MBA 202 β€” Research Methodology | Unit 3 & 4 Visual Notes
MBA 202 β€” Research Methodology

Unit 3 & Unit 4
Visual Study Notes

P.R. Pote Patil College of Engineering & Management, Amravati Β· Sem II 2025–26

Sampling Design Hypothesis Testing Z / t / F / χ² Tests Research Report Writing Data Analysis Software Ref: Beri Β· Bajpai Β· Cooper Β· Gupta Β· Brayman
Unit 3 Sampling Design & Hypothesis Testing
Big Picture: What is Sampling?
🎯 SAMPLING
πŸ“¦ Population (Universe) πŸ”¬ Sample (Subset) πŸ“‹ Sampling Frame πŸ“ Sample Size (n) ⚠️ Sampling Error πŸ“Š Parameter vs Statistic
"Sampling is a process of learning about the population on the basis of a sample drawn from it." β€” G.C. Beri, Marketing Research
🏠 Real Example
Amul wants to know if Indians prefer its new flavoured milk. Surveying all 140 crore Indians is impossible. So Amul picks 2000 consumers from different cities β†’ those 2000 = Sample. All Indians = Population.
Key Terminology
πŸ“š
Must-Know Terms at a Glance
TermOne-Line DefinitionExample
PopulationAll elements sharing a common characteristicAll MBA students in Maharashtra
SampleA smaller selected group from population200 MBA students from 5 Amravati colleges
Sampling FrameComplete list from which sample is drawnRegistered MBA students list
Sampling UnitSingle element selected in the sampleEach individual student
Parameter (ΞΌ, Οƒ)Describes the population β€” usually unknownAvg income of ALL Infosys employees
Statistic (xΜ„, s)Describes the sample β€” calculated from dataAvg income of 500 surveyed employees
Sampling ErrorDifference: sample statistic vs population parameterReduced by ↑ sample size
Non-Sampling ErrorMistakes in collection / measurement β€” not samplingRespondent misunderstands a question
Sampling BiasSystematic error making sample unrepresentativeOver/under-representing a group
CensusStudy every single element of the populationIndia's Population Census every 10 years
🧠 Memory Trick
Parameter β†’ Population (both P) Β· Statistic β†’ Sample (both S)
Greek letters (ΞΌ, Οƒ) for Population Β· English letters (xΜ„, s) for Sample
Probability Sampling Methods

Every unit has a known, non-zero chance of selection. Results CAN be statistically generalized. βœ… Preferred for quantitative, conclusive research. (Brayman & Bell)

🎰
1. Simple Random Sampling (SRS)

Every unit has an equal & independent probability of selection β€” like a lottery.

Methods
Lottery / Random Number Table
πŸ“– Example
Teacher writes 60 roll numbers on chits, picks 10 blindly. Each student = 10/60 = 16.7% chance.
βœ… Unbiased βœ… Simple ❌ Impractical for large pop.
πŸ“
2. Systematic Random Sampling

Pick first unit randomly, then every kth unit after it.

k = N Γ· n    (Interval = Population Γ· Sample Size)
πŸ“– Example
Zomato: 10,000 delivery partners, need 500 β†’ k = 20. Start at #7 β†’ pick 7, 27, 47, 67...
βœ… Quick & spread evenly ❌ Periodic pattern can cause bias
πŸ₯§
3. Stratified Random Sampling

Divide population into strata (non-overlapping subgroups), then randomly sample from each.

Proportionate
Sample from each stratum proportional to its size in population.
Disproportionate
Sample more from smaller/hard-to-reach strata for better representation.
πŸ“– Example
Reliance Retail: Jio Mart 40%, Smart Bazaar 35%, Reliance Fresh 25%. Sample of 400 β†’ 160 + 140 + 100.
πŸ—ΊοΈ
4. Cluster Sampling

Divide into clusters (geographic/natural). Randomly select clusters β†’ study ALL units within selected clusters.

