Data Processing: Editing, Coding, Tabulation, and Selecting Statistical Techniques

Raw data gathered from hundreds of questionnaires or online forms is messy, incomplete, and filled with errors. Before running statistical algorithms, researchers must process, clean, and organize the data into structured matrices.

1. The Data Preparation Workflow

   ┌────────────────────────────────────────────────────────┐
   │ THE DATA PREPARATION WORKFLOW:                         │
   │ 1. Questionnaire Checking (Accept / Reject)            │
   │                          │                             │
   │ 2. Data Editing (Field Editing & Central Editing)      │
   │                          │                             │
   │ 3. Data Coding & Codebook Creation                     │
   │                          │                             │
   │ 4. Data Cleaning (Handling Missing Values & Outliers)  │
   │                          │                             │
   │ 5. Tabulation (Univariate & Cross-Tabulation)          │
   │                          │                             │
   │ 6. Statistical Technique Selection & Analysis          │
   └────────────────────────────────────────────────────────┘

2. Editing and Coding

A. Data Editing

Editing is the inspection and correction of completed questionnaires to ensure completeness, consistency, and legibility:

Field Editing: Conducted by field supervisors the same evening interviews take place to catch missing pages or illegible handwriting while memories are fresh.
Central Editing: Conducted at the central research office by data analysts performing thorough quality checks before digital data entry.
Handling Missing Responses: Options include discarding the questionnaire, re-contacting the respondent, or substituting a neutral mean value.

B. Data Coding

Coding involves assigning numerical values or symbols to survey responses:

Example for Gender: Male = 1, Female = 2, Other = 3.
Codebook: A master document containing the variable name, description, question text, and numerical code rules for every question in the survey.

3. Tabulation of Data

Tabulation is the orderly arrangement of data into columns and rows to present summary facts clearly:

   ┌───────────────────────┐         ┌───────────────────────┐
   │ 1. Univariate Table   │         │ 2. Cross-Tabulation   │
   │ (Single Variable)     │         │ (Two or More Variables│
   │ Frequency & % Counts  │         │ Contingency Matrix)   │
   └───────────────────────┘         └───────────────────────┘

Univariate Frequency Table: Shows how frequently different categories of a single variable appear (e.g., 60% preferred Brand A, 40% preferred Brand B).
Cross-Tabulation (Bivariate / Contingency Table): Examines the relationship between two categorical variables simultaneously (e.g., comparing smartphone brand preference across age groups: 18-25 vs. 26-40).

4. Criteria for Selecting an Appropriate Statistical Technique

Selecting the right statistical test depends on four critical criteria:

   ┌────────────────────────────────────────────────────────┐
   │ SELECTION CRITERIA FOR STATISTICAL TESTS:              │
   │ 1. Number of Variables (Univariate / Bivariate / Multi)│
   │ 2. Level of Measurement Scale (Nominal/Ordinal/Metric) │
   │ 3. Sample Independence (Independent vs. Paired Samples)│
   │ 4. Nature of Population Distribution (Normal or Not)   │
   └────────────────────────────────────────────────────────┘

A. Parametric vs. Non-Parametric Techniques

Parametric Tests: Require Metric Data (Interval or Ratio scale) and assume the underlying population follows a normal distribution.
- Examples: Student’s t-test, ANOVA (F-test), Pearson Correlation, Linear Regression.
Non-Parametric Tests: Used for Non-Metric Data (Nominal or Ordinal scale) or when the population distribution is unknown/skewed.
- Examples: Chi-Square Test (Goodness of fit, Independence), Mann-Whitney U test, Kruskal-Wallis test.

5. Overview of Core Statistical Tests in Marketing

Statistical Test	When to Use in Marketing	Real-World Marketing Application
Chi-Square Test (Independence)	Analyzing the association between two nominal variables.	Testing whether car color preference is related to gender (Male vs. Female).
Independent Samples t-test	Comparing the means of two independent groups on a metric scale.	Testing whether college students spend significantly more hours on Instagram than working professionals.
Paired Samples t-test	Comparing means from the same group before and after an event.	Measuring consumer brand perception scores before and after watching a celebrity television commercial.
ANOVA (One-Way Analysis of Variance)	Comparing the means of three or more distinct groups.	Testing whether sales revenue differs significantly across three different sales territories (North, South, West).
Linear Regression	Predicting the value of a dependent variable from one or more independent variables.	Predicting quarterly sales revenue based on advertising expenditure and retail store count.

Summary Checklist for Students

Mastered the 6 stages of the data preparation workflow.
Differentiated Field Editing vs. Central Editing.
Understood Data Coding and the purpose of a research Codebook.
Contrasted Univariate Frequency tables with Cross-Tabulation matrices.
Mastered criteria for choosing between Parametric (t-test, ANOVA) and Non-Parametric (Chi-Square) statistical tests.