Working with population totals in the presence of missing data comparing imputation methods in terms of bias and precision View Full Text


Ontology type: schema:ScholarlyArticle      Open Access: True


Article Info

DATE

2016-10-24

AUTHORS

Thierry Onkelinx, Koen Devos, Paul Quataert

ABSTRACT

Missing observations in water bird censuses are commonly handled using the Underhill index or the birdSTATs tool which enables the use of TRIM under the hood. Multiple imputation is a standard technique for handling missing data that is rarely used in the field of ecology, but is a well known statistical technique in the fields of medical and social sciences. The purpose of this paper is to compare these three methods in terms of bias and variance. The bias in the Underhill method depends on the algorithm and starting values. birdSTATs and multiple imputation are unbiased in the case of missing values that are missing completely at random; more missing values implies less information, and so wider confidence intervals are expected as the missingness increases. The Underhill method and birdSTATs tool underestimate the variance; omitting data from a complete dataset and applying the Underhill index or birdSTATs tool results in smaller confidence intervals. Multiple imputation with an adequate imputation model provides wider confidence intervals. Biased parameter estimates with underestimated variance can potentially lead to incorrect management and policy conclusions. Hence, we dissuade the use of Underhill indices or the birdSTATs tool to handle missing data, rather we suggest that multiple imputation is a more robust alternative, even in suboptimal conditions. More... »

PAGES

603-615

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/s10336-016-1404-9

DOI

http://dx.doi.org/10.1007/s10336-016-1404-9

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1006321158


Indexing Status Check whether this publication has been indexed by Scopus and Web Of Science using the SN Indexing Status Tool
Incoming Citations Browse incoming citations for this publication using opencitations.net

JSON-LD is the canonical representation for SciGraph data.

TIP: You can open this SciGraph record using an external JSON-LD service: JSON-LD Playground Google SDTT

