Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

A Critical Assessment of Sampling Biases in Geometric Morphometric Analysis: The Case of Homo erectus

Zhang, Chi (2020) A Critical Assessment of Sampling Biases in Geometric Morphometric Analysis: The Case of Homo erectus. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

Download (24MB) | Preview


This study primarily explores the potential of GMA in capturing and recognizing specific morphological features of specimens commonly allocated to H. erectus. For these purposes, four surface semilandmark datasets were collected for analyzing the gross morphology of the entire crania, posterior crania, temporal bones, and frontal bones.

Results show that though dense surface semilandmarks can potentially capture many morphological features that cannot be included by using discrete landmarks, some shape differences visualized by PCA, include those in lateral cranial profiles, the trajectories of the squamosal sutures and the inclinations of the nuchal planes, do not match observed morphology. Procrustes superimposition is identified as an important, yet usually ignored, factor that may contribute to some incompatibilities between visualized shape differences and observed morphology. This is because Procrustes superimposition rotates configurations to different orientations to minimize their overall differences, while observed morphology is based on aligning specimens in the Frankfurt plane.

This study also assesses three sampling issues associated with statistical analyses commonly used for testing whether H. erectus is too variable to be a single species based on landmark data. The first two are whether densities of semilandmarks and focusing on different cranial regions can yield inconsistent results of the same statistical analyses. The third is whether including a large sample of H. sapiens may affect results of PCAs.

Results suggest that semilandmarks used in this study are oversampled, because using a much smaller number of semilandmarks can yield nearly identical results. The reason is that GMA places more weight on gross shape differences than details. Furthermore, the analyses of the temporal and frontal bones show results incompatible with those of the entire and posterior crania, likely because the former analyses focus on shapes of individual bones, while the latter analyses focus on gross cranial shapes. In addition, including a large sample of H. sapiens with fossils cause higher-ranked PCs to emphasize variations within H. sapiens. In this case, it is also important to examine lower PCs to understand shape differences between fossils.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Zhang, Chichz31@pitt.educhz31
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairSchwartz, Jeffreyjhs@pitt.edujhs
Committee MemberHanks, Bryanbkh5@pitt.edubkh5
Committee MemberJudd, Margaretmjudd@pitt.edumjudd
Committee MemberHarcourt-Smith,
Date: 8 June 2020
Date Type: Publication
Defense Date: 27 February 2020
Approval Date: 8 June 2020
Submission Date: 18 March 2020
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Number of Pages: 457
Institution: University of Pittsburgh
Schools and Programs: Dietrich School of Arts and Sciences > Anthropology
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: Paleoanthropology, biological anthropology, geometric morphometrics, Homo erectus, hominids
Date Deposited: 08 Jun 2020 17:11
Last Modified: 18 Sep 2020 13:32


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item