2: types of studies1. 2 in chapter 2: 2.1 surveys 2.2 comparative studies
Post on 21-Dec-2015
215 views
TRANSCRIPT
2: Types of Studies 3
Types of Studies• SurveysSurveys: describe population characteristics
§2.1
• Comparative studies:Comparative studies: determine relationships between variables §2.2
2: Types of Studies 4
2.1 Surveys• Goal: describe population characteristics
• Normally based on a sample (or subset) of the population
• Types of samples– Probability sample (e.g., simple random
sample)– Non-probability sample (e.g., convenience
sample)
2: Types of Studies 6
Sampling• Probability samples use chance
mechanisms to select individuals
• The most basic type of probability sample is the simple random sample (SRS)
• SRS ≡ a sample of size n selected so that all possible combinations of n individuals from the population are equally likely to comprise the sample
2: Types of Studies 7
SRS Method1. Identify each population member with the
numbers 1, 2, . . ., N
2. Pick a random spot to enter Table A3. Select n tuples, discarding inappropriate tuples
Alternatively, use a random number generator (e.g., www.random.org) to generate n random numbers between 1 and N
Keep in mind: the objective of an SRS: every possible subset is equally likely!
2: Types of Studies 9
Other Types of Probability Samples (More Advanced Methods)
• Stratified random sample: randomly sample strata (subsets) with the population
• Cluster sample: randomly sample clusters comprising varying numbers of observations
• Multistage sampling: randomly sample random samples in multiple stages
2: Types of Studies 10
Cautions Undercoverage: groups in the source
population are left out or underrepresented sampling frame
Volunteer bias: self-selected participants (volunteers) are atypical population members
Nonresponse bias: a large percentage of individuals refuse to participate or cannot be contacted
2: Types of Studies 11
§2.2 Comparative Studies
• Objective: to determine the relationship between an explanatory and response variable
• Experimental designs assign subjects to treatment or exposure groups and compare outcomes
• Nonexperimental design classified subjects in exposure groups (no intervention) and compare outcomes
2: Types of Studies 13
Women’s Health Initiative (WHI) Experimental Design
The WHI study randomly assigned about half its subjects to a group that received hormone replacement therapy (HRT). The other half received an identical looking placebo.
Subjects were followed for ~5 years to ascertain various health outcomes, including heart attacks, strokes, breast cancer and so on.Example of publication: Writing Group for the Women's Health Initiative Investigators. (2002). Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women's Health Initiative randomized controlled trial. JAMA, 288(3), 321-333.
2: Types of Studies 14
Nurse’s Health Study (NHS) Nonexperimental Design
This study classified individuals according to whether they received HRT
Subjects were followed (ongoing) for various health outcomes, e.g., heart attacks, strokes, breast cancer and so on.
Example of publication: Stampfer, M., Colditz, G., Willett, W., Manson, J., Rosner, B., Speizer, F., et al. (1991). Postmenopausal estrogen therapy and cardiovascular disease. Ten-year follow-up from the nurses' health study. New England Journal of Medicine, 325(11), 756-762.
2: Types of Studies 15
Comparison of WHI and NHS
• Both the WHI and NHS analyzed relationships between HRT (explanatory variable) and various health outcomes (response variables)
• WHI investigators assigned the exposure (HRT) experimental
• NHS investigators measured the exposure but did not assign it nonexperimental (observational)
2: Types of Studies 16
Let us focus on selected experimental design concepts
and techniquesExperimental designs provides a
paradigm for nonexperimental designs.
2: Types of Studies 17
Terminology Used in Experiments
• Subject ≡ an individual participating in the experiment
• Factor ≡ an explanatory variable being studied in an experiment
• Treatment ≡ a specific set of factors
2: Types of Studies 19
• Subjects = 100 individuals who participated in the study• Factor A = Health education (active, passive)• Factor B = Medication (Rx A, Rx B, or placebo)• Treatments = the six specific combinations of factor A
and factor B
Subjects, Factors, Treatments, Example, cont.
2: Types of Studies 21
Randomized Controlled Trials (RCT)
• Randomized: chance assigns the exposure
• Controlled: the exposed group is compared to a non-exposed group (comparative study design)
• Blinded: observations are made without knowledge subjects’ exposure status
2: Types of Studies 22
Randomization• Randomization is the second principle of
experimentation• Randomization refers to the use of chance
mechanisms to assign exposures / treatments
• Randomization balances lurking variables among comparison groups, mitigating confounding effects of lurking variables
2: Types of Studies 23
Randomization - ExampleSource: JAMA 1994;271:595-600 (link is active)• Explanatory variable: Nicotine or placebo patch• 60 subjects (30 in each group)• Response variable / outcome: Cessation of
smoking (yes/no)
Random Assignment
Group 130 smokers
Treatment 1 Nicotine Patch
CompareCessation
ratesGroup 2
30 smokers
Treatment 2 Placebo Patch
2: Types of Studies 24
Randomization – Example
• Number subjects 01,…,60
• Use Table A (or a random number generator) to select 30 two-tuples between 01 and 60
• If you use Table A, arbitrarily select a different starting point each time
• For example, if we start in line 19, we see 04247 38798 73286
2: Types of Studies 25
Randomization, cont.
• We identify random two-tuples, e.g., 04, 24, 73, 87, etc.
• Random two-tuples greater than 60 are ignored• The first three individuals in the treatment group are
01, 24, and 29• Keep selecting random two-tuples until you identify
30 unique individuals • The remaining subjects are assigned to the control
group
2: Types of Studies 26
Controlled Trial
• “Controlled” means there is non-exposed control group
• A concurrent control group is essential • You cannot judge the effects of an expose
without a concurrent control group because:– Many factors contribute to a health outcome
(so-called multi-causality; causal interdependence)
– Conditions change on their own over time– The placebo (and related) effects
2: Types of Studies 27
Blinding• Blinding: the measurement of the response
made without knowledge of treatment type• Works by balancing measurement error and
preventing observer bias• Can occur at several levels:
– Single blinding: subjects blinded– Double blinding: subjects and measurement-
takers blinded– Triple blinding: subjects, measurement-takers,
analysts blinded
2: Types of Studies 28
Ethics of Experimental Studies
• Equipoise - balanced doubt; can’t knowingly expose subjects to harm or withhold benefit
• Informed consent – subjects must be aware of study objectives
• Beneficence – must provide overall benefit to society
• Justice – must be administrated justly• Independent oversight (e.g., IRB)