Understanding Data Representation: Graphs and Charts

This cheat sheet covers various methods of representing data, focusing on statistical graphs and charts. It highlights how to construct and interpret these visuals, while also cautioning against common pitfalls that can lead to misleading representations.

Core Principles

• Data visualization is crucial for understanding patterns and trends.
• Different types of data (qualitative vs. quantitative) require different graphical methods.
• Accurate representation ensures data is not misleading.
• Understanding the components of graphs (axes, labels, scales) is key to interpretation.
• Be aware of potential biases and errors in data collection and presentation.

Formulas

• $Relative Frequency = \frac{Frequency}{Total}$

Key Terms

• Population: The entire group that we want to collect data from.
• Individual: Each specific object or person in the population.
• Variable: The information that we want to learn from the population.
• Data: The observation of the variable for each individual.
• Sample: A small group selected from the population, ideally representative.
• Statistic: A calculation made from sample data.
• Parameter: A calculation made from population or census data.
• Descriptive Statistics: Organizing and summarizing collected data.
• Inferential Statistics: Drawing conclusions or making predictions based on descriptive statistics.
• Qualitative Variable: Represents qualities or categories (e.g., color, name).
• Quantitative Variable: Numerical data that can be counted or measured (e.g., height, age).
• Frequency: The number of times a characteristic or category is observed.
• Relative Frequency: The frequency of a category in relation to the total number of observations.
• Sampling Error: Variation between samples; an expected and unavoidable error.
• Non-sampling Error: Errors due to flaws in study design or execution; leads to bias.
• Sampling Bias: Bias resulting from how the sample is chosen (e.g., incomplete frame).
• Nonresponse Bias: Bias occurring when individuals do not respond to a survey.
• Response Bias: Bias occurring due to how individuals respond (e.g., lying, pollster pressure).

Real World Examples

• A professor calculating the average exam score for a class.: Identifies population (class), individuals (students), variable (score), and data (individual scores).
• A city surveying residents about their commute.: Distinguishes between census (all residents) and sample (subset), and parameter (city-wide average) vs. statistic (sample average).
• Categorizing favorite novel genres.: Demonstrates frequency and relative frequency calculations and visualization using bar graphs and pie charts.
• Tracking office copier paper refills daily.: Illustrates organizing quantitative data using frequency tables, dot plots, and histograms.
• A study comparing soda vs. juice preference.: Highlights how manipulating the vertical axis scale can create misleading bar graphs.

Timeline

• Present: Cross-sectional studies collect data at one point in time.
• Past: Case-control (retrospective) studies use existing historical data.
• Future: Cohort (prospective/longitudinal) studies follow subjects over time.