Download presentation
Presentation is loading. Please wait.
Published byMaryann Turner Modified over 9 years ago
1
1 URBDP 591 A Analysis, Interpretation, and Synthesis -Assumptions of Progressive Synthesis -Principles of Progressive Synthesis -Components and Methods -Meta-Analysis
2
2 Causal Explanations (Ford 2000) A causal scientific explanation has four characteristics. (1)It defines causal and/or organizational processes that describe how systems function. (2) These are consistent – under the same conditions the causal and/or organizational processes will produce the same effect. (3) A causal scientific explanation provides general information about events of a similar kind. (4) When experiments are possible then a designed manipulation will produce a predictable response.
3
3 Progressive Synthesis (Ford 2000) Progressive synthesis is a methodology for scientific investigation the aknowledge the requirement to make upward inference about integrative ecology
4
4 Progressive Synthesis Philosophy (Ford 2000) The philosophy is “pragmatic realism”: Science aims to provide the best explanatory account of natural and social phenomena. Three principles: - Criticism must be applied to objectives, methods and results -Precision in definition is required to develop the coherence in breadth and detail of an explanation -Postulates must be assessed with explicit standards
5
5 Progressive Synthesis Components (Ford 2000)
6
6 Progressive Synthesis Philosophy (Ford 2000)
7
7 1.Criticism must be applied to objectives, methods and results Direct analysis Testing through direct repetition Refining through extended use Standpoint criticism
8
8 2. Postulates must be assessed with explicit standards Data statements Explanatory coherence Explaining the observed data Predicting future results Reveling results of a different kind
9
9 3.Precision in definition is required to develop the coherence of an explanation Functional classification of concepts: According to their knowledge status According to their role in a theory
10
10 Types of Explanations (Ford 2000) Causal Explanations Respond to questions that ask why or how something comes about Organizational explanations answer questions that ask what something is Unifying explanations link causal and organizational explanations in a coherent theory
11
11 Integrative Concepts (Ford 2000) Natural concepts define and/or classify measurable or observable entities or events in the ecological world. Functional concepts define properties of natural concepts or express relationships between two or more natural concepts. Integrative concepts can carry knowledge, defined by a theory, built in another (or many other) systems.
12
12 Integrative Concepts and their Domains (Ford 2000)
13
13 Integrative Concepts and their Domains (Ford 2000) The domain of an integrative concept has two components: - specification of the extent of what the concept seeks to explain - specification of the set of functional concepts that an integrative concept uses (Ford 2000 p289).
14
14 Benthic Index of Biotic Integrity (Karr 1985)
15
15 Scientific and Statistical Inference (Ford 2000)
16
16 Scientific Inference for Integrative Concepts (Ford 2000) We make scientific inference by assessing the coherence of a causal scientific explanation. This can be judged using six criteria (Ford 2000 p275-276): 1.Acceptability of individual propositions, including that they have been tested with data. 2. Concept definitions are consistent throughout the theory network. 3.Part and kind relationships are consistent throughout the theory network. 4.There are not ad hoc propositions that include special circumstances. 5.Generally, theories with fewer rather than more propositions are favored as explanations that are more coherent. 6.The explanation applies to broad questions and circumstances.
17
17 Meta-Analysis: Definition “The critical review and statistical combination of the results of previous research” “The structured and systematic integration of the results of separate independent studies: “Analysis of analyses”
18
18 A:Universe: All Conducted Studies C: Study Sample : All studies found and “included” according to determined criteria D: Data Analyzed E: Conclusions Correct Inferences in the Structured Review
19
19 Medline Searches A Medline search will reveal about 51% of published RCTs Goes down with experience of searcher -Chalmers I, Systematic Reviews, BMJ Publishing Group, 1998 p.27
20
20 The Cochrane Collaboration Aim: “To prepare, maintain, and promote the accessibility of systematic reviews in all areas of health care” Started 1974 by Archie Cochrane Today 4000 professionals participate The Cochrane Library Cochrane Database of Systematic Reviews, Database of Abstracts of Reviews of Effectivesness, The Controlled Trials Register, Cochrane Review Methodology Database See www.cochrane.orgwww.cochrane.org
21
21 Major goals of Meta-Analysis 1.Objective summaries 2.Increase power to detect true effects 3.Estimate effect Size 4.Resolve uncertainty If the studies produced dissimilar (heterogeneous) results, what is the explanation?
22
22 Ways to summarize data Get differences in percentages, grouping all people together Take averages of percentages in each group, compare groups “Head-Counting”: Count the direction of the results in the studies “Head-Counting” Statistical: Count the direction of the statistically significant results in each study Classic Meta-Analysis
23
23 Classic Meta-Analysis Analyzes RR, OR, or absolute differences in percentages between groups. Uses weighted averages of each percentage to get overall percentage in each group Uses the the inverse of the variance of the estimate provided by each participating trial for the weights. This gives a minimum variance unbiased estimate of the effect. Large trials carry more weight than small trials.
24
24 Meta-Analysis Issues All the rules get broken “Meta-analysis” often not done Retrospective study Publication Bias Heterogeneity
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.