Unit-II:

Research
Design

©shumank deep
 • A structured framework to guide research efforts.
Concept: • It outlines methods, procedures, and strategies for
data collection and analysisCreswell

• Systematic: Follows a logical sequence from defining

What Is Nature: problems to analyzing results.

• Objective-driven: Aims to address specific questions

research
Design Importance: • Ensures reliability and validity of findings.
• Aligns research objectives with methods

Example: • Determining consumer preferences for eco-friendly

packaging using surveys and conjoint analysis

©shumank deep
 • Classification based on purpose:
1. Exploratory: Understanding new areas or issues (e.g., focus groups).
2. Descriptive: Providing a detailed account (e.g., market share analysis).
3. Explanatory: Explaining causal relationships (e.g., impact of HR training on employee
performance)
• Type of Research Design
• Qualitative:
• Focuses on understanding underlying reasons, opinions, or motivations.
• Methods: Interviews, focus groups, thematic analysis.
• Example: Exploring employee perceptions of diversity policies
• Quantitative:
Types of Research • Focuses on quantifiable data and statistical analysis.
Design • Methods: Surveys, experiments, numerical modeling.
• Example: Measuring the impact of marketing campaigns on sales revenue
• Experimental:
• Establishes causal relationships by manipulating variables.
• Methods: Laboratory or field experiments.
• Example: Testing the effect of a new inventory management system on operational
efficiency
• Mixed Methods:
• Combines qualitative and quantitative approaches for comprehensive insights.
• Methods: Surveys with follow-up interviews, data triangulation.
• Example: Evaluating the success of customer loyalty programs through surveys
and focus groups

©shumank deep
 Defining Objectives
• Clearly articulate the research purpose and questions.
• Align objectives with the scope and context of the problem being investigated.
• Example: In HR, defining whether the research aims to improve employee retention or evaluate training programs.

Method Selection
• Qualitative: Explore perceptions or behaviors (e.g., interviews, focus groups).
• Quantitative: Measure variables numerically (e.g., surveys, experiments).
• Mixed Methods: Combine approaches to provide deeper insights.
• Experimental: Test cause-and-effect relationships through controlled variables.
• Example: In marketing, use a mixed-method approach by conducting surveys for numerical insights and interviews for in-depth understanding of consumer behavior.

Stages in
Data Collection
• Primary Data: Gather original data through interviews, surveys, or experiments.
• Secondary Data: Use existing data from reports, databases, or previous studies.
• Ensure tools like questionnaires, observation checklists, or technical sensors align with research goals.

Research
• Example: In business analytics, using transactional data alongside customer feedback for holistic analysis.

Sampling Design
• Define the population and sampling technique:

Design
•Probability Sampling: Random selection ensures representativeness.
•Non-Probability Sampling: Focus on specific sub-groups.
• Example: In finance, selecting 100 companies randomly from the S&P 500 for portfolio analysis.

Validating the Research Design

• Conduct pilot testing or use triangulation to confirm reliability and validity.
• Example: In marketing, pretesting a survey questionnaire with a small sample before full-scale distribution.

Data Analysis
• Choose appropriate statistical or thematic methods:
•Qualitative: Content analysis, thematic coding.
•Quantitative: Regression, hypothesis testing, machine learning models.
•Example: In operations, applying simulation models to analyze supply chain efficiency.

Reporting and Interpretation

• Present findings in a clear, concise format with visualizations like charts or tables.
• Discuss implications for stakeholders, limitations of the study, and recommendations.
©shumank
• Example: In finance, presenting risk assessment deep
results with actionable investment strategies.
 • Participants must be fully aware of the study's
Informed purpose, methods, risks, and benefits.
Consent:
• Provide clear, understandable consent forms

• Safeguard participant data through encryption

Ensuring Confidentiality:
and secure storage.
• Use anonymization where possible to protect

Ethical
identities

Voluntary • Ensure participation is free from coercion.

Compliance Participation: • Allow participants to withdraw at any stage

in Research
• Minimize physical, psychological, or emotional
risks to participants.
Avoiding Harm:
• Conduct thorough risk assessments before
starting the research

• Report findings honestly without fabrication,

Integrity: falsification, or plagiarism.
• Acknowledge all sources and contributions

©shumank deep
 Obtain approval from
Institutional Review Boards
Ethical Review: (IRBs) or ethics
committees.

Clearly communicate
Transparency: research goals and

Best
intentions to stakeholders.

Practices Cultural Sensitivity:

Respect cultural norms
and values during data
collection

Ensuring customer surveys

comply with GDPR by
Example : anonymizing responses
and using secure data
storage.
©shumank deep
 Definition:
• Original data collected directly from the source for specific research purposes

Methods:
• Surveys and Questionnaires:
• Quantitative data through structured questions.
• Example: Customer satisfaction survey in marketing.
• Interviews:

Primary
• Qualitative insights via one-on-one or group discussions.
• Example: Employee interviews to assess workplace culture in HR.
• Observations:
• Monitoring behaviors or activities without interference.

Data
• Example: Observing consumer behavior in a retail store.
• Experiments:
• Controlled environments to test hypotheses.
• Example: A/B testing on a new product feature in business analytics.
• Focus Groups:

Collection
• Group discussions to gather diverse perspectives.
• Example: Exploring customer reactions to a new marketing campaign.

Advantages:
• It can be adjusted to research objectives.
• Provides current and specific data.

Challenges:
• Time-consuming and costly.
• Requires skilled personnel for accurate data collection
©shumank deep
 Definition:

• Pre-existing data gathered from published or archived sources

Sources:

• Internal Records:
• Company reports, sales data, employee performance reviews.
• Example: Historical sales data for forecasting in operations.

Secondary
• External Sources:
• Industry reports, government publications, research databases.
• Example: World Bank data for financial market analysis.
• Online Resources:

Data • Websites, social media analytics, public datasets.

• Example: Analyzing social media trends in marketing research.
• Academic Research:

Collection
• Journals, dissertations, and conference papers.
• Example: Literature review for identifying gaps in research

Advantages:

• Time and cost-efficient.

• Accessible from diverse sources.

Challenges:

• Data may be outdated or lack relevance.

• Limited control over data quality or methodology

©shumank deep
Comparison

Aspect Primary Data Secondary Data

Source First-hand from respondents Pre-existing from other sources
Cost High Low
Relevance Adjusted to study May require adjustments
Quality depends on original
Accuracy Direct control over quality
source

©shumank deep
 Definition
•Survey: A systematic method for collecting data from a predefined group of respondents to gain information about
their opinions, behaviors, or characteristics

Key Features
•Structured Approach:
•Uses a standardized set of questions to ensure consistency.
•Large Sample Sizes:

Survey as a
•Can collect data from a wide audience efficiently.
•Quantitative Focus:
•Primarily used for statistical analysis, but can also include open-ended qualitative responses

Primary
Types of Surveys
•Online Surveys:
•Tools like Google Forms or SurveyMonkey.
•Example: Customer feedback surveys for an e-commerce business.

Data
•Face-to-Face Surveys:
•Conducted in person for in-depth understanding.
•Example: Interviews with employees during organizational change initiatives.
•Telephone Surveys:
•Suitable for reaching geographically dispersed respondents.

Collection
•Example: Polling voters before an election.
•Paper-Based Surveys:
•Delivered physically, often in academic or healthcare settings.
•Example: Patient satisfaction surveys in hospitals.

Method Advantages
•Cost-effective for large-scale data collection.
•High flexibility in design and distribution methods.
•Generates standardized data for easy comparison and analysis

Challenges
•Low Response Rates:
•Online surveys often face low engagement.
•Survey Bias:
•Poorly worded questions or unrepresentative samples may skew results.
•Time-Consuming Analysis:
•Large datasets require robust analytical methods.
©shumank deep
 Define Objectives:
Define •What do you aim to learn?
•Example: Understanding customer preferences for product features in marketing.