πŸ“– Example
Govt. of India: randomly selects 10 districts β†’ surveys ALL students in those 10 districts. Saves travel cost!
βœ… Very cost-effective ❌ Higher sampling error
πŸͺœ
5. Multi-Stage Sampling

Extension of cluster sampling β€” sampling occurs in multiple stages, progressively narrowing down. Most practical for large national surveys. (Cooper & Schindler)

1
Select States

NASSCOM: randomly pick 10 states from 28

2
Select Cities

From each state, randomly select 5 cities

3
Select Companies

From each city, randomly select 10 companies

4
Select Employees

From each company, randomly select 20 employees

Non-Probability Sampling Methods

Selection based on judgment, convenience, or criteria β€” NOT random chance. Results CANNOT be statistically generalized. βœ… Used in exploratory, qualitative research. (Brayman & Bell)

πŸ›οΈ
1. Convenience Sampling

Also called Accidental Sampling. Select units that are most easily accessible.

πŸ“– Example
Researcher stands outside Nagpur mall, stops shoppers who happen to be there.
βœ… Fastest & cheapest ❌ High bias
Best for: Pilot studies, pre-testing questionnaires
🧠
2. Purposive (Judgmental) Sampling

Researcher deliberately selects units they believe are most relevant and informative.

πŸ“– Example
Studying AI adoption β†’ deliberately selects senior IT managers from TCS, Infosys, Wipro. Only 15 people but all experts.
πŸ“Š
3. Quota Sampling

Non-probability version of stratified sampling. Researcher fills fixed quotas from each subgroup by any convenient means.

πŸ“– Example
HUL survey: collect exactly 100 male + 100 female + 75 urban + 75 rural respondents β€” fill these quotas however possible.
❄️
4. Snowball Sampling

Also called Chain Referral Sampling. Start small β†’ existing respondents refer others β†’ sample grows like a snowball.

πŸ“– Example
Studying drug addiction: start with 5 known addicts β†’ each refers 3 more β†’ sample snowballs. No sampling frame exists for this population!
Best for: Hidden groups, rare populations, social network analysis
Probability vs Non-Probability β€” Quick Comparison
Basis🎲 ProbabilityπŸ€” Non-Probability
SelectionRandom, based on chanceNon-random, judgment/convenience
Equal Chance?Known probabilityUnknown probability
GeneralizabilityCan generalize βœ…Cannot generalize statistically ❌
BiasLess biasedMore prone to bias
CostMore expensive, time-consumingCheaper, faster
Use CaseQuantitative, conclusive researchQualitative, exploratory research
ExamplesSRS, Stratified, Cluster, SystematicConvenience, Purposive, Quota, Snowball
Hypothesis Testing β€” The Core Idea
βš–οΈ
Think of it Like a Court Case
πŸ›οΈ In a Court
Accused is assumed INNOCENT until proven guilty beyond reasonable doubt.
πŸ“Š In Hypothesis Testing
Null Hypothesis (Hβ‚€) is assumed TRUE until data provides strong enough evidence to reject it.
πŸ“– Example
Amul claims new milk increases customer satisfaction to 8/10. Survey of 100 β†’ avg = 8.3. Is 8.3 truly higher than old avg 7.5 OR just due to chance? Hypothesis testing answers this scientifically.
Hβ‚€
Null Hypothesis

No effect, no difference, no relationship. The default assumption β€” the status quo. Researcher tries to disprove this.

  • Symbol: Hβ‚€ (H-naught)
  • Example: Average exam score = 60 marks (ΞΌ = 60)
  • Example: No significant difference in sales before/after ad campaign
H₁
Alternative Hypothesis

Effect exists, difference exists, relationship exists. What researcher wants to prove. Supported when Hβ‚€ is rejected.

