COM6CJ306 • Methodology for Business Research

Module 4 Notes

Module IV: Data Processing, Analysis, and Report Writing

The Analytical Architecture: From Raw Data to Scientific Deliverables

Analytical Synthesis

Data collection yields a massive volume of raw, unorganized empirical facts that cannot be directly interpreted by decision-makers. Before meaningful business intelligence can be extracted, the raw data must undergo systematic data processing—editing, coding, classification, and tabulation. Subsequently, researchers execute descriptive and inferential statistical analyses to test hypotheses and establish generalizable truths. Finally, the entire empirical investigation culminates in the preparation and presentation of a formal research report conforming to international referencing standards. This module provides an exhaustive study of data preparation, statistical analysis theory, report structure, and citation mechanics.

1. Data Processing Operations: Editing, Coding, Classification, and Tabulation

Data processing is the intermediary phase that bridges empirical fieldwork with statistical analysis. It transforms chaotic field responses into a clean, structured, and computer-readable dataset.

Stage 1

Editing

Inspecting raw questionnaires to correct errors, omissions, and illegibility.

Stage 2

Coding

Assigning numerical symbols to responses for automated software processing.

Stage 3

Classification

Grouping homogeneous data into meaningful categories based on attributes.

Stage 4

Tabulation

Systematic orderly arrangement of data into concise columns and rows.

A. Editing of Data

Editing is the examination of completed questionnaires and schedules to identify and rectify errors, omissions, illegible handwriting, and inconsistent responses.

Field Editing: Conducted by enumerators or field supervisors on the exact same day of data collection, while the interview context is fresh in memory. Aims to clarify rushed shorthand abbreviations and correct obvious recording slips.
Central Editing: Conducted at the central research office after all schedules are returned. Rigorously reviews the entire batch of schedules for logical consistency, fake response patterns (e.g., "straight-lining" identical ratings across all items), and contradictory answers (e.g., respondent reports age of 19 but reports having 20 years of work experience).

B. Coding and The Codebook Architecture

Coding is the technical process of assigning numerical numerals, alphabetical letters, or symbols to answers so that responses can be grouped into a limited number of classes and processed efficiently by statistical software (such as SPSS, R, Python, or SAS).

Specimen Research Codebook Architecture

Data Dictionary

Variable Name	Variable Label / Questionnaire Item	Data Type	Coding Rules / Value Labels	Missing Values
`RESP_ID`	Unique Respondent Identification Number	Numeric (Integer)	Sequential: 001, 002, 003, ...	None allowed
`GENDER`	Gender of the Respondent	Nominal	1 = Male, 2 = Female, 3 = Non-Binary	99 = No Response
`EDUCATION`	Highest Educational Attainment	Ordinal	1 = High School, 2 = Graduate, 3 = Post-Graduate, 4 = PhD	99 = Unstated
`PU_1`	Perceived Usefulness: Saves Time	Interval (Likert)	1 = SD, 2 = D, 3 = N, 4 = A, 5 = SA	9 = Blank
`INCOME`	Monthly Gross Household Income (INR)	Ratio (Continuous)	Actual monetary amount in INR entered directly	-999 = Refused

C. Classification of Data

Classification is the process of arranging data into groups or classes according to their common characteristics and affinities. The four fundamental types of classification are:

Geographical / Spatial Classification: Classified according to location or geographic regions (e.g., sales classified by States: Kerala, Karnataka, Maharashtra).
Chronological / Temporal Classification: Classified on the basis of time periods (e.g., corporate revenue from 2018 to 2023).
Qualitative Classification: Classified according to descriptive attributes or qualities that cannot be measured directly (e.g., religion, gender, marital status). Can be Simple (dichotomous) or Manifold (multiple sub-attributes).
Quantitative Classification: Classified on the basis of measurable numerical magnitudes forming discrete or continuous frequency distributions (e.g., workers classified by daily wage brackets: INR 500-700, 701-900).

D. Tabulation: Structure and Principles

Tabulation is the orderly and systematic presentation of numerical data in rows and columns designed to elucidate relationships and facilitate comparative analysis.