Design the Questionnaire:

Design •Use clear, concise, and unbiased questions.
•Types: Multiple-choice, Likert scale, open-ended

Steps to Select
Select Sampling Method:
•Probability sampling (e.g., random sampling) or non-probability sampling (e.g., convenience

Conduct a
sampling).

Survey
Pilot Test:
Pilot •Test the survey on a small sample to refine questions and format.

Distribute the Survey:

Distribute •Choose the appropriate medium (online, in-person, etc.).

Analyze Data:
Analyze •Use statistical tools to draw meaningful insights (e.g., SPSS, Excel).

©shumank deep
 Designing the Survey
• Define Objectives:
• Clearly outline what the survey aims to achieve.
• Question Types:
• Closed-ended (e.g., multiple-choice, Likert scale).
• Open-ended for detailed responses.
• Clarity and Neutrality:
• Use simple, unbiased language.
• Logical Flow:
• Start with easy questions and group related ones.

Designing and
• Keep it Concise:
• Avoid respondent fatigue with short, focused questions.

Administering
Administering the Survey
• Choose Mode:
• Online (Google Forms), face-to-face, phone, or paper.

Survey • Sampling:
• Define the target audience and select a representative sample.
• Pilot Test:

Questionnaires • Test the survey with a small group to refine questions.

Encouraging Responses
• Personalize invitations and explain the purpose.
• Offer incentives (e.g., discounts).
• Send polite follow-up reminders.

Ethical Compliance
• Obtain informed consent.
• Maintain
©shumankconfidentiality
deep and data security.
 Clarity and Simplicity
• Use simple, straightforward language.
• Avoid jargon, technical terms, or complex phrasing.
Question Design
• Closed-ended: Use structured questions (e.g., Yes/No, multiple choice).
• Open-ended: Allow detailed responses for deeper insights.
• Use balanced Likert scales (e.g., Strongly Agree to Strongly Disagree).
Avoid Bias
• Frame neutral questions to prevent leading responses.

Guidelines for
• Example:
• Biased: “Don’t you think this product is amazing?”
• Neutral: “How would you rate this product?”

Drafting Logical Flow

• Start with general, easy questions to engage respondents.

Effective
• Group related questions into sections for better context.
• Place sensitive questions at the end.
Length and Brevity

Questionnaires • Keep it concise to minimize respondent fatigue.

• Avoid redundant or overly detailed questions.
Pre-Test and Refine
• Pilot the questionnaire with a small group to identify errors.
• Revise based on feedback before full distribution.
Ethical Considerations
• Clearly state the purpose of the survey.
• Assure confidentiality
©shumank deep and data protection.
 Creating online surveys
Google Forms: Free and easy to use.

Choose an Online Tool

SurveyMonkey: Advanced features for large-scale surveys.
Qualtrics: Comprehensive for academic and business needs.
Typeform: Interactive and visually appealing surveys.

Distribute the Survey

Share via email, social media, or embedded links on websites.
Optimize for mobile devices for broader reach.

Analyze Results
Export data to tools like Excel, SPSS, or Google Sheets.
Use charts and tables for visualization.

©shumank deep
Interview

Definition A conversation with structured or unstructured questions to gather in-depth information directly from participants.

Structured: Predefined questions for consistency.

Types: Semi-structured: Combines predefined and open-ended questions.
Unstructured: Free-flowing conversation for exploratory insights.

Provides detailed, qualitative data.

Advantages: Enables clarification and probing for deeper insights.

Time-consuming and resource-intensive.

Challenges: Interviewer bias can influence responses.

Example: HR: One-on-one interviews to evaluate employee job satisfaction.

©shumank deep
 Definition

• Systematic monitoring of behaviors or activities in their natural setting.

Types:

• Participant Observation: Researcher actively engages in the setting.

• Non-Participant Observation: Researcher observes without
involvement.
• Structured: Predefined criteria for observation.
• Unstructured: Open-ended and exploratory.

Observation Advantages:

• Captures real-time, authentic behavior.

• Uncovers non-verbal cues and interactions.

Challenges:

• Limited control over external factors.

• Observer bias may affect interpretation.