[
  {
    "@context": "https://springernature.github.io/scigraph/jsonld/sgcontext.json", 
    "about": [
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/05", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Environmental Sciences", 
        "type": "DefinedTerm"
      }, 
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/06", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Biological Sciences", 
        "type": "DefinedTerm"
      }, 
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0502", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Environmental Science and Management", 
        "type": "DefinedTerm"
      }, 
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0608", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Zoology", 
        "type": "DefinedTerm"
      }
    ], 
    "author": [
      {
        "affiliation": {
          "alternateName": "Research Institute for Nature and Forest (INBO), Kliniekstraat 25, 1070, Anderlecht, Belgium", 
          "id": "http://www.grid.ac/institutes/grid.435417.0", 
          "name": [
            "Research Institute for Nature and Forest (INBO), Kliniekstraat 25, 1070, Anderlecht, Belgium"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Onkelinx", 
        "givenName": "Thierry", 
        "id": "sg:person.012650302401.91", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012650302401.91"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Research Institute for Nature and Forest (INBO), Kliniekstraat 25, 1070, Anderlecht, Belgium", 
          "id": "http://www.grid.ac/institutes/grid.435417.0", 
          "name": [
            "Research Institute for Nature and Forest (INBO), Kliniekstraat 25, 1070, Anderlecht, Belgium"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Devos", 
        "givenName": "Koen", 
        "id": "sg:person.012302577023.00", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012302577023.00"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Research Institute for Nature and Forest (INBO), Kliniekstraat 25, 1070, Anderlecht, Belgium", 
          "id": "http://www.grid.ac/institutes/grid.435417.0", 
          "name": [
            "Research Institute for Nature and Forest (INBO), Kliniekstraat 25, 1070, Anderlecht, Belgium"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Quataert", 
        "givenName": "Paul", 
        "id": "sg:person.0771550506.84", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0771550506.84"
        ], 
        "type": "Person"
      }
    ], 
    "citation": [
      {
        "id": "sg:pub.10.1007/s00265-010-1044-7", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1022305626", 
          "https://doi.org/10.1007/s00265-010-1044-7"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/s10841-004-1332-5", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1033148717", 
          "https://doi.org/10.1007/s10841-004-1332-5"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/s10344-013-0768-x", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1046979458", 
          "https://doi.org/10.1007/s10344-013-0768-x"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/978-0-387-21706-2", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1035613449", 
          "https://doi.org/10.1007/978-0-387-21706-2"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/s10841-014-9644-6", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1036623878", 
          "https://doi.org/10.1007/s10841-014-9644-6"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/s10336-007-0176-7", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1041727822", 
          "https://doi.org/10.1007/s10336-007-0176-7"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/s11121-007-0070-9", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1018783636", 
          "https://doi.org/10.1007/s11121-007-0070-9"
        ], 
        "type": "CreativeWork"
      }
    ], 
    "datePublished": "2016-10-24", 
    "datePublishedReg": "2016-10-24", 
    "description": "Missing observations in water bird censuses are commonly handled using the Underhill index or the birdSTATs tool which enables the use of TRIM under the hood. Multiple imputation is a standard technique for handling missing data that is rarely used in the field of ecology, but is a well known statistical technique in the fields of medical and social sciences. The purpose of this paper is to compare these three methods in terms of bias and variance. The bias in the Underhill method depends on the algorithm and starting values. birdSTATs and multiple imputation are unbiased in the case of missing values that are missing completely at random; more missing values implies less information, and so wider confidence intervals are expected as the missingness increases. The Underhill method and birdSTATs tool underestimate the variance; omitting data from a complete dataset and applying the Underhill index or birdSTATs tool results in smaller confidence intervals. Multiple imputation with an adequate imputation model provides wider confidence intervals. Biased parameter estimates with underestimated variance can potentially lead to incorrect management and policy conclusions. Hence, we dissuade the use of Underhill indices or the birdSTATs tool to handle missing data, rather we suggest that multiple imputation is a more robust alternative, even in suboptimal conditions.", 
    "genre": "article", 
    "id": "sg:pub.10.1007/s10336-016-1404-9", 
    "inLanguage": "en", 
    "isAccessibleForFree": true, 
    "isPartOf": [
      {
        "id": "sg:journal.1035137", 
        "issn": [
          "2193-7192", 
          "2193-7206"
        ], 
        "name": "Journal of Ornithology", 
        "publisher": "Springer Nature", 
        "type": "Periodical"
      }, 
      {
        "issueNumber": "2", 
        "type": "PublicationIssue"
      }, 
      {
        "type": "PublicationVolume", 
        "volumeNumber": "158"
      }
    ], 
    "keywords": [
      "multiple imputation", 
      "terms of bias", 
      "imputation model", 
      "underestimated variance", 
      "smaller confidence intervals", 
      "imputation methods", 
      "statistical techniques", 
      "missing values", 
      "robust alternative", 
      "biased parameters", 
      "imputation", 
      "population total", 
      "field of ecology", 
      "confidence intervals", 
      "standard techniques", 
      "algorithm", 
      "variance", 
      "complete dataset", 
      "terms", 
      "tool", 
      "wide confidence intervals", 
      "technique", 
      "social sciences", 
      "model", 
      "field", 
      "parameters", 
      "intervals", 
      "data", 
      "datasets", 
      "bias", 
      "less information", 
      "values", 
      "science", 
      "conditions", 
      "cases", 
      "precision", 
      "alternative", 
      "trim", 
      "use", 
      "information", 
      "index", 
      "observations", 
      "purpose", 
      "bird censuses", 
      "suboptimal conditions", 
      "ecology", 
      "policy conclusions", 
      "incorrect management", 
      "census", 
      "presence", 
      "management", 
      "conclusion", 
      "hood", 
      "increase", 
      "method", 
      "total", 
      "paper", 
      "water bird censuses", 
      "Underhill index", 
      "birdSTATs tool", 
      "use of TRIM", 
      "Underhill method", 
      "birdSTATs", 
      "missingness increases", 
      "adequate imputation model"
    ], 
    "name": "Working with population totals in the presence of missing data comparing imputation methods in terms of bias and precision", 
    "pagination": "603-615", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1006321158"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/s10336-016-1404-9"
        ]
      }
    ], 
    "sameAs": [
      "https://doi.org/10.1007/s10336-016-1404-9", 
      "https://app.dimensions.ai/details/publication/pub.1006321158"
    ], 
    "sdDataset": "articles", 
    "sdDatePublished": "2022-01-01T18:39", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-springernature-scigraph/baseset/20220101/entities/gbq_results/article/article_702.jsonl", 
    "type": "ScholarlyArticle", 
    "url": "https://doi.org/10.1007/s10336-016-1404-9"
  }
]
 

Download the RDF metadata as:  json-ld nt turtle xml License info

HOW TO GET THIS DATA PROGRAMMATICALLY:

JSON-LD is a popular format for linked data which is fully compatible with JSON.

curl -H 'Accept: application/ld+json' 'https://scigraph.springernature.com/pub.10.1007/s10336-016-1404-9'

N-Triples is a line-based linked data format ideal for batch operations.

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/s10336-016-1404-9'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/s10336-016-1404-9'

RDF/XML is a standard XML format for linked data.

