

learning

T i m e i n y e a r s


Unit of analysis l.jpg

Unit of Analysis

  • The child

  • Sources of information

    • Person most knowledgeable about the child (PMK)

    • Teacher

    • School principal

    • Child himself/herself

      • cognitive measures

      • self-administered


Unit of analysis10 l.jpg

Unit of Analysis

X

  • Caution

  • Other types of Analysis

    • Weights are designed for the child

    • Concepts like family are characteristics of the child

      • Not a domain for estimation

C

Statements like . . .

The NLSCY estimates the number of children whose families with characteristic …

The NLSCY estimates number of families with characteristic …


Slide11 l.jpg

NLSCY Content

  • Child (depending on age)

    • socio-demographic

    • health

    • perinatal information

    • development (motor, social

    • and physical)

    • temperament

    • academic performance

    • education

    • literacy

    • extracurricular activities

    • work experience

    • socialization

    • relationship with parents

    • family history and

    • legal custody of children

    • child care

    • behaviour

    • self-esteem

    • cigarettes, alcohol, drugs

    • vocabulary assessment

    • math test

    • reading comprehension test

    • sexual activity and loving

    • relationship

  • Parents

    • socio-demographic

    • education/literacy

    • labour market

    • income

    • health

    • social support

    • parental involvement at school

    • parents’ aspirations for child’s

    • education

School

number of students

discipline problems

school atmosphere

resources

characteristics

Teachers

teaching practices

demography

qualifications

  • Family

    • demography of members

    • relationships between members

    • of household

    • family functioning

    • household

    • neighbourhood

  • Principal

    • demography

    • qualifications

Note: Minor changes are made in the content from

one cycle to the next.


Classic trade off of quantity vs quality l.jpg

Classic Trade-off of Quantity vs Quality

In the NLSCY, you will find

data are less processed.

In surveys, it is normal

to clean up the data


Investment for the nlscy l.jpg

Investment for the NLSCY

  • Focus on derived variables

    • scales, cognitive measures

    • transition measures

  • Non-response adjustment

    • total non-response

  • Processing of financial data

    • family income

    • personal income

  • Dissemination within reasonable time


Data and analysis l.jpg

Data and Analysis

  • Changes in some longitudinal variables

    • improving the concept

    • different respondent

      • means a conceptual difference

    • different response

      • different PMK

      • response error

  • Unprocessed responses

    • verification and consistency

    • partial responses

Analytical

burden


Partial non response l.jpg

Partial Non-response

  • Respondent units are those which answered the key questions.

    • Not necessarily all the questions.

    • Some variables will include non-response, identified by:

      • not stated

      • don’t know

      • refusal

    • Sometimes an entire block is missing.


What should you do about partial non response l.jpg

What Should You Do About Partial Non-response?

  • Assess the extent of the partial non-response.

  • Determine the impact on your analysis

    • By examining the variables related to the variable of interest

    • See if the missing responses can form a separate category

    • Decide to do non-response processing

      • reweight for each variable to take partial non-response into account. Can be very tedious.

      • Impute


Nlscy data collection strategy l.jpg

NLSCY Data Collection Strategy

0 11

1994-95

Cycle 1

0 1

2 13

Early Years

1996-97

Cycle 2

0 1

2 3

4 15

1998-99

Cycle 3

2 3

4 5

6 Released summer 2003 17

0 1

2000-01

Cycle 4

2 3

4 5

8 19

0 1

E.Y.

Cycle 5


Sample counts l.jpg

Sample counts


Issues l.jpg

Issues

  • CROSS-SECTIONAL ANALYSIS


Issues20 l.jpg

Issues

  • CROSS-SECTIONAL ANALYSIS

    • Limitations due to the age of the sample

      • part of the sample was not selected for cross-sectional estimates

      • inherent complexity in the sample design to meet divergent needs

      • coverage problems

        • no update of the sample to reflect changes in the population (e.g., immigration); only the sampling weights have been adjusted to reflect changes

        • the older the cohort gets, the more difficult it is to adjust the sampling weights properly


Issues21 l.jpg

Issues

  • CROSS-SECTIONAL ANALYSIS

    • Limitations due to the nature of the survey

      • Some concepts were defined for the purposes of longitudinal analysis

      • Problems with sample erosion

      • Conditioning bias

    • Interpretation of the results

      • Impact on the effectiveness of estimation methods

      • Making inferences

      • Greater potential with the supplementary samples that have been added


Issues22 l.jpg

Issues

  • LONGITUDINAL ANALYSIS


One survey but actually many datasets l.jpg

One Survey but actually many datasets

A longitudinal file

0 11

1994-95

Cycle 1

2 13

1996-97

Cycle 2

4 15

1998-99

Cycle 3

6 17

2000-01

Intended for cohort analysis of 2 ages,

eg, 0-1, 2-3, 4-5, 6-7, 8-9, 10-11

Cycle 4

6-7

0-1

2-3

4-5

8-9

10-11


Sample counts24 l.jpg

Sample counts


Issues25 l.jpg

Issues

  • LONGITUDINAL ANALYSIS

    • Limitations due to sample erosion

      • sample shrinkage problems

      • representation (coverage) problems

      • Swiss cheese problems

    • Conditioning bias

    • Interpretation of results

      • impact on effectiveness of estimation

      • inferences


Issues26 l.jpg

Issues

  • MIXED ANALYSIS


Issues27 l.jpg

Issues

  • MIXED ANALYSIS (longitudinal and cross-sectional)

    • Pay attention to the differences in the population targeted by the two types of analysis

    • Sample sizes vary a lot for these two types of analysis

    • Pay attention to the conclusions drawn from these analyses

    • The problems mentioned earlier can take different forms depending on the type of analysis


Dissecting nlscy data l.jpg

Dissecting NLSCY Data

Cross-sectional Data

Repeated Surveys

0 11

In 1994-95

Cycle 1

0 1

2 13

Early Years

1996-97

Cycle 2

0 1

2 3

4 15

1998-99

Cycle 3

The sample size is very different from one cycle to the next, from one cohort to the next.