Essential Structural Parts of a Table

Table Number: Distinct identification number for referencing.
Title: Clear, concise narrative explaining What, Where, and When.
Caption (Box-Head): Column headings and sub-headings.
Stub: Row headings placed at the extreme left.
Body: The core matrix containing actual numerical frequencies.
Headnote / Prefatory Note: Explains units (e.g., "In Lakh INR").
Footnote & Source Note: Explains asterisks and original data sources.

Principles of Good Table Construction

Self-Explanatory Clarity: The table must be fully understandable without requiring reference to the text.
Proportional Sizing: Sized attractively to fit standard margins without overcrowding.
Logical Ordering: Rows and columns arranged alphabetically, chronologically, or in order of magnitude.
Totalling & Sub-Totalling: Clear grand totals to allow percentage cross-verification.

2. Analysis of Data: Descriptive and Inferential Paradigms

Data analysis involves breaking down complex data into manageable operational parts, synthesizing findings, searching for patterns, and applying statistical models.

A. Descriptive Statistical Analysis

Descriptive statistics mathematically summarize, organize, and describe the characteristics of a sample without attempting to infer or generalize beyond the sample.

Descriptive Dimension	Statistical Metrics & Descriptions	Managerial Application
Measures of Central Tendency	• Arithmetic Mean: Mathematical average sensitive to extreme outliers. • Median: Middle value of ordered dataset; robust against skewness. • Mode: Most frequently occurring value; used for nominal data.	Determining average sales, median household income, or most preferred product color.
Measures of Dispersion (Variability)	• Range: Difference between maximum and minimum values. • Standard Deviation (SD): Root mean square deviation from mean. • Variance: Square of SD; foundation of ANOVA models. • Coefficient of Variation (CV): (SD / Mean) x 100; relative risk measure.	Assessing stock price volatility, consistency of production quality, or income disparity.
Measures of Shape	• Skewness: Degree of asymmetry around the mean. Positively skewed (tail right, Mean > Median); Negatively skewed (tail left, Mean < Median). • Kurtosis: Peakedness of distribution (Leptokurtic = sharp peak; Mesokurtic = normal; Platykurtic = flat).	Evaluating financial return distributions (fat-tail risk in investment portfolios).

B. Inferential Statistical Analysis: The Logic of Hypothesis Testing

Inferential statistics enables the researcher to make scientific inferences and draw probability-based conclusions about an unknown population based on empirical sample statistics.

The Hypothesis Testing Framework

Inferential Logic

Null Hypothesis (H0): A statement of no effect, no relationship, or no difference between groups (e.g., H0: Digital advertising expenditure has no significant effect on brand awareness). It is assumed to be true until empirical evidence forces its rejection.
Alternative Hypothesis (H1): The research proposition that indicates a true relationship, difference, or effect exists (e.g., H1: Digital advertising expenditure significantly increases brand awareness).
Type I Error (Alpha - α): The probability of rejecting the Null Hypothesis when it is actually true (false positive). Commonly set at a 5% level of significance (α = 0.05).
Type II Error (Beta - β): The probability of failing to reject the Null Hypothesis when it is actually false (false negative). The Power of a Test is defined as 1 - β (typically targeted at ≥ 0.80).
Level of Significance and p-value: The p-value is the exact probability of obtaining test results at least as extreme as the results actually observed, under the assumption that H0 is true.
• If p-value ≤ 0.05: Reject H0; the finding is statistically significant!
• If p-value > 0.05: Fail to reject H0; insufficient evidence.

Specimen Statistical Analysis Tables: Bivariate and Multivariate Models

Specimen Output 1: Cross-Tabulation and Chi-Square Contingency Analysis

Bivariate Analysis

Educational Qualification	Digital Banking Adopter (Count / %)	Non-Adopter (Count / %)	Total Sample
Under-Graduate	45 (45.0%)	55 (55.0%)	100 (100%)
Graduate	90 (60.0%)	60 (40.0%)	150 (100%)
Post-Graduate & Above	120 (80.0%)	30 (20.0%)	150 (100%)
Total	255 (63.75%)	145 (36.25%)	400 (100%)

Statistical Test Result: Chi-Square Value = 31.84, Degrees of Freedom (df) = 2, p-value < 0.001. Since p < 0.05, the Null Hypothesis is rejected; there is a statistically significant association between educational qualification and digital banking adoption.