  • Symbol: H₁ or Hₐ
  • Two-tailed: ΞΌ β‰  60 (either direction)
  • One-tailed right: ΞΌ > 60 (greater than)
  • One-tailed left: ΞΌ < 60 (less than)
Type I & Type II Errors β€” The Matrix
Hβ‚€ TRUE in Reality
Hβ‚€ FALSE in Reality
We REJECT Hβ‚€
TYPE I ERROR (Ξ±)
False Positive
"We wrongly rejected a true Hβ‚€"
Probability = Ξ±
βœ… CORRECT
Power of test = 1 βˆ’ Ξ²
We FAIL TO REJECT Hβ‚€
βœ… CORRECT
Confidence = 1 βˆ’ Ξ±
TYPE II ERROR (Ξ²)
False Negative
"Missed a real effect"
Probability = Ξ²
πŸ₯ Medical Example
Type I = Telling a healthy person they have COVID (false alarm).
Type II = Telling an infected person they're healthy (missed detection) β€” MORE DANGEROUS in medicine!
🧠 Memory Trick
Type Irror = Innocent person punished. Type II = Incorrect release (guilty person goes free).
7 Steps in Hypothesis Testing
1
State Hβ‚€ and H₁

Clearly define what you are testing and in which direction.

2
Choose Level of Significance (Ξ±)

Usually Ξ± = 0.05 (5%) in business research. β†’ 95% confidence.

3
Select Appropriate Test Statistic

Z-test / t-test / F-test / Chi-square β€” based on data type and sample size.

4
Determine Critical Value

Look up statistical table using Ξ± and degrees of freedom.

5
Calculate Test Statistic

Use sample data to compute the test statistic value.

6
Compare & Make Decision

If calculated value > critical value β†’ REJECT Hβ‚€. Else β†’ Fail to Reject Hβ‚€. Same rule: if p-value < Ξ± β†’ REJECT Hβ‚€.

7
Draw Conclusions

Interpret results in the context of the research question in plain language.

The 4 Test Statistics β€” When to Use What
🧭 Quick Decision Guide
πŸ“‹ Data is CATEGORICAL (Gender, Stream, Preference)?
β†’
χ² Chi-Square Test
πŸ”’ QUANTITATIVE data, LARGE sample (nβ‰₯30), Οƒ known?
β†’
Z - Test
πŸ”’ QUANTITATIVE data, SMALL sample (n<30) or Οƒ unknown?
β†’
t - Test
πŸ”’ QUANTITATIVE data, comparing 3 or MORE groups?
β†’
F - Test (ANOVA)
Z
Z-Test β€” Large Sample, Οƒ Known

Used when n β‰₯ 30 and population standard deviation (Οƒ) is known. Uses the standard normal distribution.

Z = (xΜ„ βˆ’ ΞΌβ‚€) Γ· (Οƒ Γ· √n)
xΜ„ = Sample mean Β· ΞΌβ‚€ = Hypothesized mean Β· Οƒ = Pop SD Β· n = Sample size
Test TypeΞ±=0.05Ξ±=0.01
Two-tailedΒ±1.96Β±2.576
One-tailed Right+1.645+2.326
One-tailed Leftβˆ’1.645βˆ’2.326
πŸ“– Solved Example
HUL manager: avg daily sales of Surf Excel = β‚Ή5000 (Hβ‚€: ΞΌ=5000). Survey of 36 outlets β†’ xΜ„=β‚Ή5200, Οƒ=β‚Ή600.
Z = (5200βˆ’5000) Γ· (600÷√36) = 200Γ·100 = 2.0
Critical value (two-tailed, Ξ±=0.05) = Β±1.96. Since |Z|=2.0 > 1.96 β†’ REJECT Hβ‚€
βœ… Conclusion: Sales ARE significantly different from β‚Ή5000.
t
t-Test β€” Small Sample, Οƒ Unknown

Used when n < 30 OR Οƒ is unknown. Uses t-distribution (thicker tails = more uncertainty). As n↑, t-distribution β†’ normal distribution. (G.C. Beri)

t = (xΜ„ βˆ’ ΞΌβ‚€) Γ· (s Γ· √n)   df = nβˆ’1
s = Sample SD (used because Οƒ unknown) Β· df = degrees of freedom
Three Types:
  • One-sample t-test: Compare sample mean to known value
  • Independent samples: Compare two different groups
  • Paired t-test: Before & after on SAME group
πŸ“– Solved Example
Principal claims MBA students study 6 hrs/day. Survey of 16 students β†’ xΜ„=5.2, s=1.2.
t = (5.2βˆ’6) Γ· (1.2÷√16) = βˆ’0.8Γ·0.3 = βˆ’2.67   df=15
Critical t (df=15, Ξ±=0.05, two-tailed) = Β±2.131. Since |t|=2.67 > 2.131 β†’ REJECT Hβ‚€
βœ… Students study significantly FEWER than 6 hours per day.
F
F-Test / ANOVA β€” Comparing 3+ Groups

