date: 2025-01-02
An experimental unit. Person, thing, house, etc. Population: Set of all experimental units. Studying all experimental unit in a population may not be possible. Subset of population sample i.e., Representative sample.
For each experimental unit in the sample we record data on several variables. E.g.: House: Price, location, area etc. Univariate: One variable per unit Bivariate: 2 variable per unit Multivariate: multiple variant
Descriptive and Inferential Statistics
- Picture
- Tables
- Summary Numbers
Inferential Statistics
Make statements about the population based on sample observation. Examples:
- Coin Toss: Population
infinite tosses of the coin with unknown Sample tosses heads. is an estimate of - Suppose 1000 cola consumers a given a blind test among 2 brands (A and B)
- Population all cola consumers: All cola consumers
- Sample: 1000 consumers
- Variability of interest: Preference A/
- Inference: Preference of sample is extended to all Cola Consumers
Type of Data
Qualitative
Example:
- Preference between A or B
- Choice of elective course for student
- Gender
- Favourite
Quantitative
- Age
- Marks
- Salary
- Cost of books
Ordinal
Ordered but not numerical
- Rating like bad average good better
- Grades
Simple Random Sampling Without Replacement = SRSWOR Simple Random Sampling With Replacement = SRSWR
Measure of Reliability
Collecting Data
- Published source
- Designed experiment: Selecting experimental units and administer same treatment EG: Effect of medicine on patients
- Observational Study: What percentage of population has ever shopped online.
Error due to Sampling
- Selection bias: Sample does not cover the whole population.
- Non-response bias: People do not respond
- Response bias (lying): Sensitive question
- Measurement Error: Frame questions in a neutral non-judgemental way or Avoid leading questions
Previous: Statistics Topics
Next: