Engineering Data Analysis - Module 1

Obtaining Data
Statistics - Statistics is the science that deals with Methods of Data Collection
● Collection ➢ Primary Data - Data collected fresh and for the
● Organization first time; original in character.
● Presentation ● Characteristics:
● Analysis - Originality
● Interpretation of data - Specific to Research
➢ Purpose - To aid decision-making ● Examples: Surveys, Observations, Experiments
Main Divisions of Statistics ➢ Primary Data | Observations - Gathering data
➢ Descriptive Statistics - Summarizes and directly from the field through observation.
describes quantitative data. ● Methods: Observer's presence or visits to the
Focus: Organizing and summarizing data. field.
Purpose: Describing data. ● Advantages: Subjective bias eliminated, current
Methods: Histograms, charts, measures of information.
central tendency, etc. ● Disadvantages: Time-consuming, limited
➢ Inferential Statistics - Draws conclusions about information, unforeseen factors.
a population based on a sample. ➢ Types of Observations
Focus: Making judgments or conclusions about ● Structured Observation - Observation
populations. with standardized conditions, recorded
Method: Using findings from a sample to make styles, and defined units.
inferences. Example: Inventory analysis in a store
Importance: Decision-making support. by an auditor.
● Unstructured Observation- Observation
Understanding Data in Statistics without prior planning.
Data - are values of qualitative or quantitative Example: Observing children playing
variables, belonging to a set of items. with new toys
➢ Data can be qualitative or quantitative. ● Participant Observation - Observer is a
➢ Data belongs to a set of items, often referred member of the observed group.
to the population. Advantages: Natural behavior, better
Population: The set of objects you are interested in. understanding, closeness to the group.
● Key Characteristics ● Non-participant Observation - Observer
- Comprehensiveness - The population includes observes without interaction.
every possible member that fits the defined Advantages: Objectivity, neutrality,
criteria. respondent willingness.
- Parameter Estimation - Population parameters ● Uncontrolled Observation- Observation
are the characteristics or values that describe in natural conditions for spontaneous
the entire population. data.
- Precision - Analyzing the entire population Purpose: Capturing life and behaviors
allows for the most precise and accurate as they occur.
statistical conclusions. ● Controlled Observation - Observation
Variables: A measurement or characteristic of with pre-arranged plans and
an item. experimental procedures.
➢ Qualitative Variables - Variables that are not Context: Often conducted in a
numerical; they describe data that fits into laboratory under controlled conditions.
categories. ➢ Considerations in Observation
➢ Quantitative Variables - Variables that involve ● Factors:
numerical measurements or values. - - Setting and context
 Engineering Data Analysis - Module 1

- Observer's role ● Telephonic Interviews - Conducted over the

- Ethical considerations phone.
- Data recording methods Use: Cost-effective data collection, remote
➢ Interviews - involves oral-verbal interviews.
communication to gather information. ● Considerations in Interviews
➢ Types of Interviews Factors:
● Personal Interviews - Face-to-face ○ Interviewer skills
communication with the interviewer and ○ Respondent comfort
respondent. ○ Ethical considerations
Use: In-depth information gathering, ○ Data recording methods
rapport building. ● Disadvantages:
● Structured Interviews - Predetermined set of ○ Time-consuming
questions. ○ Potential for interviewer bias
Use: Consistency in data collection, easy ○ Limited to available respondents
analysis. ● Advantages:
● Unstructured Interviews - No set questions; ○ In-depth data
flexible conversation. ○ Clarification of responses
Use: Exploratory research, open-ended ○ Rapport building
responses. ➢ Questionnaires - The questionnaire method
● Focused Interviews - Emphasis on a specific involves sending a set of questions to
experience and its effects. respondents, who are expected to answer in
Use: Understanding specific aspects deeply. writing.
● Clinical Interviews - Explores broad feelings, ➢ Use of Questionnaires:
motivations, or life experiences. ● Popularity in big surveys
Use: Psychological or therapeutic contexts. ● Written responses from respondents
● Group Interviews - Interviews with 6 to 8 ● Independent response generation
individuals simultaneously. ➢ Advantages of Questionnaires
Use: Collecting diverse perspectives, group ● Cost-effectiveness for large geographic areas
dynamics. ● Answers in respondents' words, reducing bias
● Qualitative and Quantitative Interviews - Based ● Adequate time for thoughtful responses
on subject matter (qualitative vs. quantitative). ● Ability to reach non-approachable
Use: Depending on research objectives and Respondents
data type. ➢ Disadvantages of Questionnaires
● Individual Interviews - One-on-one interviews ● Low return rate of completed questionnaires
with a single person. ● Slow data collection process
Use: In-depth exploration of an individual's ● Difficulty in verifying respondent identity
experiences or opinions. ● Potential for incomplete or inaccurate
● Selection Interviews - Interviews for job responses
recruitment or candidate selection. ➢ Types of Questions
Use: Assessing qualifications and suitability for ● Open-ended questions
a position. ● Closed-ended questions
● Depth Interview - Aims to elicit unconscious ● Multiple-choice questions
and personality-related material. ● Likert scale questions
Use: Psychological research, personality ➢ Questionnaire Design
analysis. ● Factors:
○ Clarity of questions
 Engineering Data Analysis - Module 1

