Lecture 5 - Data Visualisation

Question 1

What makes a good graph?

Answer 1

• induce’s the reader to think about the data being presented
• doesn’t distort data
• presents many numbers with minimum ink
• makes a large dataset coherent
• encourages reader to compare different pieces of data
• not in 3D
• good colour scheme

Question 2

What must a graph include?

Answer 2

• a scale that starts at 0
• good title
• axis labelled

Question 3

What does a normal distribution look like?

Answer 3

• aka as the Gaussian distribution or bell curve
• highest and lowest score are symmetrically distributed around the mean
• (look up graph)

Question 4

What happens if data is not normally distributed?

Answer 4

• can be positively skewed = looks like your right foot (look up)
• can be negatively skewed = left foot (look up)

Question 5

What are the 2 different shapes of distribution?

Answer 5

1. Leptokurtic (spiked, look up) = when majority of scores fall in the centre of distribution with very little scores away from the centre of distribution
2. Platikurtic (flat, look up) = scores are widely distributed around the mean = more flat and broader distribution

Question 6

What is bimodal distribution?

Answer 6

• when there is 2 modes
• look up a graph

Question 7

Why might data not be normally distributed?

Answer 7

• outliers
• insufficient data, sample size might be too small
  -> central limit theorem = the larger your sample size the more normally distributed your data will be
• multiple distributions; bimodal or multimodal
• measurement issues
• silly mistakes

Question 8

boxplot and violin plots?

Answer 8

• an outlier is a data point that is located outside the whiskers of the box plot
• a violin plot is when the boxplot ‘marries’ the histogram, it has a frequency distribution
• look up e.g.

Question 9

column/bar charts?

Answer 9

• bar graphs are rectangular bars represented horizontally
• column graphs are rectangular bars represented vertically
• use for categorical data where the length of the bar is proportional to the quantity of data
• useful to show categorical comparisons

Question 10

When should you use a histogram?

Answer 10

• for quantitative + continuous data e.g. frequencies/ scores

Question 11

What is a line graph?

Answer 11

• data points are plotted on a graph then connected by a line to show a trend over time

Question 12

what figure should you use for a categorical design?

Answer 12

• bar graph

Question 13

what figure should you use for an experimental design?

Answer 13

• use a bar chart if the IV is nominal
• use a line graph if the DV is nominal

Question 14

What figure should you use for a correlational design?

Answer 14

• scatter graph

Question 15

What do we mean by a figure?

Answer 15

• any visual display other than a table
• e.g. graphs, images of experiment set up, drawings

Question 15

What are the principles of figure creation?

Answer 15

• All figures should be numbered e.g., Figure 1 which is written in bold above the figure
• Below it, goes the title of that Figure in Italics
• Titles should be clear and informative
• Left alignment
• Any notes (optional) go underneath the Figure
• look at e.g

Question 16

When should you use a table?

Answer 16

• to present a lot of numerical information
• should be numbered independently from figures e.g. table 1
• below should be the title of the table in italics
• left alignment
• look at e.g.