Tables and Graphs - L4

What are the 8 guidelines for producing good tables?

• Clear title

• Well labelled columns or rows

• Includes units of measurement

• Include notes of explanation if needed

• Include a source of the data

• Be well laid out and presented

• Have consistent use of a decimal place within a variable

• Include relevant totals/subtotals that add up

What are percentage tables useful for?

Data where the categories have different sample sizes

What are 3 reasons to use graphs?

• They make it easier to visualise patterns and trends

• It is quicker for the reader to see the data

• It is an effective way of highlighting particular aspects of the findings

What are the 3 types of graphs to use for categorical data?

• Bar charts

• Stacked bar charts

• Pie charts

What are key features of a bar chart (including multiple bar charts)?

• The bars are rectangular, of the same width, with the height proportional to frequency or percentage

• Bars should not touch, but can touch in multiple bar charts if the bars correspond to the same x-axis category

• The bars should be ordered by logical order or by the height of the bars

• Try not to include more than 3 sets of bars on a multiple bar chart

What do stacked bar charts show?

Categories within categories. It can be displayed as a count or as a percentage (must add up to 100%)

How do you convert a percentage into a proportion for a bar chart?

Divide by 100 then multiply by 360

What are the limitations of pie charts?

• Not recommended when the data has many categories

• Can be confusing when used to compare outcomes of 2 different surveys or experiments

• Can be misleading if they are based on a small number of observations

• Avoid 3D pie charts as they distort the relative proportions of each category in the chart

What are the 3 graphs used for continuous data?

• Stem and leaf plots

• Histograms

• Scatterplots

What are the positives of a stem and leaf plot?

• They make frequency distribution more visual

• They give an idea of the shape of the distribution

• They can be used to identify distribution skew and outliers

• They display all of the data

How do stem and leaf plots work?

The column on the left is the stem, representing the tens of the data values and the numbers on the right are the leaf, representing the digits of the data values. Stems can have multiple digits but leaves only have 1.

What is the limitation of stem and leaf plots?

They are time consuming with large samples

How do histograms work?

The bar height represents frequency or percentage and the bar width represents the category width. All of the categories are continuous, so the bars should touch

What is a limitation of histograms?

They only show counts or percentages falling within a category, not all of the data

What should you do when there are multiple categories of continuous data?

• Categories should be on the x-axis

• Make sure the scales aren’t misleading

• Distinguish different categories with colours or markers

• Use a column chart or a boxplot

• Add error bars or individual values

How do you represent a time series graphically?

• Use a line graph for data on a variable over time

• Time should be on the x-axis

• Distinguish different lines with different colours or markers

• Include error bars of standard deviation or individual lines or standard error

How do you calculate standard error?

standard deviation/sqrt number of data values

What happens to standard error as sample size increases?

It decreases

What is standard error used for?

To show confidence in a calculation of the mean

What do scatterplots show?

2 continuous variables and their relationship

What are the 7 guidelines for graphs?

• The graph should not distort the data

• It should not contain unnecessary adornments (decorations)

• Has a scale for each axis

• y-axis should begin at 0, unless it is justified

• All axes should be properly labelled

• Contains a title

• The simplest possible graph should be used for each data set