Specimen Output 2: Multiple Linear Regression Model Summary & Coefficients

Multivariate Regression

Dependent Variable: Continuance Intention to Use FinTech (CI)
Model Fit: R = 0.762, R-Square = 0.581, Adjusted R-Square = 0.577, ANOVA F-Statistic = 182.45 (p < 0.001).

Model Predictors	Unstandardized B	Std. Error	Standardized Beta	t-statistic	p-value (Sig.)
(Constant)	0.482	0.142	-	3.394	0.001
Perceived Usefulness (PU)	0.415	0.048	0.432	8.645	< 0.001
Perceived Ease of Use (PEOU)	0.284	0.052	0.278	5.461	< 0.001
Perceived Security Risk (PSR)	-0.218	0.041	-0.224	-5.317	< 0.001

Managerial Interpretation: 58.1% of the variance in customer continuance intention is explained by the three predictors. Perceived Usefulness exerts the strongest positive relative impact (Beta = 0.432), while Perceived Security Risk exerts a significant negative deterrent effect (Beta = -0.224).

Dimension	Parametric Tests	Non-Parametric Tests
Distributional Assumptions	Strict assumption of Normal Distribution of the population; homogeneity of variance (homoscedasticity).	Distribution-Free: No assumption regarding the underlying shape of the population distribution.
Level of Measurement	Requires continuous data measured on Interval or Ratio Scales.	Suitable for categorical data measured on Nominal or Ordinal Scales (or non-normal continuous data).
Statistical Power	Higher statistical power; more capable of detecting true differences when assumptions are met.	Slightly lower statistical power; robust against severe outliers and small sample sizes.
Prominent Statistical Tests	• Independent Samples t-test: Compares means of two independent groups. • Paired Samples t-test: Compares means of two related groups (pretest vs. posttest). • One-Way ANOVA (F-test): Compares means across three or more groups. • Pearson Correlation & Linear Regression: Linear relationship modeling.	• Chi-Square Test of Independence: Association between two nominal variables. • Mann-Whitney U Test: Non-parametric equivalent of independent t-test. • Wilcoxon Signed-Rank Test: Non-parametric equivalent of paired t-test. • Kruskal-Wallis H-Test: Non-parametric equivalent of one-way ANOVA.

3. The Research Report: Types, Functions, and Preparation Stages

A research endeavor is incomplete until its findings are formally communicated in a well-written, structured Research Report. The report represents the tangible deliverable that executives and academic peers examine to evaluate the validity and practical utility of the entire investigation.

Technical / Academic Report

Target Audience: Academic scholars, dissertation committees, journal peer-reviewers, and research scientists.
Tone & Style: Formal, academic, exhaustive, with rigorous methodological precision and theoretical citations.
Key Emphasis: Deep theoretical background, mathematical modeling, full statistical tables, psychometric reliability checks, and research limitations.

Popular / Executive Report

Target Audience: Corporate board members, marketing directors, financial executives, and policy stakeholders.
Tone & Style: Clear, concise, jargon-free, action-oriented, and highly visual (dashboards, graphs, charts).
Key Emphasis: The "Executive Summary," strategic business implications, cost-benefit trade-offs, and actionable policy recommendations.

Stages in Research Report Preparation

Drafting an authoritative research report requires a disciplined five-stage progression:

1. Logical Analysis of the Subject Matter: Organizing findings into a logical, coherent narrative sequence—either chronologically or from simple foundational concepts to complex multivariate relationships.
2. Preparation of the Comprehensive Outline: Establishing an exhaustive chapter-by-chapter framework with sub-headings to ensure complete coverage without unnecessary repetition.
3. Drafting the Rough Working Draft: Writing the initial complete text without over-polishing grammar, focusing on capturing all empirical findings, tables, and theoretical linkages.
4. Rewriting, Refining, and Polishing: Critically reviewing the rough draft to eliminate ambiguity, enhance readability, format data visualizations, and verify that every conclusion is supported by data.
5. Final Proofreading and Production: Checking grammar, spelling, table numbers, cross-references, and ensuring adherence to academic formatting and citation standards.