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/s10336-016-1404-9'


 

This table displays all metadata directly associated to this object as RDF triples.

173 TRIPLES      22 PREDICATES      99 URIs      82 LITERALS      6 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/s10336-016-1404-9 schema:about anzsrc-for:05
2 anzsrc-for:0502
3 anzsrc-for:06
4 anzsrc-for:0608
5 schema:author N5aa280f93727433f95c4a7899ed887b0
6 schema:citation sg:pub.10.1007/978-0-387-21706-2
7 sg:pub.10.1007/s00265-010-1044-7
8 sg:pub.10.1007/s10336-007-0176-7
9 sg:pub.10.1007/s10344-013-0768-x
10 sg:pub.10.1007/s10841-004-1332-5
11 sg:pub.10.1007/s10841-014-9644-6
12 sg:pub.10.1007/s11121-007-0070-9
13 schema:datePublished 2016-10-24
14 schema:datePublishedReg 2016-10-24
15 schema:description Missing observations in water bird censuses are commonly handled using the Underhill index or the birdSTATs tool which enables the use of TRIM under the hood. Multiple imputation is a standard technique for handling missing data that is rarely used in the field of ecology, but is a well known statistical technique in the fields of medical and social sciences. The purpose of this paper is to compare these three methods in terms of bias and variance. The bias in the Underhill method depends on the algorithm and starting values. birdSTATs and multiple imputation are unbiased in the case of missing values that are missing completely at random; more missing values implies less information, and so wider confidence intervals are expected as the missingness increases. The Underhill method and birdSTATs tool underestimate the variance; omitting data from a complete dataset and applying the Underhill index or birdSTATs tool results in smaller confidence intervals. Multiple imputation with an adequate imputation model provides wider confidence intervals. Biased parameter estimates with underestimated variance can potentially lead to incorrect management and policy conclusions. Hence, we dissuade the use of Underhill indices or the birdSTATs tool to handle missing data, rather we suggest that multiple imputation is a more robust alternative, even in suboptimal conditions.
16 schema:genre article
17 schema:inLanguage en
18 schema:isAccessibleForFree true
19 schema:isPartOf N3dc28cc4071e45989be62d9479526622
20 N58dceb2e0896429394b87983c7d5cc5b
21 sg:journal.1035137
22 schema:keywords Underhill index
23 Underhill method
24 adequate imputation model
25 algorithm
26 alternative
27 bias
28 biased parameters
29 bird censuses
30 birdSTATs
31 birdSTATs tool
32 cases
33 census
34 complete dataset
35 conclusion
36 conditions
37 confidence intervals
38 data
39 datasets
40 ecology
41 field
42 field of ecology
43 hood
44 imputation
45 imputation methods
46 imputation model
47 incorrect management
48 increase
49 index
50 information
51 intervals
52 less information
53 management
54 method
55 missing values
56 missingness increases
57 model
58 multiple imputation
59 observations
60 paper
61 parameters
62 policy conclusions
63 population total
64 precision
65 presence
66 purpose
67 robust alternative
68 science
69 smaller confidence intervals
70 social sciences
71 standard techniques
72 statistical techniques
73 suboptimal conditions
74 technique
75 terms
76 terms of bias
77 tool
78 total
79 trim
80 underestimated variance
81 use
82 use of TRIM
83 values
84 variance
85 water bird censuses
86 wide confidence intervals
87 schema:name Working with population totals in the presence of missing data comparing imputation methods in terms of bias and precision
88 schema:pagination 603-615
89 schema:productId N2ffbf8b8766f4ea9b6ce61cd75023cac
90 N42d943c20a014f5b95d6009973f2c8ba
91 schema:sameAs https://app.dimensions.ai/details/publication/pub.1006321158
92 https://doi.org/10.1007/s10336-016-1404-9
93 schema:sdDatePublished 2022-01-01T18:39
94 schema:sdLicense https://scigraph.springernature.com/explorer/license/
95 schema:sdPublisher Nb903f8f6485a4547a82dffdd87fa8993
96 schema:url https://doi.org/10.1007/s10336-016-1404-9
