10.07.2015 Views

ICCS 2009 Technical Report - IEA

ICCS 2009 Technical Report - IEA

ICCS 2009 Technical Report - IEA

SHOW MORE
SHOW LESS
  • No tags were found...

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Chapter 11:Scaling procedures for <strong>ICCS</strong> test itemsWolfram Schulz and Julian FraillonIntroductionThis chapter describes the procedures used to analyze and scale the <strong>ICCS</strong> international andregional test items that were administered to measure students’ civic knowledge. The chaptercovers these topics:• The scaling model used to analyze and scale the test items;• Test coverage and item dimensionality;• Assessment of item fit;• Assessment of scorer reliabilities for open-ended items;• Differential item functioning by gender;• Review of cross-national measurement equivalence;• International item adjudication;• International item calibration and test reliability;• International ability estimates (plausible values and weighted likelihood estimates);• Estimation of changes in civic content knowledge between 1999 and <strong>2009</strong>; and• Regional test items for the European and Latin American modules.The development of the <strong>ICCS</strong> test items was described in Chapter 2 and was guided by the<strong>ICCS</strong> assessment framework (see Schulz, Fraillon, Ainley, Losito, & Kerr, 2008).The scaling modelItem response theory (IRT) scaling methodology was used to scale the test items.Use of the one-parameter (Rasch) model (Rasch, 1960) for dichotomous items means that theprobability of selecting Category 1 instead of 0 is modeled asP i (q) = exp(q n– d i )1+exp(q n – d i ) ,where P i (q) is the probability for person n to score 1 on item i, q n is the estimated ability ofperson n, and d i is the estimated location of item i on this dimension. For each item, itemresponses are modeled as a function of the latent trait q n .In the case of items with more than two (k) categories (as, for example, with Likert-type items),this model can be generalized to the partial credit model (Masters & Wright, 1997), whichtakes the form ofexpS(q n – d i + t ij )k=0P xi (q) = m i kx i =0,1,…,m i .S expS(q n – d i + t ij )h=0xk=0Here, Px i (q) denotes the probability of person n scoring x on item i, and q n denotes theperson’s ability. The item parameter d i gives the location of the item on the latent continuum; t ijdenotes an additional step parameter.ACER Conquest, Version 2.0 software (Wu, Adams, Wilson, & Haldane, 2007) was used toscale the <strong>ICCS</strong> test data.129

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!