Lecture 4 (Sep 27) Flashcards
Misclassify
Very common error in health research
Information (measurement) error leads to
misclassification
Non-differential classification (the same in all study groups)
Usually weakens associations – i.e. brings effect estimates (RR, OR, AR) closer to the null value
But not always…
May have misclassified 1.4 % of them but the vast majority was not misclassified
Differential (different in different study groups)
Effect estimates may change in any direction, depending on the particular error
Misclassified to much of the data to a point of having no idea whether you should
Ex. Low Birthweight mothers a biased to assume that it is their fault that and the over represent the amount of pesticides they were around
Causal Inference
The process of determining whether a cause-and-effect relationship exists between two variables. It aims to answer questions like, “Does X cause Y?” For example, in medicine, researchers might ask, “Does a new drug cause better health outcomes?”
Building Blocks:
Measures of Disease Frequency, Various Study Designs
Results of Research:
Measures of Association
Once you have calculated a measure of association, you need to determine if the observed association is
valid and if it is causal
Research Evidence Strong evidence is
Strong evidence is
- Of the lowest possible random sampling error (a statistically significant exposure/outcome association)
- Based on a good design
Free of selection and information biases
Under minimal influence of confounding (next session)
Internal validity
refers to the extent to which a study or experiment accurately establishes a causal relationship between the treatment (or independent variable) and the observed outcome (or dependent variable), without being affected by other confounding variables or biases. In simpler terms, it addresses whether the effects observed in a study can be confidently attributed to the intervention or treatment itself, rather than to external factors or flaws in the research design.
Generalizability
also known as external validity, refers to the extent to which the findings of a study can be applied or generalized beyond the specific conditions of the research. In other words, it addresses whether the results of a study hold true across different populations, settings, time periods, or variations in the treatment. Requires internal validity.
If a study lacks internal validity, external validity
is irrelevant
We do not compromise internal validity in an effort to
achieve external validity (generalizability)
Internal validity is when
when the effect estimated from the analytic sample is equal to the true causal effect in the study sample
External validity is when
the true causal effect in the study sample is equal to the true causal effect in the target population.
Four Hallmarks of Health Studies
- A research question/plausible theory
- A well thought design to address the research question
- Measurement of exposure and outcome
- Analysis to compare groups (measured association)
Validity is…
Having fewer errors
Error=Measured value-True value
Sources of error (CBC):
Chance (random sampling error)
Bias Systematic error in selection of participants and/or measurement
Confounding (next week)
Threats to Validity
Chance (random sampling error)
Bias Systematic error in selection of participants and/or measurement
Confounding (next week)
Random Sampling Error is not to be confused with
1) Random error in measurement
Information bias (previous session)
2) Randomization (a process in experimental studies)
Nov 8, 2024
Type I error:
jumping the gun
Concluding that there is a treatment effect, or an exposure/outcome association (in the population from which you sampled), when there is NOT
Type II error:
missing the boat
Failing to detect a REAL treatment effect or an exposure/outcome association (in the population from which you sampled)
Bias
Bias refers to a systematic error in the design or conduct of a study
When bias occurs in a study the observed association between the exposure and outcome will be different from the true association
Most biases relate to the study design and procedures and can be classified into two categories:
Selection bias
Information bias (due to measurement error, last week)
Information Bias vs Selection Bias:
Information bias occurs when there is a systematic error in the way data on the variables (e.g., exposure or outcome) is measured, collected, or classified in a study. This can lead to incorrect or misleading information being used in the analysis, potentially affecting the results.
Selection Bias: Selection bias occurs when the participants included in a study (or excluded from it) are not representative of the target population. This can happen during the selection or retention of participants and leads to systematic differences between those who are studied and those who are not.
Why does Selection Bias Happen?
NOT an Error Associated with Sampling, recall week 2
Types of Selection Bias
Inappropriate Control Selection (Control-Selection Bias) > Case control
Differential Participation > Case control, cohort
Differential Loss to Follow Up > Cohort Experimental
