Children's Interactions Across Preschool Contexts: An Observational Assessment

1
Childrens Interactions Across Preschool
Contexts An Observational Assessment Olivia
Kathryn Lima, Jason Downer, Emily Gerber, Anne
Henry, Tracy Nishida, Bonnie Yoder Robert Pianta
BACKGROUND
RQ 1 Can classroom observations capture
variability in childrens interactions throughout
two mornings?
RQ 3 Are observations associated with teacher
ratings of the same constructs?
RQ 4 Are childrens behaviors different across
varying activity settings?

• Readiness to function competently in early
  schooling is perhaps best understood in terms of
  the nature and quality of childrens interactions
  in classrooms with adults, peers, and learning
  activities.
• Standardized observations are recognized as
  best practice for assessing these
  context-specific, relational processes, yet few
  psychometrically sound observational measurement
  tools exist.
• To address this gap, an observational measure,
  the Classroom Assessment Scoring System Child
  Version (CLASS-C), was developed to reflect
  patterns of adaptation within 4 developmental
  tasks competent interaction with adults, peers,
  and activities, and self-state regulation.

RESEARCH QUESTIONS

• RQ 1 To what extent can classroom
  observations capture variability in childrens
  interactions throughout two mornings?
• RQ 2 Can childrens interactions with adults,
  peers, and activities, as well as self-state
  regulation, be reliably observed across raters?
• RQ 3 Are observations of these classroom
  interactions associated with teacher ratings of
  the same constructs?
• RQ4 Are childrens behaviors different across
  activity settings?

CONCLUSIONS

• Observations of childrens behavior largely fit a
  normal distribution, except
• Conflict scales were skewed due to low incidence
  of these behaviors. and,
• Self-state regulation scales were skewed,
  suggesting the need for improved
  operationalization of these internal processes in
  observable, behavioral terms.
• Strong inter-rater agreement suggests that it is
  possible to reliably observe these behaviors
  across coders, after intense coder training.
• Observations were associated with teacher ratings
  of similar behaviors, though the moderate
  magnitude of correlations suggests that these
  methods capture related but distinct constructs.
• Childrens behavior was relatively stable across
  activity settings. Peer interaction scores were
  lower when teachers were present, for both whole
  group and free choice activities.

METHODS

• Participants
• 39 lead teachers in 44 preschool classrooms
• 164 children observed
• Procedures
• Data were collected in Fall 2006.
• Each classroom was visited over 2 mornings,
  typically a week apart.
• Up to 4 children per classroom were observed in
  alternating cycles across the morning.
• Each observation cycle covered 15 minutes (10
  minutes observing, 5 minutes coding).
• Measures
• Observation
• Classroom Assessment Scoring System Child
  Version (CLASS-C)
• See Table 1 for specific domains and dimensions.
• Teacher ratings

FUTURE DIRECTIONS
RQ 2 Can childrens classroom behaviors be
reliably observed across raters?

• Growth and Context-Specific Test-Retest
  Reliability
• Change in childrens competence from the fall to
  the spring will be evaluated.
• Reliability across time and within activity
  settings will be investigated with test-retest
  analysis.
• Scale Refinement
• The Self-State Regulation scales will be revised
  to increase range and better capture the
  variability within and between children.
• An exploratory factor analysis will be conducted
  to determine whether the proposed four-factor
  conceptual structure fits the data.

• Inter-rater reliability was first conducted
  after observer training using 6 master-coded
  videotapes
• Mean 84.67 (within 1 of the master code)
• Live inter-rater reliability was then calculated
  across 20 of all observations (n18 double-coded
  visits)
• Mean 86.54 (within 1 of double coder)
• Range across dimensions 70 - 99

ACKNOWLEDGEMENTS
This study is part of the Interagency Consortium
for School Readiness Outcome Measures (ICSROM)
and co-funded by the U.S. Departments of Health
and Human Services and Education. Many thanks
also to the teachers and children who invited us
into their preschool classrooms this past fall.
Poster presented at the AERA Annual Meeting and
Exhibition April 10, 2007 (Chicago) For more
information, please contact Jason Downer at
jdowner_at_virginia.edu