4. Standard Structural Anatomy of a Comprehensive Research Report

An authoritative business research report follows an internationally recognized structural anatomy organized into three major sections:

Standard Research Report Architecture

Structural Anatomy

I. Preliminary Section:

Title Page (Research title, author name, institution/organization, submission date).
Declaration of Originality and Supervisor / Institutional Certificate.
Acknowledgments (Honoring mentors, sponsors, and field enumerators).
Table of Contents, List of Tables, List of Figures, and List of Abbreviations.
Executive Summary / Abstract: A concise 250 to 500-word standalone synthesis summarizing problem, methodology, major statistical findings, and primary business recommendations.

II. Main Body of the Report:

Chapter 1: Introduction: Background of the study, organizational/industry context, statement of the research problem, research questions, specific SMART objectives, hypotheses, scope, and limitations.
Chapter 2: Literature Review & Conceptual Framework: Critical review of past studies, theoretical model formulation, identification of the research gap.
Chapter 3: Research Methodology: Research design, target population, sampling frame, sampling techniques, sample size calculations, instrumentation, operationalization of variables, reliability and validity evidence, data collection procedure.
Chapter 4: Data Analysis, Findings, and Interpretation: Demographic analysis, descriptive statistics, inferential hypothesis testing outputs, regression models, cross-tabulations, accompanied by detailed managerial interpretation.
Chapter 5: Summary of Findings, Conclusions, and Recommendations: Synthesis of major discoveries, strategic implications for corporate management, actionable recommendations, and directions for future research.

III. End Matter / Back Matter:

References / Bibliography: Complete, alphabetized list of all scholarly sources cited in the text.
Appendices: Supplementary materials including the complete survey questionnaire, interview guides, complex mathematical proofs, and unedited statistical computer printouts.

5. Referencing and Bibliography Styles in Business Research

Citing intellectual sources is an ethical and legal imperative that prevents plagiarism, acknowledges intellectual debt to previous scholars, and enables readers to verify the empirical foundation of the research.

In-Text Citations vs. Bibliography

In-Text Citation: A brief marker inserted within the narrative text immediately adjacent to a borrowed idea or quote, pointing directly to the full reference (e.g., Author-Date format: (Cooper & Schindler, 2014)).

Reference List vs. Bibliography: A Reference List contains ONLY the specific sources cited directly in the body of the paper. A Bibliography contains all cited sources plus background literature consulted during research, even if not explicitly quoted.

Prominent Citation Styles

APA Style (7th Edition): Standard in business, management, economics, and social sciences (Author-Date system).
Harvard Style: Widely used in UK and Commonwealth business schools (Author-Date system).
MLA Style (9th Edition): Author-Page system; primarily used in the humanities.
Chicago / Turabian Style: Uses footnotes/endnotes; common in corporate consulting and history.

APA 7th Edition Referencing Rules & Examples

Citation Standards

1. Peer-Reviewed Scholarly Journal Article:

Format: Author, A. A., & Author, B. B. (Year). Title of article. Title of Periodical, Volume(Issue), pages. DOI

Example:
Bajpai, N., & Sharma, S. (2021). Determinants of mobile payment adoption among rural micro-enterprises. Journal of Business Research, 45(2), 112–128. https://doi.org/10.1016/j.jbusres.2021.04.015

2. Authored Academic Book:

Format: Author, A. A. (Year). Title of work: Capital letter for subtitle (Edition). Publisher.

Example:
Kothari, C. R., & Garg, G. (2019). Research methodology: Methods and techniques (4th ed.). New Age International Publishers.

3. Edited Book Chapter:

Format: Author, A. A. (Year). Title of chapter. In E. E. Editor (Ed.), Title of book (pp. xx–xx). Publisher.

Example:
Kumar, R. (2020). Designing experimental models in consumer behavior. In M. V. Patel (Ed.), Advanced Business Analytics (pp. 45–68). Pearson Education.

4. Corporate / Government Web Report:

Format: Organization Name. (Year, Month Day). Title of report. Website Name. URL

Example:
Reserve Bank of India. (2023, September 15). Report on digital financial inclusion in India. Reserve Bank of India Publications. https://www.rbi.org.in/reports/dfi2023.pdf