scientific method and health research Flashcards
Scientific method - rules for:
how knowledge should be acquired
the form in which knowledge should be stated
how the truth or falsity of the knowledge should be evaluated
other methods used: Authority
Other methods used:
Authority – knowledge considered true because of tradition, or said by experienced or distinguished person
eg. Linus Pauling – vitamin C
can lead to contradictions/conflicts
Rationalism
. Rationalism – reasoning/logic
But: eg. a) all persons with heart disease are males
b) person x has heart disease
c) therefore person x is male
c only true if a and b are true
Need way to establish truth of premises (a and b)
- Intuition
. Intuition – insights without conscious reasoning (“Eureka!” experience)
sometimes valid/useful
sometimes proven false under empirical test
new scientific discoveries sometimes resisted when counterintuitive
eg. Semmelweiss – 1848 introduced antiseptic procedures in his wards
- colleagues offended that physicians considered carriers of disease
The Scientific Method and the Positivist View
Scientific method based on 3 elements:
Scientific method based on 3 elements:
Skepticism – any proposition or statement open to doubt and analysis
Determinism – events occur according to natural laws and causes (not magical, not haphazard)
Empiricism – enquiry should be through observation and verified through experience (ie. tested)
Developed over several centuries – modern western science traced to 16th century
Philosophers- “positivism|”, “repductionism”
Observation, description and measurement
Observations – descriptions of phenomena from precise, unbiased recording
verbal descriptions (eg. questionnaires)
sets of measurements (instruments)
key attributes: accuracy, replicability
need to be appropriately summarized
When confirmed by others form bases of scientific knowledge
Generalization and induction
Induction – logic underlying scientific generalizations
assertions about a phenomena based on a limited number of observations
eg. penicillin cured pneumonia in “n” patients
penicillin cures pneumonia (in all patients)
Hypotheses
statements that specify the relationship between two or more observations
eg. giving penicillin vs observations or measurements of pneumonia
terms must have clear-cut, observable referents
eg. Saturated fat in the diet raises bloof cholesterol
when have strong empirical support may be called laws
Theories
integration of hypotheses into more general explanatory systems
eg. heliocentric theory of solar system
DNA theory of inheritance
clarify relationships between diverse observations and hypotheses
eg. diets high in saturated fat are risk factor for heart disease
physiology - increase sat. fat -> increase blood cholesterol
epidemiology - increase blood cholesterol -> increase heart disease
provide conceptual means for predicting and influencing events
eg. Change heart diseases by changing diet
Deduction
theories should lead to verifiable statements or hypotheses
hypotheses can be deduced from relationships postulated by theories
Eg. Decrease dietary sat. fat -> decrease stroke
Controlled observation
hypotheses need to be tested under controlled conditions to exclude competing hypotheses that might explain the predicted phenomenon
conversely need to show the specificity of the hypothesis to the phenomenon (ie. doesn’t lead to host of unpredicted phenomena)
NOTE: in nutrition, control is oftern very difficult
Verification and falsification
after collecting evidence, do findings support hypothesis?
If yes, the theory from which the hypothesis was deduced is strengthened, or verified
If no, the theory is falsified, and becomes less useful (alternative: experiment questioned)
- usually later discarded in favor of new more powerful theories
often advances in instrumentation/measurement challenge existing theories and/or spur creation of new theories
Scientific theoriries not absolute truths, but provisional explanations of available evidence
Controversies on the Nature of the Scientific Method
scientific method is set of rules/conventions devised by philosophers and scientists
- also open to criticism, debate and change
The theory-dependence of observation
observer has pre-conceived notions or theories
may be selective in what is recorded as evidence
observations are “theory-dosed”
specify which observations are important
may ignore other aspects
sometimes used to “expertly” interpret evidence
eg. Identify signal from noise in an ECG trace
The validity of induction
questions the validity of generalizations based on limited evidence
information may emerge that challenges generalization
eg. increase dietary fat -> increase heart disease
- now known that not all fats behave this way
What constitutes falsification
theories in practice not so readily modified or discarded by scientists
often inner core of propositions protected by outer belt of auxiliary, modifiable hypotheses
eg. germ theory of infectious disease – predicts that antibiotic will cure bacterial infections
- but find when give penicillin to patient – no improvement
- auxiliary hypothesis development of penicillion- resistant bacteria
- methodological controversy – logical basis for discarding one theory and accepting its rival?
Science and the Cultural Context
Science as a human activity
scientific enquiry influenced by social and interpersonal conditions/environment
formulation of hypotheses and theories are creative acts, not automatic application of induction
influenced by culture of scientist
The social context of health care research
Medical model – illness is lesion or dysfunction in body
- patient = locus
Biopsychosocial model – holistic approach
increasingly used due to:
increase roles of other health professionals (nurses, physiotherapists, nutritionists)
increase cost of health care ->:
increase focus on prevention
improve cost effectiveness
self-help and community health initiatives
Quantitative and Qualitative Methods
increase holistic approach brings new methodologies
- patient/client not just biological system, but needs to make sense of and cope with health care problems
medical model – requires reductionist or quantitative approach to research and knowledge
holistic model
– brings in qualitative or interpretive approach
eg. cancer patient
quantitative – measure variables, apply therapy, measure effectiveness by physical outcomes/survival
qualitative – considers patient’s view, family/social setting
radical therapy might extend life but quality of life and give other burdens
- in clinical practice and research, may want to integrate evidence from quantitative and qualitative enquiries
The Research Process
planning hypotheses or aims research design data collection organissation and presentation of Data Data Analysis Interpretation and conclusion
Planning:
consider previous research findings, ethical and economic factors
define hypothesis or aim
select strategy to answer questions or test hypotheses
Research design
choose sampling method
Ensure controlled observation
Data collection (quantitative and/or qualitative)
get the data
Organization and presentation of the data
descriptive statistics – summarize and describe data
table, graph, condense, communicate findings
Data analysis
apply principles of probability to calculate confidence and test hypotheses
inferential statistics statistics - determine if data support hypotheses
Interpretation and utilization of research
data support/question existing theories or practices?, suggest new?
findings rarely definitive, often suggest need for further research
Publication – journals, conferences
need to pass methodological/scientific evaluation
become part of knowledge
conventions for preparing/presenting
Qualitative research may use more flexible sequence
empirical data not analyzed through statistics but through use of language and analysis of narratives