biostats

Graph Type: one numerical variable

Histogram

Graph Type: categorical data

Bar Graph

Graph Type: two numerical variables

Scatterplot

Graph Type: data over time

Line Graph

Stat Test: one categorical variable

Chi-Square GOF

Stat Test: two categorical variables

Chi-Square TOI

Stat Test: categorical (two groups) + numerical variable

Two-Sample t-Test

Stat Test: paired numerical data

Paired t-Test

Stat Test: two numerical variables

Correlation

Stat Test: predicting one numerical variable from another

Regression

Graph that correlates to: Chi-Square GOF

Bar Graph

Graph that correlates to: Chi-Square TOI

Grouped Bar Graph

Graph that correlates to: Two-Sample t-Test

Box Plot

Graph that correlates to: Correlation

Scatterplot

Graph that correlates to: Linear Regression

Scatterplot w/ Trendline

Graph that correlates to: Paired t-Test

Box Plots

Assumptions for Chi-Square Tests

1. Individuals in the data set are a random sample from the whole population

2. Samples are independent of each other

3. None of the categories should have an expected frequency < 1

4. No more than 20% of the categories should have expected frequencies < 5

Assumptions for t-Test

1. Variances/SDs of the two groups are about the same

• use Welch’s t-test if SDs are unequal

2. Normal distribution

• Transformation of data or Mann-Whitney U Test if not normal

What are the hypothesis testing steps?

1. State hypotheses

2. Compute test statistic

3. Determine p-value

4. Draw conclusions