{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/79972"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/79972","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Implementation of Optimal Design for Item Calibration in Computerized Adaptive Testing (Cat)","abstract":"Item calibration in small sample conditions is an important research and practical issue. This thesis aims to solve this issue by implementing a promising method, optimal design. Contrary to traditional random item selection, optimal design matches examinees and pretest items for maximal information on item parameters according to certain optimal design criterion. For example, in L-optimal design, item selection is based on correct item response probabilities equal to 25% or 75% with equal weights. A two-stage design is adopted for applying optimal design: random item selection as the first stage, and optimal design as the second stage with initial item parameter estimation obtained from the first stage. For controlling the measurement error in ability estimate which is an important issue when pretest items are seeded at the earlier stages of a CAT, optimal design uses two types of MLEs of ability levels: using current ability estimates in calculating item response probabilities at the item selection step, and using the final ability estimates at the item parameter estimation step. Random item selection uses the moment reconstruction (MR) method in logistic regression estimation. The results show that both of the methods of measurement error control are effective. Optimal design provides less biased item parameter estimation than estimation methods under random item selection (including MML estimation). The implications for item calibration, the research limitations, and future research directions are discussed. This exploratory study builds a solid basis for the future application of optimal design for accurate item calibration in small sample conditions in CATs.","abstract_html":"Item calibration in small sample conditions is an important research and practical issue. This thesis aims to solve this issue by implementing a promising method, optimal design. Contrary to traditional random item selection, optimal design matches examinees and pretest items for maximal information on item parameters according to certain optimal design criterion. For example, in L-optimal design, item selection is based on correct item response probabilities equal to 25% or 75% with equal weights. A two-stage design is adopted for applying optimal design: random item selection as the first stage, and optimal design as the second stage with initial item parameter estimation obtained from the first stage. For controlling the measurement error in ability estimate which is an important issue when pretest items are seeded at the earlier stages of a CAT, optimal design uses two types of MLEs of ability levels: using current ability estimates in calculating item response probabilities at the item selection step, and using the final ability estimates at the item parameter estimation step. Random item selection uses the moment reconstruction (MR) method in logistic regression estimation. The results show that both of the methods of measurement error control are effective. Optimal design provides less biased item parameter estimation than estimation methods under random item selection (including MML estimation). The implications for item calibration, the research limitations, and future research directions are discussed. This exploratory study builds a solid basis for the future application of optimal design for accurate item calibration in small sample conditions in CATs.","abstract_has_math":false,"creators":["Zhu, Rongchun"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"Ph.D.","degree_level":"Dissertation","degree_discipline":"Education","degree_department":null,"school":null,"contributors":["Douglas, Jeffrey"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2015,"date_issued":"2015-09-25T19:54:45Z","date_published":"2015-09-25T19:54:45Z","updated_at":"2026-07-22T22:26:13Z","subjects":["Education, Technology of"],"languages":["eng"],"rights":[],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["(MiAaPQ)AAI3250359"],"render_values":[{"text":"(MiAaPQ)AAI3250359","href":null,"code":true}]}]},"links":{"outbound_url":"http://hdl.handle.net/2142/79972","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Douglas, Jeffrey"]},{"key":"dc:creator","label":"Author","values":["Zhu, Rongchun"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2015-09-25T19:54:45Z","10000-01-01","2006"]},{"key":"dc:type","label":"Dc Type","values":["text"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Education"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Dissertation"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Ph.D."]