97 sgo:license sg:explorer/license/
98 sgo:sdDataset articles
99 rdf:type schema:ScholarlyArticle
100 N2ffbf8b8766f4ea9b6ce61cd75023cac schema:name dimensions_id
101 schema:value pub.1006321158
102 rdf:type schema:PropertyValue
103 N31334c1359d547f6814c329517019286 rdf:first sg:person.012302577023.00
104 rdf:rest N61f0ab4d0af342d0bbc9c47352632eac
105 N3dc28cc4071e45989be62d9479526622 schema:issueNumber 2
106 rdf:type schema:PublicationIssue
107 N42d943c20a014f5b95d6009973f2c8ba schema:name doi
108 schema:value 10.1007/s10336-016-1404-9
109 rdf:type schema:PropertyValue
110 N58dceb2e0896429394b87983c7d5cc5b schema:volumeNumber 158
111 rdf:type schema:PublicationVolume
112 N5aa280f93727433f95c4a7899ed887b0 rdf:first sg:person.012650302401.91
113 rdf:rest N31334c1359d547f6814c329517019286
114 N61f0ab4d0af342d0bbc9c47352632eac rdf:first sg:person.0771550506.84
115 rdf:rest rdf:nil
116 Nb903f8f6485a4547a82dffdd87fa8993 schema:name Springer Nature - SN SciGraph project
117 rdf:type schema:Organization
118 anzsrc-for:05 schema:inDefinedTermSet anzsrc-for:
119 schema:name Environmental Sciences
120 rdf:type schema:DefinedTerm
121 anzsrc-for:0502 schema:inDefinedTermSet anzsrc-for:
122 schema:name Environmental Science and Management
123 rdf:type schema:DefinedTerm
124 anzsrc-for:06 schema:inDefinedTermSet anzsrc-for:
125 schema:name Biological Sciences
126 rdf:type schema:DefinedTerm
127 anzsrc-for:0608 schema:inDefinedTermSet anzsrc-for:
128 schema:name Zoology
129 rdf:type schema:DefinedTerm
130 sg:journal.1035137 schema:issn 2193-7192
131 2193-7206
132 schema:name Journal of Ornithology
133 schema:publisher Springer Nature
134 rdf:type schema:Periodical
135 sg:person.012302577023.00 schema:affiliation grid-institutes:grid.435417.0
136 schema:familyName Devos
137 schema:givenName Koen
138 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012302577023.00
139 rdf:type schema:Person
140 sg:person.012650302401.91 schema:affiliation grid-institutes:grid.435417.0
141 schema:familyName Onkelinx
142 schema:givenName Thierry
143 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012650302401.91
144 rdf:type schema:Person
145 sg:person.0771550506.84 schema:affiliation grid-institutes:grid.435417.0
146 schema:familyName Quataert
147 schema:givenName Paul
148 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0771550506.84
149 rdf:type schema:Person
150 sg:pub.10.1007/978-0-387-21706-2 schema:sameAs https://app.dimensions.ai/details/publication/pub.1035613449
151 https://doi.org/10.1007/978-0-387-21706-2
152 rdf:type schema:CreativeWork
153 sg:pub.10.1007/s00265-010-1044-7 schema:sameAs https://app.dimensions.ai/details/publication/pub.1022305626
154 https://doi.org/10.1007/s00265-010-1044-7
155 rdf:type schema:CreativeWork
156 sg:pub.10.1007/s10336-007-0176-7 schema:sameAs https://app.dimensions.ai/details/publication/pub.1041727822
157 https://doi.org/10.1007/s10336-007-0176-7
158 rdf:type schema:CreativeWork
159 sg:pub.10.1007/s10344-013-0768-x schema:sameAs https://app.dimensions.ai/details/publication/pub.1046979458
160 https://doi.org/10.1007/s10344-013-0768-x
161 rdf:type schema:CreativeWork
162 sg:pub.10.1007/s10841-004-1332-5 schema:sameAs https://app.dimensions.ai/details/publication/pub.1033148717
163 https://doi.org/10.1007/s10841-004-1332-5
164 rdf:type schema:CreativeWork
165 sg:pub.10.1007/s10841-014-9644-6 schema:sameAs https://app.dimensions.ai/details/publication/pub.1036623878
166 https://doi.org/10.1007/s10841-014-9644-6
167 rdf:type schema:CreativeWork
168 sg:pub.10.1007/s11121-007-0070-9 schema:sameAs https://app.dimensions.ai/details/publication/pub.1018783636
169 https://doi.org/10.1007/s11121-007-0070-9
170 rdf:type schema:CreativeWork
171 grid-institutes:grid.435417.0 schema:alternateName Research Institute for Nature and Forest (INBO), Kliniekstraat 25, 1070, Anderlecht, Belgium
172 schema:name Research Institute for Nature and Forest (INBO), Kliniekstraat 25, 1070, Anderlecht, Belgium
173 rdf:type schema:Organization
 




Preview window. Press ESC to close (or click here)


...