UNIT 5 : Data Literacy -Data Collection to Data Analysis

1. Data Literacy

• Definition:
  Data literacy refers to the ability to locate, understand, analyze, and use data effectively. It includes a range of skills from collecting and organizing data to interpreting results and applying insights ethically. It is essential in the age of AI, where raw data is transformed into actionable insights for various purposes.
• Importance of Data Literacy:
  • Critical Thinking: It enhances critical thinking by allowing students to distinguish between different types of data and their reliability.
  • Decision Making: Data-driven decision-making is key in fields like education, healthcare, business, and government. Being data literate enables individuals to make informed decisions based on solid evidence.
  • AI and Machine Learning: As AI and machine learning (ML) are heavily data-driven, understanding data is crucial for anyone involved in these fields. AI depends on converting large datasets into usable knowledge.
• Skills in Data Literacy:
  • Data Collection: Gathering relevant and accurate data.
  • Data Organization: Structuring data in meaningful ways (tables, charts, etc.).
  • Data Analysis: Using statistical methods or AI tools to interpret the data.
  • Data Ethics: Ensuring that data is used responsibly and ethically.
• Example Question: “Can you categorize the information you see online, in books, and from friends? Is all of this information the same?”
  • This primes students to think about data as information in different forms and how it can be used for various purposes.

2. Data Collection

• Definition:
  Data collection refers to the process of gathering information from multiple sources for analysis, prediction, or further use. It is the foundational step in any AI or machine learning project.
• Importance:
  Collecting accurate and relevant data is critical for creating predictive models in AI. High volumes of data are often required to develop reliable algorithms, especially in complex projects such as medical AI.
• Primary vs. Secondary Data:
  • Primary Data: Collected specifically for a particular purpose. This can be gathered through:
    • Surveys: Collecting opinions or feedback through questionnaires.
      Example: A researcher uses a questionnaire to understand customer preferences for a new product.
    • Interviews: Direct communication with individuals or groups to gather information.
      Example: An organization conducts interviews to collect employee feedback on job satisfaction.
    • Observation: Watching and recording behaviors as they occur naturally.
      Example: Observing children’s play patterns in a schoolyard to understand social dynamics.
    • Experiments: Manipulating variables to observe outcomes and establish cause-effect relationships.
      Example: Testing the effectiveness of two different advertising campaigns on consumers.
  • Secondary Data: Data that has already been collected by others and is available for reuse. This includes:
    • Books, Journals, News Articles: Information already compiled and analyzed.
    • Web Scraping: Using automated tools to extract data from websites (e.g., scraping product prices from an e-commerce site).
    • Social Media Tracking: Gathering and analyzing user behavior on social platforms.
      Example: Analyzing social media comments to understand public opinion on a new product.
    • Precompiled Datasets: Databases like Kaggle that offer ready-made datasets for analysis.
• Key Considerations in Data Collection:
  • Diversity: Ensure the data collected is diverse enough to cover various scenarios the AI model might encounter. For example, in training a robot to sort recyclable materials, the data should include many types of materials to improve the robot’s performance.
  • Volume of Data: The amount of data needed depends on the complexity of the model. Simple models like license plate detection require less data, whereas advanced AI systems in healthcare demand vast amounts of data.
  • Question for Students: “Think about your favorite movie recommendation platform. How do you think they use data to suggest movies you might like?”
  • This helps students connect the concept of data collection with a familiar real-world application, such as Netflix or YouTube.

3. Exploring Data

• Definition:
  Data exploration is the process of understanding the data, identifying patterns, and cleaning it before detailed analysis. This involves getting familiar with the values in the data and understanding whether they are typical, extreme, or require correction.
• Levels of Measurement:
  • Nominal: Categories with no inherent order, such as colors or car brands.
    • Example: The color of a student’s eyes or the model of a smartphone.
  • Ordinal: Categories with a specific order, but the differences between them cannot be measured.
    • Example: Restaurant ratings like “unpalatable,” “just okay,” “tasty,” and “delicious.”
  • Interval: Ordered data where the differences between values are meaningful, but there is no true zero.
    • Example: Temperature measured in Celsius or Fahrenheit. A 20-degree difference is meaningful, but 0 degrees does not represent “no temperature.”
  • Ratio: Ordered data with a true zero point, allowing for meaningful ratios between values.
    • Example: Weight or exam scores. A score of 80 is four times greater than a score of 20.
• Key Statistical Concepts:
  • Mean: The average of a dataset.
  • Median: The middle value of a dataset, useful for skewed data.
  • Mode: The most frequently occurring value in a dataset.
  • Variance and Standard Deviation: Measures of how spread out the data is from the mean
  • Question: “Imagine you’re collecting data on students’ favorite movie genres. Could you rank the genres from most to least popular (ordinal), or would you just say which genre is the favorite (nominal)?”
  • This exercise helps students understand the difference between nominal and ordinal data and how different types of data are analyzed.

4. Statistical Analysis of Data

• Definition:
  Statistical analysis involves using mathematical techniques to summarize and interpret data. In AI, statistics help transform raw data into insights that can guide decisions.
• Measures of Central Tendency:
  • Mean: The arithmetic average of a dataset.
  • Median: The value that divides a dataset into two equal halves.
  • Mode: The most common value in a dataset.
• Example:
  • Mean: For the dataset {5, 10, 15, 20, 30}, the mean is (5+10+15+20+30)/5 = 16.
  • Median: In the dataset {10, 11, 15, 17, 20, 21, 27, 28, 30, 32, 32, 35, 40}, the median is 27.
  • Mode: In the dataset {22, 24, 17, 18, 17, 19, 18, 21, 20, 21, 22, 22}, the mode is 22.
• Variance and Standard Deviation: These measures tell us how spread out data points are from the average. A small variance indicates that data points are close to the mean, while a large variance shows that data points are more spread out.
  • Example: If the heights of five dogs are 600mm, 470mm, 170mm, 430mm, and 300mm, the variance and standard deviation can help us understand how much the heights deviate from the mean height.

5. Representation of Data

• Definition:
  Data representation involves visualizing data to make it easier to interpret. This can include graphs, charts, and diagrams, which help simplify complex data into a more understandable format.
• Types of Graphical Representations:
  • Line Graphs: Show trends over time.
  • Bar Charts: Compare categories or groups.
  • Pie Charts: Show proportions of a whole.
  • Scatter Plots: Show the relationship between two variables.
  • Histograms: Represent the frequency distribution of data.