Compare means of 3 or more groups simultaneously. Doing multiple t-tests instead would increase Type I error! (Cooper & Schindler)

F = Mean Square Between Groups Γ· Mean Square Within Groups

F-value is always POSITIVE. Larger F = larger group differences.

Types of ANOVA:
  • One-Way ANOVA: 1 independent variable
  • Two-Way ANOVA: 2 independent variables
  • Also used in regression analysis
πŸ“– Example
Tata Motors: Are monthly sales different across North, South, West regions?
Hβ‚€: ΞΌNorth = ΞΌSouth = ΞΌWest   H₁: At least one is different.
⚠️ Important: F-test only tells you THAT a group is different β€” NOT WHICH groups differ. Use post-hoc tests (Tukey's HSD / Bonferroni) to find which ones.
χ²
Chi-Square Test β€” Categorical Data

Most important non-parametric test. Works with CATEGORICAL data (not numbers β€” categories). "One of the most useful tools in attribute data analysis." β€” G.C. Beri

χ² = Ξ£ [ (O βˆ’ E)Β² Γ· E ]
O = Observed frequency Β· E = Expected = (Row Total Γ— Col Total) Γ· Grand Total
df = (rowsβˆ’1) Γ— (colsβˆ’1)
Two Main Uses:
  • Goodness of Fit: Do observed frequencies match expected? (1 variable)
  • Test of Independence: Are 2 categorical variables associated? (Most common)
Requirement: All expected frequencies β‰₯ 5
πŸ“– Example β€” Your AI Research!
Survey 200 students: 1) Use ChatGPT? (Yes/No) 2) Stream? (MBA/Engineering/Commerce)
Hβ‚€: ChatGPT usage and stream are independent. H₁: They are associated.
df = (2βˆ’1)Γ—(3βˆ’1) = 2. Critical χ² at Ξ±=0.05 = 5.991
If calculated χ² > 5.991 β†’ REJECT Hβ‚€ β†’ ChatGPT usage IS associated with stream of study!
Master Comparison β€” All 4 Tests
Feature Z-Test t-Test F-Test (ANOVA) χ² Chi-Square
Data TypeQuantitativeQuantitativeQuantitativeCategorical
Sample SizeLarge (nβ‰₯30)Small (n<30)AnyLarge (Eβ‰₯5)
Οƒ Known?YesNoN/AN/A
Groups1 or 21 or 23 or moreAssociation
DistributionNormal (Z)t-distributionF-distributionχ² distribution
Formula (xΜ„βˆ’ΞΌβ‚€)/(Οƒ/√n) (xΜ„βˆ’ΞΌβ‚€)/(s/√n) MSB/MSW Ξ£(Oβˆ’E)Β²/E
ExampleSales vs targetPre vs post trainingSales across regionsGender vs preference
Unit 4 Research Report Writing & Presentation
What is a Research Report & Why Does it Matter?
"The research report is the medium through which the researcher communicates his work to others." β€” G.C. Beri
"A research report is a written document that describes the research process and findings." β€” Brayman & Bell
Purpose 1
πŸ“’ Communicate Findings
Purpose 2
πŸ—‚οΈ Permanent Documentation
Purpose 3
🎯 Decision Making
Purpose 4
πŸ“š Knowledge Contribution
Types of Research Reports
πŸ”¬
Technical Report

Expert audiences β€” researchers, academics, specialists. Very detailed, full statistical analysis, technical jargon.

PhD Dissertation Journal Article
Length: 100–500 pages
πŸ“°
Popular Report

Non-expert audiences. Simple language, lots of visuals (charts, infographics), avoids technical jargon.