Example:

• Marketing: Observing customer behavior in a retail store to analyze

purchasing patterns.
©shumank deep
 Internal Sources
•Definition: Data generated within an organization.
•Examples:
•Financial statements and reports.
•Sales and customer databases.
•Employee performance records.

External Sources
•A. Government Publications:
•Census reports, economic surveys, trade statistics.
•Example: Using labor market reports for HR planning.

Sources of
•B. Industry Reports:
•Published by trade associations or market research firms.
•Example: Nielsen reports for marketing analysis.

Secondary Data
•C. Academic Research:
•Journals, theses, and dissertations.
•Example: Literature review for identifying research gaps.

Collection •D. Online Sources:

•Websites, blogs, and social media analytics.
•Example: Analyzing trends using Google Trends or LinkedIn data.

Commercial Databases
•Subscription-based platforms like Statista, IBISWorld, or Bloomberg.
•Example: Accessing financial data for market analysis.

Media Sources
•Newspapers, magazines, TV reports.
•Example: Analyzing public opinion trends from news archives.

Public Data Archives

•Open data platforms and repositories.
•Example: WHO health statistics or UN data for global research.
©shumank deep
 What is it?
• A method used to pick a small group (sample) from a larger population to save time,
effort, and resources during research.

Type of Sampling
• Probability Sampling
• Non-Probability Sampling

How to Identify the Population?

Sampling • The population is the group you want to study. Example: If you're studying university
students' eating habits, the population could be "all students at XYZ University."

Why Does It Matter?

Design • It helps ensure the small group reflects the larger population, so results are reliable
and meaningful.

Why Should You Care?

• Saves time and avoids surveying everyone.
• Makes research manageable and efficient.
• Leads to better decisions and insights in any field—business, health, or everyday
problem-solving.

Takeaway:
• Sampling design is like choosing a few people to represent a crowd—it’s smart,
efficient, and practical!
©shumank deep
 In probability sampling, every member of the population has a known and
non-zero chance of being selected. This approach is ideal for obtaining a
representative sample.

Simple Random Sampling: Every individual in the population has an equal

chance of being selected. This is achieved using random selection
methods, such as drawing lots or using random number generators.

Systematic Sampling: Selection is done at regular intervals from an

Probability
ordered population list, e.g., every 10th person.

Sampling Stratified Sampling: The population is divided into strata or subgroups

based on shared characteristics (e.g., age, income), and random samples
are drawn from each stratum.

Cluster Sampling: The population is divided into clusters (e.g., schools,

cities), and a random sample of clusters is selected. All or a subset of
individuals within selected clusters may then be studied.

Multistage Sampling: A more complex form of cluster sampling where

sampling is performed at multiple levels (e.g., selecting states, then cities
within states, then individuals within cities).

©shumank deep
 In non-probability sampling, individuals are selected based on non-
random criteria, which might lead to bias but is practical in
exploratory research.
Convenience Sampling: Participants are chosen based on
availability or ease of access, such as using students in a nearby
classroom.

Non- Judgmental or Purposive Sampling: Researchers select

participants based on their judgment of who would best contribute
to the study.
Probability Snowball Sampling: Existing study participants recruit future

Sampling
participants from their network, commonly used in studies
involving hidden populations.
Quota Sampling: Researchers ensure the sample reflects certain
characteristics in proportion to their prevalence in the population
(e.g., gender, age groups).
Haphazard Sampling: Participants are selected in an unstructured
and informal manner, often without a defined plan.

©shumank deep
 Definition
• Converting raw data into a structured format using numerical or symbolic
values.

Steps:
Preparation • Identify Variables: List all variables (e.g., gender, age, responses).
• Assign Codes:

of data:
• Use numbers or symbols for categories.
• Example: Gender → Male = 1, Female = 2.
• Create a Codebook: Document all codes and their meanings for future

How to
reference.
• Apply Consistently: Ensure all data entries follow the same coding rules.

