Unit 4: Introduction to Data Processing and Data Analysis

1. Introduction to Data Processing

Data processing is the crucial "middle step" between collecting data (Unit 3) and analyzing it. It involves converting raw, messy data into a clean, organized, and machine-readable format.

Key Steps in Data Processing:

1. Editing:
   • Checking the raw data (questionnaires, schedules) for errors, inconsistencies, or missing answers.
   • Example: Finding a respondent who ticked both "Yes" and "No," or who listed their age as "150," and deciding how to handle it (e.g., discard, or mark as "missing").
2. Coding:
   • The process of assigning numerical codes to non-numerical answers. This is essential for statistical analysis.
   • Example:
     • Question: "What is your gender?" (Male = 1, Female = 2)
     • Question: "How do you feel about the economy?" (Very Good = 1, Good = 2, Bad = 3, Very Bad = 4, Don't Know = 99)
3. Tabulation:
   • Organizing the coded data into tables to see the patterns.
   • Simple Tabulation: A simple frequency count (e.g., "How many people said 'Good'?").
   • Cross-Tabulation: A more complex table that compares two variables (e.g., "What was the opinion of *men* vs. *women* on the economy?").

2. Introduction to Data Analysis

Data analysis is the process of *interpreting* the processed data to find patterns, answer research questions, and discover insights.

There are two broad types of analysis: **Quantitative** (deals with numbers) and **Qualitative** (deals with words, meanings, and interpretations).

3. Content Analysis

Definition: Content analysis is a research method used to analyze the content of communications in a systematic and quantitative way. It turns qualitative text data into quantitative, countable data.

• What it does: It *counts* the frequency of specific words, themes, or images.
• The Question it answers: "How many times?" or "How much?"
• The Process:
  1. Define your research question (e.g., "How do newspapers cover female vs. male politicians?").
  2. Select your media (e.g., 100 articles from 4 newspapers).
  3. Develop "categories" (e.g., Category 1: "Words about appearance," Category 2: "Words about policy").
  4. "Code" the articles by counting how many times words from each category appear.
  5. Analyze the numbers (e.g., "Articles on female politicians mentioned 'appearance' 70% of the time, vs. 10% for male politicians").
• Example: Analyzing party manifestos to *count* how many times they use the words "development," "poor," or "security."

4. Discourse Analysis

Definition: Discourse analysis is a research method used to study language *beyond* the level of the sentence. It is a qualitative method that analyzes *how* language is used in social contexts to create meaning and power.

• What it does: It doesn't just count words; it interprets the *meaning*, *assumptions*, and *power dynamics* behind the words.
• The Question it answers: "What does this mean?" or "What is being *done* with this language?"
• The Process:
  1. Select a "discourse" (e.g., a political speech, a media debate).
  2. Analyze the language, metaphors, and narratives used.
  3. Ask: What "worldview" is this text promoting? Whose interests are being served? What is being left *unsaid*?
• Example:
  • Content Analysis would *count* how many times a news report uses the word "terrorist."
  • Discourse Analysis would ask *why* the person is being called a "terrorist" instead of a "freedom fighter" or "militant," and how that choice of word influences the reader's understanding and supports a particular political position.

5. Comparison: Content vs. Discourse Analysis

6. Exam Corner: Key Distinctions