○ Question sequence ● Validity concerns with closed-ended

○ Language and tone questions
○ Length of the questionnaire ● Data errors due to non-responses
➢ Case Studies - A case study is an intensive ● Wide-reaching data collection
investigation of a particular unit or subject Ethical Considerations
under study. ● Informed consent
Purpose: Explore the use, advantages, and ● Privacy and data protection
disadvantages of case studies in research. ● Ensuring respondent well-being
Uses : Secondary Data - Data collected by someone else and
● In-depth investigation have been through the statistical process.
● Applicable in various fields (e.g., ● Characteristics:
psychology, medicine, business) - Not Original
● Ideal for complex or unique scenarios - Collected for Other Purposes
Advantages: ● Sources:
● Cost-effective and less time-consuming Government publications | Examples | Uses
● Suitable for hard-to-obtain exposure data - Census data, economic reports, health
● Effective for studying dynamic populations statistics
● Rich qualitative data - Comprehensive and reliable data from
Disadvantages: government sources.
● Subject to selection bias Technical and trade journals | Ex. | Uses
● Limited ability to calculate incidence - Industry-specific data, research
(absolute risk) findings
● Findings may not be generalizable - Valuable for industry analysis and
● Time and resource-intensive for data specialized research.
collection Books, magazines, newspapers | Ex. | Uses
Ethical Considerations Factors: - Historical data, public opinions
● Informed consent - Historical context and public sentiment
● Privacy and confidentiality analysis.
● Respect for participants Reports from industry, banks, stock exchanges
➢ Surveys - are common methods for diagnosing - Annual reports, market research studies
and addressing social problems. - Business analysis and market trends.
Purpose: Explore the use, advantages, and Academic research reports | Examples | Uses
disadvantages of surveys in research. - Theses, dissertations, research papers
Use of Surveys - Scholarly insights and findings.
● Wide applicability in social research Public records | Examples | Uses
● Ideal for collecting data from diverse - Birth records, property records, legal
populations documents
● Versatility in question types (e.g., - Legal and historical data.
closed-ended, open-ended) Consideration Factors:
Advantages ● Nature, Scope, and Object of Inquiry
● Easy administration ● Availability of Funds
● Relatively quick development time ● Time Factor
● Cost-effectiveness (depending on ● Precision Required
survey mode) ➢ Designing a Survey - Surveys gather data
Disadvantages through a series of questions.
● Potential for respondent bias Purpose: Explore the steps involved in
● Accuracy and honesty of responses may vary designing an effective survey.
 Engineering Data Analysis - Module 1
Steps in Designing a Survey Key Concepts:
1. Determine the goal ● Control groups and baseline measurements
2. Identify the sample population ● Treatment variables and conditions
3. Choose an interviewing method ● Randomization to reduce bias
4. Decide on question order and phrasing ● Replication for result reliability
5. Conduct the interview and collect data Basic Principles of Design of Experiments (DOE)
6. Analyze results ➢ Randomization - Random assignment of
➢ The Scientific Method - A systematic approach treatments to experimental units.
to answering questions and solving problems. ● Importance: Eliminating biases and ensuring
Steps: the validity of comparative experiments.
1. Observation, ➢ Replication - The practice of repeating
2. Hypothesis, experimental conditions to assess the
3. Experimentation, consistency of results.
4. Data Collection, ● Significance: Enhancing the reliability and
5. Analysis, robustness of experiment outcomes.
6. Conclusion. ➢ Blocking - Incorporating additional factors
Number of Factors in an Experiment (blocking factors) to control undesirable
● Independent variables - are factors that variation.
researchers intentionally manipulate to observe ● Application: Controlling sources of noise in
their effect on the dependent variable. experiments, such as age and gender in human
● Extraneous variables - are other factors that studies.
may unintentionally affect the dependent ➢ Multi-factor Designs - Experiments involving
variable but are not the primary focus of the multiple factors or variables.
study. ● Benefits: Studying interactions and main
Design of Experiments (DOE) - A structured approach effects simultaneously for comprehensive
to experimentation for optimizing processes and insights.
solving problems. ➢ Confounding - The intermingling of two or
Purpose: Enhance the efficiency and effectiveness of more effects in an experiment.
experiments. ● Examples: Avoiding confounding in
Key Concepts in DOE: comparative studies; strategic confounding in
● Factorial experiments complex experiments.
● Control factors Steps for Experimentation
● Interaction effects ● Recognition and statement of the problem
● Response surface methodology ● Choice of factors, levels, and ranges
Advantages: ● Selection of the response variable(s)
- Improved process understanding, optimized ● Choice of design
results, reduced experimentation time and cost. ● Conducting the experiment
Engineering Experiments Purposes: ● Statistical analysis
● Reduce time to develop new products and processes. ● Drawing conclusions and making recommendations
● Enhance the performance of existing processes. Factors in Experimentation:
● Improve product reliability and performance. Treatment factors - main variables of interest
● Achieve product and process robustness. Nuisance factors - secondary variables to control
● Evaluate materials,design alternatives,tolerances, etc. Experimental factors - manipulated
Experimental Design - The systematic approach to Classification factors - categorization criteria
planning and conducting experiments. Quantitative vs. Qualitative - Factors can be numerical
Components: Control groups, treatment groups, (quantitative) or categorical (qualitative), influencing
randomization, replication. the choice of statistical analysis methods.