The Proceedings of the Eighth International Conference on Creationism (2018)

three genera ( Cyclopes , Myrmecophaga , and Tamandua ) from two families in the anteater kind Vermilingua, but our results support only Myrmecophaga and Tamandua in holobaramin Myrmecophaginae. These results will need further evaluation in future studies to clarify these differences. Wise’s (2009) estimation of mammalian ark kinds differs substantially from all others in the creationist literature. His estimates are based on the Post-Flood Continuity Criterion (PFCC), which in turn is based on the proposal that an ark kind should leave a continuous fossil record from the Flood to the present. His minimal estimate places a large number of species into as few as 97 ark kinds, and his maximal estimate places fewer species into as many as 234 ark kinds. Not surprisingly, our putative holobaramins do not match his estimated ark kinds (Table 3). Instead, we find tentative evidence of discontinuity within 33 of his minimal ark kinds and 18 of his maximal ark kinds. Sixteen of our putative holobaramins match closely sixteen of Wise’s maximal ark kinds. Resolving the discrepancy between Wise’s results and ours will require assessments of the completeness of the post-Flood fossil record as well as further evaluations of our baraminology results. Our results should also help creationists evaluate baraminology methods, which have recently come under scrutiny (Wood 2016b; O’Micks 2016; Wood 2017; O’Micks 2017). Previously, Wood’s (2009) survey of 73 statistical baraminology studies examined taxon samples that should have revealed discontinuity as well as those that should not (such as within a single genus). Wood found that 61.6% of the studies were “successful” in either revealing or not revealing predicted discontinuity. Our survey differs in that we focused exclusively on datasets that should have revealed discontinuity around a taxonomic family, but we found 69.5% of studies revealed evidence of the expected discontinuity. This is an improvement from the previous survey, and the question remains: Why do some datasets reveal evidence of discontinuity while others do not? We attempted to determine what sort of feature of the datasets might correlate or predict successful discontinuity detection by examining the number of taxa and characters, the stress and k min of the MDS, and the median bootstrap value and F 90 of the BDC results, but we found no significant correlations, as in previous studies. Future studies will definitely want to examine these results more closely to determine the type of dataset best suited to these studies. Future studies should more closely evaluate individual taxonomic groups named herein, especially those where discrepancies between different studies are noted. The current best practice in statistical baraminology is to use multiple holistic character sets compiled from different sources to evaluate the baraminology of a single group, as in hominin baraminology (e.g., see Wood 2016c; 2017). Some of our studies add to baraminology studies of previous groups, while most provide only a first approximation of a group’s baraminology. Nevertheless, the present results provide an important advance in baraminology work and in the study of mammalian created kinds. Thompson and Wood ◀ A survey of Cenozic mammal baramins ▶ 2018 ICC 219 Character type Frequency in all datasets Percentage in all datasets Frequency in holobaramin datasets Percentage in holobaramin datasets Frequency in inconclusive datasets Percentage in inconclusive datasets Craniodental 36 43.9% 25 43.9% 11 50.0% Craniodental + postcranial 23 28.0% 15 26.3% 7 31.8% Dental 8 9.8% 6 10.5% 2 9.1% >3 types 9 11.0% 6 10.5% 1 4.5% Cranial + postcranial 4 4.9% 4 7.0% 0 0.0% Dental + postcranial 1 1.2% 0 0.0% 1 4.5% Cranial 1 1.2% 1 1.8% 0 0.0% Table 1. Character types in the datasets used in this study. “Holobaramin datasets” refer to datasets in which a putative holobaramin was identified, and “Inconclusive datasets” are datasets which produced inconclusive BDC and MDS results. Table 2. Comparison of results between datasets for which a holobaramin was identified and datasets which produced inconclusive results. Welch’s unequal variance t -test was used to compare the holobaramin and inconclusive datasets. Holobaramin dataset mean Inconclusive dataset mean Welch’s t P-value Taxa in analysis 21.9 26.6 -1.08 0.2915 Characters in analysis 71.4 66.1 0.38 0.7029 3D stress 0.16 0.17 -0.45 0.6528 k min 6.2 6.0 0.19 0.8473 Median bootstrap 86.1 82.0 1.80 0.0800 F 90 0.44 0.34 1.98 0.0537