Biostatistics - Acads – HASSLE MABUHAY · •Pictogram / Cartogram •Cluster Diagram . Horizontal Bar Graph . Vertical Bar Graph . Histogram . Component Bar Graph . Grouped Bar

Biostatistics

Quantitative Qualitative

Observational Studies Descriptive

Case Study/Report Case-Series

Analytical Ecological Study Cross-sectional Study Longitudinal Study

Case Control (Retrospective) Nested Case Control Cohort (Prospective) Historical Cohort

Descriptive Qualitative Case Study Case Report/Series

Traditional Qualitative

Historical research Phenomenological research Grounded theory research Ethnographical research

Meta-synthesis

Experimental Studies Pre-Experimental Quasi-experimental True Experimental

Meta-analysis

Correlational Studies

Multi-Method /Mixed Method

Component Design Integrated Design

QUANTITATIVE Pre-experimental

Uncontrolled Experimental Study

(Single Group Post Test Only)

Self Controlled Experimental Study

(Within Group Pre-Post Test)

Quasi-experimental

Controlled Study

• Between Groups Non-equivalent Post Test Only

• Between Groups Non-equivalent Pre-Post Test

Trials with External Controls

Factorial Design

True experimental Randomized Controlled Trial (RCT)

Completely Randomized Design (CRD)

Randomized Complete Block Design (RCBD)

Onset of Study

Time

Sequential Control (Crossover Design)

X X X X

Intervention

X X X X Washout

period

Experimental subjects

Controls

Pre Clinical Trial

Phases of a Full Clinical Trial

• Phase I: finalizes the treatment (e.g., to determine things like drug dose and safety)

• Phase II: seeks preliminary evidence of effectiveness

• Phase III: fully tests the treatment (randomized clinical trial or RCT)

• Phase IV: focuses on long-term consequences of the treatment

LEVELS OF MEASUREMENT

DESCRIPTIVE STATISTICS

• Descriptive Measures

• Tabular Presentation

• Graphical Presentation

• Textual Presentation

Descriptive Measures

• Measures of the Middle (Central Tendency)

• Measures of Variability (Dispersion)

• Measures of Position

• Measures of Association

• Measures of Ratios and Proportions

• Measures of Disease Frequency and Association

Central Tendency • Mean

– Arithmetic Mean

– Weighted Mean

• Median

• Mode

Variability • Range = (Highest – Lowest)

• Variance = Mean of Squared Deviations

• Standard Deviation = Square Root of Variance

• Coefficient of Variation = standard deviation divided by mean times 100

• Kurtosis

• Skewness

KURTOSIS

Position • Percentile

• Deciles

• Quartile

• Z score (Standard score)

Association Numerical & Numerical Pearson’s Product Moment Correlation Coefficient

Numerical & Binary Point-Biserial Correlation Coefficient

Ordinal and Ordinal Spearman’s Rank Correlation Coefficient

Kendall’s Tau Rank Correlation

Ordinal and binary Rank Biserial Correlation

Binary and Binary Phi Coefficient

Strength of Linear Association

0

1

0.5

0.4

0.8

weak relationship (0.01 – 0.39)

strong relationship (0.80 – 0.99)

moderate relationship (0.40 – 0.79)

perfect linear relationship

no linear relationship

Positive Correlation

Negative Correlation

r = 0.95 r = - 0.95

Ratios and Proportions • Proportions

• Percentage

• Ratios

• Rates

Disease Frequency • Frequency of Occurrence

• Incidence Rate

• Prevalence Rate

• Risk Ratios

Descriptive Measures

• Central Tendency: Mean, Median, Mode

• Dispersion: Variance, Standard Deviation, Coefficient of Variation, Kurtosis, Skewness

• Position: Percentiles, Deciles, Quartiles, Z Score

• Association: Correlation Coefficients

• Ratios and Proportions

• Disease Frequency: Incidence and Prevalence

tables

• Simple Frequency Tables • Contingency Table / Cross-Tabulation • Dummy Tables • Master Tables • Summary Tables

