ASSIGNMENT

Course Name: EpidemilogyAnd Bio Statistics

Module 5: Statistical Analysis Concepts
1. Causal Pathways

Causal pathways are theoretical models that explain how certain factors (exposures) lead to
specific outcomes, typically in the context of health or social sciences. These pathways identify
and illustrate the sequence of events or mechanisms by which an exposure influences an
outcome.

 Components of Causal Pathways:

o Direct Effects: The immediate effect of an exposure on an outcome without

intermediaries.

o Indirect (Mediated) Effects: Effects that operate through one or more intermediary
variables, known as mediators. For example, smoking leads to lung damage
(mediator), which then leads to respiratory disease (outcome).

o Confounding Variables: External factors that may influence both the exposure and
outcome, potentially biasing the observed effect.

 Purpose of Causal Pathways:

o Understanding causal pathways is essential for developing interventions, as it

allows researchers to target specific points in the pathway.

o It also helps in adjusting for confounders and identifying potential mediators that
influence the exposure-outcome relationship.

Causal pathways are essential for public health research and clinical trials, as they help clarify
how and why certain factors lead to specific health outcomes.

2. Stratification Adjustment
Stratification adjustment is a statistical method used to control for confounding by dividing study
participants into strata or groups based on confounding variables (e.g., age, sex). This approach
ensures that comparisons within each stratum are more homogeneous.

 Purpose:

o By stratifying data, researchers can reduce the effect of confounding variables and
produce a more accurate estimate of the relationship between the exposure and the
outcome.

 Process:

o Participants are grouped based on levels of a confounding variable (e.g., age

groups) and then analyzed within each stratum.

o After stratification, results from each stratum are combined to get an adjusted
overall estimate, known as a weighted average.

 Example: In studying the effect of a drug on blood pressure, stratifying by age allows
researchers to control for age as a confounding variable, as blood pressure naturally
varies with age.

Stratification adjustment is valuable for studies where confounding is a concern, improving the
validity of the observed associations.

3. Propensity Analysis

Propensity analysis, often referred to as propensity score matching, is a statistical technique used
to reduce selection bias in observational studies by mimicking the effects of randomization. It
involves calculating the probability (propensity score) of each participant being assigned to a
specific treatment group based on observed characteristics.

 Purpose:

o Propensity scores help to create comparable groups by matching participants with

similar scores, reducing the impact of confounding variables.

 Steps:

o Calculate the propensity score for each participant, based on characteristics that
could influence both treatment selection and outcome.
 o Match participants with similar propensity scores from different groups (e.g.,
treated and untreated groups).

o Analyze the outcome between matched groups to estimate the treatment effect.

 Application: In medical studies, propensity analysis allows for a more accurate

comparison of treatment effects in non-randomized settings, such as retrospective studies.

Propensity analysis is useful when randomization is not feasible, allowing researchers to control
for confounding and make fairer comparisons.

4. Confidence Limits and Confidence Intervals

Confidence intervals (CIs) are a range of values, derived from the sample data, that are likely to
contain the true population parameter. Confidence limits are the lower and upper boundaries of
this interval.

 Interpretation:

o A 95% confidence interval means there is a 95% chance that the interval contains
the true parameter value if the study were repeated multiple times.

o Confidence Limits: These are the endpoints of the confidence interval (e.g., if the
CI for a treatment effect is 1.2 to 2.5, then 1.2 and 2.5 are the confidence limits).

 Importance:

o Confidence intervals provide insight into the precision and reliability of an

estimate. Narrow intervals suggest precise estimates, while wide intervals indicate
more variability or uncertainty.

 Calculation: CIs are calculated based on sample size, variability, and the desired
confidence level (e.g., 95% or 99%).

Confidence intervals and limits are critical in reporting statistical results, as they indicate both
the estimate’s value and its possible range.

5. Degrees of Freedom
Degrees of freedom (df) represent the number of values in a calculation that are free to vary. In
statistical analysis, they help to determine the shape of the sampling distribution and influence
the precision of estimates.

 Purpose in Statistics:

o Degrees of freedom are used in various statistical tests, such as the t-test and chi-
square test, to interpret the results accurately.

 Calculation:

o In a sample, degrees of freedom are typically calculated as the sample size minus
the number of estimated parameters. For example, in a t-test comparing means, df
= N - 1, where N is the sample size.

 Examples of Use:

o t-Tests: Degrees of freedom affect the critical value needed to determine

significance.

o Chi-Square Tests: In contingency tables, df depends on the number of rows and

columns.

o ANOVA: Each group and factor has its own degrees of freedom, which are used
to calculate the F-statistic.