Introduction to graphical data representation

In psychological research, raw data refers to the unprocessed information collected from participants, such as scores or measurements. This data, along with measures of central tendency like means or medians, can be presented graphically to make patterns and relationships easier to understand. Graphs help researchers visualise differences, distributions, or associations, which supports clearer analysis and interpretation. Choosing the right graph depends on the type of data and the research question.

Key principles for creating effective graphs

• Full title - Every graph needs a descriptive title that summarises what it shows, including the variables involved.
• Axis labelling - The horizontal x-axis typically shows the independent variable or categories, while the vertical y-axis displays the dependent variable or frequency. Both should include units where relevant.
• Scale and clarity - Use appropriate scales to avoid distortion, and ensure the graph is easy to read without overcrowding.

Bar charts for categorical data

Bar charts are graphical tools used when data falls into distinct categories, such as different experimental groups or conditions. They are particularly useful for displaying differences in measures of central tendency, like the mean, between these categories. This helps researchers compare outcomes across groups, revealing how variables might influence behaviour or performance.

When to use bar charts

Bar charts suit nominal data, which consists of categories without a numerical order, or when summarising group averages. For example, they can illustrate how an independent variable, like the presence of a stimulus, affects a dependent variable, such as recall accuracy.

Structure of a bar chart

• X-axis (horizontal) - Labels the categories or conditions, such as different experimental groups.
• Y-axis (vertical) - Shows the scale or frequency, often the mean value of the dependent variable.
• Bars - Represent the values for each category; they are separated by spaces to emphasise the categorical nature.

Example of a bar chart in psychology

Consider a study measuring word recall in two groups: one with background noise and one without. The bar chart would have the x-axis labelling "Noise" and "No noise", while the y-axis would show the mean number of words correctly recalled out of 20.

Histograms for continuous data distributions

Histograms are used to represent data on a continuous scale, where values can fall anywhere along a numerical range without distinct categories. They display the distribution of scores, showing how frequently values occur within intervals. This is valuable in psychology for understanding patterns in variables like reaction times or test scores, as it reveals the shape of the data, such as whether it is normally distributed or skewed.

When to use histograms

Choose histograms for interval or ratio data, which involves measurable quantities that can be divided into ranges. They help identify central tendencies and variability in a dataset, aiding in the analysis of group performance or individual differences.

Structure of a histogram

• X-axis (horizontal) - Divides the continuous variable into intervals, such as time ranges in seconds.
• Y-axis (vertical) - Indicates the frequency, or how many observations fall into each interval.
• Bars - Touch each other to show the continuous nature of the data, with heights representing frequency.

Example of a histogram in psychology

Imagine a study timing students running 100 metres. The x-axis would show the time taken (in seconds), and the y-axis would show the number of students.

Scatter graphs for correlations

Scatter graphs, also known as scatterplots, illustrate the relationship between two continuous variables, often to explore correlations. A correlation refers to the extent to which changes in one variable are associated with changes in another, which could be positive (both increase together), negative (one increases as the other decreases), or none. In psychology, this helps investigate links, such as between sleep and cognitive performance, without implying causation.

When to use scatter graphs

Use them for co-variables, which are two related but non-manipulated variables measured to assess association. They are ideal for correlational research designs, where experiments are not feasible.

Structure of a scatter graph

• X-axis (horizontal) - Represents one co-variable, scaled appropriately.
• Y-axis (vertical) - Represents the second co-variable.
• Data points - Each point marks the intersection of values for a pair of observations; a line of best fit can be added to show the trend.

Example of a scatter graph in psychology

Suppose a study examines the link between hours of sleep deprivation and words recalled from a list. The x-axis would show the number of hours deprived of sleep, and the y-axis would show the number of words correctly recalled.