Preliminary Data Processing and Data Analysis

Data analysis methods

1. Qualitative data analysis
2. Quantitative data analysis

Qualitative Data Analysis

Qualitative data analysis focuses on interpreting non-numerical data (e.g., text, audio, video) to
understand concepts, experiences, or social contexts.

Qualitative data analysis methods

a. Narrative analysis
b. Thematic analysis
c. Discourse analysis
d. Grounded theory
a. Narrative Analysis

Narrative analysis is a way of studying the stories people tell to understand how they see and
make sense of their own experiences.

Example:

If someone tells a story about how they overcame a challenge in school, researchers use
narrative analysis to:
 Look at what happened
 Understand how they felt
 See what meaning they gave to the experience
b. Thematic Analysis

Thematic analysis is a method used to find, analyze, and report patterns or themes in data. It's often
used to understand common ideas, experiences, or issues across multiple pieces of data (e.g.,
interviews, surveys, or focus groups).

Example:

Imagine you’re conducting interviews with students about their study habits. After reading through all
the interviews, you might notice several common themes, such as:
1. Time Management: Many students mention struggling to manage their time effectively.
2. Use of Technology: Several students talk about using apps or websites to help with studying.
3. Group Study: Many students highlight the benefits of studying in groups.

These themes would be identified and analyzed through thematic analysis to understand the most
important topics or patterns in the data.

Thematic analysis helps researchers focus on the big ideas that appear across different responses or
c. Discourse Analysis

Discourse analysis is the study of how language (in texts or speech) is used to create meaning
and influence social contexts. It looks at how words and phrases shape our understanding of the
world, ideas, and relationships.

Discourse analysis helps uncover how language influences the way people think, act, and interact
within a society.
Discourse analysis studies how language is used to build meaning and impact social realities.
d. Grounded Theory

Grounded Theory is a research method where theory or ideas are developed from data, rather
than starting with a pre-existing theory. This method involves collecting data, analyzing it, and
then developing theories based on what the data reveals.
2. Quantitative Data Analysis

Quantitative analysis involves numerical data and aims to quantify variables and relationships.

a. Descriptive Statistics

Summarizes or describes the features of a data set.

Common Tools:
 Frequency
 Mean
 Median
 Mode
 Percentages

b. Inferential Statistics

Uses sample data to make inferences or generalizations about a population.

Common Tests:
 t-test: Compares means between two groups.
Nature and Scope of Data Preparation

Data preparation is the process of getting raw data ready for analysis. It helps reduce errors,
improves accuracy, and makes it easier for people to use the data.

Goals of data preparation

One of the main goals of data preparation is to make sure that data is accurate and consistent, so
the results of analysis are trustworthy and useful.

Raw data often has:

 Missing values
 Errors
 Different formats (especially when combining data from different sources)
Steps in data preparation
1. Data Collection
2. Data Editing
3. Data Coding
4. Data Cleaning
5. Data Validation and Publishing

Data Collection

This is the first step where data is gathered from various sources. Sources can include
surveys, transaction records, online databases, or external platforms.

The goal of data collection is to gather all relevant data that will be needed for analysis.
Data Editing

Reviewing the collected data to correct obvious errors or inconsistencies.

This step helps remove issues such as incomplete responses, incorrect formatting, or duplicate entries.

Example: Fixing spelling errors in survey responses or ensuring all dates are in the same format.

The goal is to make the data:

 Accurate
 Consistent
 Well-organized
 Ready for analysis

Why Data Editing Is Important:

 Fixes wrong or unclear answers
 Makes sure responses match the questions
 Helps prepare the data for coding and analysis
Types of data editing
1. Consistency editing
2. Duplication editing
3. Validity editing
4. Statistical editing
5. Historical editing

Consistency Editing

Ensures that answers in the data do not contradict each other.

Example.

If someone says they are 10 years old but also says they have a full-time job, the responses are
inconsistent and need checking.
Duplication Editing

Removes duplicate or repeated entries.

Example.
If the same survey was accidentally entered twice for the same person, one copy
should be removed.
Validity Editing

Makes sure data entries are within acceptable or logical ranges.

Checks whether the values provided are possible or allowed.

Example.

If a question asks for age and someone enters "150", it’s likely invalid and should be corrected or
removed.
Statistical Editing

Identifies outliers or unusual values using statistical techniques.

Flags data that looks very different from the rest and may be an error.

Example: If the average monthly income is $2,000 and one entry says $50,000, it may need to be
reviewed.

Historical Editing

Compares new data with past data to find inconsistencies or unexpected changes.

Helps identify unusual patterns based on historical records.

Example.

If a student’s grades suddenly drop from A’s to F’s without explanation, historical editing might
flag that for review.
Data Coding

Transforming qualitative or categorical data into a numerical or standardized format for easier
analysis.

Data coding is the process of changing text or category responses into numbers so that software
like SPSS (Statistical Package for the Social Sciences) can easily process and analyze the data.

The goal is to simplify data processing by assigning codes or numbers to categories.

How Encoding Works:

Assign numbers or symbols to each answer

Example.
For gender: "1" = Female, "2" = Male
Data Cleaning

Removing or correcting incorrect, missing, duplicate, or irrelevant data entries.

This is one of the most time-consuming but crucial parts of data preparation.

The goal is to improve the quality of data and ensure accuracy in analysis.

Why data cleaning is important?

 Dirty data = wrong results

 Errors in the data can make your conclusions and analysis unreliable
What Data Cleaning Involves:
1. Fixing errors
2. Removing duplicates
3. Checking consistency
4. Handling missing data
• If someone skips a question, the blank answer is called a missing value
Data Validation and Publishing

Final checks to ensure the data is accurate, complete, and meets the required standards. Once
validated, the data is published or made available for analysis.

The goal is to confirm the reliability of the data before it is used in reporting or analytics