Measurement of Growth
Damian W. Betebenner
Center for Assessment
NCME 2010
Denver, Colorado
Discussions of growth lie at the intersection of
three topics
Longitudinal Data Analysis/Applied Statistics
Accountability/Education Policy/Data Use
Measurement/Psychometrics
Four Related yet Diverse Papers
• Measurement:
– Monitoring Growth with Vertical Scaling or Vertically Moderated Standards
Setting (McJunkin, Poggio, Glasnapp)
• Longitudinal Data Analysis & Data Use/Accountability
– Measuring Academic Growth with Growth Norms (Betts)
– Growth on an Aggregate Level (Furgol)
• Measurement & Longitudinal Data Analysis
– A Unified Model for the Analysis of Change (Hsieh, von Eye, Maier)
Vertical Scales or Vertically Moderated
Standards Setting
• Construction of a developmental scale:
– The Holy Grail of scale construction? See Briggs and Rekase.
• Scale requirements for student growth?
– Do we need VSS/VMSS to discuss student growth?
• Questions:
– VSS/VMSS are designed to answer very different questions about student growth
and the two are mutually exclusive. (Later in the same paragraph: “If it is not
possible for a state to implement vertical scaling measures, VMSS may be used
to track student growth across years”)
– If a state has a VSS, can’t they add VMSS on top?
– Lissitz and Huynh argued that if a state is considering implements a system that
focuses on progression within achievement levels and desires a system that will
provide meaningful information on classroom instruction and teacher adaptation,
then vertical scales would not be appropriate.
– “The most useful method for measuring student growth across all grades would
be vertically equated scores, scaled to yield equal interval scores” (Clemens
Growth Norms and RTI
• Gain scores as the basis of growth norms
– Scale considerations (VSS)
– Measurement error
• Without a VSS, how might these results be
extended?
• Growth above/below expectation and RTI.
How might these results be extended to
account for growth-to-standard.
Growth for Aggregates
• Describing student growth and “group
growth” (e.g., “school growth”)
A Unified Model for the Analysis of Change
• Synthesize two of the three Longitudinal Data Analysis/Applied Statistics
circles into one: MGRM-
ALGC
• Time invariant latent
construct(s)?
• A priori number of latent Accountability/Education Policy/Dat
Measurement/Psychometrics
constructs or determined as
part of analysis?
• Applicable to content
measures like reading &
math?
• Computation time?
Accountability/Public Policy/Data Use/Validity
Consequential validity is the ultimate
criterion by which we should judge an
accountability system.
Braun (2008)
Bibliography
❑ Braun, H. (2008). Viccissitudes of the Validators.
Presented at the Reidy Interactive Lecture Series,
Portsmouth, New Hampshire.
About this copy
Measurement of Growth — NCME Annual Meeting, May 2010.
This is an image-based copy of the original PowerPoint slides; builds and animations are flattened.