Project 2577: M. A. O'Leary, K. Alphonse, M. Arce Hernandez, D. Cavaliere, A. Cirranello, T. Dietterich, M. Julius, S. Kaufman, E. Law, M. Passarotti, A. Reft, J. Robalino, N. B. Simmons, S. Smith, D. Stevenson, E. Theriot, P. M. Velazco, R. Walls, M. Yu, M. Daly. accepted. Crowds replicate performance of scientific experts scoring phylogenetic matrices of phenotypes. Systematic Biology. 67 (1):49-60. (In Press)
Specimen: Kerivoula hardwickii (AMNH/Mammalogy:107483)
View: Skull-dorsal

Abstract

Scientists building the Tree of Life face an overwhelming challenge to categorize phenotypes (e.g., anatomy, physiology) from millions of living and fossil species. This biodiversity challenge far outstrips the capacities of trained scientific experts. Here we explore whether crowdsourcing can be used to collect matrix data on a large scale with the participation of the non-expert students, or “citizen scientists.” Crowdsourcing, or data collection by non-experts, frequently via the internet, has enabled scientists to tackle some large-scale data collection challenges too massive for individuals or scientific teams alone. The quality of work by non-expert crowds is, however, often questioned and little data has been collected on how such crowds perform on complex tasks such as phylogenetic character coding. We studied a crowd of over 600 non-experts, and found that they could use images to identify anatomical similarity (hypotheses of homology) with an average accuracy of 82% compared to scores provided by experts in the field. This performance pattern held across the Tree of Life, from protists to vertebrates. We introduce a procedure that predicts the difficulty of each character and that can be used to assign harder characters to experts and easier characters to a non-expert crowd for scoring. We test this procedure in a controlled experiment comparing crowd scores to those of experts and show that crowds can produce matrices with over 90% of cells scored correctly while reducing the number of cells to be scored by experts by 50%. Preparation time, including image collection and processing, for a crowdsourcing experiment is significant, and does not currently save time of scientific experts overall. However, if innovations in automation or robotics can reduce such effort, then large-scale implementation of our method could greatly increase the collective scientific knowledge of species phenotypes for phylogenetic tree building. For the field of crowdsourcing, we provide a rare study with ground truth, or an experimental control that many studies lack, and contribute new methods on how to coordinate the work of experts and non-experts. We show that there are important instances in which crowd consensus is not a good proxy for correctness.


Read the article »

Article DOI: https://doi.org/10.1093/sysbio/syx052

Project DOI: 10.7934/P2577, http://dx.doi.org/10.7934/P2577
This project containsMatrices
  • 96 Media
  • 1 Matrix
  • 1 Document
  • 6 Taxa
  • 15 Specimens
  • 20 Characters
Total size of project's media files: 318 MB

Download Project SDD File
Total scored cells: 140
Total media associated with cells: 122
Total labels associated with cell media: 148
Characters
Total characters: 20
Total characters with associated media: 20
Total characters with media with labels: 0
Total character states: 55
Total character states with associated media: 52
Total character states with media with labels:0
Total unordered/ordered characters:20/0
Currently Viewing:
MorphoBank Project 2577

    This research
    supported by



    Members

    member name taxa specimens media chars character
    media
    labels
    cell scorings
    (scored, NPA, "-")
    cell
    medialabels
    rules
    Andrea Cirranello
    Project Administrator
    Last logged in 05/30/17
    0000000
    (0, 0, 0)
    000
    Maureen O'Leary
    Full membership
    Last logged in Never
    0000000
    (0, 0, 0)
    000
    Nancy Simmons
    Full membership
    Last logged in Never
    0000000
    (0, 0, 0)
    000
    Paul Velazco
    Full membership
    Last logged in Never
    0000000
    (0, 0, 0)
    000


    Taxonomic Overview for Matrix 'M24371' (6 Taxa)

    taxon unscored
    cells
    scored
    cells
    no cell
    support
    NPA
    cells
    "-" cellscell images labels on
    cell images
    member
    access
    [1] Cynopterus sphinx
    Last Modified in 12/21/16
    0400002022none
    [2] Saccopteryx bilineata
    Last Modified in 12/21/16
    0200002125none
    [3] Trachops cirrhosus
    Last Modified in 12/21/16
    0200002125none
    [4] Kerivoula hardwickii
    Last Modified in 12/21/16
    0200002027none
    [5] Miniopterus schrebersii
    Last Modified in 12/21/16
    0200002026none
    [6] Erinaceus europaeus
    Last Modified in 12/21/16
    0200002023none


    Project views

    type number of views Individual items viewed (where applicable)
    Total project views15757
    Project overview849
    Matrix views344Matrix landing page (317 views); Bat Skull Matrix (27 views);
    Media views9393Media search (694 views); M435071 (87 views); M435131 (79 views); M435137 (84 views); M435118 (68 views); M435083 (99 views); M435129 (89 views); M435095 (91 views); M435084 (84 views); M435053 (94 views); M435086 (96 views); M435107 (78 views); M435135 (87 views); M435141 (84 views); M435127 (90 views); M435113 (85 views); M435054 (100 views); M435064 (107 views); M435146 (95 views); M435124 (109 views); M435136 (89 views); M435089 (86 views); M435076 (86 views); M435085 (105 views); M435058 (109 views); M435068 (96 views); M435057 (99 views); M435106 (92 views); M435130 (95 views); M435104 (102 views); M435087 (143 views); M435116 (81 views); M435122 (86 views); M435142 (90 views); M435088 (102 views); M435060 (96 views); M435133 (86 views); M435070 (94 views); M435072 (106 views); M435063 (80 views); M435145 (80 views); M435081 (95 views); M435069 (95 views); M435138 (76 views); M435126 (73 views); M435147 (87 views); M435056 (115 views); M435108 (86 views); M435140 (76 views); M435091 (103 views); M435062 (86 views); M435111 (85 views); M435143 (87 views); M435082 (97 views); M435094 (90 views); M435080 (86 views); M435128 (74 views); M435059 (102 views); M435066 (96 views); M435125 (79 views); M435144 (96 views); M435067 (84 views); M435114 (106 views); M435061 (109 views); M435109 (73 views); M435090 (82 views); M435134 (89 views); M435077 (80 views); M435098 (81 views); M435123 (72 views); M435100 (95 views); M435119 (77 views); M435115 (89 views); M435096 (81 views); M435078 (92 views); M435074 (80 views); M435105 (87 views); M435112 (104 views); M435097 (94 views); M435148 (91 views); M435110 (91 views); M435055 (97 views); M435099 (87 views); M435093 (99 views); M435065 (99 views); M435121 (92 views); M435073 (84 views); M435132 (78 views); M435102 (104 views); M435092 (95 views); M435075 (89 views); M435120 (89 views); M435101 (98 views); M435117 (91 views); M435079 (88 views); M435139 (80 views); M435103 (79 views);
    Taxon list1727
    Specimen list2224
    Views for media list273
    Documents list281
    Bibliography666




    Project downloads

    type number of downloads Individual items downloaded (where applicable)
    Total downloads from project66
    Project downloads53
    Matrix downloads11Bat Skull Matrix (11 downloads);
    Document downloads2Supplementary material (2 downloads);