3 data cycles

for children aged 0 to 11

2 data cycles

for children aged 12 and 13


Dissecting nlscy data29 l.jpg

Dissecting NLSCY Data

Cross-sectional Data

Repeated Surveys

0 11

In 1994-95

Cycle 1

0 1

2 13

Early Years

1996-97

Cycle 2

0 1

2 3

4 15

1998-99

Cycle 3

NOTE: The sample units are not independent of one another.

Whereas these units are independent


Issues30 l.jpg

Issues

  • CROSS-SECTIONAL ANALYSIS (REPEATED)

    • Same limitations as noted earlier

    • The sample overlaps from one cycle to the next.

    • Independence or interdependence of samples

      • There is sample interdependence when the sample is made up of the same respondents

      • Involves a covariance factor

      • Sample independence is possible only for certain domains (e.g., children aged 0-1)


One survey but actually many datasets31 l.jpg

One Survey but actually many datasets

An early childhood file

0 11

0 5

1994-95

Cycle 1

0 1

2 13

2 5

Early Years

1996-97

Cycle 2

0 1

2 3

4 15

4 5

1998-99

Cycle 3

2 3

4 5

6 17

0 1

2000-01

Cycle 4


Dissecting nlscy data32 l.jpg

Dissecting NLSCY Data

Early Years

Cross-sectional Data

1,793

0 5

12,333

10,465

1,453

0 1

2 5

12,333 children aged 0 to 5 in Cycle 1 in a cross-sectional study.

The cross-sectional cohort can also be used in Early Years research.

20,210

5,420

0 1

2 3

4 5

10,465 C2 respondent children

The sample size is very different from one cycle to the next, from one cohort to the next.

Or, once again, repeated analysis of the three data cycles

20,210 C3 respondent children

6,390 1-year-olds

5,420 five-year-olds


Analysis of overlapping domains l.jpg

Analysis of Overlapping Domains

Analysis of five-year-olds.

1,793

5

C1

Born in 1989

  • There are actually 8,666 five-year-olds in the 3 cycles.

  • The reference period becomes an analytical variable.

  • Inference for a prescribed population is in context.

1,453

5

C2

Born in 1991

5,420

5

C3

Born in 1993

8,666

For example, analysis of children at a particular age


Other issues l.jpg

Other issues

  • Age of the child:

    • Definition of effective age introduced in cycle 3.

    • Effective age = Ref. Year – year of Birth

    • Example, cycle 4: 2002 – Y. of B. = Eff. Age

  • Codebooks vs Master files:

    • Counts in the codebooks are the cross-sectional counts, but the Master files have all the records


Where does the nlscy sample come from l.jpg

Where Does the NLSCY Sample Come From?

  • Why ask the question?

    • Issues concerning analysis

      • impact on statistical/analytical software applications

      • impact on subject areas being studied

STATA

SAS

SUDAAN

SPSS


The nlscy sample l.jpg

The NLSCY Sample

  • A large part of the sample comes from the LFS

    • geographic stratification

    • multi-stage, with the primary sampling unit being a geographic cluster

    • with a systematic sample of households

    • After the LFS interview, households are identified as containing or not containing the population of interest (children)

    • The in-sample unit is the child (not the household)

  • Constitutes the initial frame of children selected in 1994

  • Main source of newborns and cross-sectional samples


Lfs structure l.jpg

LFS Structure

WARNING!

Applications such as SAS and SPSS

can use sampling weights for estimation

purposes, but they do not use them

correctly for

variance estimation.

Some applications can use sampling

weights for estimation purposes and can

estimate

the variance correctly. But

they cannot take into account

the NLSCY’s special

problems.


Where does the nlscy sample come from38 l.jpg

Where Does the NLSCY Sample Come From?

  • Other reasons for asking the question

    • Issues concerning the type of analysis that can be done

      • cross-sectional, longitudinal, other

    • Issues concerning interpretation of the results

      • impact on the effectiveness of estimation methods

      • inferences


Implications for analysis l.jpg

The structure favours analysis for geographic areas

Loss of effectiveness for other subject areas

Implications for Analysis

  • The advantage is reflected in a gain in operational efficiency

  • Can use a larger sample for the same cost

  • Target the analysis to take advantage of this structure

  • Some estimation methods can improve efficiency


The survey vehicle l.jpg

The Survey Vehicle

NLSCY

We have seen that the survey is loaded with information


The survey vehicle41 l.jpg

The Survey Vehicle

NLSCY

We can lighten the load by targeting our analysis


The survey vehicle42 l.jpg

The Survey Vehicle

NLSCY

We can see what’s possible and what’s not


The survey vehicle43 l.jpg

The Survey Vehicle

NLSCY

We can greatly improve the survey’s effectiveness

by taking advantage of the way it’s constructed


The survey vehicle44 l.jpg

The Survey Vehicle

NLSCY

When we adjust our analytical approach, the vehicle

becomes lighter and more manoeuvrable


The survey vehicle45 l.jpg

The Survey Vehicle

NLSCY

And we know we’re not the only ones doing analysis


The survey vehicle46 l.jpg

The Survey Vehicle

NLSCY

Questions


My cordinates l.jpg

My cordinates

Charles Tardif

Room 2500, Main Building, Statistics Canada

Tunney’s Pasture, Ottawa, Ontario

K1A 0T6

[email protected]

Tel: (613) 951-4353

NLSCY - National Longitudinal Survey of Children and Youth


  • Login