Customer Survey Summary
Length: 5–20 pages
πŸ’Ό
Management Report

Written for decision-makers. Concise, action-oriented, focuses on implications & recommendations. Technical details in appendix.

Consulting Firm to Wipro
Length: 15–50 pages
⏳
Interim / Progress Report

Written during research to update clients/supervisors. No final conclusions yet.

Monthly progress update
πŸŽ“
Academic / Thesis

Most rigorous. Strict institutional guidelines. Includes all components in detail. For degree requirements.

MBA Dissertation Β· PhD Thesis
Complete Structure of a Research Report
πŸ“‹
Section A: Preliminary (Front Matter)
01
πŸ“Œ
Title Page
Title, Author, Institution, Date
02
✍️
Declaration
Originality & non-plagiarism
03
πŸ…
Certificate / Approval
Signed by guide/supervisor
04
πŸ™
Acknowledgement
Thanks to supporters, supervisor
05
πŸ“‘
Table of Contents
All chapters + page numbers
06
πŸ“Š
List of Tables & Figures
Sequential index of visuals
07
πŸ“
Abstract / Executive Summary
Written LAST, placed FIRST. 150–300 words (abstract) / 1–3 pages (exec summary). Covers: problem, objectives, methodology, key findings, recommendations.
🧠 Memory Trick
"D-C-A-T-L-A" = Declaration β†’ Certificate β†’ Acknowledgement β†’ Table of Contents β†’ List of Tables β†’ Abstract
πŸ“–
Section B: Main Body (Chapters)
1
Chapter 1: Introduction

Background Β· Statement of Problem Β· Research Objectives Β· Research Questions Β· Significance Β· Scope & Limitations Β· Chapter Plan

2
Chapter 2: Review of Literature

Cover all major previous studies Β· Critically analyse (not just summarise!) Β· Identify gaps Β· Establish theoretical framework Β· Justify why current research is needed Β· Cite properly (APA/MLA/Chicago)

3
Chapter 3: Research Methodology

Research Design Β· Population & Sample Β· Data Collection instruments Β· Variables (Dependent/Independent) Β· Data Analysis Methods Β· Reliability & Validity Β· Limitations

4
Chapter 4: Data Analysis & Interpretation

Demographic Analysis Β· Analysis per objective Β· Tables & Figures (numbered, titled, sourced) Β· Statistical Tests (with calculated value, critical value, decision) Β· Interpretation in plain language after every table

πŸ“– Correct Interpretation Example
"Since calculated χ² of 8.23 exceeds critical value 5.991 at 5% significance (df=2), we reject Hβ‚€ and conclude there IS a significant association between gender and online shopping preference."
5
Chapter 5: Findings, Conclusions & Recommendations
Findings
WHAT was found. Factual, numbered. Each must correspond to an objective.
Conclusions
WHAT IT MEANS. Higher-level interpretation. Links findings to research problem.
Recommendations
WHAT TO DO. Specific, actionable, practical suggestions for stakeholders. Must be justified.
🧠 Memory
Findings = WHAT Β· Conclusions = SO WHAT Β· Recommendations = NOW WHAT
πŸ”š
Section C: End Matter (Back Matter)
Bibliography / References

Complete list of all sources cited. Consistent format throughout.

APA 7th (Most common in business)
Cooper, D.R., & Schindler, P.S. (2019). Business Research Methods (13th ed.). McGraw-Hill.
Appendices
  • Research Questionnaire
  • Statistical tables used
  • Raw data / Computer outputs
  • Maps, Photographs
  • Legal permissions obtained
Report Formatting Standards
ElementStandard
Paper SizeA4 (210mm Γ— 297mm)
MarginsTop/Bottom: 1 inch Β· Left: 1.5 inches (for binding) Β· Right: 1 inch
Font (Body)Times New Roman 12pt or Arial 11pt
Font (Headings)H1: 16–18pt Bold Β· H2: 14–16pt Bold Β· H3: 12–14pt Bold
Line Spacing1.5 lines for body Β· Single for tables, footnotes, references
AlignmentJustified (both left and right margins)
Page NumberingRoman numerals (i, ii) for preliminary Β· Arabic (1, 2, 3) from Introduction
Tables & FiguresTable 4.1 Β· Title ABOVE table Β· Source BELOW Β· Caption below figure
🧠 Writing Style Rules
Third Person ("The researcher found..." not "I found...") Β· Passive Voice Β· Precise numbers (not "many" or "some") Β· Objective tone Β· Consistent terminology
Research Paper Structure β€” IMRaD

Internationally accepted structure for scientific research papers. Used in Harvard Business Review, Journal of Marketing Research, IJMS.