Graphical Presentation

• Bar Graphs

• Line Graphs

• Scatter/Dot Plots

• Pie/Circle Graphs

• Pictogram / Cartogram

• Cluster Diagram

Horizontal Bar Graph

Vertical Bar Graph

Histogram

Component Bar Graph

Grouped Bar Graph

Range Bar Graph

Box and Whisker Plot

Waterfall Chart, Flying Bricks Chart, Mario Chart or Bridge Chart

Line Graph

SHAPES OF DISTRIBUTION

• Symmetric (Normally distributed)

SHAPES OF DISTRIBUTION

• Asymmetric (skewed) Distributions

Area Chart

Dot Plot

Scatter Plot

Pie Chart

Exploded Pie Chart

Polar Area Chart

Radar Chart, Web chart, Spider chart, Star chart, Cobweb chart, Star plot, Irregular polygon, Polar chart,

or Kiviat diagram

Nightingale’s Rose/Coxcomb Chart

Spie Chart

Multi-level Pie, Radial tree, or Ring chart

3-D Pie

Doughnut Chart

Pictogram

Cartogram

Timeline

Stem and Leaf Plot

Cluster Diagrams

• Network

• Flow chart

• Organizational chart

• Tree map

• Fan chart

Network Diagram

Flowchart

Organizational Chart

Tree Diagram / Dendrogram

Fan Chart / Geneology Chart

Inferential Statistics 1. Student’s T-Test

2. Chi-squared Test

3. ANOVA

T Tests

• Small sample test

• With sample size <30

• For large samples use Z Test

• Assumptions include

– Random sampling

– Sufficient sample size (do sample size estimation)

– Normal distribution of data

T Tests and Alternatives

One sample T Test Sign Test

Paired (Dependent) sample T Test

Wilcoxon Signed Rank Test or Mann-Whitney U Test

Two (Independent) samples T Test

Wilcoxon Rank Sum Test

Parametric Tests Non Parametric Tests

Chi Square Test

Types - • One Random Sample: Test of Goodness of Fit • One Random Sample, Two Group Comparison: Test of independence • Two Random Samples: Test of Homogeneity

Alternatives to Chi square Test

• Merge Columns or Rows

• Fisher’s Exact Test (applicable only to 2X2 table, more than 20%

less than 5 and with zero)

• McNemar’s Test (Dependent samples)

One way ANOVA Compare two or more groups

Example: Compare the waist-hip ratio among sedentary, semi-active and active people

Require Post Hoc Tests when significant

• Example: Tukey’s HSD

Alternatives • One-way ANOVA : Kruskal-Wallis

• Dependent Samples: Friedman

Correlational Statistics

• Correlation Coefficients

• Regression Analysis

– Correlation Coefficients + Prediction

• Note: only establishes associations (functional relationships)

• Limited by data of sample

Multivariate Statistics * Many Groups of Data * Many Variables

1. Inferential Methods: ANOVA & ANCOVA

2. Regression Methods

3. Classification Methods

4. True Multivariate Methods:

MANOVA & Canonical Correlation

5. Meta-analysis

ANOVA Multiple Independent Single Dependent

ANCOVA Single or Multiple Independent Single Dependent Co-variables (Confounders)

Linear Regression

Independent Dependent

Many

Numerical or

Categorical

One

Numerical

Multiple Regression

One or Many

Numerical or

Categorical

One

Binary

Binary Logistic/Logit Regression

One or Many

Numerical or

Categorical

One

Categorical

Multinomial Logistic Regression

One or Many

Categorical

One

Categorical

Log Linear Analysis

Multivariate Statistics

Classification Methods

• DISCRIMINANT FUNCTION ANALYSIS

• Factor Analysis (example: Principal Component Analysis)

• Cluster Analysis

• CLASSIFICATION AND REGRESSION TREE ANALYSIS (CART)

MULTIVARIATE ANALYSIS OF VARIANCE

• Simply called as “ANOVA with many dependent variables”

CANONICAL CORRELATION ANALYSIS

• the correlation of two canonical (latent) variables, one representing a set of independent variables, the other a set of dependent variables

META-ANALYSIS

• using statistical procedures to combine the results from different studies.

CORRELATIONAL

STATISTICS

INFERENTIAL

STATISTICS

MULTIVARIATE

STATISTICS