🔬

1. Statistics

Topic

Basic Science

Podcast

001

Updated

Types of Data - Normal vs Non-Parametric
Box and Whiskers Graph/ Box plot
Sensitivity and Specificity, PPV, NPV/ Screening/ likelihood ratio
Receiver Operator Characteristics ROC Curve
Meta-Analyses: Forest Plot
Meta-Analyses: Vertical Funnel Plots
Horizontal Funnel Plots: Mortality and Revision Rates
Survival Curves
Reading RCTs/ Error
Designing RCTs: Power, Randomization, Bias, and Blinding
Outcome measures/ Grades of evidence
Interobserver Variability

Types of Data - Normal vs Non-Parametric

‣

What types of data are there?

‣

How do you analyze data?

‣

How do you determine if data is normally distributed?

‣

How are categorical data summarized and analyzed?

Normal/ Parametric Measurement Data

‣

What does this image show?

‣

Why is this considered a normal distribution?

‣

What is the significance of a bell-shaped curve?

‣

What are the measures of central tendency? [Descriptive Stats]

‣

What are the measures for dispersion? [Descriptive Stats]

‣

What are measures used in inferential statistics?

‣

What are standard deviations?

‣

What is the purpose of a statistical test?

‣

What tests are used for normal (parametric) data?

‣

How do you test correlation for normal data?

Non-Parametric Measurement Data

‣

What kind of data distribution is shown in the image? Which side is it skewed to? What do the different colored lines represent?

‣

What statistical tests are used for non-parametric data?

‣

How do you test correlation for non-parametric data?

‣

Can you transform non-parametric data to normal (parametric) data?

[Back to Top]

Box and Whiskers Graph/ Box plot

‣

What is this graph?

‣

What do the lines represent?

‣

What is an outlier?

‣

What does the size of the box indicate?

‣

Why might the median not be in the middle of the box?

[Back to Top]

Sensitivity and Specificity, PPV, NPV/ Screening/ likelihood ratio

‣

What are sensitivity, specificity, PPV, and NPV?

‣

Definitions of PPV, NPV, Sensitivity, and Specificity

‣

Which is more relevant to the patient: Sensitivity or PPV?

‣

How can you increase sensitivity or specificity?

‣

What kind of test will you choose to "diagnose as many cases as possible"?

‣

What kind of test will you choose to "confirm a diagnosis"?

‣

What is the likelihood ratio?

‣

What is accuracy?

‣

What makes a successful screening test? (e.g., for DDH, Scoliosis)

‣

Example: Is ultrasound screening for DDH a good screening test?

‣

What biases are screening tests susceptible to?

[Back to Top]

Receiver Operator Characteristics ROC Curve

‣

What is this curve?

‣

How is the ROC plotted?

‣

What are the X and Y axes?

‣

What is the area under the curve?

‣

Based on this curve, which test is good for screening and which for confirmation?

‣

How can you improve the test?

Meta-Analyses: Forest Plot

‣

What is this?

‣

What is MH?

‣

What influences the weight?

‣

What is a confidence interval?

‣

What is homogeneity/heterogeneity? How is it tested?

‣

How to address heterogeneity?

‣

What's the difference between a systematic review and a meta-analysis?

[Back to Top]

Meta-Analyses: Vertical Funnel Plots

‣

What is this?

‣

What is standard error?

‣

What does asymmetry show?

‣

What are funnel plots used for?

[Back to Top]

Horizontal Funnel Plots: Mortality and Revision Rates

‣

What is this? [Mortality Rate]

‣

What is this? [Revision Rate]

‣

How are the curves risk-adjusted?

‣

What are the terms for outliers and those within the normal range?

‣

What are the criticisms of such analysis?

[Back to Top]

Survival Curves

‣

What is this?

‣

What are some assumptions in this curve? (3)

‣

How do you determine if there is a significant difference between the groups?

‣

What is the downside of this analysis?

‣

What statistical test do you use to compare these curves?

‣

Can you derive mean revision time or rate from this curve?

‣

Reading RCTs/ Error

‣

What is a confidence interval?

‣

How does CI differ from P-values? What is a P-value?

‣

Do you prefer to know the CI or the P-value?

‣

What is the null hypothesis?

‣

What is a type 1 (alpha) error? How can it be reduced?

‣

What is a type 2 (beta) error? How can it be reduced?

‣

What is a type 3 (gamma) error?

‣

What are risk ratio (relative risk) and absolute risk reduction?

‣

What is the Number Needed to Treat (NNT)?

‣

What is an Odds Ratio?

‣

What can you interpret from the PROSPECTIVE study result?

‣

If it is RETROSPECTIVE, what can you interpret from here?

[Back to Top]

Designing RCTs: Power, Randomization, Bias, and Blinding

‣

How do you design a Randomized Controlled Trial?

‣

How do you conduct a literature review before designing an RCT?

‣

What are the levels of evidence?

‣

What characterizes a good Randomized Controlled Trial?

‣

What are the advantages of randomization?

‣

How do you implement randomization?

‣

Types of blinding

‣

How to calculate sample size and power

‣

What is bias?

‣

Types of Bias and How to Address Them

‣

How to eliminate bias?

‣

What is a confounder?

‣

How to address confounders?

‣

What are key considerations for inclusion and exclusion criteria?

‣

What are the approaches to analysis of RCTs?

‣

Which type of analysis is considered the gold standard for RCTs? Why?

[Back to Top]

Outcome measures/ Grades of evidence

‣

What are the types of outcome measures?

‣

What is a validated outcome measure?

‣

How do you evaluate outcome measures?

‣

Examples of Outcome Scores

‣

What are grades A, B, C recommendations?

[Back to Top]

Interobserver Variability

‣

How to measure interobserver consistency or variability? What tests are used?

[Back to Top]

📖

A false balance is an abomination to the Lord, but a just weight is his delight. Proverbs 11:1