Researchers and publishing guidelines recommend reporting confidence intervals (CIs) not just along with null hypothesis significance testing (NHST), but for many other statistics such as effect sizes and reliability coefficients. Although CI and standard errors (SEs) are closely related, examining standard errors alone in simulation studies is not adequate because we do not always know if a standard error is small enough. Overly small SEs may lead to increased probability of Type-I error and CIs with lower coverage rate than expected. Statistical simulation studies generally examine the magnitude of the empirical standard error, but it is not clear if they examine the properties of confidence intervals. The present study examines confidence interval investigating and reporting practices, particularly with respect to coverage and bias as diagnostics in published statistical simulation studies across eight psychology journals using a systematic literature review. Results from this review will inform editorial policies and hopefully encourage researchers to report CIs.
| Identifer | oai:union.ndltd.org:unt.edu/info:ark/67531/metadc1538784 |
| Date | 08 1900 |
| Creators | Kabakci, Maside |
| Contributors | Natesan, Prathiba, Henson, Robin, Boesch, Miriam, Pavur, Robert |
| Publisher | University of North Texas |
| Source Sets | University of North Texas |
| Language | English |
| Detected Language | English |
| Type | Thesis or Dissertation |
| Format | v, 89 pages, Text |
| Rights | Public, Kabakci, Maside, Copyright, Copyright is held by the author, unless otherwise noted. All rights Reserved. |
Page generated in 0.0019 seconds