1.5 The Large Data Set (STATISTICS AND MECHANICS 1- Chapter 1: Data collection)

Summary

This section of the Stats and Mechanics Year One book covers data collection using a large dataset focused on weather information from locations like Jacksonville, Beijing, Perth, and the UK. The dataset includes variables like temperature, rainfall, wind direction, and visibility, recorded between May and October in 1987 and 2015. The UK data also includes specific measurements such as gusts, humidity, cloud cover, and pressure. Various sampling methods and statistical techniques such as mean, median, and random sampling are explained. Additionally, the importance of sample size and the limitations of small samples in drawing conclusions is emphasized.

Takeaways

• 😀 The large data set consists of weather data collected from various locations during May and October in 1987 and 2015.
• 😀 Data includes temperature, rainfall, sunshine, wind speed, and wind direction measurements over 24-hour periods.
• 😀 The term 'tr' in the rainfall data represents a trace amount of rain (less than 0.05 millimeters).
• 😀 The Beaufort scale is used to measure wind speed, which can be described in terms like 'calm,' 'moderate,' or 'gale.'
• 😀 The data set covers locations such as Jacksonville (Florida), Beijing (China), and Perth (Australia), each with characteristic weather conditions.
• 😀 For the UK, additional data is provided, including daily gusts, humidity, cloud cover, visibility, and pressure.
• 😀 The UK’s weather includes detailed measurements like relative humidity and cloud cover, with units like knots and hectopascals.
• 😀 High pressure generally brings settled, dry weather, while low pressure brings unsettled, wet weather.
• 😀 Sample selection methods in the script include random sampling, where Alison generates random numbers to select data points.
• 😀 Missing or unavailable data is marked as 'N/A' in the data set, and this could affect the sample size for analysis.
• 😀 Alison's sampling method may not yield a full sample of five if randomly selected days contain missing data (marked as 'N/A').
• 😀 The weather in Perth is generally dry, and a small sample of data might not be sufficient to conclusively support or reject claims about weather patterns across different regions.

Q & A

What is the purpose of the large data set described in the script?

The large data set aims to provide weather-related information for various locations, recorded between May and October of 1987 and 2015, to assist in analyzing patterns and drawing conclusions about weather conditions.

What types of data are recorded in the large data set?

The data includes temperature (mean temperature over 24 hours), total rainfall, total sunshine, wind speed, wind direction, humidity, cloud cover, visibility, and pressure for various locations.

What is the significance of 'TR' in the data cells?

'TR' stands for 'trace,' meaning that the amount of rainfall recorded is less than 0.05 millimeters, which is considered too small to be significant.

How are wind speeds measured in the data set?

Wind speeds are measured using the Beaufort scale, which categorizes wind speed based on observable effects such as the movement of trees or the force of the wind.

What are some key locations included in the data set and their general weather conditions?

The data includes locations such as Jacksonville (Florida), Beijing (China), and Perth (Australia), with Jacksonville and Beijing experiencing hot and humid weather, and Perth experiencing hot and dry conditions.

What extra data is recorded specifically for the UK locations?

In addition to the common weather data, the UK locations include daily mean gust, humidity, cloud cover (measured in octaves), visibility (measured in decameters), and daily mean pressure (measured in hectopascals).

Why might Alison’s sampling method not generate a full sample of five?

Alison’s sampling method could fail to generate a full sample of five because two of the randomly selected dates could have 'NA' (data not available) entries, making it impossible to gather data for those dates.

What is the median daily rainfall for the 14th to 20th of June in the UK data set?

The median daily rainfall for the 14th to 20th of June in the UK data set is 0.1 millimeters, with one of the days having a 'trace' amount of rainfall.

How does the rainfall data in Perth compare to that of the UK for June, and what can be concluded?

The median daily rainfall in Perth for June is 19 millimeters, which is significantly higher than the UK data. While this might support the idea that southern countries experience more rainfall in June, the sample size is too small to make a definitive conclusion.

How does the script suggest weather patterns in the UK differ from those in Perth or Jacksonville?

The script suggests that UK weather is more variable, with cooler, wetter winters and warmer, drier summers, whereas Perth and Jacksonville tend to experience more consistent hot and dry or hot and humid conditions year-round.