-
Notifications
You must be signed in to change notification settings - Fork 0
/
scoping study summary.Rmd
97 lines (53 loc) · 8.02 KB
/
scoping study summary.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
---
title: "Scoping study summary"
output: pdf_document
knit: (function(inputFile, encoding) {
rmarkdown::render(inputFile, encoding = encoding, output_dir = "../man") })
---
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```
Meeting time: 24.03.2022
Present
- Frode Singsaas
- Matt Grainger
- Anders Kolstad
# Research question
We started by formulating a primary research question. We decided on a PO type question (ref), with two levels of outcome (O1 and O2).
Then we worked to identify related terms and to constrict the search so that we get an acceptable level of precision.
```{r question, echo=F, out.width="100%", fig.cap="The primary research question for the Re-scalable Indicator Review and how the consepts link to the search terms."}
knitr::include_graphics("../figures/researchQuestion.png")
```
## litserachr
Matt searched web of science (WoS) using `ecosystem condition assessment indicator` and downloaded the first 1000 hits as a RIS file which he then imported into [`litseachr`](https://elizagrames.shinyapps.io/litsearchr/).
The app gives a list of words or phrases that typically occur together within the title, abstract and author defined key words.
It then condense this list to identify key words that you could use in order to obtain 25, 50, or 80% of the papers.
The most condensed list (25%) contained around 170 phrases.
Based on this we updated Figure 1 with some additional search terms.
Single terms were restricted when needed, using the NEAR statement. Peripheral terms are excluded from the [initial search string](data/initial search terms.docs), but kept for later sensitivity analyses (crossed though word in FIgure 1).
# Initial search
The initial search string yielded 49 577 hits (WoS; 24.03.2022):
`(ecosystem* OR ecologic* OR environment* OR natur* OR biologic*) AND (condition OR state NEAR/3 (ecosystem* OR ecologic* OR environment* OR natur* OR biologic*) OR capital NEAR/2 (natur*) OR integrity) AND (indicator* OR bioindicator* OR ((index* OR indec*) NEAR/2 (ecosyst* OR ecologic* OR environment* OR natur* OR biologic* OR condition* OR integrity)) OR assessment* NEAR/1 (ecosyst* OR ecologic* OR environment* OR natur* OR condition* OR integrity) OR account* NEAR/1 (ecosystem* OR ecologic* OR environment* OR natur* OR biologic*))`
We then excluded all but the following article types:
- Articles or Reviews
- Early access
- Proceedings Papers
- Book chapters
`[...] AND (DT==("ARTICLE" OR "REVIEW" OR "EARLY ACCESS" OR "PROCEEDINGS PAPER" OR "BOOK CHAPTER")`
This reduced the number of hits to 49 348 hits
We then removed references that were tagged with WoS categories belonging to clearly non-relevant resaerch fields:
https://www.webofscience.com/wos/woscc/summary/b3531dbb-90e4-4816-b8b5-58b79372e762-2c35adc2/relevance/1
`[...]... NOT TASCA==("PUBLIC ENVIRONMENTAL OCCUPATIONAL HEALTH" OR "GREEN SUSTAINABLE SCIENCE TECHNOLOGY" OR "ENGINEERING CIVIL" OR "COMPUTER SCIENCE INFORMATION SYSTEMS" OR "ENDOCRINOLOGY METABOLISM" OR "MINERALOGY" OR "PSYCHOLOGY CLINICAL" OR "EDUCATION EDUCATIONAL RESEARCH" OR "HEALTH CARE SCIENCES SERVICES" OR "NUTRITION DIETETICS" OR "ASTRONOMY ASTROPHYSICS" OR "MEDICINE RESEARCH EXPERIMENTAL" OR "NUCLEAR SCIENCE TECHNOLOGY" OR "CLINICAL NEUROLOGY" OR "TELECOMMUNICATIONS" OR "IMMUNOLOGY" OR "NANOSCIENCE NANOTECHNOLOGY" OR "BUSINESS FINANCE" OR "SPORT SCIENCES" OR "METALLURGY METALLURGICAL ENGINEERING" OR "ONCOLOGY" OR "ELECTROCHEMISTRY" OR "SPECTROSCOPY" OR "ENGINEERING PETROLEUM" OR "OPTICS" OR "HEALTH POLICY SERVICES" OR "DEVELOPMENT STUDIES" OR "RADIOLOGY NUCLEAR MEDICINE MEDICAL IMAGING" OR "SOCIAL SCIENCES BIOMEDICAL" OR "DENTISTRY ORAL SURGERY MEDICINE" OR "TROPICAL MEDICINE" OR "ENGINEERING OCEAN" OR "ENGINEERING BIOMEDICAL" OR "PEDIATRICS" OR "NURSING" OR "PSYCHOLOGY EXPERIMENTAL" OR "REHABILITATION" OR "ENGINEERING MANUFACTURING" OR "PSYCHOLOGY APPLIED" OR "SURGERY" OR "ERGONOMICS" OR "PHYSICS MULTIDISCIPLINARY" OR "PSYCHOLOGY BIOLOGICAL" OR "REPRODUCTIVE BIOLOGY" OR "ENGINEERING AEROSPACE" OR "LAW" OR "MEDICAL INFORMATICS" OR "GERONTOLOGY" OR "MATHEMATICS APPLIED" OR "CHEMISTRY INORGANIC NUCLEAR" OR "PHYSICS ATOMIC MOLECULAR CHEMICAL" OR "CHEMISTRY MEDICINAL" OR "MATERIALS SCIENCE COMPOSITES" OR "MEDICINE LEGAL" OR "AREA STUDIES" OR "OBSTETRICS GYNECOLOGY" OR "PUBLIC ADMINISTRATION" OR "ARCHITECTURE" OR "PSYCHOLOGY DEVELOPMENTAL" OR "ORTHOPEDICS" OR "AUDIOLOGY SPEECH LANGUAGE PATHOLOGY" OR "COMPUTER SCIENCE THEORY METHODS" OR "MATERIALS SCIENCE BIOMATERIALS" OR "CHEMISTRY ORGANIC" OR "COMPUTER SCIENCE SOFTWARE ENGINEERING" OR "MATHEMATICS" OR "PATHOLOGY" OR "PHYSICS MATHEMATICAL" OR "RESPIRATORY SYSTEM" OR "GASTROENTEROLOGY HEPATOLOGY" OR "HEMATOLOGY" OR "PHYSICS CONDENSED MATTER" OR "DEMOGRAPHY" OR "DERMATOLOGY" OR "MEDICAL LABORATORY TECHNOLOGY" OR "CARDIAC CARDIOVASCULAR SYSTEMS" OR "OTORHINOLARYNGOLOGY" OR "HISTORY PHILOSOPHY OF SCIENCE" OR "DEVELOPMENTAL BIOLOGY" OR "PHYSICS FLUIDS PLASMAS" OR "SOCIAL WORK" OR "INFORMATION SCIENCE LIBRARY SCIENCE" OR "OPHTHALMOLOGY" OR "ALLERGY" OR "CRITICAL CARE MEDICINE" OR "PSYCHOLOGY EDUCATIONAL" OR "ACOUSTICS" OR "ENGINEERING MARINE" OR "INTERNATIONAL RELATIONS" OR "MATERIALS SCIENCE CHARACTERIZATION TESTING" OR "POLYMER SCIENCE" OR "SUBSTANCE ABUSE" OR "AUTOMATION CONTROL SYSTEMS" OR "HOSPITALITY LEISURE SPORT TOURISM" OR "MATHEMATICAL COMPUTATIONAL BIOLOGY" OR "POLITICAL SCIENCE" OR "GERIATRICS GERONTOLOGY" OR "INTEGRATIVE COMPLEMENTARY MEDICINE" OR "MATERIALS SCIENCE PAPER WOOD" OR "ART" OR "COMPUTER SCIENCE HARDWARE ARCHITECTURE" OR "MATERIALS SCIENCE TEXTILES" OR "UROLOGY NEPHROLOGY" OR "SOCIAL ISSUES" OR "PERIPHERAL VASCULAR DISEASE" OR "PSYCHOLOGY SOCIAL"))`
This reduced the number of hits to 39 099.
We coul not find a way to increase precision further without compromising.
As a next step, Matt will do topic modelling to classify the references into research fields based on natural language processing. We might then be able to excluded more references.
# Scope
40 000 references is too much to go through and we need to find a way to cut this down further, even after the topic modelling has done its part. Here are some of the thoughts we have had around that.
1. We can stratify our review to only consider terrestrial ecoystems. Most of the references belong to either the freshwater or marine realms, so this would reduce the number of references considerably, and we might be able to process all the references. We call this the **biased but complete** approach. However, there is no good way to do this filtering in the search. For example, using the `NOT` statement is not recommended. But, we should be able to judge such an exclusion criteria quite rapidly in a title or abstract screening.
1. We can randomly select a managemen number of references from the complete list, as it stands now, covering all the realms. We call this the **Un-biased but incomplete** approach, because we would not be able to process all the references.
1. We have not been able to include concepts such as `reference state/condition` or `rescaled` in our serach string. The largest part of the litterature will be presenting variables in their raw units, and not as re-scaled indicators, so by having these concepts in the search terms we would constrict the output considerably. However, these are concepts that are rarely mentioned in the title or abstract.
1. We cold also consider stratifying based in geographhy (e.g. Europe only). Same as for theo other options, this is not possible at the search stage, but can be implemented at the title and abstract screening stage.
**Conclusion and recomendation:**
We think that option 1 (**biased but complete**) is the best. Knowledg on indicators of ecological condition in terrestrial ecosystems are, presumably, lagging behing compared to freshwater and marine ecosystems, due to the longer history and tradition for ecosystem monitoring in these relams (e.g. WFD and MSFD). Another issue with option 2 is that our project team doesn't include any aquatic ecologists.
# Citation chasing
We will do forward and backwards citation chasing on a subset of the references which will be decided on later.