L9 - Data Analysis 1

What does biological research rely on?

• Biological research relies on accurate and precise measurements of various biological parameters e.g., length, mass, concentration, time, or genetic sequence. 

What is the scientific process?

The scientific process is:

1. Observe

2. Question

3. Hypothesise

4. Test (collect data)

5. Analyse data

6. Conclude

What is data in the scientific process?

Data are the evidence at the core of this cycle: they are what let us test a hypothesis rigorously, and they are what let us trust the conclusions we draw from an investigation.

What are categorical variables?

Categorical variables are factors that sort data into groups or levels. For example species or treatment group.

What are continuous variables?

Continuous variables are variables where quantities are measured or counted on a numeric scale, allowing for calculation. For example, heigh in cm or cell count.

What is a statistical observation?

A statistical observation is a record of a single individual at a single point in time, that we can collect information about. For example, persons, specimens, or tissues and cells etc.

What are statistical variables?

Statistical variables are variables that we can measure or categorise in an observation. For example, height, weight etc (measurement) or sex, nationality, and education level etc. (category).

What is the statistical term for a variable that is measured or counted?

The statistical term for a variable that is measured or counted is covariate. Covariates are quantitative variables consisting of number values.

What may continuous covariates have values of?

Continuous covariates may have values of (in principle) unlimited decimal degradation. In practice, the actual numbers recorded are limited by measuring equipment and/or sensible decimal rounding.

What do discrete covariates have values of?

Discrete covariates have only integer values, often representing counts. E.g, a patch of grassland may have 10 or 11 species, but not 10.5.

What is the statistical term for a categorical variable?

The statistical term for a categorical variable is factor. A factor is a qualitative variable consisting of two or more categories.

Categories are known as levels of factor, what are these?

Categories are known as levels of factor. These are:

• Nominal factors, which consist of levels with no inherent hierarchy (they are simple different, e.g frog, newt, and toad).

• Ordinal factors, which have levels that are ordered categories (They cannot be arranged into any order, e.g. none, light, moderate, and severe).

What is the mean?

The mean is the sum of all the values divided by the number of values.

What is the median?

The median is the middle value when data are sorted (or the average of the middle two).

What do measures of central tendency aim to do?

The measures of central tendency aim to summarise data points as a ‘typical’ middle value. A central value for each group allows us to quantify the size difference, generally, between groups

What information may mean and median omit, using the example of height?

There can be two groups of people with the same mean height, but different degrees of variation around the mean.

What is the range?

The range is the difference between the highest and lowest values. Range = max - min

What is standard deviation?

Standard deviation is the typical distance of each value from the mean - how spread out the data are.

Why is it valuable to visualise data?

It is valuable to visualise data because:

• It reveals patterns. These can be trends that are hard to see in a table of numbers.

• It flags outliers. Unusual values stand out visually, prompting a closer look.

• It enables visual comparison. Groups, treatments, or time points can be compared at a glance.

• It communicates results. A clear figure can be easier to interpret than a paragraph of text.

What is the best graph for comparing categories?

The best graph for comparing categories is a bar chart.

What is the best graph for distribution of one numerical variable?

The best graph for distribution of one numerical variable is a histogram.

What is the best graph for showing a relationship/trend between two continuous variables?

The best graph for showing a relationship/trend between two continuous variables is a histogram.

What is the best graph for comparing spread across groups?

The best graph for comparing spread across groups is a box plot.

What is the best graph for showing change over time?

The best graph for showing change over time is a line graph.