BUSS1020 Chapter Notes - Chapter 1: Design Of Experiments, Sampling Error, Statistical Inference
BUSS1020 – QUANTITATIVE BUSINESS ANALYSIS
CHAPTER 1: DEFINING AND COLLECTING DATA
• Data: the observed values or outcomes of one or more variables
• Variable: characteristics of an item or individual represented in data
• Population: all the items or individuals about which you want to draw a conclusion à N
• Sample: portion of a population selected for analysis à results used to estimate characteristics of pop. à n
• Parameter: a numerical measure that describes a relevant characteristic of a population à purpose of the survey
• Statistic: a numerical measure that describes a characteristic of a sample à estimates a parameter
• Framework for conducting statistical analyses à DCOVA
o Define: the problem or objective, and the data required
o Collect: required data
o Organise: prepare for analysis, tabulate and summarise
o Visualise: the data
o Analyse: the data
• Statistics: methods that collect, describe and transform data into useful insights for decision makers
o Descriptive statistics: collecting (e.g. survey), summarising, presenting (tables, graphs) + organising data
o Predictive: using a model + data to make forecasts of future outcomes
o Inferential: using data collected from a small group to draw conclusions about a larger group
§ Estimation (e.g. estimate pop. using sample average) + hypothesis
testing
• Types of Variables:
o Categorical (qualitative): defined + specific characteristics
§ E.g. marital status, yes/no, gender etc.
o Numerical (quantitative): values represent actual quantities
§ Discrete: counting process
§ Continuous: measuring process
LEVELS OF DATA MEASUREMENT:
• Categorical variables:
o Nominal: lowest level of measurement à distinct categories with NO rank
§ E.g. profession, favourite soft drink
o Ordinal: distinct categories with rank implied
§ E.g. 1. Not helpful 2. Moderately helpful 3. Extremely Helpful
• Numerical variables:
o Interval: Data is numerical and differences b/w values have a consistent meaning
§ No true 0 = no meaning
§ E.g. temperature, time, scaled marks
o Ratio: highest level of measurement
§ Zero has a true meaning = absence of thing being measured
§ E.g. Height, weight, price, profit, age, salary
SOURCES OF DATA:
• Primary Source: analyst collects data
• Secondary Source: analyst is not data collector
o Data distributed by orgs: e.g. industry data
o Data from designed experiment: e.g. consumer product testing
o Survey data: e.g. Political polls, census
o Data from observational studies: e.g. measuring time taken for customers to be served
o Automated data: e.g. Mobile phone + data usage
• Data Format:
o Structured: follows some organising plan à e.g. tables
o Unstructured: follows no repeating pattern à not storable in excel, several locations etc.
TYPES OF SAMPLING METHODS:
• Non-Probability Samples: chosen w/o regard to probability of occurrence à can’t be used for statistical inference
o Convenience: easy selection, inexpensive, quick
o Judgement: experts select most appropriate items/people, by convenience
o Self-Selection: individuals choose to participate
Document Summary
Chapter 1: defining and collecting data: data: the observed values or outcomes of one or more variables, variable: characteristics of an item or individual represented in data. Population: all the items or individuals about which you want to draw a conclusion n. Sample: portion of a population selected for analysis results used to estimate characteristics of pop. Parameter: a numerical measure that describes a relevant characteristic of a population purpose of the survey. Statistic: a numerical measure that describes a characteristic of a sample estimates a parameter. Framework for conducting statistical analyses dcova: define: the problem or objective, and the data required, collect: required data, organise: prepare for analysis, tabulate and summarise, visualise: the data, analyse: the data. Statistics: methods that collect, describe and transform data into useful insights for decision makers: descriptive statistics: collecting (e. g. survey), summarising, presenting (tables, graphs) + organising data, predictive: using a model + data to make forecasts of future outcomes.