# Stats

Get Started. It's Free
Stats

## 1. Continuous vs Categorical data

### 1.1. Categorical

1.1.1. Male vs Female

1.1.2. Red light, blue light, green light

1.1.3. Organic fertiliser vs artificial fertiliser

### 1.2. Continuous

1.2.1. Height, weight, pH, ppm, kPa

1.2.2. Could be at any point in a range of values

1.2.2.1. E.g. 2.15cm vs 2.18cm

### 1.3. Discrete or ordinal

1.3.1. Can only have a particular value- e.g. numbers

## 2. Independent variable - What you changed

### 2.1. Continuous e.g. temperature, pH

2.1.1. Dependent variable- what you measured?

2.1.1.1. Continuous - e.g. temperature, CO2 volume produced

2.1.1.1.1. Pearson's regression

2.1.1.1.2. Eg. measured with a ruler/probe/sensor

2.1.1.2. Discrete - e.g. counted

2.1.1.2.1. Counted? Frequencies? Measured without being a continuous?

### 2.2. Categoric - e.g. drug 1 vs drug 2 e.g. boy v girls or discrete

2.2.1. Dependent variable?

2.2.1.1. Continuous? E.g. pH/length

2.2.1.1.1. Comparing means of two groups?

2.2.1.1.2. Comparing means of more than two groups (3 or more)

2.2.1.1.3. Counted things where there isn't clear theory

2.2.1.2. Discrete/Ordinal data - e.g. number or count or frequency

2.2.1.2.1. Counted or frequency?

## 3. When to use error bars and what type

### 3.1. Is you data normal?

3.1.1. Ie would most of it likely fall into normal distribution

3.1.1.1. E.g. Weight, height, activity of enzyme at a particular pH etc

3.1.1.1.1. Standard deviation error bars

### 3.2. Is your data non-normal?

3.2.1. Things where you aren't sure of any theory behind it

3.2.1.1. Use a max-average, average minus min bar

3.2.1.1.1. Min-max bars

3.2.1.1.2. Max is 50, average is 25, max- average = 25. Positive error bar is 25

3.2.1.1.3. Min 22, average is 25m average - min = 3 so negative error bar = 3

3.2.2. Use range bars to show max and min

3.2.3. Distribution

3.2.4. Counted data