A reasonable project would be to develop one or more statistical methods to determine which of the cranial characteristics in Howells' craniometric series are useful for distinguishing the 28 populations (or a reasonable subset) in the database. For simplicity, focus on males. If you use discriminant analysis you should determine which populations are clearly separated on the various canonical discriminant functions. Furthermore, it would be useful to assess which populations are relatively difficult to distinguish, and whether there is any geographic or biological explanation for this. Here is a reminder on the database. A copy of the raw data is available under course projects. DATABASE DOCUMENTATION for W.W. HOWELLS' (1973, 1989) CRANIOMETRIC SERIES -------------------------------------------------------------------------- INTRODUCTION This database consists of the raw variates for W.W. Howells craniometric series (1973, 1989) in ascii format. It consists of 47 of the original 77 craniometric measurements for over 2500 individuals from 28 populations. DATA MATRIX ROWS: are data on individuals. COLUMNS: columns are variables (indicator variables and measurement variates). Variables 1-2 are indicator variables, and variables 3-47 are measurement variates (refer to Measurement Notes below regarding measurement scale etc.). All columns are defined below: Variable No. Label Indicator or Measurement 1. POP Population affiliation for each individual, coded as 1-28. See Populations (below) for coding designations. 2. SEX Designation of male or female for each individual: Male = 1, Female = 2. See Sex (below). 3. GOL Glabello-occipital length 4. NOL Nasio-occipital length 5. BNL Basion-nasion length 6. BBH Basion-bregma height 7. XCB Maximum cranial breadth 8. XFB Maximum frontal breadth 9. STB Bistephanic breadth 10. ZYB Bizygomatic breadth 11. AUB Biauricular breadth 12. WCB Minimum cranial breadth 13. ASB Biasterionic breadth 14. BPL Basion-prosthion length 15. NPH Nasion-prosthion height 16. NLH Nasal height 17. OBH Orbit height, left 18. OBB Orbit breadth, left 19. JUB Bijugal breadth 20. NLB Nasal breadth 21. MAB Palate breadth, external 22. MDH Mastoid height 23. MDB Mastoid breadth 24. ZMB Bimaxillary breadth 25. SSS Zygomaxillary subtense 26. FMB Bifrontal breadth 27. NAS Nasio-frontal subtense 28. EKB Biorbital breadth 29. DKS Dacryon subtense 30. DKB Interorbital breadth 31. NDS Naso-dacryal subtense 32. WNB Simotic chord (least nasal breadth) 33. SIS Simotic subtense 34. IML Malar length, inferior 35. XML Malar length, maximum 36. MLS Malar subtense 37. WMH Cheek height 38. SOS Supraorbital projection 39. GLS Glabella projection 40. FOL Foramen magnum length 41. FRC Nasion-bregma chord (Frontal chord) 42. FRS Nasion-bregma subtense (Frontal subtense) 43. FRF Nasion-subtense fraction 44. PAC Bregma-lambda chord (Parietal chord) 45. PAS Bregma-lambda subtense (Parietal subtense) 46. PAF Bregma-subtense fraction 47. OCC Lambda-opisthion chord (Occipital chord) POPULATIONS As noted above, population affiliation for each individual is denoted in column #2. Each population is coded by values 1-28 specified below: Population code Population 1. Norse: (Medieval), Europe, Oslo 2. Zalavar: Central Europe, Hungary 3. Berg: Central Europe, Carinthia, Austria 4. Teita: East Africa, Kenya 5. Dogon: West Africa, Mali 6. Zulu: South Africa 7. Lake Alexandrina Tribes: South Australia 8. Tasmanian: Tasmania 9. Tolai: Melanesia, New Britain 10. Mokapu: Oahu, Hawaii, Polynesia 11. Easter Island: Polynesia 12. Moriori: Chatham Islands, Polynesia 13. Arikara: (Early) North America 14. Santa Cruz Island: California, N. America 15. Yauyos: Peru, South America 16. Hokkaido: North Japan 17. North Kyushu: South Japan 18. Hainan: Haikou City, China 19. Atayal: Taiwan Aboriginals 20. Phillipine: Phillipine Islands 21. Guam: Latte Period 22. Egypt: Gizeh, 26th-30th Dynasties 23. San: South Africa 24. Andaman Islands: Andaman Islands 25. Ainu: S. and SE. Kokkaido, Japan 26. Buriat: Siberia 27. Eskimo: Inugsuk, Greenland 28. Anyang: Shang Dynasty, China SEX All populations are represented by both males and females except: #20 Phillipines and #28 anyang, which have only males in the samples. MEASUREMENT NOTES 1. There are very few missing data entries; these are denoted with periods (.). 2. Columns 3-47 are linear measurements. All linear measurements are recorded to the nearest millimeter except: columns 32 (WNB) and 33 (SIS), which are recorded to the nearest 1/10 millimeter. SAMPLE N In a couple of instances (e.g. pop. #3 Berg males and pop. #27 Eskimo females) sample sizes in the present data base are off by 1 individual from those published in Howells (1973, 1989). It is unclear whether these are additions or omissions in the original study, or the result of database transformation to the present format. However, given the relatively large N for both of these samples, this is unlikely to significantly affect the outcome of any analysis. BIBLIOGRAPHY Howells WW (1973) Cranial variation in man: a study of multivariate analysis of patterns of difference among recent human populations. Peabody Museum Papers 67:1-259. Howells WW (1989) Skull shapes and the map: craniometric analyses in the dispersion of modern Homo. Peabody Museum Papers 79:1-189.