I
Introduction
What is the problem? Why does it matter? What has been done? What gap does this fill?
M
Methods
How was it conducted? Enough detail for replication. Past tense, passive voice.
R
Results
Actual findings without interpretation. Tables, graphs, stat outputs with test statistic + df + p-value.
aD
Discussion
What do results mean? Compare with prior studies. Practical implications. Limitations.
Oral Presentation Tips
🎀
Structure (Timing)
  • Opening: 2–3 min β€” Greet, introduce, state title
  • Problem Statement: 3–5 min β€” What & why
  • Methodology: 5–8 min β€” How conducted
  • Key Findings: 10–15 min β€” Most important results + visuals
  • Conclusions & Reco: 5–8 min β€” What does it mean & what to do
  • Q&A: 10–15 min β€” Be prepared!
πŸ’‘
Key Rules
  • 6Γ—6 Rule: Max 6 bullets per slide Β· Max 6 words per bullet
  • Eye Contact: Look at audience, not screen
  • Voice Modulation: Vary pitch, pace, volume
  • Handling Questions: Listen β†’ Repeat β†’ Answer clearly
  • Prep: Practice 3–4 times before delivery
Chart Types β€” When to Use Which
Chart TypeBest Used ForExample
πŸ“Š Bar ChartComparing categories (most common)Sales comparison across 5 regions
πŸ₯§ Pie ChartProportions & percentages of a wholeMarket share: Tata, Hyundai, Maruti
πŸ“ˆ Line GraphTrends over timeMonthly sales Jan to Dec
πŸ“‰ HistogramFrequency distribution of continuous dataDistribution of exam scores
πŸ”˜ Scatter PlotRelationship between two variablesPrice vs demand, Ad spend vs sales
πŸ“¦ Box PlotSpread and outliers in dataSalary range across departments
Data Analysis Software
πŸ“Š
SPSS
IBM Β· Paid
MBA / PhD

GUI-based. Z/t/F/χ², Correlation, Regression, Factor Analysis, Cluster. Most used in social science.

πŸ“‹
Microsoft Excel
Microsoft Β· Paid
MBA / UG

AVERAGE, STDEV, CORREL, Pivot Tables, Data Analysis ToolPak. Perfect for MBA minor projects!

πŸ”΅
R Software
Free & Open Source
PhD / Advanced

Requires coding. 15,000+ packages. Most advanced statistical capabilities. Popular in academia.

🐍
Python
Free & Open Source
Tech / Data Science

Pandas, NumPy, Matplotlib, Scikit-learn. Big data, ML. Most popular in AI research.

🏦
SAS
Very Expensive Β· Enterprise
Corporate Β· Banking

Used by ICICI, HDFC, Pharma companies. Handles millions of records. Risk analytics.

πŸ“
Google Forms + Sheets
Free Β· Google
Minor Project βœ…

Forms auto-creates charts. Sheets has basic stats. Perfect for MBA minor research projects!

⚑
Software Quick Reference
SoftwareCostEasePowerBest For
SPSSPaidEasy (GUI)HighSocial research Β· MBA/PhD
ExcelPaidVery EasyMediumBusiness analysis Β· MBA
RFreeDifficult (code)Very HighAcademic research Β· PhD
PythonFreeModerateVery HighBig Data / ML Β· Tech
SASVery ExpensiveModerateVery HighBanking / Pharma Β· Corporate
Google SheetsFreeVery EasyLowSmall surveys Β· Minor Project
Research Agencies in India
🏒
Full-Service Agencies

Complete research from problem to final report.