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of Illinois at Urbana-Champaign"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Education, Technology of"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/2142/79972","(MiAaPQ)AAI3250359"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Item calibration in small sample conditions is an important research and practical issue. This thesis aims to solve this issue by implementing a promising method, optimal design. Contrary to traditional random item selection, optimal design matches examinees and pretest items for maximal information on item parameters according to certain optimal design criterion. For example, in L-optimal design, item selection is based on correct item response probabilities equal to 25% or 75% with equal weights. A two-stage design is adopted for applying optimal design: random item selection as the first stage, and optimal design as the second stage with initial item parameter estimation obtained from the first stage. For controlling the measurement error in ability estimate which is an important issue when pretest items are seeded at the earlier stages of a CAT, optimal design uses two types of MLEs of ability levels: using current ability estimates in calculating item response probabilities at the item selection step, and using the final ability estimates at the item parameter estimation step. Random item selection uses the moment reconstruction (MR) method in logistic regression estimation. The results show that both of the methods of measurement error control are effective. Optimal design provides less biased item parameter estimation than estimation methods under random item selection (including MML estimation). The implications for item calibration, the research limitations, and future research directions are discussed. This exploratory study builds a solid basis for the future application of optimal design for accurate item calibration in small sample conditions in CATs.","Made available in DSpace on 2015-09-25T19:54:45Z (GMT). No. of bitstreams: 2 license.txt: 4848 bytes, checksum: 96035ab3f5e1c23cc7138a224ce498bd (MD5) 3250359.pdf: 1838772 bytes, checksum: 6f315a2dc634be99a3258501c1837869 (MD5) Previous issue date: 2006","Embargo set by: Seth Robbins for item 81254 Lift date: Forever Reason: Restricted to the U of I community idenfinitely during batch ingest of legacy ETDs","Restricted to the U of I community idenfinitely during batch ingest of legacy ETDs","U of I Only","72 p.","Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2006."]},{"key":"dc:title","label":"Title","values":["Implementation of Optimal Design for Item Calibration in Computerized Adaptive Testing (Cat)"]}]}],"canonical_facts":{"dc:contributor":["Douglas, Jeffrey"],"dc:creator":["Zhu, Rongchun"],"dc:date":["2015-09-25T19:54:45Z","10000-01-01","2006"],"dc:description":["Item calibration in small sample conditions is an important research and practical issue. This thesis aims to solve this issue by implementing a promising method, optimal design. Contrary to traditional random item selection, optimal design matches examinees and pretest items for maximal information on item parameters according to certain optimal design criterion. For example, in L-optimal design, item selection is based on correct item response probabilities equal to 25% or 75% with equal weights. A two-stage design is adopted for applying optimal design: random item selection as the first stage, and optimal design as the second stage with initial item parameter estimation obtained from the first stage. For controlling the measurement error in ability estimate which is an important issue when pretest items are seeded at the earlier stages of a CAT, optimal design uses two types of MLEs of ability levels: using current ability estimates in calculating item response probabilities at the item selection step, and using the final ability estimates at the item parameter estimation step. Random item selection uses the moment reconstruction (MR) method in logistic regression estimation. The results show that both of the methods of measurement error control are effective. Optimal design provides less biased item parameter estimation than estimation methods under random item selection (including MML estimation). The implications for item calibration, the research limitations, and future research directions are discussed. This exploratory study builds a solid basis for the future application of optimal design for accurate item calibration in small sample conditions in CATs.","Made available in DSpace on 2015-09-25T19:54:45Z (GMT). No. of bitstreams: 2 license.txt: 4848 bytes, checksum: 96035ab3f5e1c23cc7138a224ce498bd (MD5) 3250359.pdf: 1838772 bytes, checksum: 6f315a2dc634be99a3258501c1837869 (MD5) Previous issue date: 2006","Embargo set by: Seth Robbins for item 81254 Lift date: Forever Reason: Restricted to the U of I community idenfinitely during batch ingest of legacy ETDs","Restricted to the U of I community idenfinitely during batch ingest of legacy ETDs","U of I Only","72 p.","Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2006."],"dc:identifier":["http://hdl.handle.net/2142/79972","(MiAaPQ)AAI3250359"],"dc:language":["eng"],"dc:subject":["Education, Technology of"],"dc:title":["Implementation of Optimal Design for Item Calibration in Computerized Adaptive Testing (Cat)"],"dc:type":["text"],"thesis:degree_discipline":["Education"],"thesis:degree_level":["Dissertation"],"thesis:degree_name":["Ph.D."],"thesis:institution_name":["University of Illinois at Urbana-Champaign"]},"updated_at":"2026-07-22T22:26:13Z"}