1996
DOI: 10.1111/j.1745-3984.1996.tb00490.x
|View full text |Cite
|
Sign up to set email alerts
|

Simulation Studies of the Effects of Small Sample Size and Studied Item Parameters on SIBTEST and Mantel‐Haenszel Type I Error Performance

Abstract: Two simulation studies investigated Type I error performance of two statistical procedures for detecting differential item functioning (DIF): SIBTEST and Mantel‐Haenszel (MH). Because MH and SIBTEST are based on asymptotic distributions requiring “large” numbers of examinees, the first study examined Type 1 error for small sample sizes. No significant Type I error inflation occurred for either procedure. Because MH has the potential for Type I error inflation for non‐Rasch models, the second study used a marke… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

14
198
5
10

Year Published

1997
1997
2019
2019

Publication Types

Select...
8

Relationship

0
8

Authors

Journals

citations
Cited by 201 publications
(227 citation statements)
references
References 11 publications
14
198
5
10
Order By: Relevance
“…and Chang ( 1994) showed that when the studied item discrimination parameter is substantially higher than the item discrimination parameters of the matching items, the Mantel method and other observed-score procedures can produce distorted results, including unacceptably high Type I error rates. These results parallel those of Roussos and Stout (1993) and Uttaro and Millsap (1994) for the dichotomous case.…”
Section: J_n Equation 3 Which Is Known As the Item Category Responsesupporting
confidence: 82%
See 1 more Smart Citation
“…and Chang ( 1994) showed that when the studied item discrimination parameter is substantially higher than the item discrimination parameters of the matching items, the Mantel method and other observed-score procedures can produce distorted results, including unacceptably high Type I error rates. These results parallel those of Roussos and Stout (1993) and Uttaro and Millsap (1994) for the dichotomous case.…”
Section: J_n Equation 3 Which Is Known As the Item Category Responsesupporting
confidence: 82%
“…Some simulation studies that have shown considerable robustness of observed-score methods are Donoghue, Holland, and Thayer (1993) and Shealy and Stout (1993) in the dichotomous case and Grima (1993a, 1993b) in the polytomous case. Other researchers have found very large Type I error rates for the observed-score methods under certain conditions, including Shealy (1989), Roussos and Stout (1993), and Uttaro and Millsap (1994) in the case of the MH and Allen and Donoghue (in press) and Chang, Mazzeo, and Roussos (in press) in the case of the Mantel or SMD procedures. Further discussion of the robustness issue appears in Section 4.3.1.…”
Section: Errors In Matchini Examinees In Dif Assessmentmentioning
confidence: 95%
“…Eğer referans grubun doğru yanıtlama olasılığı büyük ise MH>1 ve eğer odak grubun doğru yanıtlama olasılığı büyük ise MH<1 olacaktır. Eğitimsel Test Hizmetleri (Educational Testing Service-ETS), DMF etki büyüklüğüne ait bazı değerler önermiş ve bu değerler Roussos and Stout (1996) tarafından modifiye edilmiştir. |ΔMH| < 1 ise ihmal edilir düzeyde; 1 ≤ |ΔMH| < 1,5 ise orta düzeyde, |ΔMH)| ≥ 1,5 yani sıfırdan büyük ölçüde farklılık gösteriyor ise yüksek düzeyde DMF olduğunu gösterir.…”
Section: Mantel-haenszel (Mh) Yöntemiunclassified
“…Tip hata oranlarının, 20 maddelik testte 40 maddelik testten daha fazla olduğunu belirtmiştir. Roussos ve Stout (1996), yaptıkları simülasyon çalışmasında DMF'yi belirlemek için MH ve SIBTEST yöntemlerini kullanmış, örneklem büyüklüklerini ve yetenek dağılımlarını manipüle etmiş ve bu yöntemlerdeki 1. Tip hatayı araştırmışlardır.…”
Section: Introductionunclassified