Research Block I

Front 1

Steps of the scientific method (8)

Back 1

1. Observation
2. Literature review
3. Define a question
4. Hypothesis
5. Perform experiment, collect data
6. Analyze data
7. Conclusions/ interpret data
8. Communicate rsults

Front 2

Steps to evidence-based medicine practice (5)

Back 2

1. Ask answerable question
2. Search
3. Critically evaluate evidence
4. Integrate appraisal with clinical expertise
5. Evaluate and improve personal effectiveness

Front 3

Clinical question: PICO

Back 3

o Patient/population/problem: age, gender, ethnicity, disease
o Intervention: exposure, risk behaviors, treatment, diagnosis
o Comparison: no disease, absence of risk, placebo, dif drug, control
o Outcome: risk of disease, diagnosis accuracy, occurrence, adverse outcomes
o For (population), what is the effect of (intervention) compared with (comparison) on (outcome).

Front 4

Levels of evidence

Back 4

1. Randomized control trials
2. Non randomized controlled trial
3. Observational study with controls
4. Observational study without controls

Front 5

pyramid thing

Back 5

1. cochrane, systematic reviews
2. other SRs and meta-analyses
3. evidence guidlines
4. evidence summaries
5. rcts case cohorts, control studies
6. clincial research critiques
7. other reviews of the literature
8. case reports, case series, practice guidelines
9. clinical reference texts

Front 6

limitations of the scientific method

Back 6

o Addresses only what is observable
o Limited by measuring instruments
o Cause/effect but not purpose
o Only repeated events
o No morality

Front 7

Sources of error

Back 7

o Assumptions
o Failure to trust data not consistent with hypothesis
o Failure to quantitatively estimate systematic errors
o Sampling errors
o Confusing correlation with causation
o Bias

Front 8

Bias

Back 8

o Confirmation bias
o Sampling bias
o Bias in data gathering
o Processing bias: cargo-cult bias

Front 9

Purpose of a research problem

Back 9

o Implies possibility of empirical investigation
o Identifies need for research
o Provides focus
o Provides overview of research
o General overview of the problem

Front 10

Criteria for Research question

Back 10

• F - Feasible
• I - Interesting
• N - Novel
• E - Ethical
• R - Relevant

Front 11

Strong Hypothesis (3)

Back 11

• Insight into research question
• Testable and measurable
• Logical

Front 12

Causes of error (7)

Back 12

• Small sample size
• Poor planning
• Bias
• Not following protocol
• Lack of randomization
• Small effect
• High variance

Front 13

Controlling variance (5)

Back 13

• Sample size
• Quality of study
• Blocking
• Protocols
• Preliminary experiments

Front 14

Controlling bias (3)

Back 14

• Double-blind
• Controls
• Randomization

Front 15

Minimizing type I error (1)

Back 15

• p-value

Front 16

Minimize type II error (4)

Back 16

• Sample size
• Effect size
• Reduction of variance
• Reduction of bias

Front 17

Good study is…(4)

Back 17

• Double blind
• Randomized
• Large n
• Protocol

Front 18

Power is affected by... (4)

Back 18

• Sample size
• Variance
• Alpha
• Effect size

Front 19

What to do if you have too few subjects (5)

Back 19

• Quit
• Less stringent demands: increase alpha or beta
• Change hypothesis
• Re-evaluate effect size
• Reduce variability

Front 20

Reducing variability (8)

Back 20

• Use more at-risk subjects
• Lengthen follow up
• More precise equipment
• Beter data collection
• Continuous variable
• Different DV
• Composite variable
• pairing

Front 21

Steps of study design (3)

Back 21

• Choices
• Controlling variance
• Controlling bias
• steps
• Formulate hypothesis
• Design experiment
• Choose a test measure
• Summarize findings

Front 22

Principles of design (3)

Back 22

• Randomization
• Replication
• Control of variability

Front 23

alpha

Back 23

type I error
1-specificity
false-positive

Front 24

• Beta

Back 24

• Beta: type II error: 1-power; false negative

Front 25

• Blocking

Back 25

• Blocking: form groups of similar subjects prior to randomization and randomize each block

Front 26

• Cargo cult bias:

Back 26

• Cargo cult bias: refusal to accept results contrary to your hypothesis

Front 27

• Categorical

Back 27

• Categorical: measured and assigned to gropus on the basis of specific characteristics

Front 28

• Conceptual definition:

Back 28

• Conceptual definition: use of words to define a variable

Front 29

• Confounding variable

Back 29

• Confounding variable: vary systematically with the independent variable and exert influence on the dependent variable; can't be controlled for

Front 30

• Continuous variable

Back 30

• Continuous variable: measured on a scale that theoretically can take on an infinite number of values

Front 31

• Effect size:

Back 31

• Effect size: magnitude of the relationship between the IV and the DV

Front 32

• Evidence-based medicine:

Back 32

• Evidence-based medicine: conscientious, explicitn, and judicious use of current best evidence in making use of current best evidence in making decisions about the care of individual patients

Front 33

• Extraneous variables

Back 33

• Extraneous variables: affect the dependent variable but are not controlled for adequately by the researcher (but could be)

Front 34

• Fact:

Back 34

• Fact: observation that has been repeatedly confirmed

Front 35

• Goal:

Back 35

• Goal: vision of what research seeks to accomplish in the long run

Front 36

• Hypothesis

Back 36

• Hypothesis: tentative explanation of observations; a testable statement that can be used to build more complex inferences and explanations

Front 37

• Meta-analysis problem

Back 37

• Meta-analysis problem: difficult to properly combine numerous studies into one analysis

Front 38

• Methodology

Back 38

• Methodology: techniques for making observations about causal relationships

Front 39

• Multiplicity

Back 39

• Multiplicity: if multiple tests are run, eventually one will be significant just due to chance

Front 40

• Null hypothesis:

Back 40

• Null hypothesis: purposes a default position that there is no relationship

Front 41

• Objective

Back 41

• Objective: more immediate outcome of the proposed studies

Front 42

• Ontology

Back 42

• Ontology: controlled vocabulary with terms arrange from most general to most sepcific; premises about the deep causal structures from which analysis begins

Front 43

• Operational definition

Back 43

• Operational definition: specification of the manner in which a variable will be measured, categorized, or controlled

Front 44

• p-value

Back 44

• p-value: percent chance of obtaining observed or more extreme result if null is true

Front 45

• Research question

Back 45

• Research question: a question that identifies the problem to be studied

Front 46

• Research problem

Back 46

• Research problem: a set of conditions needing discussion, a solution, and information

Front 47

• Taxonomy

Back 47

• Taxonomy: techniques for assigning definitions according to pre-determined criteria

Front 48

• Test statistic

Back 48

• Test statistic: summary measure that evaluates the ypotheses

Front 49

• Theories

Back 49

• Theories: set of statements or principles devised to explain a group of facts or phenomena especially one that has been repeatedly tested or is widely accepted and can be used to make predictions about natural phenomena

Front 50

• Type I error

Back 50

• Type I error: false positive; rejecting a true null hypothesis

Front 51

• Type II error

Back 51

• Type II error: fail to reject a false null hypothesis

Front 52

• Variables

Back 52

• Variables: represent a concept of characteristic that varies