Research into the order of the items of the Revised Checklist of the Dutch version of the Portage Pr

1
Research into the order of the items of the
Revised Checklist of the Dutch version of the
Portage Program

• - Results -

University of Groningen A.T. Hoekstra (Ph.D.
student) e-mail A.T.Hoekstra_at_rug.nl B.F. van
der Meulen (Professor) G.G.H. Jansen (Associate
Professor) C.E. Oenema-Mostert (Assistant
Professor) A.J.J.M. Ruijssenaars (Professor)
2
Portage in the Netherlands

• Portage Programma Nederland (PPN) (van der
  Meulen, Sipma Feenstra, 1993)
• Based on Bluma, Shearer, Frohman and Hilliard
  (1976)

3
Previous research

• Focus more on families (Sipma, 1996 Van Gennep,
  Procee, Vermeer, van der Meulen, Janssen de
  Graaf, 1997 Meisels Shonkoff, 2000
  Oenema-Mostert 2006)
• Revise the Checklist (White East, 1983 White,
  Bungay Gabriel, 1994 Pol Ruiter, 1998
  Faber, 2002)

4
Research University of Groningen

• Revision of the Dutch version of the Portage
  Program
• - focus more on families
• - revise the Checklist

5
Revision number of items

• Adding new items (from 441 to 901)
• Limitation to first five years of life
• Remove area of Infant Stimulation

6
Revision assumption about order

• Testing procedure assumes correct order of the
  items
• but
• this has never been empirically validated

7
Pilot study

• Parents are able to complete the Checklist by
  themselves
• Division into age groups is necessary
• The age groups have to overlap

8
Main study, sample model

• Stratified sample
• - counties by region level of
  urbanization
• - children by age level
• - children by sex

9
Main study, sample model

• Recruition also via
• - Child Health Centers
• - Other research projects University of
• Groningen
• - Vaccination Organization of the
  Netherlands

10
Main study, sample model

• 2375 Checklists
• 736 completed
• 732 used in analysis

11
Main study
Mothers Chi² 382.70 p .00 Fathers Chi²
299.69 p .00
12
Main study
Chi² 0.02 p 0.87
13
Main study, IRT-model

• IRT-modeling framework
• - Latent variable determines response
  behavior
• - Probability correct answer increases
• monotonically with increasing ability

14
Main study, IRT model - OPLM

• OPLM
• - category parameters determine order
• of items
• - two category parameters because of
• three categories (0, 1, 2)

15
Main study, IRT model - OPLM

Probability
Ability
16
Results Social area

• 123 items
• After analysis 79 items
• R1c in total 604.43 (df 560), p-value .09

17
Social area, ß1 -values
18
Social area, reordered
19
Social area
20
Results - Cognitive area

• 226 items
• After analysis 158 items
• R1c in total 111.01 (df 1039), p .06

21
Cognitive area, ß1-values
22
Cognitive area, reordered
23
Cognitive area
24
Discussion

• The items of the Checklist are successfully
  reordered
• Cross-validation