Example
Code Data • Survey Question: "How satisfied are you with the service?"
• Very Satisfied = 5
• Satisfied = 4
• Neutral = 3
• Dissatisfied = 2
• Very Dissatisfied = 1

©shumank deep
 Definition

• Reviewing and correcting data for completeness,

accuracy, and consistency.
Preparation Steps:
of data: • Check for Missing Data: Identify and address

How to Edit
unanswered questions or blanks.
• Resolve Ambiguities: Clarify unclear responses
or replace with logical estimates.
Data • Correct Errors: Look for typos, logical
inconsistencies, or mismatched data (e.g., age
200).
• Verify Against Original Sources: Cross-check
with the original responses if possible.

©shumank deep
 Definition

• Removing or correcting inaccurate, inconsistent, or

duplicate data to ensure reliability.

Preparation of Steps:

data: How to
• Detect Outliers: Identify extreme values that may
distort results.
• Address Missing Data:
Clean Data • Use techniques like mean substitution, regression,
or deletion.
• Remove Duplicates: Check for repeated entries.
• Standardize Formats: Ensure consistency (e.g.,
date format: DD/MM/YYYY).
• Validate Final Dataset: Double-check for errors
after cleaning.

©shumank deep
Introduction to SPSS
SPSS (Statistical Package for the Social Sciences) is a software tool for data analysis, management, and
What is SPSS? visualization.

User-Friendly Designed with a point-and-click interface for ease of use, making it accessible to beginners and advanced
Interface users alike.

Versatile Data Supports a wide range of statistical methods, including descriptive statistics, regression, ANOVA, and factor
Analysis analysis.

Data Management Allows efficient data entry, editing, cleaning, and transformation to prepare datasets for analysis.

Graphical Generates high-quality charts and graphs for data visualization and presentation.
Representation

Widely Used Popular in academia, research, healthcare, and business for data-driven decision-making.
©shumank deep
 Step 1: Open SPSS

•Launch SPSS software and select "New Dataset" or start with a blank file.

Step 2: Define Variables

•Switch to the "Variable View" Tab:

•Each row represents a variable (e.g., name, gender, age).
•Set Variable Names:
•Provide short, meaningful names (e.g., Age, Gender).
•Define Variable Properties:
•Type: Choose data type (e.g., numeric, string).

Creating an •Label: Add descriptive labels (e.g., "Age in Years").

•Values: Assign value labels for categorical data (e.g., 1 = Male, 2 = Female).
•Missing Values: Specify codes for missing data, if any.

SPSS Data
•Measure: Define the level of measurement (Nominal, Ordinal, Scale).

Step 3: Enter Data

File •Switch to the "Data View" Tab:

•Each row represents a case or participant.
•Input Data:
•Enter values in the corresponding cells under each variable.

Step 4: Save the File

•Click File → Save As.

•Choose a location, provide a file name, and save it as an SPSS data file (*.sav).

Step 5: Validate Data Entry

•Double-check entries for errors or inconsistencies.

©shumank deep
 Step 1: Prepare the Data File

•Ensure your data file is clean and in a supported format (e.g., .csv, .xlsx, .txt, .sav).
•The first row should contain variable names for easy recognition.

Step 2: Open SPSS

•Launch SPSS software.

Step 3: Select the Import Option

•Go to File → Open → Data.

•In the "Open Data" dialog box:

Importing
•Choose the file type (e.g., Excel, CSV, or Text).
•Locate and select your data file.
•Click Open.

Data into Step 4: Configure Import Settings

•Follow the import wizard for your file type:

SPSS
•For Excel Files (.xlsx):
•Specify the sheet to import.
•Indicate whether the first row contains variable names.
•For CSV or Text Files:
•Define delimiters (e.g., commas, tabs).
•Check if the first row contains variable names.
•For SPSS Files (.sav):
•Files are directly opened without additional configuration.

Step 5: Review Imported Data

•The data will appear in the Data View tab.

•Use the Variable View tab to adjust variable properties (e.g., labels, types, missing values).

Step 6: Save the SPSS File

•Go to File → Save As.

•Save the imported data as an SPSS file (*.sav) for future use.

©shumank deep