Nielsen IMRB / Kantar
πŸ”
Syndicated Research

Conduct research once, sell to multiple clients who share the cost.

NASSCOM Nielsen Retail
πŸ’»
Online Research

Surveys exclusively through digital platforms. Large online panels.

SurveyMonkey Qualtrics
πŸ›οΈ
Government Agencies

Large-scale official research for policy-making.

NSSO NCAER NITI Aayog RBI
⚠️ Common Mistakes β€” Avoid These!
❌ Common Mistakeβœ… How to Avoid
Vague research problemBe specific: WHO, WHAT, WHERE, WHEN. Not "Study of sales" but "Impact of Instagram ads on SME sales in Pune (2024)"
Too many objectivesLimit to 3–5. Each objective must have a corresponding analysis section.
Poor literature reviewDon't just summarise β€” critically analyse. Identify contradictions & gaps.
Methodology not justifiedAlways explain WHY you chose a method. Why convenience? Why n=50?
Data without interpretationAfter every table/test, write what it MEANS in plain language.
Missing citationsEvery borrowed idea/stat must be cited. Use Zotero or Mendeley.
OvergeneralizationDon't claim 50-person sample findings apply to the entire world!
Findings = ConclusionsFindings = WHAT Β· Conclusions = SO WHAT Β· Recommendations = NOW WHAT
Informal languageNo contractions (don't, can't), no slang, formal academic tone throughout.
βœ… Pre-Submission Checklist
  • Title page complete with all required information
  • Declaration and certificate signed
  • All objectives have corresponding analysis sections
  • Every table has a number, title, and interpretation
  • All statistical tests include calculated value, critical value & decision
  • All sources cited in bibliography
  • Questionnaire attached as appendix
  • Consistent font, spacing, and page numbering throughout
  • Abstract/executive summary written LAST, is accurate
  • Word count is within required limits
πŸ”’ All Key Formulas at a Glance
Z-Test
Z = (xΜ„ βˆ’ ΞΌβ‚€) Γ· (Οƒ Γ· √n)

Large sample, Οƒ known, nβ‰₯30

t-Test
t = (xΜ„ βˆ’ ΞΌβ‚€) Γ· (s Γ· √n)   df = nβˆ’1

Small sample / Οƒ unknown, n<30

F-Test / ANOVA
F = MSB Γ· MSW

Mean Square Between Γ· Mean Square Within. 3+ groups.

Chi-Square
χ² = Ξ£[(Oβˆ’E)Β² Γ· E]   df = (rβˆ’1)(cβˆ’1)

E = (Row Total Γ— Col Total) Γ· Grand Total

Systematic Sampling Interval
k = N Γ· n

N = Population size Β· n = Sample size

Sample Size (Proportions)
n = ZΒ² Γ— p Γ— (1βˆ’p) Γ· EΒ²

Z=1.96 at 95% confidence Β· p=0.5 (worst case) Β· E=margin of error

πŸ“ Final Exam Strategy
Reference Book Emphasis:
  • G.C. Beri: Define BEFORE you explain
  • S.L. Gupta & H. Gupta: Always give Indian company examples
  • Cooper & Schindler: Practical application & decision-making
  • Naval Bajpai: Step-by-step processes
  • Brayman & Bell: Research design & methodology rigor
Perfect Answer Formula:
1
Define the concept
2
Explain it clearly
3
Give a relevant Indian company example
4
State advantages & disadvantages
πŸ”‘ Most Important Decision to Remember
Categorical β†’ χ² Β· Quantitative + Large + Οƒ Known β†’ Z Β· Quantitative + Small/Οƒ Unknown β†’ t Β· 3+ Groups β†’ F(ANOVA)
MBA 202 β€” Research Methodology Β· Unit 3 & Unit 4 Visual Notes Β· Vikas Bhusare Β· Sem II 2025–26
P.R. Pote Patil College of Engineering & Management, Amravati
Ref: Brayman & Bell Β· G.C. Beri Β· Naval Bajpai Β· S.L. Gupta & H. Gupta Β· Donald Cooper & Pamela Schindler
Scroll to Top