Describing data

Describing data is a fundamental skill in scientific investigation, focused on making results easier to understand and enabling the identification of trends or patterns. While quantitative data is initially recorded in organized tables, describing it involves extracting and communicating key insights from these tables or their visual representations, such as graphs.

• State the Overall Trend: Begin by outlining the general relationship or pattern between the variables presented. This provides an immediate understanding of what the data suggests. For example, if observing the number of adult males who smoked and the male lung cancer mortality rate, you would note that both decreased between 1990 and 2012.

• Include Precise Details and Quantities: Supplement the general trend with specific numerical values from the data, including their units. This adds precision and evidence to your description. For instance, you might state that "the number of new cases of asthma in the UK fell between 1996 and 2000, from 87 to 62 per 100,000 people".

• Describe Changes in Gradient or Key Points: If the data is displayed graphically, highlight any significant changes in the curve's steepness or direction, quoting the coordinates (both x and y values) at these points.

• Highlight Patterns and Relationships: Clearly articulate any observed patterns or links within the data. For example, describing that "the data shows that diabetes is more prevalent in inactive than active people, regardless of BMI".

• Make Calculations from Data: Descriptions can be enhanced by performing relevant calculations, such as percentage change, to quantify differences or trends. It is very important to show every step of your calculation clearly.

• Distinguish Between Correlation and Causation: A crucial aspect of describing data in biology is to recognize correlations but be cautious about inferring causal relationships. A correlation indicates a link between two things, but not necessarily that one causes the other; there could be other influencing factors. Unless there is substantial evidence, it's best to state correlation rather than causation.

• Avoid Temporal Language (unless time is a variable): When describing graphs where time is not represented on an axis, refrain from using words that imply speed or change over time, such as 'faster,' 'slower,' or 'rapidly'. Instead, use terms like 'steep decrease' or 'gentle decrease'.

• Address Anomalous Results: If anomalous (unusual or outlier) results are present, they should be identified and potentially ignored when describing overall trends or calculating means, especially if a reason for their deviation is known.

• Explain Results Using Scientific Knowledge: Beyond merely stating what the data shows, you may be asked to explain why the observed relationships exist by drawing upon your scientific knowledge and understanding.

By following these guidelines, the description of data provides a clear, precise, and scientifically sound foundation for drawing conclusions and further analysis.